This PDF file contains the front matter associated with SPIE Proceedings Volume 13160, including the Title Page, Copyright information, Table of Contents, and Conference Committee information.

There are certain safety hazards in construction production that can cause economic losses and personnel injuries. This paper aims to develop an RFID-based construction restricted area intrusion detection system to keep construction personnel outside the restricted area. The system uses RFID technology to realize real-time intrusion detection of the construction restricted area, and combines with other auxiliary monitoring methods and audio-visual alarm devices to effectively manage the entry and exit of personnel in the construction restricted area, and upload alarm information to the IoT platform for subsequent management. This paper introduces the hardware and software design of the system and verifies its performance and feasibility through experiments. The experimental results show that the system can accurately and timely detect the intrusion of personnel in the construction restricted area, and issue alarms in a timely manner when personnel intrusion occurs, effectively improving the safety and work efficiency of on-site personnel. The system has advantages such as strong real-time performance, low cost, easy installation and maintenance, and has broad application prospects.

Aiming at the selection of refueling ports, the amount of refueling, and the speed selection faced by ships in container liner transportation, taking into account constraints such as carbon emissions, fuel consumption, vessel tank capacity, speed selection, and arrival time window, a fuel replenishment optimization model considering liner speed selection was established with the goal of minimizing vessel fixed costs, fuel costs, and carbon tax costs. An improved linear approximation method is proposed to transform the model into an equivalent linear programming model to obtain the optimal solution. Finally, using a real case study of a shipping company as an example, this paper analyzes the influence of error precision change, arrival time window and fuel price on the decision-making of shipping companies, verifying the effectiveness and practicality of the model and algorithm constructed in this article. The results show that the new method can effectively improve the efficiency and accuracy of calculations, and reduce the operating costs of shipping companies.

This paper analyzes the traffic noise characteristics of the long tunnel and underground interchange based on field monitoring and investigation of traffic noise and frequency in the area where they are located. Numerical simulation methods were used to study the noise characteristics of the long tunnel and underground interchange, and a finite element model of the noise in the long tunnel and underground interchange was established. The results showed that the simulated noise field model effectively reflected the actual noise level. This lays the foundation for future research on traffic noise reduction methods for the long tunnel and underground interchange based on this model. The research results of this project can provide theoretical and technical support for the prevention of similar engineering traffic noise and the promotion and application of related technologies.

With the continuous development of drone technology, it is favored by the logistics field because of its advantages of environmental protection, flexibility, reducing the inherent cost and time cost of distribution, and reducing interpersonal contact. However, it is also limited by endurance and load capacity. In order to overcome this limitation and give better play to the advantages of drone distribution, the current mainstream drone distribution mode is the base station supply mode along the route and the long-distance driving mode of the vehicle equipped with drone, but both modes have the disadvantages of high cost or small service radius. In this regard, the model of Vehicle-carrying delivery drone base station is proposed, and its operation process is confirmed through mechanical design and system design. Taking its application in emergency logistics and rural terminal logistics as an example, it is found that it not only improves the distribution efficiency and service radius on the basis of ensuring the cost, but also realizes automation and intelligence, and improves the space utilization.

Green supply chain is one of the important means to achieve the sustainable development of the automobile industry. This paper aims to improve the current evaluation index system of green supply chain management in the automobile industry, so as to evaluate the environmental impact, resource utilization efficiency and sustainability in the whole process of the supply chain of the automobile industry. The improved evaluation index system can provide a basis for the long-term development of enterprises to evaluate the level of their own green supply chain, and also provide an important tool for the government to evaluate the green degree of the supply chain of the automobile industry. In the future, with the progress of environmental protection requirements and the development of the automobile industry, the evaluation index system also needs to be further improved and improved.

