This PDF file contains the front matter associated with SPIE Proceedings Volume 13064, including the Title Page, Copyright information, Table of Contents, and Conference Committee information.

There exist lots of problems with combined loading and distribution plans with truck stability in practical applications. For example, some goods are heavy in the electric materials industry, so the limitation of truck stability should be considered during the operation of distribution systems. In this paper, an issue about combined loading and distribution plans with truck stability is put forward, and through analyzing the properties of goods’ loading and unloading, a multi-objective programming model is established to achieve truck stability and the minimum times of goods reverse placement. Further, a heuristic method is presented to solve this problem through the proposed four rules.

In order to prevent and reduce the human error of high-speed railway train dispatchers, the method of TOmada de Decisão Interativa Multicritério (TODIM) and Preference Ranking Organization Method for Enrichment Evaluations (PROMETHEE-II) was used to identify the human error risk of train dispatchers. In this paper, the risk attribute set is constructed from the three dimensions of human error probability, human error severity, and human error detection degree. Considering the fuzziness and uncertainty of risk attributes, it is represented by 2-tuples. At the same time, the entropy weight method is used to calculate the attribute weight. In order to reduce the defects of using TODIM and PROMETHEEII methods alone, the two methods are integrated and a new risk identification model is constructed. And the model is applied to the risk identification of human errors in dispatching command operations. The results show that the model can effectively identify the human error risk of train dispatchers, and the top three human error modes are the inversion of the operating procedure, the wrong handling decision and the issuance of invalid orders.

In high-speed trains, aerodynamic forces can have a significant impact on the performance of the brake plate. This paper propose an aerodynamic brake plate for a high-speed train and analyses the high-speed train's aerodynamic performance assessed by the Reynolds averaged Navier–Stokes (RANS) equations based on the k−ω turbulence model. The results show that the effect of the opening of the plate increases the turbulence of the downstream flow field behind the plate, which creates a recirculation zone. The affected area is mainly concentrated in the flow field around the location of the plate for the pressure field and the whole flow field behind the plate for the velocity field. The effect of pressure difference near the plate creates vortices. Additionally, the plate affects the velocity field by causing a significant decrease in velocity upstream, increase in the airflow separation from the plate, as well as a decrease downstream due to a recirculation zone and turbulence. Indeed, the plate also decreases turbulent kinetic energy near it. These findings highlight the importance of aerodynamic design elements, such as the braking plate in optimizing train performance and controlling velocity. Further research in this area can contribute to developing more efficient and streamlined high-speed train designs.

Integrated commissioning and testing is a necessary section of the construction process of China’s high-speed railways, which provides important technical support to ensure the opening and operation. Although the subgrade is a crucial component of the integrated commissioning and testing process for high-speed railways, there has been a lack of systematic analysis regarding the typical problems identified during previous practice. Based on the past testing data of our country’s high-speed railway foundation, this paper systematically summarizes the main typical problems found in integrated commissioning and testing, analyzes the causes behind these problems, and proposes corresponding countermeasures. The objective is to provide guidance for the integrated commissioning and testing of foundations during grading, as well as serve as an important reference for the operation and maintenance of high-speed railway foundations.

With the rapid development of logistics industry, the demand for environmental logistics becomes more and more important. This paper identifies and extracts 16 influencing factors of environmental logistics capability from four aspects: customer demand, environmental protection, development strategy and corporate governance; firstly, the mechanism model is constructed by interpretative structural model (ISM) to obtain the five-level recursive relationship among the influencing factors. Secondly, the cross-influence matrix multiplication for classification model (MICMAC) is applied to explore the driving force and reliability degree of the influencing factors. Finally, the environmental logistics optimization strategy is proposed.

The trend and periodicity in metro passenger flow series data can provide some predictive power. Using AFC data, correlation analysis is used to verify the degree of correlation between current and historical passenger flow data. Through analysis of the original sample data and extraction of passenger flow features, Random Forest (RF), Long Short-Term Memory Networks (LSTM) and extreme gradient boosting (XGBoost) can be constructed respectively. The prediction model was used to forecast the future day rail passenger flow, and the accuracy of the model was evaluated by the mean absolute error, root mean square error and coefficient of determination. The model prediction and evaluation results show that the XGBoost model performs better, and the errors of the evaluation indicators are smaller than those of other models, and the algorithm is more accurate.

It is a feasible attempt to use drones to deliver some precious agricultural products (such as matsutake,) in mountainous, rural or plateau areas. And how to plan a safe and feasible trajectory for drones is the key. Therefore, this article focuses on the path planning problem of drones in three-dimensional (3D) mountainous environments, aiming to improve the efficiency of drone delivery of agricultural products. Firstly, a corresponding 3D map was established based on the mountainous environment. Furthermore, a digital twin model of the logistics drone was established based on the maximum load limit, maximum turning angle, maximum pitch angle, and other limitations. A multi-objective mathematical model was established based on the objective function of drone transportation of fresh agricultural products, aiming to shorten delivery time while improving the safety of drone flight. The fire-fly algorithm (FA) is affected by the mobile strategy in the process of optimization, so it is easy to fall into local optimization when dealing with large-scale optimization problems such as UAV path planning. Therefore, combined with the optimization strategy of particle swarm optimization (PSO) algorithm, this paper designs an improved heuristic algorithm, the improved Firefly algorithm (IFA). Finally, a simulation experiment was conducted to demonstrate the effectiveness of the algorithm.

In recent years, emergencies occur frequently in China. In order to better protect the safety interests of the affected people and improve the level of emergency management, this paper puts forward a more specific and advantageous path planning method compared with the past. Focus on material distribution in smaller areas after a disaster. Taking the time spent in the rescue process and the urgency of each rescue point as indicators, the emergency path planning model is constructed, and the genetic algorithm is used to verify the solution. Take Jiuzhaigou area in Sichuan province as an example to see the feasibility of the method.

The platform detection device is an important device for assisting the train to stop at the station. It is connected to the car body through the mounting bracket, and the entire structure bears both aerodynamic loads from the air and vibration loads from the car body. The bracket, as a cantilever structure, has the risk of cracking. Once the bracket breaks, it will affect the train’s entry into the station and even endanger traffic safety. This paper combines experiments and simulations to study the fatigue strength of the mounting bracket of a platform detection device installed on a certain type of EMU. Firstly, modal and line tests are carried out on the bracket, using the rainflow counting method, referring to IIW standard, and calculating fatigue life of stress measurement points according to Miner cumulative damage theory. It is found that the bracket has not reached the design requirement of 9 million kilometers, and it is found that it is because the first-order mode of the bracket is excited by analyzing test data. Then, based on the experimental modal information, an accurate finite element model is established, and IRF is obtained by using finite element simulation with measured aerodynamic and vibration loads as input. The stress at the measurement point is calculated by using the time-domain convolution integral method, and the test results are reproduced to verify the accuracy of the simulation method. Finally, structural optimization is carried out on the bracket structure, and fatigue life prediction is carried out on the new structure using verified simulation methods. The fatigue life of the new structure is 129 billion kilometers, which meets the design requirements.

To gain a more accurate understanding of maritime traffic flow, this paper proposes a ship traffic flow prediction method based on GTO-CNN-LSTM. The method utilizes ship traffic flow data, which is preprocessed and used as input for CNN-LSTM. By establishing a connection between the input and output through high-dimensional mapping, ship traffic flow is predicted. To further optimize the chosen model, the Artificial Gorilla Troop optimizer (GTO) is employed to adjust the model's hyperparameters, selecting the optimal ones for maximizing the model's performance. The method is validated using actual data from the vicinity of the Port of Los Angeles. The results demonstrate that the model achieves a prediction accuracy of 95.82%. The experiment confirms that the CNN-LSTM prediction model optimized with the GTO exhibits higher accuracy.

Accurate prediction of traffic flow is the basis for realizing intelligent transportation systems. It is challenging to achieve accurate prediction of highway traffic flow because of the characteristics of highway traffic state evolution such as temporal non-linearity and spatial heterogeneity. In this paper, a Conv-LSTM freeway traffic flow prediction model based on an attention mechanism is proposed, which can automatically extract the inherent features of historical traffic flow data. First, the convolution and Long Short-Term Memory model are combined to form a Conv-LSTM module based on the attention mechanism, which could take out the time-space features of the traffic flow data. The attentional mechanisms are designed to identify the importance of different flow series. In addition, the Bi-LSTM module is used to analyze the historical traffic flow data to capture the trend of traffic flow in the forward and backward directions to extract the daily and weekly traffic flow cycle features. Finally, the results show a better prediction performance realized by the proposed integrated model compared to other available approaches.

Road flood damage is one of the common disasters along rivers or in mountainous areas. In recent years, finite volume methods, finite difference methods, discrete element methods, and methods that involve coupling different approaches have been used for simulating road flood damage. Some of these methods have been validated through practical applications in engineering projects. This paper categorizes types of road flood damage, summarizes their causes and introduces some widely used simulation method for road flood damage. A comparative analysis was conducted to show the advantages and weaknesses of the different methods, which may be useful to future research on road resilience under flooding.

In order to scientifically determine the traffic management measures of mixed traffic sections, it is necessary to quantitatively analyze the safety and efficiency characteristics of mixed traffic flow, which includes exploring the main influencing factors and changing rules of traffic safety and efficiency. This paper took the mixed traffic flow on the mixed road section as the research object, and used video recognition to obtain the trajectory of motor vehicles and nonmotor vehicles. The traffic flow operation characteristics were analyzed with the number of traffic conflicts as the safety index and the traffic specific power as the efficiency index. By constructing the stepwise regression model, the main influencing factors of the running state of the mixed traffic flow were determined. The change point theory was introduced to analyze the change trend and state change point of traffic safety and efficiency, and on this basis, the traffic improvement measures were put forward. The results show that the characteristics of mixed traffic flow mainly affect the traffic efficiency of the mixed road section, the characteristics of non-motorized traffic flow and road conditions mainly affect the traffic safety status. The critical non-motorized vehicle permeability of the road sections with the actual operation management level of secondary trunk road and branch road is 0.5 and 0.6, respectively. Before reaching this threshold, the increase of non-motorized vehicle permeability will lead to an increase in the number of traffic conflicts and a rapid decline in traffic efficiency. The recommended speed limit for motor vehicles in the non-mixed road section is 25 km/h, at which the traffic safety and traffic efficiency can reach the maximum. The analysis method used clarifies the influencing factors and changing rules of the operation state of mixed traffic flow, and provides the basis for the planning and management of mixed traffic flow.

The increasing speed and operation density have raised higher challenge to train trajectory tracking problem. In this paper, an event-triggered Tube Model Predictive Control (ETTMPC) controller is proposed, which can handle large frequent external disturbances with lower computational complexity. First, a classic MPC model of High-Speed Train (HST) trajectory tracking is established, then a Tube MPC (TMPC) model is introduced using the robust reachable set method, which improves computational time and space efficiency comparing to the traditional minimum robust positive invariant set (mRPI) method. In order to reduce the trigger frequency, an event-triggered strategy and the ETTMPC control structure are introduced, which only triggers computation when the trajectory deviation exceeds the threshold set. The simulation result shows that the proposed TMPC has higher robustness in large disturbance scenarios, while the ETTMPC has similar tracking performance as the TMPC but with a lower online computational cost.

To ensure the safety of China's grain import sea transportation, the risks affecting China's grain import marine transportation channel are divided into channel condition factors, political environment factors, economic and trade environment factors, and military environment factors, and a risk evaluation index system for China's imported grain marine transportation channels is constructed. After weighting the indicators through the improved CRITIC method based on the entropy weight method, using the principle of system dynamics, considering the COVID-19 and Sino-US trade friction as exogenous factors, a risk assessment model of China's soybean import marine transportation channel is constructed. The risk of China's imported soybean sea channel was simulated, and the risk level of each marine transportation channel for China's soybean import was analyzed and a short-term forecast was made. Finally, the corresponding countermeasures and suggestions are put forward according to the simulation results affecting the risk level of the channel.

The train operations are subject to several uncertain factors, which pose challenges to achieving the train’s multi-objective optimal control. To improve the service quality and transportation efficiency of high-speed railways, this paper proposes a control method to realize the multi-objective operation of high-speed trains in terms of safety, punctuality, energy efficiency, and driving smoothness considering the uncertainty. Firstly, by studying the kinematic model under uncertainty and operational constraints of the train, and combining them with multi-objective optimization theory, an optimal control model for the multi-objective operation of high-speed trains is established. Secondly, to solve the optimal control problem under the uncertainty, the nonlinear and non-convex optimal control problem for multi-objective operations of the train is reformulated as a robust economic model predictive control framework, which is then convexified into a computationally tractable convex model, and then the corresponding algorithm is designed to give the solutions. Finally, the proposed method is validated through simulation on a real high-speed rail line, and compared with tracking-type robust model predictive control. The simulation results demonstrate that the proposed method is capable of achieving safe, punctual, energy-efficient, and smooth operation of high-speed trains in the presence of uncertain factors.

Traffic flow prediction is very important to urban traffic management and public safety, and is also a key link in the development of intelligent transportation system. However, the long-term traffic flow data has complex time characteristics, which brings severe challenges to the traffic flow forecasting research. Deep learning methods are mainly used in time series prediction tasks such as traffic flow prediction. Recently, Transformer-based approaches have been able to improve traffic flow forecasting results, but there are still limitations such as high computational costs and difficulty capturing overall trends in traffic flow data. We propose a multi-scale sensing neural network (MSPFormer) based on Transformer, which uses multi-layer perceptrons to achieve more efficient feature extraction in step prediction, while designing a progressive decomposition architecture to preserve historical information authenticity as much as possible. Detailed experimental studies have shown that our approach can improve the accuracy of traffic flow prediction in Transformer architecture by 4% to 24% without consuming additional computing resources.

With the improvement of urbanization level and the continuous expansion of urban scale, the traditional single-line operation mode has gradually been unable to adapt to the existing diversified passenger flow needs, and various urban rail transit is also transforming to networked and interconnected operation. How to develop a matching train operation plan based on the needs of passenger flow has become the primary problem to be solved. Based on the cross-line OD passenger flow, this paper analyzes the travel choice behavior of different OD passengers to establish a model of urban rail cross-line train operation. Finally, combined with the actual data of two lines in a city that meet the cross-line operation, the feasibility of the model is verified, and the passenger flow conditions required to operate the cross-line train are determined, compared with the single-line independent operation scheme, the implementation of cross-line operation can reduce the operating costs and passenger travel costs of enterprise.

Due to the closed internal environment of the tunnel and frequent traffic accidents, it is difficult for workers to carry out accident rescue work. With full consideration of traffic flow, environmental monitoring data, the ant colony algorithm is improved, and the tunnel topology diagram is established for simulation experiments to form the optimal emergency evacuation path, effectively guiding vehicles to evacuate in an orderly manner.

Traffic flow characteristics is one of the important factors affecting the passage capacity of the waterway, for the Jinzhou port waterway to the port ship traffic flow increase and the development trend of large ships, combined with the waterway, berths, anchorage, and natural conditions and other factors, the use of Anylogic software on the passage capacity of the main channel of the Jinzhou port and the level of service and other simulation analyses. The simulation results show that: when the traffic flow increases to 1.3 times of the current one, the increase of the proportion of large ships has a greater impact on the service level of the port, and it is necessary to increase the berths of large ships or improve the efficiency of berth operation to improve the passage capacity of the waterway.

Graph convolution is a deep learning method for graph data, which can process unstructured traffic network data, which is very suitable for traffic flow prediction tasks. Researchers have introduced graph neural networks to model the graph structure and contextual information within transportation systems, achieving state-of-the-art performance in various traffic prediction problems. In this survey, we provide an overview of the rapidly growing body of research utilizing diverse graph neural network methods across different domains. The paper begins with a review of the fundamental principles of graph convolutional models, emphasizing their advantages in capturing node features and local neighborhood information. Subsequently, the application process of graph convolution in traffic flow prediction is discussed. By utilizing graphs as inputs, certain models based on GCNs have demonstrated improved performance in tasks such as road traffic flow and speed prediction compared to previous methods. Building upon this foundation, various variants of graph convolution are explored, including residual graph convolution, multi-graph convolution, etc., aimed at progressively enhancing the model's representational capabilities. Furthermore, this paper focuses on the experimental results and application effectiveness of the research. Numerous studies have validated the performance of graph convolutional models on realworld traffic datasets. The experimental outcomes demonstrate that, in comparison to conventional methods, graph convolutional models have achieved significant improvements in traffic flow prediction. Nevertheless, challenges such as data quality, real-time applicability, and model interpretability still need to be addressed. Future research can delve into these issues to further propel the application and development of graph convolutional methods.

The loss of containers at sea not only causes serious economic losses, but also causes varying degrees of pollution to the accident area. However, there is a lack of detailed research on the causes of container falling into the sea accidents; Therefore, in order to better understand the causes of such accidents, this article will use 50 reports of container accidents that occurred in different regions as samples, and the 24 model is used to obtain the influencing factors of container falling into the sea accidents; Then establish evaluation indicators for the influencing factors of container falling into the sea accidents, use Analytic Hierarchy Process and Coefficient of Variation method to qualitatively and quantitatively analyze the relevant factors, obtain the subjective and objective weights of different factors, and then calculate the combined weights. The ultimate conclusion is that the factors that have the highest impact on these accidents are individual and management factors. Therefore, in order to reduce the occurrence of such accidents, it is necessary to control human behaviors and continuously improve the safety management system.

With the acceleration of urbanization, the scale of urban population continues to increase, which brings great pressure to the operation and management of urban rail transit, especially the large passenger flow during holidays will affect the normal operation of the station. Taking Chengdu East Railway Station National Day passenger flow as an example, this paper uses AnyLogic to model, analyzes the causes of congestion and puts forward corresponding optimization measures, and then simulates the optimized passenger transport organization to verify the feasibility of the scheme.

The prediction of port channel flow velocity is very important for offshore operation, navigation safety and coastal engineering construction. This paper focuses on the prediction of short-term flow velocity in port waters. By considering the characteristics of time dependence and feature dependence in flow velocity prediction, a hybrid model of CNN, Bi _ LSTM and self-attention mechanism is integrated, and a SA-CNN-Bi _ LSTM model is proposed to improve the performance of traditional models in flow velocity prediction. The port flow velocity prediction model based on deep neural network established in this paper can take advantage of CNN and Bi _ LSTM to extract features

With the rapid urbanization and the exponential growth of the urban population in China, urban road transportation has become the primary mode of travel for residents. However, motor vehicles have led to an alarming increase in greenhouse gas emissions, posing a serious environmental challenge for China. Traditional carbon emission calculation methods are generally used to estimate the total carbon emissions over a macro-region within a specific time period, but they have limitations when calculating the carbon emissions per unit of time for road sections where vehicles pass through. To address this issue, we focus on a typical urban road section and combine it with the carbon emission calculation formula. Under the framework of deep learning, we employ a Convolutional Neural Network (CNN) to study the motor vehicle driving parameter detection method, and conduct research on the carbon emission of motor vehicles on urban roads. The system has a low investment cost and enables the analysis of regular carbon emission patterns at specific locations, providing a solid theoretical basis for establishing an early warning mechanism for excess carbon emissions. In the future, it can evolve into an information exchange platform for transportation and environmental protection departments, offering promising applications and practical value.

Due to the growing importance of maritime container transportation in the logistics system, berth and quay crane scheduling issues have been a focus of research until now. In this paper, we construct a continuous berth and quay crane joint scheduling optimization model considering carbon emission cost. Minimizing the vessel carbon emission cost, terminal operation cost and the cost of time in port for vessel are taken as multiple objectives in the model. After that, the improved immune particle swarm algorithm is used to solve the example of a port in Zhejiang to verify the effectiveness of the model and algorithm in this paper. This paper provides a reference for constructing green port.

In this paper, based on the six-axis inertial measurement unit, the driving behavior data of 30 drivers on a certain bus line were collected for a period of 46 days, and the driving road conditions and working conditions were identified in units of trips. The Mann-Whitney U rank sum test method is used to test 20 features in turn to discuss the driving behavior that affects energy consumption under different road conditions and working conditions. The results show that the features of driving behaviors with significant differences in energy consumption under different road conditions and working conditions are different (such as the upper quartile of acceleration, the mean value of angular velocity difference), and with the increase of intense driving behaviors during the rush hour, the differences of some driving behavior features in different energy consumption levels are more obvious. In addition, based on the analysis results of different road conditions, working conditions and time periods, this paper puts forward corresponding energy-saving driving suggestions for driving behavior factors that affect energy consumption. For example, the driver should try to keep the average acceleration less than 1.42 m/s2 when accelerating to start and decelerating to stop, so as to keep energy consumption low.

Considering the dynamics and uncertainty of passenger and freight demand, this paper proposes an optimization method of rolling stock circulation plan for passenger-freight co-transportation. Based on the flexible composition mode, a twostage stochastic programming model is proposed. The first stage considers the constraints of train composition and rolling stock circulation plan, with the objective of minimizing the train operation cost. The second stage considers the rules of passenger and cargo flow allocation, with the objective of minimizing the delay time of cargo groups and the number of passenger delays. A series of numerical experiments based on the Beijing Subway Yizhuang line are implemented to show the effectiveness of the proposed method, which are solved by Gurobi and variable neighborhood descent (VND) algorithm. The results show that in comparison with the fixed composition mode, the flexible composition mode can better cope with the uncertainty of demand, which can reduce the train operation cost and improve the quality of transportation service.

The requirement for intelligent trains to enable real-time sensing of multi-source information throughout the entire operational process has become vital as the government aggressively encourages the digitalization, scalability, intensification, and synergistic development of rail transportation. This work introduces a fusion method that combines millimeter-wave radar and cameras in order to accurately detect obstacles inside restricted zones and anticipate their direction in real-time. First, during the millimeter-wave radar data collecting process, the threshold approach is used to filter out extraneous obstacle data outside the restricted area. This makes it possible to extract speed and distance data pertaining to obstacles located within the restricted area in front of the train's operation. The Kalman filter algorithm is subsequently used to track the obstacles and forecast their motion. Second, the information obtained by the camera and millimeter-wave radar is combined in time and space. The data is then transformed into a standardized coordinate system after the fusion procedure. It is suggested to use an algorithm based on data fusion to make target matching for the same detected barrier easier. Finally, numerous trials are carried out in various settings and with various lighting levels. Comparisons using only one sensor are part of these experiments. The outcomes show that implementing the millimeter-wave radar and camera fusion method considerably raises the rate of success of obstacle identification in front of the train while it is operating.

Accurate urban traffic management relies on precise traffic predictions. Spatial-temporal graph neural networks, combining graph neural networks with time series processing, have gained popularity for traffic prediction. However, traditional graph neural networks fall short in modeling high-order interactions among multiple nodes. To address these challenges, we present a novel approach, the Heterogeneous Fusion Graph Neural Network (HFGNN). In our model, we integrate hypergraph convolution which can model higher-order relationships between nodes with graph convolution to capture more complex spatial dependencies in traffic network structure. Experiments on real-world traffic data confirm the superiority of our model over state-of-the-art methods.

With the acceleration of urbanization and the continuous improvement of per capita vehicle ownership rate, traffic congestion and traffic accidents have become a global problem, especially highway congestion has become a national problem. The expressway is a bridge to the modernization of a country, and it is the only way to develop the modern transportation industry. The traffic congestion phenomenon on the expressway not only brings great inconvenience to people's travel, but also restricts the improvement of the service quality of the expressway, and affects the regional economic and social development. Predicting highway traffic flow one or more days in advance can not only assist highway management personnel to arrange deployment in advance, reasonably induce vehicle diversion and evacuation, but also provide reference for the public to select travel routes in advance, and is also an effective way to alleviate highway congestion. Aiming at the problems of low fine reading, poor real-time performance, insufficient adaptability and robustness of the current traffic flow prediction model in the field of intelligent traffic prediction, this study proposed an improved graph convolutional neural network for traffic flow prediction in intelligent traffic prediction. The model is composed of GCN and LSTM. Self-attention module is embedded in GCN backbone network to improve the ability of network to extract spatial features. The ECA attention module is embedded in the backbone network of LSTM to improve the ability of the network to extract time features.

Long-term traffic flow prediction plays a crucial role in intelligent transportation systems. However, the temporal nonstationary nature of data from traffic detector nodes and the interdependence among nodes pose significant challenges in this prediction task. Therefore, we propose a traffic prediction transformer framework, called TPformer, to learn the spatiotemporal correlations of traffic features among traffic detector nodes. The success of applying the transformer to traffic flow prediction relies on the following key factors: 1) The length of both the input sequence and the predicted sequence is equal to the number of detector nodes, facilitating the construction of a node-node attention mechanism. 2) A node triplet encoding is introduced to allocate additional local contextual structural features to each node. 3) The topological structural features of the transportation network are incorporated into the self-attention mechanism, allowing the model to focus more on neighboring nodes when encoding the node traffic feature. Experimental results on two real-world datasets from the Performance Measurement System (PeMS) demonstrate that the proposed model outperforms state-of-the-art baselines in both short-term and long-term prediction tasks.

Conducting effectiveness evaluation and analysis of the key elements of rescuing the overboard personnel by search and rescue ships, an evaluation index system for rescuing the overboard personnel is constructed. Considering the randomness and fuzziness of various evaluation indicators for rescuing the overboard personnel, a qualitative concept and quantitative calculation are transformed based on cloud models to establish an effectiveness evaluation model for rescuing the overboard personnel. Analytic Hierarchy Process is used to calculate the weights of each evaluation indicator and combined with the evaluation indicators, upper level indicators parameter of cloud model are calculated. The effectiveness evaluation conclusion of the overall goal of rescuing the overboard personnel is obtained by comparing it with the evaluation standard cloud model.

In view of the high risk of passenger ro-ro ships navigating in the Qiongzhou Strait and the serious consequences of accidents, factors such as channel passage conditions, ship organization and dispatching, and the influence of natural conditions on navigation safety are integrated, and the risk factors of passenger ro-ro transportation in the Qiongzhou Strait are identified based on accident cause theory, risk identification theory, and risk factor checklist. According to the safety risk factors, a relatively complete safety risk assessment index system for passenger and ro-ro transportation in the Qiongzhou Strait is constructed from the five aspects of ships, cargo transportation, transportation organization management, natural environment and personnel management, and game theory is used to combine the hierarchical analysis method and the entropy weighting method to determine the weight of the safety risk assessment index. The risk analysis method of LEC operating conditions is used to comprehensively evaluate and calculate the risks of all aspects of passenger ro-ro transportation in the Qiongzhou Strait from the three aspects of the probability of occurrence of safety risks, the degree of loss of safety risks, and the frequency of exposure to the risk environment. Combined with the research on the emergency rescue systems and mechanisms of the two shores and three places, research conclusions are given and relevant countermeasures are proposed to provide emergency protection for the construction of the “Safe Strait”.

As a key technology for maritime applications, trajectory prediction can effectively help ships reduce risks such as collisions and groundings at sea. Currently, although the combination of rich automatic identification system (AIS) data and deep learning brings new possibilities for ship trajectory prediction, ship trajectory prediction is still hugely challenging due to the complexity of ship motion. In this paper, we improved a new ship prediction model based on TrAISformer. On the one hand, we sparse the multi-dimensional data through dictionary coding, map it into probability space, and use a new loss function to measure network performance. On the other hand, we proposed a An MLP algorithm enables it to effectively learn the timing characteristics of AIS data while avoiding destroying encoding correlation. Finally, we conducted experiments on public AIS data, and the experiments showed that the improved model improved the prediction performance by about 12.5 % compared with TrAISformer.

This paper takes Z hazardous chemicals enterprise as an example, analyzes its sales and transportation problems, and optimizes the sales and transportation. The optimal hazardous chemicals sales plan is obtained with the goal of minimum transportation cost. According to the obtained sales plan results as input, a heuristic algorithm is designed to minimize the time for the last carrier to leave the refinery. Provide decision support for the sales and transportation of hazardous chemicals enterprises, and provide feasible solutions for the sales and transportation of hazardous chemicals.

In order to reveal the ship collision causation propagation law, a ship collision causation propagation model is proposed in this paper based on the complex network dynamics theory. The model improved the Susceptible—Infected—Recovered (SIR) model and makes it applicable to the dynamic process of risk propagation. In order to apply the model to the actual network, we construct a ship collision causation network. The network is used as the simulation object of this study. The simulation results show that the causative nodes of different importance have the ability to propagate the risk to the whole network. The impact of accident causation can be reduced by shortening the inspection time of accident causation, reducing the risk propagation rate, increasing the risk control rate, and selecting appropriate initial control factors. The model can help to simulate the propagation of the causes of ship collisions, and can quantitatively analyze the impact of the causes on ship collisions.

In the coal scheduling process of the rail-sea intermodal coal supply chain for large energy enterprises, trains are easy to accumulate in front of the port. The train dispatching efficiency is low, and coal railway transportation has weak anti-risk capabilities. In order to solve these problems, an optimization method of coal transport trains scheduling is proposed. Considering the current situation of long coal railway lines and long train scheduling cycles, a segmented optimization method is proposed to optimize train scheduling. Using the intersection of railway lines as the dividing point between the upstream and downstream of the railway, a segmented optimization model for coal transport train scheduling is constructed. The model aims to increase railway transportation efficiency and comprehensively considers constraints such as railway capacity, train operation rules, station loading and unloading capacity etc. The results of train scheduling optimization in rail-sea intermodal coal supply chain of a large power generation group in China show that: This model can effectively address railway congestion and high port inventory issues caused by uneven traffic flow. Through the scheduling optimization of trains, the congestion time of trains is effectively reduced, the accumulation of trains in front of the port is prevented, and the transportation efficiency is improved.

With the development of computer and cloud computing related technologies, cloud products play a more important role in the economic and cultural activities of enterprises, organizations and individuals. In recent years, numerous cloud service providers have emerged globally to offer cloud products to consumers. Cloud infrastructure is the physical computers that cloud service providers rely on in order to provide cloud products, and cloud infrastructure provides resources such as CPU, GPU, memory, hard disk, and network for cloud products. It can be said that cloud infrastructure is a particularly important aspect of cloud computing. When cloud service providers manage the supply chain of cloud infrastructure, they often have too much or too little inventory, which results in a waste of resources or a failure to meet user demand. Therefore, how to accurately predict the inventory demand of cloud infrastructure supply chain has become a problem that cloud service providers need to solve. In this paper, we propose a GA-LightGBM model for predicting the inventory demand of cloud infrastructure supply chain by comparing multiple models. In this paper, GA-LightGBM is experimentally verified and analyzed with the control model, and it is found that the average RMSE of GA-LightGBM is 358.8792, which is significantly higher than that of each model in the control group; in the process of multiple training and validation, the range of the RMSE and the standard deviation of the RMSE of GA-LightGBM are significantly smaller than that of each model in the control group. It can be seen that the GA-LightGBM model has higher prediction accuracy and stability than the models in the control group. It is recommended that cloud service providers adopt this model to optimize the management of cloud infrastructure supply chain inventory demand forecasting, so as to improve enterprise benefits. The GA-LightGBM model proposed in this paper will also complement research in related areas.

Aiming at the problem of reducing the instantaneous noise of vacuum circuit breaker closing, the noise reduction system for rail transit vehicles is innovatively designed by using TRIZ theory. Through function analysis and cause-effect chain analysis, this paper finds the root cause of the problem, solves the problem and puts forward the scheme by establishing the technical contradiction, physical contradiction and su-field model, and finally puts forward the design of the noise reduction system, so as to provide a theoretical basis for the development of subsequent products.

Based on the entropy weight method (EWM), the coupling theory and synergetics theory, this paper aims to evaluate the synergy degree of the composite system of cold chain logistics and fresh agricultural product circulation in a quantitative way. Through the construction of the evaluation index system, development index model, the coupling coordination degree model of the composite system, the synergy degree model of the composite system and synergy degree evaluation standard, this paper puts forward an evaluation model and takes a certain province as an example to carry out empirical research. The evaluation model is of great reference significance for quantitatively measuring the synergy degree of cold chain logistics and fresh agricultural product circulation and for determining the type of synergetic development. By doing so, this paper helps decision-making for further promoting the synergetic development of the composite system.

With the rapid construction of China's high-speed rail, the gradual improvement of network layout, and the increasing track connections in the station, issues emerge such as the difficulty in selecting the stop train strategy of the station and the ambiguity of checking the calculation of station carrying capacity. In view of this situation, a carrying capacity calculation model based on multi-objective optimization and simulation verification is proposed for the railroad station integration scheme. Firstly, a multi-objective optimization calculation model is constructed to determine the objective function and constraints related to the issues. Secondly, the train tracking scenario simulation is applied and realized, and then the train operation plan is inferred based on the algorithm to realize the construction of the simulation environment for the cluster operation of the EMU. Finally, a high-speed railway intermediate station in southern China is taken as the simulation target, and the track connection scheme of the station is used as the simulation basis, so as to map the line information and abstract it for verification based on the real railroad network environment and station equipment data. The test results show that the multi-objective optimization and simulation verification model can effectively help develop a reasonable stop train strategy for the station and thereby obtain the station carrying capacity.

In order to improve the use of roads and solve the huge pressure brought by the influx of a large number of private cars on the existing traffic and transportation network due to the development of the economy and the improvement of people's living standards1 , thus leading to more and more large and medium-sized city roads facing traffic congestion and increasingly serious status quo, to achieve the purpose of reducing traffic congestion, this paper uses an intelligent algorithm - this paper uses an intelligent algorithm - the seagull optimisation algorithm - to optimise the driving route of in-car navigation, so as to achieve the shortest possible vehicle driving time and driving path, reduce traffic pressure and improve the current situation of traffic congestion.

Vehicles and roadside of the guardrail terminals collision is extremely dangerous and complex accident. This paper presents a new type of the guardrail terminals structure, utilizing software PRO/E and ANSYS establish a finite element model of guardrail terminals subject to impact loading, and a reasonable simulation test. The results show the new guardrail structure has a perfect energy-absorbing ability, security and reliability, can absorb most of the collision energy and avoid the serious occurrence of guardrail terminals to insert the interior.

To address the challenge of determining the optimal median opening length in the multi-lane conversion of expressway renovation and expansion projects, this study presents a novel universal calculation model. The model is developed based on an analysis of vehicle driving stability and the geometric relationship of the road alignment at the median opening conversion section. Furthermore, a fuzzy comprehensive evaluation model is established to assess the traffic operational status of vehicles traversing the median opening conversion section, utilizing the developed median opening length model. Taking into consideration both safety and efficiency, key indicators, including time occupancy rate, speed consistency, and travel delay, are selected for evaluation. The research results demonstrate that the developed model for the length of the median opening in multi-lane transition can ensure the operational status of vehicles in the transition section. This contribution has significant implications for the planning and execution of expressway renovation and expansion projects, providing valuable insights for determining the optimal median opening length.

To improve the rationality of the connection between large appertaining parking facilities and urban road traffic, and to solve the problem of rapid access, this paper proposes a method for optimizing the layout of parking entrances and exits that focuses on the synergy of safety and efficiency. Firstly, the number of entrances is restricted and optional locations are determined based on specifications. Then, an objective function is constructed to minimize the total layout cost as the sum of conflict risk and travel time cost, considering road flow and service level constraints. And a parking lot entrances and exits layout optimization model is established from the perspective of safety and efficiency synergy. The genetic algorithm is used to solve the multivariate non-linear objective function and an optimization scheme for the layout of entrances and exits of an appertaining underground parking facility is proposed, with model parameters and traffic flow sensitivity analysis carried out. Finally, VISSIM simulation experiments are conducted to evaluate the effect of the optimization scheme in terms of safety and efficiency improvement. The results show that the optimized scheme has good adaptability to real-life traffic scenarios, reducing the average queue length at entrances and exits by 47.33% and overall vehicle travel time by 4.72%, as well as reducing the total number of conflicts by 10.38%. This disperses conflict points and improves safety, while reducing delays and improving parking efficiency. The method can provide a reference for the layout of entrances and exits of parking facilities in the planning stage, further improving the coordination between static and dynamic traffic in the city.

Fatigue Risk Management (FRMS) for flight crew is a new data-driven approach to fatigue risk management that has been promoted by international civil aviation in recent years, and the collection of crew sleep data is an important task. The use of reliable wearable devices for sleep monitoring is a more objective and convenient recommended practice than manual recording, so it is necessary to conduct validation experiments on such devices in practical application scenarios. To verify the effectiveness of a HUAWEI smartwatch with a sleep monitoring module applied to FRMS-targeted pilot sleep monitoring, a comparative validation experiment was conducted in this thesis among 18 pilots on four ultra-longrange routes. Subjects wore both a certain HUAWEI smartwatch and an ACTIGRAPH somatic tachograph and recorded subjective sleep diaries. The comparative analysis of the 94 sets of sleep data collected showed that there was good agreement between HUAWEI and subjective sleep log results, and the deviation of HUAWEI and ACTIGRAPH did not meaningfully affect the fatigue prediction results of the bio-mathematical model in FRMS. In addition, the in-flight sleep data comparison results demonstrated that the sleep monitoring accuracy of HUAWEI and ACTIGRAPH in the cabin environment was comparable to that in the ground environment.

With the continuous expansion of shipping demand, the problem of environmental pollution is becoming increasingly severe. An effective strategy for mitigating carbon emissions is the implementation of well-designed shipping routes. To address this concern, we construct a branch transportation network under the hub-and-spoke route structure and then formulate a mixed-integer programming model with the aim of minimizing overall operational costs. A genetic algorithm is designed to solve the model and determine the optimal combination of liner routes. Taking the main ports in the Bohai Rim region as an example, a series of numerical experiments are carried out. The results demonstrate that the deceleration of navigation is a feasible approach for carbon emission reduction.

To study the effect of autonomous vehicles on the characteristics of mixed traffic flow, and make up for the lack of actual scenarios on the road, the simulation approach is proposed. Firstly, we chose the Intelligent Driver Model and Longitudinal Control Model as the autonomous and manual vehicle car-following models, and set the detailed model parameters. In addition, we take highway as the research object, and used the simulation of urban mobility (SUMO) software to model three different types of scenarios. Finally, each scenario is simulated after setting different parameters, the autonomous vehicles proportion increases gradually from 10% to 90% at the rate of 10%, and collected the corresponding data to compute the travel time. Conclusions can be drawn through analysis, the travel time decreased as the proportion of autonomous vehicles increased under the same traffic volume, this result further proves that autonomous vehicles can improve road efficiency.

With the rapid development of shared bicycles in recent years, it not only facilitates urban residents, but also brings several problems to urban traffic, mainly in disorderly parking. In order to solve the disorderly parking of shared bicycles, this paper takes the Yangguang Campus of Wuhan Textile University for example and considers the trip demand of teachers and students. We choose the cosine distribution time satisfactory function as the service quality level function to quantify the quality of service, which can reflect the distance between demanding points and parking sites to users’ satisfaction. Then we establish a multi- objective integer planning location-selection model based on gradual coverage method, which aims at maximizing service quality and minimizing the number of parking sites. We use LINGO to calculate the model with actual survey data. The results show that the solution of the location-selection model is close to the reality and of certain practical significance.

Train braking capacity is an important performance index of train, which directly affects the safety performance of the train. The choice of braking strategy has a direct impact on the braking ability of the train. Existing braking strategies often do not balance availability and safety. This paper analyzes the advantages and disadvantages of existing braking strategies according to these two evaluation methods. And a new braking strategy is proposed on the basis of the existing braking strategy, which is based on permanent magnet linear eddy current braking. This new braking strategy increases the availability and safety of train braking systems at the same time, which can effectively ensure the safety of the train after increasing the speed.

Aiming at the problem that excessive temperature of brake disc may cause traffic accidents due to thermal decay, thermocouples are used to measure the instantaneous temperature of vehicle brake during braking, and the accurate instantaneous temperature after interference is obtained through analysis and processing. The heating model and cooling model are established based on the heat conduction principle with friction instantaneous temperature as the parameter, and the final brake temperature monitoring model is obtained. By bench test, the maximum relative error between the predicted result and the test data is 9.3%, which shows that the model is accurate and feasible.

Reasonable traffic organization and signal-coordinated control scheme is one of the valid ways to improve the efficiency of arteries. Through the investigation of the current geometric layout, traffic flow, phase setting, traffic operation and control scheme of the three intersections in the west section of Longmen Road in Tianjin. Conduct an arterial green wave control study for these three intersections. Greenwave control signal timing scheme design by Synchro software. Based on the control program, the VISSIM software simulates the road sections and intersections before and after optimization, and outputs the travel time, average queue length and delay time. The results of the study show that trip times and average queue lengths at all three intersections were reduced during the morning peak, flat peak, and evening peak hours. Some evaluation metrics improve significantly. Minimum improvement of 10.6% and a maximum of 66.48% in total roadway delay, which proves the effectiveness of green wave control in improving the running conditions of the road section.

Subway bogie is a key infrastructure. The frame is an important bearing part of the subway bogie, and its mechanical properties have a wide impact on supporting the car body and ensuring the stable operation of subway vehicles. In order to further improve the fatigue reliability of the bogie frame to provide data support, the article is based on the bogie of a certain subway vehicle, analysis of its frame in the design and manufacture of the strength of the reliability of the bogie frame according to the EN13749 and UIC615-4 standard bogie frame static strength and fatigue strength test. First of all, through the examination of the static strength of the frame in the corresponding load, and then analyze the fatigue strength under a certain number of fatigue cycles. According to the analysis of the test results, the static strength of the frame meets the test requirements, and the fatigue strength is assessed based on the results of the static strength test conducted by simulating the operating conditions, and the stress amplitude at each measurement point is within the envelope of the fatigue limit Goodman diagram of the material, which meets the fatigue strength requirements. After 10 million cycles of loading, no cracks were found in the non-destructive flaw inspection, and the fatigue test of the frame was qualified.

Shipping industry is the lifeline that responsible for 80% of the total global trade. At the same time, environmental pollution and greenhouse gas emissions caused by the port and shipping industry have become the focus of attention of the international community. In order to promote green, low-carbon and sustainable development of waterway transportation, a port-ship multi-energy integration system has been constructed by using three renewable energy including wind power, solar energy, hydrogen, and one energy storage technique. Taking China as an example, this paper analyzes the adaptability of 4 ways of energy supply in the design of multi-energy integration system of port-ship theoretically and then the AHPTOPSIS method is adopted to evaluate their adaptability from policy and standard support, technological development, energy availability, economic efficiency, and environmental friendliness perspectives. The methodology of this paper can be extensively employed on performance evaluation of multi-energy integration system from national or port level.

In order study and determine the better traffic organization mode of double-decker highway, and make up for the lack of actual scenarios, the simulation approach is applied. Firstly, according to the basic conditions of the highway, the upper is passenger car type, the lower is passenger and freight mixing type, and the upper and lower are passenger and freight mixing type are selected. The harmonic average speed and time to collision (TTC) are selected as the efficiency and safety indicators, and give the detailed calculation formula. What’s more, combined with the actual road structure and parameters, the simulation model is built by using the simulation of urban mobility (SUMO) software, and set the parameters. Finally, each traffic organization mode is simulated, and collected the corresponding data to compute the indexes. Conclusions can be drawn through analysis, although there are some shortcomings in the traffic organization mode 1, it has advantages in traffic efficiency and traffic flow stability.

EMU spare parts are important materials to ensure the safe and stable operation of EMUs, which need to maintain a relatively high level of inventory. Besides EMU spare parts center inventory management is relatively independent, while the spare parts replenishment lead time is volatile, such as imported parts are easy to supply untimely. In this paper, we study the problem of inventory control of EMU spare parts, formulate a reasonable inventory level to ensure supply, and reduce the risk of stock-outs. Introduce transshipment to realize inventory sharing among EMU spare parts centers to reduce inventory costs, and we consider the impact of replenishment cycle differences on the stochastic demand for EMU spare parts. The study shows that there is a unique optimal inventory level under the transshipment model, and the corresponding inventory levels should be set according to different replenishment lead times. The transshipment inventory model under centralized decision-making reduces the out-of-stock quantity while reducing the total cost compared to the non-transshipment inventory model under decentralized decision-making.

Considering the limited effectiveness of a single traffic control strategy in alleviating traffic congestion at merging bottleneck on highway, this paper establishes a merging bottleneck model on a multi-lane cell transmission model (CTM) and proposes an integrated control method of variable speed limit (VSL) and lane change control (LC). Feedback-based VSL limits the upstream traffic demand and harmonizes the traffic operation state, while the LC uses lane change logic to change the distribution among the lanes of the bottleneck upstream, mitigating the mutual disturbance between the mainline and ramp traffic and minimizing the negative impact of capacity drop. Numerical simulations demonstrate the effectiveness of the proposed integrated control method.

This paper studies the integrated optimization of two-way truck scheduling for import and export container collection and distribution of port and surrounding hinterland with time window constraints. A multi-dimensional time-space-state network was constructed based on the physical network to dynamically reflect the spatio-temporal trajectory and loading conditions of trucks. Aiming at maximizing the container volume of collection and distribution, minimizing the total transport cost and the amount of trucks dispatched, a multi-objective integer programming model was formulated to optimize the truck scheduling for port container collection and distribution based on service network. Taking the short distance import and export container collection and distribution in Shanghai Port of the Yangtze Delta as an example, the Gurobi9.5.2 was used to solve the optimization model. The results indicated that the multi-dimensional time-space-state network can rationally plan the truck routes and collecting and distributing process, thus achieving the integrated optimization of two-way truck scheduling for port container collection and distribution with time window constraints while dynamically reflecting the spatio-temporal trajectory and loading conditions of trucks. These inferences are expected to serve as a theoretical support for sustainable source of port collection and distribution.

Cold cities are located in the Northern Hemisphere, where travel conditions in winter are often harsh, mainly due to weather conditions and seasonal restrictions. Although there are efforts to plan and construct subways, light rail, and other rail transportation, as well as promote emerging public transportation methods like Bus Rapid Transit in cold cities, the progress is relatively slow. In cold cities, traffic congestion is more severe and complex than in non-cold cities due to significant surface transportation pressure. Therefore, to accurately understand the traffic characteristics and needs of cold cities, it is necessary to formulate effective transportation development strategies to enhance the performance of the transportation system and meet the travel needs of residents. This research takes transportation in cold cities as its subject, focusing on current problems and travel characteristics. The study covers planning, design, and operation phases, proposing indicators for the development of transportation in cold cities. The goal is to explore methods and indicators to solve the current problems. In addition, this research aims to provide guidance and technical support for planning managers, policy makers, and urban managers to better cope with the transportation challenges in cold cities.

Disasters generate a large amount of material and human damage. Evacuation planning plays a crucial role in disaster management by minimizing the adverse effects of disasters on communities. It is acknowledged as a challenge of muti-objective optimization, encompassing conflicting spatial objectives and constraints within a decision-making context. Addressing such problems through traditional means is arduous. Yet, metaheuristic methods have shown to provide effective solutions. This paper delves into the examination of designing and analyzing evacuation routes within transportation networks during disaster situations. Furthermore, the paper introduces an optimized method for planning evacuation routes in context of disasters. The technique utilizes an iterative bio-inspired heuristic algorithm, which outlines paths from disaster areas to each shelter site. This aids in allocating vehicle flow for efficient evacuation planning, all while considering travel time as a key parameter.

With the development of the new generation of information technology, the traditional ship field is increasingly integrated with artificial intelligence, big data, communication and navigation technologies. It is obvious that the Maritime Autonomous Surface Ships (MASS) is currently in focus world-wide and the research on this topic is growing. However, the fully autonomous operation of MASS still faces many challenges of key technologies, such as target detection, path planning, collision avoidance, etc., among which path planning has become one of the most critical problems. In this paper, a novel path planning method for ships, named Sector-AStar (S-AStar), is proposed. Taking the characteristics of ship motion into consideration, this method improves the search strategy based on the ship heading angle. It can be seen from the simulation results that the proposed Sector-AStar method can effectively shorten the decision time and increase the safety of ship navigation. Moreover, the path turning angle and the path tracking deviation of Sector-AStar are both smaller than AStar, which is more conducive to ship motion control.

In the context of considering the Emission Control Area (ECA) of a ship, the navigation cost of a ship is optimized through speed optimization to reduce the operating cost of the ship. A speed optimization model is constructed to minimize the operating cost when the ship passes through the ECA. Meanwhile, the dung beetle algorithm is adopted to find the optimal solution for the speed optimization model, and a case study is conducted to verify the feasibility of the model by taking a container ship as an example. It is found that a low speed in the ECA area is conducive to the reduction of the cost, but it is necessary to control the speed reasonably, and too low a speed will cause the operating cost to rise instead.

To reduce the costs associated with empty container repositioning and to lower the operational expenses of shipping companies, this research developed a model that accommodates multiple container types and factors in the costs associated with either premature or delayed arrivals of empty containers at yards and ports, encompassing storage fees and opportunity loss. Furthermore, it delved into the practical issue of container type substitution, establishing a model for sea-land transportation under multiple container type substitutions. The total costs realized under container substitution proved to be lower than those without such a strategy, underscoring the efficacy of employing

Logistics distribution center is an important part of logistics system, whether its location is reasonable will directly affect the operational efficiency and economic benefits of logistics system. Scientific and reasonable location distribution can effectively reduce transportation costs, save distribution time and improve logistics service quality. In this paper, taking the location of distribution center of R enterprise as an example, the combination of hierarchical analysis and gray correlation method is used to clarify the relative relationship between the elements and the target decision by the gray correlation method, while the relative importance between the elements is chosen to analyze by the hierarchical analysis method, and finally the comprehensive correlation degree is calculated. The fusion of these two methods is an advanced algorithm.

In view of the shortcomings of the traditional A* algorithm, such as multiple expansion nodes, a large angle of path inflection point and an uneven path, the evaluation function is improved from the heuristic function and the dynamic weighting method is used on the basis of the traditional A* algorithm and the Bessel curve is used to smooth the planned path. The traditional A* algorithm and the improved A* algorithm are simulated and analyzed in Python environment and the improved A* algorithm is applied to the path planning of the left turn condition of the driverless vehicle at the intersection. The results show that the improved A* algorithm reduces the search expansion nodes, so that the search speed is improved, the search time is shortened and the path is smoother.

Under the operating conditions of urban rail transit network, the scale of the network is constantly expanding, and sudden events such as large passenger flow impact and equipment failure are easy to affect the stable operation of the rail transit network, resulting in disruption events. The operation disruption event will not only cause the train delay on the current line, but also affect other lines, forcing a large number of passengers to delay their journey, choose detour or even give up travel. Taking passengers affected by disruption events as the research object, this paper adopts cumulative prospect theory to describe passengers' path selection behavior under operation disruption events, establishes a scenario prospect probability model by using value function and cumulative decision weight function, and proposes a passenger path selection model under interval disruption scenarios to ensure passengers' travel to the maximum extent and reduce the impact of interval disruption events. It has a certain reference value for the emergency management of urban rail transit under section interruption events.

The selection and matching of electric vehicle power system is a key issue in the development process of the whole vehicle. According to the vehicle parameters and performance indicators, the key parameters of the motor and the power battery pack are determined by matching and calculating the power system. Based on the cruise simulation software, build the electric vehicle simulation analysis model, and take the power and economic simulation analysis. According to the comparison between the simulation results and the experimental data, the rationality of the simulation model is further determined, which lays a foundation for the subsequent simulation optimization.

Urban waterlogging is often accompanied by rainfall, to study the impact of urban waterlogging on evacuation traffic optimization, this paper introduces the equation of "rainfall intensity-waterlogging depth-free-flow speed", and proposes a dynamic loading link transmission model(LTM) considering the impact of rainfall and waterlogging, and designs a phased algorithm to address the problem of time-varying free-flow speed caused by differences in rainfall intensity and dynamic changes in waterlogging depth. Taking the Nguyen-Dupius road network as an example, the results show that: with the increase in rainfall intensity, the growth rates of total travel time and network clearance time increase gradually, while the growth rates gradually decreased with the increase in waterlogging depth; The saturation in water-logged links is higher than that in water-free links, and congestion is more likely to occur in case of heavy rainfall and severe waterlogging.

The research and development technology field of 5G vehicle networking traffic safety mechanism project is a new information technology industry supported by the state. Artificial intelligence, vehicle networking, and intelligent transportation are the core areas of China's "national 2025 manufacturing" strategy^[1]. Safety early warning mechanism plays a very important role in preventing vehicle collision accidents and improving road traffic safety. By studying the anti-collision system and safety early warning system of networked vehicles at home and abroad, focusing on the collision problem of vehicles at traffic crossings, this study proposes a vehicle collision early warning algorithm based on trajectory prediction, and establishes vehicle trajectory models, including straight-line driving trajectory model and turn-around driving trajectory model respectively^[2,3]. The time and position of the collision are calculated and the threshold is set to issue the collision warning. The Matilab experiment simulation is used to evaluate the algorithm, and a visual and auditory integrated vehicle network safety warning system is designed combined with the driver characteristics and warning results.

In the context of the increasing unbalanced passenger flow in both time and space, the conventional fixed formation mode is no longer able to cater to the dynamic demands of passengers. Thus, this paper presents a mixed-integer nonlinear programming model that aims to optimize train scheduling, rolling stock utilization planning, and passenger flow control strategy by incorporating virtual formation mode, in which the virtual formation enables trains to modify their composition through coupling/uncoupling operations at stations. The model with nonlinear constraints is subsequently converted into a model with linear constraints, making it amenable to direct solution using Gurobi solver. Finally, a series of numerical experiments are conducted using actual data from the Beijing Changping metro line. The computation results show that the proposed method effectively meets passenger demands while simultaneously ensuring reasonable reductions in operational costs.

With the continuous growth of the oversea automobile market, the freight rate rise and capacity shortage have occurred in the commercial vehicle market. In order to meet the export demand, many automobile companies began to order Rollon/Roll-off (Ro-Ro) Ships shipbuilding, hoping to control the transport capacity by establishing their own fleet. Stowage planning for Ro-Ro Ships is critical to ensure safety at sea. In order to improve the efficiency and safety of the stowage planning for Ro-Ro Ships, this paper presents an automatic stowage planning model. This model combines the improved discrete particle swarm optimization (DPSO) algorithm and the lowest horizontal line search algorithm as the encoding and decoding operations. The algorithm efficiently provides the deck stowage plan of the specialized ship, meeting different loading requirements. Finally, the algorithm proposed in this paper is compared with the improved genetic algorithm (GA) and simulated annealing (SA) algorithm through simulation, proving that the algorithm can make reasonable stowage plans while ensuring the safety requirements of ship navigation, and preliminarily solve the automatic loading problem of large-scale Ro-Ro ships.

In order to address the issue of passenger flow imbalance when metropolitan commuter trains cross-operate onto urban rail transit lines, this paper adopts a demand-driven approach. Optimization models are constructed with the goals of minimizing the waiting time for passengers on urban rail transit and minimizing the total travel time for passengers on the metropolitan line. An improved Harris Hawk algorithm is utilized for solving these models. To validate the effectiveness of the proposed models and algorithm, actual operational data from a specific metropolitan line and urban rail transit line are examined. The results indicate that by uniformly adjusting the departure intervals of both the metropolitan and urban rail transit trains, the objective function is reduced by 7.20%, the waiting time for passengers on the urban rail transit line is decreased by 4.96%, and the onboard travel time for metropolitan line passengers is reduced by 11.28%.

After the wheels of rail vehicles have been running for a long time, the polygon effect often causes a series of changes in the dynamic response of the vehicle-rail system, and the vibration response of the vehicle and the track also becomes more intense. This will cause the degradation of the vehicle dynamics performance, and cause intense noise of the wheels and the track, which in turn affects the comfort of the ride and the stability of driving. In order to study the impact of wheel polygon on the dynamic performance of high-speed trains, based on the theory of vehicle system dynamics, this paper takes the CRH3 rolling stock as a prototype, and studies its vehicle dynamic performance under the condition of wheel polygon wear. Firstly, a vehicle-rail coupled dynamics model considering rail, wheelset and axle-box flexibility is established to study and analyze the vehicle dynamics performance under wheel polygonal wear, which mainly includes vehicle stability, ride comfort, axle-box vibration acceleration, wheel-rail force and axle dynamic stress. It can be seen through the analysis that low-order wheel polygonal abrasion has a small impact on the stability and ride smoothness of the vehicle; high-order wheel polygonal abrasion has a high excitation frequency of the wheelset, and its transverse component makes the vehicle dynamics deteriorate slightly at high speeds. The high-frequency wheel-rail vertical force due to wheel polygonal wear causes large axle-box vibration acceleration and axle stress fluctuation.

The swift growth of e-commerce has intensified the pressure on the express delivery industry in terms of timely delivery. Therefore, it becomes imperative to explore the implementation of a joint distribution model that enables more efficient and sustainable delivery of express shipments. Considering factors such as the capacity of delivery stations and the number of transport vehicles, this study proposes a two-level delivery location selection model under the joint distribution mode, aiming to minimize the total cost and carbon emissions of the delivery station location selection. The proposed model is solved using the Improved Multi-objective Particle Swarm Optimization (IMOPSO) algorithm. According to the research results, by adopting the joint distribution model, the total cost can be significantly reduced to 63,867.05 yuan while also achieving a substantial reduction in carbon emissions to 997,945.81 kg. Furthermore, the facility utilization rate exceeds 90%. Notably, the IMOPSO algorithm exhibits notable advantages in terms of both accuracy and computational efficiency.

In order to solve the problem of insufficient quantitative analysis of the impact of air passengers' travel characteristics on airport landside transport mode choice behavior, we deconstructed air travel characteristics based on three aspects: personal attributes, travel attributes, and transfer scenarios, elucidated the impact of travel characteristics on passengers' actual transfer behaviors, and utilized an artificial neural network (ANN) model to discriminate the degree of importance of travel characteristics variables, and established a passenger transfer transport mode choice multinominal logit (MNL) model, and finally calibrated and tested the model using passenger questionnaire data from a large Chinese airport in 2019, revealing the behavioral characteristics and marginal effects of airport passengers' transfer choices. The results show that: passenger gender, education, and travel purpose differences all have a weak effect on interchange transportation mode choice, and flight time has a non-significant effect on passenger choice; when the probability of taking airport buses decreases continuously with the gradual increase in the number of annual air trips of the passengers; as the time point of the passenger's transportation mode decision is moved further back in the future, the more they are inclined to take cabs, the more they exclude the airport buses; when there are people traveling with them, the passenger's The probability of choosing a cab will become 2.2 times when passengers change from leaving directly to planning to leave after a stopover at the airport, the tendency of choosing airport buses versus online taxis will be more obvious.

During a large-scale epidemic outbreak, the strategic planning of storage and distribution locations for essential life supplies becomes of paramount importance. This article focuses on the scientific supply of epidemic-related goods, taking the case of Changchun city as an example. It considered the human resource costs involved in the supply chain and constructed an organized network diagram for the supply of essential life supplies to residents. The results indicate that through the analysis of the shortest paths, we derive the optimal number of distribution points for essential life supplies in various administrative regions and outlines detailed plans for the provision of these supplies to residents. This article establishes a solid theoretical and practical foundation for future emergency management strategies.

This paper proposes an evaluation method for the level of bus and slow traffic system connection around rail stations, considering the necessity, feasibility, accuracy and other attributes of the indicators, and determines seven constraint-guaranteed indicators and five quality-improved indicators, which can reflect the level of bus-slow traffic connection in a more comprehensive way; and gives the calculation method of each indicator, so that we can get the quantitative results of the calculation; and then determines the weight value of each indicator. Then, according to the entropy weighting method and the hierarchical analysis method improved by spherical fuzzy set, the weighting value of each index is determined. Finally, the fuzzy comprehensive evaluation method is used to get the evaluation level of corresponding stations. Finally, the targeted evaluation of different types of rail stations is realised, which can provide the basis for the subsequent adjustment of bus lines and the construction of connecting facilities.

The second question of the first question of the second question of the Chinese 2022 Mathematical Modeling Contest for University Students requires how to use the minimum signal to obtain the accurate positioning of the UAV when the UAV cluster is flying in formation. The paper uses the theory of elementary mathematics, uses the angle and side length relationship between multiple UAVs, and applies the sine theorem and cosine theorem to establish the bearings-only passive positioning model. The model can be solved by only two transmitters for accurate positioning, Through practice, the model has a high fitting degree, and finally realizes the positioning of UAV.

The 16-line LiDAR-based lane detection method is inexpensive and widely used, but it has the following problems. The point cloud distribution is sparse and unable to distinguish between solid and dashed lines when the number of detection frames is too small; when the number is too large, the data processing efficiency is low. Therefore, it is necessary to find the minimum number of detection frames required to distinguish between solid and dashed lines. To address this problem, this paper established a model to calculate the minimum number considering the factors such as lane, LiDAR installation height and vehicle speed. Based on the simulation results, this paper summarized a calculation method of the minimum number suitable for a variety of common working conditions.

With the increasing passenger flow in urban rail transit and the design capacity of existing train control technology tends to be saturated, virtual train formation technology has become one of the research focuses of new urban rail train control technology due to its ability to significantly improve the capacity of existing lines and effectively solve the problem of passenger flow imbalance during a specific time span. But the actual running environment of urban rail transit trains is complex. Trains are affected by basic resistance and unknown additional resistances in the lines. At the same time, the end-to-end communication mode adopted between train consists causes communication delays, making trains within the consist unable to obtain real-time data from currently communicating trains, which affects consist running safety. To solve the above problems, this paper proposes a train consist cooperative control method based on the "leader-follower" consist control model and dynamic consist interval control strategy with fixed headway. A state delay correction observer is designed to estimate the actual state of the trains to ensure consist running safety, in view of communication delays within the consist. Finally, a distributed sliding model controller is designed to ensure the coordinated running of the consist.

In order to be able to accurately assess the Human Error Probability (HEP) of high-speed rail train dispatchers. a HEP quantitative evaluation model based on Improved Weighted Cognitive Reliability and Error Analysis (CREAM) was constructed, which corrected the shortcomings of traditional CREMA. Using bipolar 2-tuples as the Common Performance Condition (CPC) evaluation linguistics, the subjective and objective weights of CPC are calculated through the Analytic Hierarchy Process and the Criteria Importance Though Intercriteria Correlation (CRITIC), and then the combined weighting method is used to further obtain the comprehensive weight of CPC ; At the same time, the idea of group decision making is used to reduce the subjectivity of CPC evaluation, and the weight of experts is calculated by using the adaptive dynamic weight adjustment method of grey correlation degree; on the basis of obtaining CPC and expert weights, the evaluation value of CPC is calculated. The weighted operation then obtains the Context Influence Index (CII) value; finally, the quantitative calculation of HEP is obtained by using the CII value. The model is validated by taking the high-speed train dispatcher's normal and abnormal train reception as examples. The research results show that: under normal conditions, the HEP of high-speed rail train dispatchers is 9.1586×10^-5, and under abnormal conditions, the HEP of highspeed rail train dispatchers is 2.1189×10^-3. Under abnormal circumstances, the human error probability of high-speed train dispatchers is more than 20 times that under normal circumstances. At the same time, compared with other methods, the model has certain advantages in terms of weight sensitivity, data utilization sufficiency, and model solution domain. The improved CREAM model can better reflect the impact of different CPCs on high-speed rail train dispatchers, and can provide a calculation method for the HEP quantification of high-speed rail train dispatchers.

The BPR function is a type of road impedance function, as one of the impedance functions of road sections, is mainly used to allocate traffic according to the calculated impedance value, and decide the path selection in the allocation process. However, the model parameters of the traditional BPR function are derived from fitting a large amount of traffic flow data of intercity highways in the United States, which are not fully applicable to China. Considering the great similarity between traffic flow and pedestrian flow, the optimized BPR function can be applied to Barkhor Street in Lhasa to calculate the road impedance. In this paper, firstly, based on the pedestrian flow data of Barkhor Street, the Greenhill model and reduction coefficient method are adopted to optimize the BPR function, so that the model parameters suitable for Barkhor Street are obtained. Secondly, the optimized BPR function is used to evaluate the impedance of each road on Barkhor Street. Finally, a preliminary crowd evacuation plan is developed with the impedance as the evaluation index.

Accurate identification of urban functional areas is of great significance to the scientific management and intelligent services of cities. Aiming at the problems of low accuracy and subjectivity of current identification methods, a quantitative identification method combining random forest and frequency density algorithm is proposed. Firstly, the frequency density calculation method is constructed by assigning weights to 19 POI types through random forest, and then the spatiotemporal feature analysis is performed on the distribution of urban functional areas identified in 2016 and 2021, respectively; secondly, the influence of rank scaling is considered, and Jingjiang Palace Street is used as an example for functional area identification; finally, a test set is selected for validation and comparison experiments. The results show that the accuracy rate of the method in this paper reaches 85.55%.

In order to achieve secure and efficient data sharing and exchange between railway and other industries, based on the business characteristics of the railway industry,we established a blockchain-based railway data sharing and exchange system. In this paper, we proposed a data storage and aggregation scheme that combines on-chain and off-chain data, and an improved consensus algorithm for data sharing based on the characteristics of railway blockchain node structures. The railway industry data exchange and sharing blockchain model proposed in this paper ensures data security while meeting the needs of secure and efficient data exchange between railways and other industries.

With the continuous integration of new technologies and the transportation industry, the high-precision spatial-temporal trajectory data of motor vehicles obtained by traffic data collection technologies such as Radar Vision Integration and Holographic Perception contains abundant traffic state information, which provides solutions for traffic state cognition and active traffic management. This paper aims to realize the continuous perception of the multi-level operation state of the intersection by using the high-precision spatial-temporal trajectory data, so as to provide the basis for the low latency optimization control of the intersection. This paper proposes about extracting indicators of control delay, queue length, and number of stops of intersection based on highly precise map and trajectory data. The distribution characteristics of traffic operation parameters extracted by the algorithm are analyzed. The analysis results show that the calculation results of the indicator extraction algorithm in this paper are in accordance with the time-varying characteristics of traffic demand at the intersection and the spatiotemporal characteristics of intersection signal control scheme and channelization scheme. Taking the intersection of Jintong East Road and Guanghua Road in Beijing as an example, using the evaluation method in this paper and the simulation evaluation method based on VISSIM, the evaluation indicators are extracted and the evaluation grade of operation index is output from the four periods of trajectory data. Through the comparative analysis of the evaluation results of the two methods, the feasibility and validity of the evaluation method in this paper are verified.

The boundaries of typical scenes need to be extracted from the original scene data occurring in real roads, but the manual scene labeling method is inefficient and costly. Therefore, it is important to study the acquisition of autonomous driving scenes in real roads for the testing of autonomous vehicles, and the cut-in scenes are typical hazard scenes. In this paper, a scenario acquisition system with multi-source heterogeneous sensors is modified based on a vehicle equipped with L2- level autonomous driving system for acquiring natural road driving scenario data. Subsequently, a cut-in scenario automatic recognition algorithm based on risk assessment is proposed, and a recognition algorithm evaluation test platform is constructed, and a large number of test case generation procedures are designed to simulate and test the algorithm. And a large amount of natural driving data is used to verify the proposed recognition algorithm. The results show that the algorithm proposed in this paper can effectively identify the cut-in scenes occurring in front of the collection vehicle and has high accuracy.

In order to ensure the safe operation of cloud-based safe computing platform in train control system, it is crucial to accurately predict cloud platform resources. Due to the different number of trains at different times and sections, the signal software load in cloud based security computing platforms also varies. Resource data often accompanies high randomness and non-stationary characteristics, increasing the difficulty of prediction. A hybrid prediction model based on Sparse Search Algorithm (SSA), Variational Mode Decomposition (VMD), SE module, and Time Convolutional Network (TCN) is proposed to address the low prediction accuracy of traditional prediction methods and the lack of research on cloud platform in train control system. First, the original data is decomposed using the VMD optimized by SSA, and then each component is predicted using the SE-TCN model. Finally, integrate the prediction results of each component to obtain the final cloud resource usage prediction result. The experimental results show that, compared with other models, the Mean absolute error (MAE) of the proposed model on the memory data are reduced by 27%, and the CPU data are reduced by 24%, with high prediction accuracy.

The variation in lighting conditions significantly affects the performance of deep learning-based visual object detection algorithms. Under strong light, weak light, and dark conditions, the algorithm's effectiveness tends to decrease. This paper proposes a neural network module based on luminance attention mechanism, which adaptively adjusts the network's attention weights for different regions of an image based on their luminance. It enhances the network's attention on areas with significant differences in luminance and background, thereby improving the robustness of the object detection algorithm to lighting changes. Experimental results on the DAIR-V2X vehicle-road cooperative dataset demonstrate that the average mAP@.5 of the YOLO series object detection network is improved by 4.54% with the addition of the proposed luminance attention mechanism module. Qualitative and quantitative analyses validate that the proposed algorithm module enhances the performance of YOLO series object detection networks in scenarios with lighting variations.

This paper addresses the capacitated hub-and-spoke airline network design problem under disruption. Specifically, this paper extends the classical hub-and-spoke airline network design problem by considering hub (airport) capacity and various disruption risks. In particular, this paper proposes a backup hub strategy to protect the designed airline network against the disruption risk. This paper formulates a mixed-integer nonlinear programming (MINLP) model for the considered problem for minimizing the sum of the construction cost of the backup hubs and the expected total transportation cost of the new airline network formed by backup hubs after disruption. Through a series of numerical experiments based on a simulated example, this paper verifies the practical value for the capacity constraints of the hub airports to enhance the ex-ante resilience. This paper also conducts sensitivity analysis on inter-hub discount factor and probability of hub disruption to provide managerial implications for airline company.

This paper proposed the architecture of Chinese digital highway standard system. Firstly, the connotation of the digital highway was analyzed. Secondly, the standard system of other fields was analyzed, and the current situation and demand for the standard system of the digital highway were studied. Next, the model and architecture of the digital highway standard system were proposed. Finally, recommendations for the construction of a digital highway standard system were given. The research results of this paper can help to promote the standards development, and further enable the construction of digital highways.

In order to solve the problem of high inventory cost existing in the EMU spares SCN, this paper takes the selection of spares center location and EMU model assignment as the optimization content, and takes the total cost of supply chain as the optimization goal. In addition, we propose the network structure of spares supply chain of EMU considering inventory pooling of the same model and establishes a mixed integer programming model. At the same time, based on the nonlinear characteristics of the objective function, the Generalized Benders Decomposition algorithm is used to solve the model, and the numerical experiment and case analysis are done subsequently. The results show that the proposed SPs SCN structure can significantly reduce the total spare part supply chain cost while maintaining a relatively high supply service level to satisfy demand from rolling stock depots and railway bureaus.

Roadside intelligent sensors can monitor road traffic status, and the installation method of sensing equipment greatly affects the richness and accuracy of traffic data information collection. The radar-video integrated machine integrates radar and camera, and is a new type of perception device. This paper conducted experimental analysis on the installation position, rotation angle, and other aspects of the radar-video integrated machine. Based on the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS), an optimal solution was proposed. Firstly, before designing the installation plan, analyzed the main factors affecting the installation plan, which were used as a single variable to design 5 sets of operating conditions. Next, data collection and analysis were conducted based on the built the Radar-Video integrated machine platform, and 9 indicator parameters were selected to construct an evaluation system. Finally, comprehensive evaluation and optimization were conducted on each installation plan. This paper selected two on-site scenarios for experimental verification, and the results show that the proposed evaluation technology can successfully optimize the installation plan. However, according to different installation scenarios, such as lane width and installation height, the optimal installation plan may be adjusted.

The traditional Douglas Pucker (DP) ship trajectory compression algorithm usually suffers from high trajectory distortion and the loss of important trajectory "inflection points". The compressed data cannot restore the true traffic flow characteristics. In order to improve the quality of ship trajectory compression, this paper proposes a trajectory compression algorithm based on SH-DP, which adds two important latitude information, namely ship speed and heading, to the traditional DP compression algorithm. Firstly, by traversing all trajectory points on the trajectory, the distance, mean speed difference, and maximum steering angle are calculated and compared with their respective thresholds to segment the trajectory. Secondly, trajectory feature points are filtered and retained for the segmented sub trajectories. Finally, compressed trajectory point data is output in chronological order, which not only preserves more trajectory feature points but also improves trajectory compression rate. The experiment was conducted using ship trajectory data from the Baoshan Waterway in Shanghai. By comparing with traditional compression algorithms, the results showed that the ship trajectory compressed by the method in the article was closest to the original trajectory.

Traffic safety problem has been highly concerned by people all over the world. Predicting potential traffic accidents can help to improve traffic facilities or emergency system, and give people alerts to potential dangers. As a random event, the occurrence of traffic accidents has obvious seasonal and spatial characteristics. The accuracy of traffic accident prediction can be improved by combining the seasonal and spatial characteristics of traffic accidents. In the study, we selected Philadelphia car accident databases during 2008 to 2012 on Opendata Philly. On the basis of data grouping preprocessing, we tested three machine learning algorithms to predict the count of accidents. The results indicate that LSTM(Long Short-Term Memory) algorithm has the best performance relatively, and XGBoost (eXtreme Gradient Boosting) model does not perform better than ARIMA(Autoregressive Integrated Moving Average model. In addition, we identify the important features of traffic accidents in Philadelphia city: the count of accidents on the same road segment, traffic control device, automobile type, roadway surface condition, weather condition. The datasets in this study has obvious spatial characteristics also, and the occurrence of accidents is closely related to the accident segment.

The selection of the highway corridor is very significant in the early stage of highway construction, and its rationality determines the economy, environmental protection, and safety of the whole life cycle of the construction project. Existing corridor selection methods are mainly based on the subjective experience of designers and lack intelligent corridor selection technology with strong operability. Therefore, a corridor selection method for new highways based on Nondominated Neighbor Immune Algorithm (NNIA) and Affinity Propagation (AP) clustering algorithm was proposed. Firstly, the mathematical model of the corridor was built. Secondly, considering the environmental and economic factors during the corridor selection stage, the optimization objective function was established. Lastly, the corridor was generated by using the NNIA and AP clustering algorithms. This method was applied to the project from Dezhou to Shangrao Highway from Qimen to the boundary of Anhui and Jiangxi provinces. The corridor generated was consistent with the trend of the actual route of the project, which had certain engineering feasibility.

To solve the problems of low recognition accuracy and slow detection of crew fatigue driving behavior in the cockpit of ships during the process of sailing in and out of the port, the SSD model was studied. By replacing its backbone network and improving the prior frame generation mechanism, the MV-SSD model is proposed. Replace the backbone network VGG16 in the original SSD model with MobileNetV3, reducing the network parameters of the backbone network. Using the K-means algorithm to cluster the real detection boxes in the face area dataset, the prior box allocation mechanism of the SSD model was redesigned, reducing the number of prior box generation by nearly half, and the ERT algorithm in the Dlib library is combined to locate the face key points, and finally, the PERCLOS criterion is used to determine whether the driver is fatigued. Experimental results show that the average accuracy (mAP value) of the MV-SSD model is 7.15% higher than that of the original SSD model, and the detection speed (FPS value) is increased by 98frames/s, which is more suitable for the detection of the crew face area, and the average accuracy of the constructed fatigue detection algorithm for fatigue features is more than 94%.

To analyze the robustness of the Chongqing metro network under emergencies, this paper takes the Chongqing metro network as the research object. Utilizing complex network theory, a Space-L model of the Chongqing metro network is constructed. The topological characteristics of the network are quantitatively analyzed, and the robustness of the network under different attacks is examined through simulation experiments. The results indicate that the Chongqing metro network exhibits strong robustness against random failures, while its connectivity declines rapidly under deliberate attacks. Therefore, it is recommended to enhance the inspection and maintenance of key stations in the Chongqing metro and formulate backup plans to ensure the service level of the network to the greatest extent.

In order to further respond to the national policy and improve the monitoring capability of the sealing status of inland bulk carriers on the Yangtze River, this paper proposes the C-YOLOv5 model on the basis of the YOLOv5 algorithm. The elkan K-Means clustering algorithm is introduced to optimize the target candidate frame so that it can adapt to the detection environment of small targets while ensuring the recall and accuracy of recognition. In order to improve the focus on the detected targets, the SE focus mechanism module is introduced in the head of the model. In addition, some of the ordinary convolutions in the structure are replaced with depth-separable convolutions to further improve the detection speed. The experimental results show that the detection accuracy and speed are improved by using the C-YOLOv5 model, with an accuracy as high as 91.1% and a speed as high as 66.5 f/s. The detection accuracy and retrieval rate of the C-YOLOv5 algorithm both reach 90%. The C-YOLOv5 algorithm's ability to recognize the target state is also significantly improved under poor line-of-sight and dark light test conditions.

Collision accidents of container ships occur occasionally, in the past decades. Due to the fast speed and large opening deck of container ships, it is easy to generate serious consequences in the case of collision. This paper studies the dynamic response, structure failure and energy absorption of container ships in collision. The material failure criterion affects the response and failure prediction of hull structure under collision scenario. The hull structures are subjected to complex stress state under collision load, so the failure forms are different. In this study, two material failure criteria are investigated in the container ship collision simulation. One of them is a constant failure criterion that is simple in form and widely used (Hereinafter referred to as CS criterion). The other one is a dynamic failure criterion which considers the biaxial stress and strain rate effect (Hereinafter referred to as D-BWH criterion). According to the simulation results, it is found that the response of collision force, energy absorption and damage degree of the side structure are different. The CS criterion underestimates the crashworthiness of the ship structure.

In order to quickly mine hotspot areas from taxi trajectory data, so as to help taxi drivers improve the efficiency of passenger search and economic income, this paper proposes a grid-optimised DBSCAN algorithm. Firstly, the study area is divided into a number of grids of the same size, then the trajectory data is mapped to the grid by a function, and finally clusters are generated based on the width-first search, which reduces the search range of the boundary points of each point, and forms clustered hotspots based on the clustering of high-density grid cells. In this paper, we use taxi trajectory data in Chengdu city to conduct experiments and compare with three clustering algorithms, DBSCAN, GSCAN and K-DBSCAN, and the results show that the algorithm in this paper has a smaller time complexity, and the algorithm's DBI index is reduced by an average of 35.25%, and the average running time is reduced by 19.14%.

With the rapid development of China's high-speed rail, efficient train operation adjustments are crucial for enhancing railway service levels and capacity utilization. To address this, this paper introduces and develops a train operation adjustment model and algorithm, considering a comprehensive set of service level evaluation indicators. Initially, based on the actual characteristics and adjustment requirements of railway train operations, a series of service level evaluation indicators were established. The fuzzy comprehensive evaluation method was adopted to determine the weights among different indicators, establishing a holistic evaluation system, which allows for a comprehensive and accurate assessment of train operation adjustment quality. Moreover, to handle potential individual train delays and clustered train delays in real operations, this study utilizes the event-activity network to construct the train operation adjustment model. This model takes the key indicators from the service level evaluation as its objective function, considering constraints like train operation timings, track crossover, and station capacity. The paper designs a concurrent solving strategy to swiftly determine objectives for different adjustment schemes. Using a major Chinese high-speed rail line as the study case, results demonstrate that, when facing unexpected delays, the proposed models and algorithms can rapidly generate required train adjustment plans, restoring the train operations to normal order.

The Neural Controlled Differential Equation (NCDE) elegantly fuses dynamical systems with deep learning, unveiling profound potential for time series modeling. Harnessing the power of neural networks to sculpt the vector fields inherent to differential equations, this methodology introduces a seamless perspective for emulating spatial-temporal dynamical paradigms. In our research, the NCDE serves as the foundational architecture. Within this construct, we adeptly integrate both Temporal Convolutional Networks (TCNs) and Graph Neural Networks (GNNs) into a framework of continuous state representation to capture long-term spatial-temporal dependencies. This avant-garde modeling approach illuminates new avenues for nuanced modeling of spatial-temporal traffic dynamics, markedly augmenting the fidelity of traffic forecasting. Empirical evaluations conducted on three publicly accessible traffic flow datasets further underscore the superior efficacy of our proposed model in traffic forecasting.

Weather is one of the important factors affecting the use of shared bicycles. This paper uses the initial data of shared bicycles and the geographical location relationship of public transportation stations to extract the transfer volume between shared bicycles and public transportation, and builds an autoregressive model of weather factors and transfer volumes. The research results show that the weather has a greater impact on the transfer volume between shared bicycles and public transportation than the total travel volume; the impact of weather factors on short-distance transfers of shared bicycles is greater than that of long-distance transfers. The above findings are helpful for the operation and management of shared bicycles, rationally dispatching shared bicycles in adverse weather, and promoting low-carbon travel.

With the rapid development of the rail transit industry, the running speed of vehicles has been continuously improved, and the operating environment of different types of vehicles is complex and changeable, which makes the load on the vehicle structure randomly variable, and the possibility of fatigue failure is also increasing, which seriously affects the driving safety. Therefore, it is of great practical significance to study the fatigue life of rail vehicle structural parts by vibration fatigue life evaluation method. In this paper, the random vibration fatigue life prediction method is studied in depth based on the application background of rail vehicles and the vehicle body with large bearing structure. The rigidflexible coupling dynamic model and longitudinal dynamic model of high-speed train with flexible body are established. And a finite element model of the high-speed train body was also established. Given the limitations of the measured load spectrum of the vehicle body, a low-frequency load spectrum inversion method for the measured acceleration of the vehicle body is proposed, providing a new approach for the inversion of the vehicle body load spectrum. Based on the proposed load spectrum inversion method, under the excitation of the Wuhan Guangzhou Railway track spectrum track, the load spectra of the center of gravity of the vehicle body, including pitching, swaying, expansion, nodding, rolling and yawing, were obtained. The fatigue life assessment of the vehicle body was calculated under various line conditions under this load spectrum. The results indicate that the cumulative damage to the door and window angles of the vehicle body is significant, and maintenance should be emphasized.

Urban rail transit, characterized by safety, efficiency, punctuality, and speed, plays a significant role in urban transportation, facilitating people’s mobility. However, the operation of urban rail transit consumes a consider- able amount of energy, resulting in high operating costs and impeding its further development. In this study, focusing on the operation within a single train interval, an improved genetic algorithm is employed to optimize the train’s speed profile, aiming to minimize the traction energy consumption within the interval. Additionally, the deviation in speed profiles due to variations in train’s basic resistance parameters is analyzed, and corresponding correction schemes are proposed. Simulation results demonstrate that a single traction/braking scheme in the train correction plan leads to lower traction energy consumption and better alignment with the operational requirements of urban rail transit.

The traditional Hausdorff distance only considers the position information when calculating the similarity of ship trajectory, and rarely considers the characteristic information such as ship heading and speed. Secondly, the traditional Hausdorff distance is easily affected by the loss of track points when measuring ship trajectory, resulting in the increase of the distance between ship trajectories. In order to measure the distance between ship trajectories more accurately, this paper proposed a trajectory similarity measurement method based on improved Hausdorff distance. The algorithm took two important behavioral characteristics of ship heading and speed into account. It makes the distance measurement between ship trajectories more accurate and solves the problem that the distance between ship trajectories increases due to the loss of ship trajectory points in the traditional Hausdorff distance measurement. The ship trajectory data from Shanghai South Channel are used in the experiment. Compared with the traditional hausdorff distance, the results show that the clustering results of the proposed method can effectively distinguish the differences in the behavior characteristics of the trajectory

With the continuous expansion and development of cities, the traffic congestion in cities is becoming more and more serious, which leads to a series of urban diseases. Among them, improving the efficiency of urban traffic and alleviating the traffic congestion becomes the important problems to maintain the sustainable development of urban social economy. Based on the volunteer travel data provided by GeoLife project, the circuity index of travel path is calculated according to GPS track data of each trip, and the path efficiency of each trip is quantitatively evaluated. Then, the relationship between path efficiency and travel distance is analyzed according to different travel modes, and the most efficient travel mode is found for different travel distance. Through the study, it can be found that subway is the most efficient travel mode in the average path, and the relationship between path efficiency and travel distance is complex and nonlinear. Specifically, when the urban travel distance is 2 ~ 18km, 18 ~ 20km, 20 ~ 22km and more than 22km, the most efficient travel mode is subway, taxi, bus and private car. Therefore, we need to vigorously advocate the public transport travel modes to ensure the sustainable development of cities. The results confirm that "giving priority to the development of urban public transport " (including subway, bus, taxi, etc.) can relieve the pressure of urban traffic and improve travel efficiency.

The distribution of passenger car flow on intercity highways is crucial for traffic planning and management, and analyzing its intrinsic influencing factors is also urgent for solving various traffic problems. In this paper, the Random Forest algorithm is used to identify the key factors affecting the distribution of different passenger car flows, and the degree of dependence of the distribution of different types of passenger car flows on different factors is further analyzed through visualization. The results show that: tertiary gross product and urban population are the significant features affecting the traffic volume of small-sized buses; gross product and urban population are the significant features affecting the traffic volume of medium-sized buses; tertiary gross product and urban population are also the significant features affecting the traffic volume of large-sized buses.

The Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) based on results of Parallel Simulated Annealing Algorithm (PSA) is extended to explore satisfactory solutions with relatively low costs for trading off benefit of multi-stockholder while aircraft landing. The Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) based on results of Parallel Simulated Annealing Algorithm (PSA) is extended to explore satisfactory solutions with relatively low costs for trading off benefit of multi-stockholder while aircraft landing. PSA is developed to get a more ideal solution than the regular Simulated Annealing (SA). The numerical simulation shows that our approach is effective and efficient.

Real-time hydrometeorological data can provide meteorological early warning services for ships, ensure navigation safety, and improve the navigation capacity of the channel. The meteorological conditions in mountainous areas are complex, with significant changes in water levels and foggy conditions. Real-time monitoring of hydrological and meteorological conditions is necessary. To solve the problem of limited data transmission in mountainous areas, this article plans to use the short message communication function of the BeiDou system to achieve real-time data transmission. Based on the characteristics and format of BeiDou short message communication, protocol extensions have been made based on BeiDou 4.0 protocol. In addition, to improve the reliability of data transmission, message error correction and data retransmission mechanisms have been added to the communication process of the BeiDou system's short message communication function. Research has shown that utilizing the short message communication function of the BeiDou system can achieve and be suitable for the transmission of hydrological and meteorological data in key sections of mountainous areas with poor mobile network communication.

In recent years, the state pays more and more attention to emergencies. When emergencies occur, quick and efficient emergency rescue is the most concerned issue of the state and people. In many emergencies, casualties and property losses are mostly caused by inefficient logistics distribution and incomplete emergency treatment system. The design of material distribution scheme has become a key link in the emergency management system. It is of great significance to study the layout and distribution path of emergency resource center. In this paper, the distribution method and design of class A materials were taken as an example to analyze the distribution of materials. That is, the emergency resource center was modeled first, and the lowest total cost of the emergency resource center was taken as the objective function to build the optimization model of facility layout. Then, the optimal path was selected again considering the damage of different sections suffered from emergencies and the degree of traffic jam at different time nodes. Finally, the reasonableness of the model construction is evaluated through the case analysis of field research in Region B. The research of this topic has practical significance to solve the distribution of class A materials. Starting from the selection of emergency resource center, this paper gradually explores the reasonable layout of emergency resource center and the re-selection of distribution path.

Precipitation creates many uncertainties in travel times on urban road networks. Travel time prediction involves a multiparameter non-linear mapping relationship, so artificial neural networks are widely used in prediction. Artificial Intelligence This paper designs a travel time calculation method based on non-minimum sections, integrating floating car data and precipitation data. Accordingly, we propose a travel time prediction model based on Kalman Filter and design its effective algorithm. Finally, we use ten-day GPS data gathered from taxis in Zhongguancun West district in Beijing to verify the effectiveness of our proposed approach. These results show that the accuracy of this prediction algorithm in guaranteeing a prediction accuracy of more than 90% is 0.75 and the improved Kalman filter travel time prediction has higher accuracy than traditional.

Real-time intersection control based on actual traffic flow detection data is one of the basic measures to optimize traffic operation, and in the process of green light dissipation of queued vehicles, if the unsaturated headway time distance cannot be effectively judged, it will lead to early interruption of the green light in real-time response control decision, and then affect the control effect. In this paper, we propose a method to determine the unsaturated headway properties of vehicles by using a multi-parameter combination of joint probability, and establish a joint probability model based on the unsaturated headway and average vehicle speed to realize the judgment of unsaturated headway properties in real-time response control. The experimental results show that the method can accurately detect the abnormal headway time distance, and the difference between the calculated results and the real value is within 5%, which is more accurate and reliable compared with the single headway time distance detection method.

The road traffic in a certain period of time in the future will be composed of Human-Driven Vehicle (HDV) and Connected and Autonomous Vehicle (CAV) as heterogeneous traffic flow, and the lane change decision model of CAV in the interweaving area is established for the heterogeneous traffic flow environment. The CAV-HDV lane-changing decision model is established based on the free and forced lane-changing behaviors of CAV in the intertwined area, and the CAVCAV lane-changing decision model is established by considering the minimum safety distance, and introducing the idea of game theory. The CAV lane change decision model is validated by using SUMO traffic flow simulation software, and the results show that the CAV lane change decision model established in this paper has higher passing efficiency and lane change success rate compared with the traditional gap threshold acceptance model under high CAV permeability.

In the paper, the simulation approach is proposed to analyze traffic flow characteristics under different mixed traffic scenarios. Firstly, the Intelligent Driver Model and Krauss models are chosen as the autonomous vehicle car-following model and manual vehicle one, and the models are described in detail. What’s more, the simulation of urban mobility (SUMO) software is applied to build three different types of highway scenarios, the two models are distinguished by setting different parameters. Finally, simulation program is run for each scenario, and the autonomous vehicles proportion increases gradually from 10% to 90% at the rate of 10%. The analysis results show that the average speed on the road increased as the proportion of autonomous vehicles increased, the conclusion proves that autonomous vehicles can improve road efficiency.

Autonomous vehicles improve the safety and efficiency of vehicles in complex traffic scenarios through autonomous decision-making intelligence technology. To address the requirements of the self-driving vehicle lane change scenario for the accuracy of vehicle lane change trajectory prediction, in this paper, we propose a lane change trajectory prediction method for self-driving vehicles based on inverse reinforcement learning. We model the inverse reinforcement learning process through a maximum entropy mechanism to learn the optimal reward function that infers the potential end targets during the vehicle lane change. This reward model is used to construct the optimal policy that can be sampled for planning in the grid world. Conditioned on the sequence of state actions sampled by this maximum entropy policy, we generate vehicle lane change prediction trajectories. We conduct training experiments on lane change scenario data from the publicly available nuScenes dataset for autonomous driving, which shows that our method can meet the vehicle lane change requirements in real scenarios and validate the accuracy and reasonableness of the lane change trajectories.

In order to solve the problems of insufficient sharing and multiplexing of integrated transportation planning services, weak ability of multi-user large-scale concurrent services of high-frequency data and spatial data, lack of customized sub-scene collaborative application functions, design and develop Transport planning analysis and data service platform(i.e., “the platform”) based on flexible architecture technology. The platform bases on SpringCloud mature distributed micro-service technology framework, integrates advanced digital and intelligent technologies of comprehensive transportation planning space-time big data governance, land space resource elements and visualization, initially builds a basic data center for transportation planning, a technical support experimental platform for transportation space planning and an electronic map for space planning that integrates data analysis, platform support, and model application. The construction and application of the platform comprehensively improved the scientific nature of the evaluation of comprehensive transportation spatial planning, and has initially played an important role in assisting the connection of transportation sector planning and territorial spatial planning.

Traditional testing methods no longer meet the requirements of ICV testing. Scenario based testing can perfectly solve this problem. The testing scenarios must come from real traffic data, and how to extract valuable scenarios from the massive actual collected data is a key issue. This paper proposes an automated scenario extraction method that accurately identifies five typical scenarios based on LiDAR target data. Based on the fragments of the right of way competition, determine whether the different key parameters of various scenarios have reached the threshold, in order to intercept and output valuable typical scenarios. Finally, the extraction results are verified by stratified sampling method, and the scenario recognition accuracy of the data segment is obtained by weighted calculation. The validation results indicate that the method designed in this paper has extremely high extraction accuracy and can effectively and correctly extract five typical scenarios.

During vehicle operation, occlusion phenomena occur, and the sensing device acquires incomplete image information, posing a new challenge to the recognition of vehicle information. Based on the convolutional and anti-convolutional structure of deep learning, we designed a contextual self-encoder to propose a vehicle de-occlusion recognition method based on contextual information. We constructed a vehicle dataset with 11 categories. Occlusions were generated using the PIL library, and the occlusion region was removed with a contextual self-encoder, improving the recall from 45.16% to 86.25%. To test the generalization performance of the algorithm, we designed experiments to analyze the effects of special weather, occlusion area, and occlusion location on the algorithm's generalization performance, respectively. After de-occlusion, the recall improved from 18.44% to 75.31% in special weather; increasing the occlusion area improved the recall from 33.59% to 85.63%; changing the occlusion location improved the recall from 44.38% to 86.56%. The experiments show that the method has good generalization performance and provides technical support for recognizing obscured vehicles.

This paper aims to realize precise stop control of high-speed trains with the air-electric combined characteristics under complex scenarios, including external disturbances and the time delay of the braking system. To describe the train braking process, a dynamic model and a braking system model which refine the air and electric braking methods as well as their three working modes are established. For better control performance, multiple disturbances and uncertainties during the actual stopping process of high-speed trains are analyzed and accurately modeled. Then, applying the principle of terminal sliding mode control, meanwhile, combining with parameter adaptive and fuzzy control methods, the refined adaptive terminal sliding mode controller (RATSMC) is designed. The simulation results show that the stopping precision of RATSMC is 0.14m, which is improved by 71%, and the passenger comfort is improved by 68% compared with the classic terminal sliding mode controller (TSMC).

It is very important to consider the influence of shoreline development on the navigation and operation of surrounding fishing vessels during port planning. Taking planning of Ningbo-Zhoushan Port as an example, this paper proposes a method for evaluating the impact of port shoreline planning on fishing vessels based on automatic identification system (AIS) cluster analysis through existing AIS and fishing port distribution data. The research results can provide decision support for port shoreline selection and other coastal development activities.

This paper aims parking accuracy problem of trains in ATO (Automatic Train Operation) mode, in order to improve the parking accuracy can improve passengers taking on and off efficiency, reduce the dwell time in the station, improve the operation efficiency, as well improve the quality of passenger service as. In the practical application, it has a strong application value. This paper proposes a new balise arrangement scheme for Nanning Rail Transit line 4, in order to improve the parking accuracy. This paper uses the balise information and the non-learning parking algorithm based on Newton mechanics, the arrangement spacing of the balise of Nanning rail Transit line 4 is reset. After optimizing and adjusting the balises spacing of line 4, the accuracy of train automatic driving parking is simulated by using Matlab simulation software. By comparing and analyzing the results of Matlab simulation testing with the results before optimizing the layout of the responder, after optimization, the parking accuracy during the automatic train operation parking process has been improved by 3.33%, which is 1cm higher than the required range of parking accuracy (± 30cm).

With the continuous advancement of technology, intelligent aids to navigation have become an integral part of modern maritime transportation. This article analyzes the application of intelligent aids to navigation in ports and waterways and explores their positive impact on maritime transportation. The study reveals that intelligent aids to navigation can provide efficient navigation and navigational assistance services, thereby improving the safety and efficiency of transportation. Furthermore, intelligent aids to navigation can achieve precise management of waterways, provide accurate navigation information, reduce the probability of accidents, and save human and material resources. Finally, this article discusses the impact of intelligent aids to navigation on economic benefits, highlighting their ability to promote the development of port economy and enhance the quality of transportation services.

In this paper, we set up a shared parking allocation method in central urban areas with the goal of improving the comprehensive benefits generated by parking allocation. Firstly, conduct problem analysis to determine assumptions, model parameters, and variables; Next, select parking lots based on factors such as travel time, walking distance, and parking costs; then, determine the optimization objective and establish a parking allocation model; Finally, genetic algorithm is used to solve the model and obtain the optimal allocation scheme. This method can effectively improve the efficiency of shared parking in the central urban area, optimize parking resources, alleviate parking pressure, while ensuring the efficiency of public parking lots, minimizing travel costs and improving travel quality as much as possible.

Amidst the vigorous endeavor towards the monumental development of our nation's infrastructural foundation, the expansion of road mileage on an annual trajectory has magnified, elevating the timely detection and upkeep of road afflictions into an exigent imperative. In the present exposition, a novel methodology is introduced, predicated upon the sagacity of ailment discernment in urban thoroughfare preservation decision-making. This approach's principal objective is to augur the efficiency and scientificity inherent in road preservation determinations. Anchored in the outputs of a road ailment recognition model, propelled by the profundities of machine learning, the raw findings of road affliction identification engender an initial corpus. This is further processed through automated milepost parsing and correlation, coupled with the shrewd evaluation of roadcraft and preservation status, culminating in the astute generation of informative summative documents germane to the requisites of road quality oversight. The fruition of this endeavor manifests in the precise synthesis and comprehensive synthesis of road ailment intelligence. The feasibility and applicability of this approach are substantiated through meticulous experimentation and the ensuing analytical exegesis, thereby corroborating its utility as a Provide important reference information for road maintenance decision-making. A vital bedrock is thereby furnished, from which the precincts of maintenance planning for roadways and the expeditious amelioration of road maladies derive a resounding impetus, engendering a realm of augmented road safety and resilience, impeccably supported.

This article adopts the Fault Tree Analysis (FTA) method for analyzing the causes of failure modes of the 25Hz phase-sensitive track circuit system. A fault prediction and health management (PHM) architecture for the 25Hz phase-sensitive track circuit is established. Proposed data collection and fault partitioning schemes, and constructed a fault diagnosis model using decision trees to output maintenance decisions, which improves the maintenance efficiency of 25Hz phase-sensitive track circuits, reduce maintenance costs, and enhance the reliability and availability of track circuits, playing a positive role in ensuring the safe operation of trains.

In order to evaluate the impact of communication distance, vehicle speed and line-of-sight on the communication performance of CVIS (Cooperative Vehicle Infrastructure System) based on broadcast Wi-Fi, packet loss rate and communication delay are selected as representative indicators. Via constructing actual road traffic scenes divided into dynamic and static ones according to vehicle speed, and designing 4 test schemes under different communication distance and line-of-sight conditions, the OBU (On-board Unit) and RSU (Roadside Unit) are deployed in the closed test field to communicate with each other. The results show that communication distance and line-of-sight are obvious influencing factors of packet loss rate, but the speed has no great influence on the packet loss rate; the speed, communication distance and line-of-sight have little influence on the communication delay. It can be seen that communication distance and line-of-sight are the important factors that lead to the degradation of communication performance of CVIS.

In the context of world economic globalization, with the rapid growth of maritime traffic, the shipping industry has undergone tremendous changes, and more and more ship trajectories have appeared in various maritime activities. This increases the load on the waterway and congests the channel. Such a complex maritime environment is a severe test for China's maritime management. The main purpose of this study is to develop an automatic prediction system for ship motion based on artificial intelligence technology, which can monitor the position, course, speed and other information of the ship in real time, and predict the future trajectory of the ship through the analysis of historical data, and provide real-time decision-making support for port management, route planning, navigation safety and other aspects. At the same time, this study also aims to explore and improve the automatic prediction algorithm of ship motion situation, improve the prediction accuracy, and enhance the stability and practicality of the system. Through research, we hope to solve some practical problems in ship management while promoting the application and development of artificial intelligence technology in the maritime field.

The transportation network is the foundation and skeleton of regional development, and it is essential to promote the efficient flow and rational allocation of various resource factors. The "Vision and Actions for Promoting the Construction of the Silk Road Economic Belt and the 21st Century Maritime Silk Road" emphasizes that infrastructure interconnection is a priority area in constructing the "Belt and Road." Therefore, this paper explores the accessibility and spatial structure evolution of the country's westward opening based on the 30-year transportation network of the core area of the Silk Road Economic Belt and provides primary data for exploring the depth of the Belt and Road. The results of the study show that: in terms of accessibility evolution, there is a continuous step transition with Urumqi, Changji, and Korla as the center, and the center is high, and the border is low; the accessibility of the Kazakhstan-Mongolia-Russia triple junction has evolved most significantly and achieved a "triple jump" within the study period; while the China-Pakistan Economic Corridor has yet to continue and develop, and the southwest section is still accessible. The southwest section is still a relatively lagging section in terms of accessibility. As for the evolution of spatial structure, the density of the transportation network doubled during the study period, and the overall increase of central potential was 19.3%, indicating that the spatial scope of structural adjustment of the transportation network becomes more and more extensive as the level of regional connectivity continues to expand.

Accurate traffic flow prediction is crucial for making informed decisions regarding travel route selection and mitigating traffic congestion. This paper introduces a WGCG model that addresses the impact of traffic flow noise by combining wavelet transform with GCN and GRU models. The Sym6 wavelet basis is utilized to decompose traffic flow into two layers, effectively reducing noise. The road network's topological structure features are described using an undirected graph and adjacency matrix, with GCN extracting spatial rules and GRU mining hidden time correlation information from historical traffic flow data. The WGCG model integrates these modules to capture the dynamic patterns of traffic flow comprehensively. The model's performance is evaluated on a real dataset from Beijing's Second Ring Road, comparing prediction results with baseline models like HA, SVR, GRU, and GCN. Experimental findings indicate that the WGCG model achieves a significant increase in prediction accuracy, reducing errors by 27.9%, 22.3%, and 21.7% respectively, compared to the second-best model SVR. Ablation experiments further demonstrate that the WGCG model outperforms combined models utilizing only specific modules, confirming its feasibility and superiority.

The management of the rail service life is directly related to the safety of railway operations. This paper divides continuous rail line into 1 km grids, and uses the spatial and temporal data of rail damage, inspection and maintenance to identify and quantify the heterogeneous factors. Then BP neural network is used to construct a rail service life prediction model. Finally, the effectiveness of the model was verified using actual production data of ShenShuo Railway. The results show that the established model can accurately predict the rail service life, which has important guiding significance for the refined management of the rail condition.

The comprehensive inspection trainset for urban rail transit is a type of large-scale inspection equipment that conforms to the green equipment system and intelligent inspection concept, equipped with multiple types of inspection systems. The existing control mode of the trainset is that each inspection system operation is handled by a dedicated person during the inspection task, which requires a large number of following personnel and high trainset usage costs. A new control mode has been explored, which is based on automated perception and processing technology and proposes automated operation logic for comprehensive inspection trainset. It utilizes this logic to design a vehicle-mounted centralized control system that automatically senses task information, sets data collection parameters, and automatically turns on and off the inspection systems. At the same time, for periodic inspection services, or by utilizing the inspection tasks of electronic tag readers and encoders, a series of logic is set to replace the role of humans to determine whether the inspection system is working properly, and to obtain information on the activation end of the trainset operation to ensure the correctness of inspection data collection. The vehicle-mounted centralized control system system provides an automated inspection process for the comprehensive inspection trainset, enabling the trainset to have the function of being unmanned during data collection, achieving cost reduction and efficiency increase in trainset use.

Ship accident risk research is often affected by insufficient data and cannot be analyzed accurately. In this paper, to make up for the defect of not being able to accurately identify the interrelationships among the influencing factors in the case of small samples, an information diffusion- based method is proposed to diffuse the information that can be collected at present, so as to obtain a sufficient sample size to realize the accurate mining of the influencing factors. Secondly, considering the defect that traditional Bayesian network assumes that the factors are indep intent of each other, a tree Bayesian network is introduced to further identify the interrelationships among the influencing factors. The dataset adopts the grounding, reefing, explosion, collision, and touchdown accidents of China Maritime Safety Administration in 2014- 2022, and 21 risk fact ors are identified from four perspectives, namely, ship, environment, management, and human factors, and the results of the study indicate that the failure of the shipping company to implement the safety training program, the insufficient professional knowledge of crew members, and the insufficient lookout of crew members are the main factors leading to the ship accidents. Finally, the validity of the data is verified by the mutual information value and belief variance among the indicators, and the five important indicators with the highest correlation are analyzed separately. In order to verify the effectiveness of the model, this paper randomly selects 30 test samples to bring in by means of scenario simulation, and the accuracy of the test results are all above 65%, which indicates that the model has a high degree of confidence, and it can provide reference suggestions for the shipping enterprises and the related maritime safety management departments.

Ship route and speed optimization has become an important research area in the maritime industry, aiming to minimize fuel consumption and reduce operational costs. The method proposed in this study is based on rolling meteorological data and genetic algorithms, and includes the following steps. First, historical route data is used to train the model to obtain the relationship between fuel consumption and meteorological conditions and ship speed. Then, predicted meteorological data is obtained, including wind, waves, swells, currents, and their velocities, periods, and directions. Genetic algorithms are used to optimize ship routes and speeds based on these data. Finally, due to insufficient meteorological data to support full-route optimization, when new meteorological forecasts become available, rolling meteorological data and the ship's current position are used for the next optimization until the route is completed. The results show that intelligent decision design based on genetic algorithms significantly reduces fuel consumption compared to traditional methods of finding the shortest path and maintaining a constant speed. The reduction exceeds 3.5% while keeping the sailing time unchanged, and the decision-making time is less than ten minutes, verifying the practicality of the method used in this paper.

Intelligent transportation system plays an important role in the field of logistics transportation, which can improve the efficiency of logistics transportation, reduce costs and reduce traffic congestion. Logistics transportation planning and simulation is a key problem in the field of intelligent transportation. It can help logistics enterprises to carry out reasonable path planning and resource allocation, predict the possible problems in the transportation process in advance, and optimize the transportation scheme to improve the transportation efficiency. Based on the logistics and transportation planning and simulation in the field of intelligent transportation, this paper studies how to optimize the path selection, resource scheduling and transportation planning arrangement in the process of logistics and transportation through the modeling and simulation of the existing logistics and transportation system. The experimental results of this paper show that the logistics simulation system based on the field of intelligent transportation has advantages over the traditional logistics system in terms of transportation time, cost and safety indicators. Specifically, the average transportation time of this system is 8.5 hours, which is shorter than 10.5 hours of the traditional logistics system; the average cost is 4280 yuan, lower than 5000 yuan of the traditional logistics system, and the average safety index is 93.3%, which is higher than 88.5% of the traditional logistics system. It shows that the logistics simulation system based on the field of intelligent transportation has potential and feasibility in logistics transportation, which is expected to promote the development of logistics industry and improve the efficiency of logistics transportation.

With the advancement and reform of the automotive industry, intelligent driving vehicles have gradually penetrated the automobile market and entered a phase of peak development. Concurrently, the significant progress has been made in testing technology for intelligent driving vehicles. However, designing extreme testing scenarios remains a crucial and challenging problem in the field of vehicle testing. Therefore, this paper proposes an optimal method based on time window to create extreme scenarios for intelligent driving vehicle testing. By analyzing vehicle movement trajectories and incorporating the information such as size and volume of vehicles, more rational and realistic test scenarios are optimized. And simulation results demonstrate that the proposed method is reliable and effective.

Accurate traffic flow prediction can effectively alleviate the traffic pressure on expressways. Compared with general cities, the traffic flow on expressways in regional central cities is highly unbalanced, which makes accurate prediction of traffic flow very difficult. For this issue, we observed that traffic flow mainly varies greatly in different time periods, and the proportion of high flow time periods is relatively small. The prediction model will treat high flow data as outliers, resulting in poor prediction performance. Therefore, this paper proposes a time segment based expressway traffic flow prediction model, which maps the high and low traffic flows to different time segments. Divide the training dataset into multiple time segments, and then train the model separately based on the partitioned dataset. The experiment shows that the expressway traffic flow prediction model based on the method proposed in this paper has better performance than similar models.

In recent years, with the gradual improvement of China's waste classification policy, the problem of sorting and recycling of bulky waste has attracted increasing attention. At present, there are relatively few studies on the use of pure electric vehicles for waste collection and transportation operations in China. In view of this problem, this paper proposes a path optimisation study of door-to-door collection and transportation of bulky waste by pure electric vehicles considering the loading rate, with the goal of reducing the cost of bulky waste collection and transportation, constructs a door-to-door collection and transportation path optimisation model for pure electric vehicles considering the loading rate of bulky waste and the constraints of a hard time window, and conducts optimisation experiments through the whale optimisation algorithm based on the greedy crossover mechanism. The optimisation experiments are carried out through the whale optimisation algorithm based on the greedy crossover mechanism, and the experimental results verify that the proposed greedy whale optimisation algorithm can efficiently optimise the collection and transportation paths of pure electric vehicles and save the logistics and transportation costs.

Environment map construction is one of the key research directions of autonomous driving technology. Autonomous vehicle driving in unknown environments need to detect obstacles and identify passable areas. Traditional methods usually use two-dimensional planar maps without vertical information, and cannot exclude the influence of dynamic obstacles. In order to reflect the unknown environment information more accurately, this paper proposes a 3D map construction method based on Apollo autonomous driving system. Firstly, the point cloud data and pose data were obtained by the sensors mounted on the vehicle. Secondly, the 3D environment space was divided into different grids by the occupancy grid map algorithm. Then, the coordinates of different state grids are updated by ray casting algorithm. Finally, the Point Pillar algorithm is used to identify the dynamic obstacles and eliminate the residual shadows to obtain a more accurate 3D map. The experimental results show that the proposed method can effectively construct the unknown environment map, accurately reflect the three-dimensional information in the environment, and avoid the influence of dynamic obstacles.

In order to evaluate the dynamic resilience of container shipping network, this paper constructs a resilience evaluation system considering both operational capacity and recovery capacity. In order to consider the dynamic propagation of network failure node traffic, designs two node load distribution methods, namely port skipping over and port replacement, and four different load distribution ratios when port skipping over are selected. Based on this, constructs a cascade failure model of container shipping network. Taking RCEP regional container shipping network as an example, this paper explores the changes of network resilience under single point attack, deliberate attack and random attack, and the influence of different parameters on network resilience. The results show that once Singapore Port fails, it will have the greatest impact on network resilience. Comprehensive distribution strategy and local load distribution strategy can reduce the risk of cascading failure propagation after the failure of high node strength ports and small ports respectively, which is conducive to maintaining network resilience. In addition, appropriately increasing the capacity of the port and choosing port skipping over can also effectively slow down the decline rate of network resilience.

This article introduces an intelligent trackless laying system for urban rail transit projects. The track laying system mainly consists of mechanical systems, intelligent control systems, deflection correction systems, obstacle avoidance systems, electrical systems, and hydraulic systems. The overall equipment can achieve free span change, lifting and height adjustment, suspension angle adjustment, and trackless movement in circular, horseshoe shaped, rectangular and other tunnels. The lifting platform also has functions such as lifting, traversing, and turning, and can lift various components such as track bed boards, track panels, ash hoppers, and concrete mixing tanks. Under electrical and hydraulic control, the equipment having strong adaptability, and multifunctional integration can achieve stepless speed change and wireless remote control, and also has automatic deviation alarm, obstacle alarm and speed reduction, and Fault self-diagnosis function. The developed equipment has improved the overall construction efficiency of track laying operations, saved project costs, and also solved the safety hazards caused by traditional construction methods.

In order to realize the autonomous navigation of ships in the complex waters under the traffic separation system (TSS), taking the Chengshanjiao TSS waters as an example, an improved autonomous collision avoidance algorithm is proposed based on the artificial potential field (APF) method. By establishing five kinds of potential fields and their force functions, including the gravitational function of the goal point, the repulsion function of the target ships (TS), the gravitational function of the centerline of the traffic lane, the velocity function of the ship, and the boundary repulsion function of the navigation lane, finally the resultant force of them is calculated to guide the navigation direction of the ship, during the navigation process in this TSS water. Multiple groups of comparative experiments demonstrated that under the improved algorithm, the ship could avoid collisions with TS in four typical scenarios, according to the requirements of the 1972 International Regulations for Preventing Collisions at Sea (COLREGS) and good seamanship practice, simultaneously the ship could proceed in the navigable lane and keep safe distance with the boundaries and separation zones. The method proposed in this paper is expected to provide basic theoretical basis and technical support for the safe navigation of autonomous ships in TSS waters.