Smart waterway, as an important part of smart cities and public transport systems, have a profound impact on the construction of some coastal cities. This paper reviews the current research status of Smart waterway, including its definition, development history and system architecture. Through the construction of perception layer, network transmission layer, data layer and application layer, it reveals the key technologies of Smart waterway in realizing the smartness of water transport. The development status and achievements of each technology layer and its role in the synergistic development of Smart waterway with Smart cities and public transport are introduced in depth. Finally, the concept of waterway hierarchy is proposed, highlighting its importance in future research and looking forward to the development of Smart waterway, which provides a new perspective for the optimization of the transport system in coastal cities.

In this paper, saturation, maximum queue length, average delay are selected to comprehensively evaluate the operation efficiency of the intersection. A comprehensive evaluation model based on fuzzy algorithm is established, the operation efficiency of intersections is divided into six levels, and the trapezoidal membership function of evaluation index is constructed based on HCM evaluation manual. Taking "Changjiang Road - Nankai Sanma Road" intersection as an example, the Vissim simulation model before and after intersection optimization is constructed, the input parameters are calibrated, and the required evaluation indicators are obtained. The evaluation model constructed is used to verify of the evaluation method, and the results show that the operation efficiency of the intersection rises from level C to level B. Especially the evaluation results of the improvement effect of intersection delay is the most obvious.

To improve the level of roughness detection in the construction process of mining tunnels, scholars domestic and abroad have combined 3D laser scanning with other information technology to achieve certain research results. However, the existing research mainly calculates roughness by the distance from the maximum concave and convex points to the fitting plane, which does not meet the requirements of the current regulations for the rule method. To address these issues, this paper uses algorithms such as cutting, denoising, complementation, and downsampling for pre-processing of laser scanning point clouds. Using the K-D tree and Octree algorithms, the rule method prescribed in the regulations is simulated, and 3D roughness is calculated. Through application verification of actual engineering projects, automation and intelligentization of roughness detection during the construction phase of mining tunnels are achieved.

In the course of ship navigation, sudden events such as collisions and grounding may occur, leading to damage and flooding of ship compartments, impacting the safe navigation of the vessel. Different states of compartment flooding have varying effects on the vessel's motion and resistance, consequently influencing the remaining navigational performance of the ship. Utilizing Computational Fluid Dynamics (CFD) methodology, this study conducts dynamic analysis of compartment flooding in different water conditions for a container ship. The research investigates the hydrodynamics of the vessel during compartment flooding, revealing the process and three degrees of freedom motion states. Simulation results indicate that the motion generated by compartment flooding under first-order waves is more intense compared to instantaneous motion under calm water conditions, although the overall motion trends are similar once stability is achieved. Different volumes of compartment flooding exhibit varying effects on vessel motion. The findings of this study provide valuable insights for estimating the remaining navigational performance of ships under compartment flooding conditions during actual voyages.

This paper aims to discuss and design an engineering risk early warning, prevention, and control system based on genetic algorithm (GA), with a specific focus on its application in water resources and civil engineering project management. By incorporating the flexibility and global search capabilities of GA, the system is designed to accurately capture and monitor complex risk relationships in water resources and civil engineering projects in real time. In terms of system design, it utilizes adaptive algorithms for optimal parameter adjustment, enabling the system to adeptly handle the complexities and variabilities inherent in these types of projects. The real-time monitoring mechanism ensures prompt detection and response to changes in engineering risks, thus enhancing the timeliness and accuracy of warnings. Additionally, the paper compares the performance of this GA-based system with systems based on the BP neural network (BPNN). The results demonstrate that the GA-based system is more effective in handling high-dimensional, nonlinear, and dynamic risk factors, providing an innovative approach to risk management in project management and offering decision-makers a more scientific and reliable basis for decision-making

Effective sign layouts are essential for guiding driving in underground construction caverns and improving transportation safety. While previous studies concentrated on evaluating drivers' gaze behavior in tunnels, the absence of a theoretical framework for visual perception of sign groups impedes comprehensive perception measurement and layout optimization. This paper aims to bridge this gap, which study explores drivers' visual cognition by analyzing eye movement and EEG indicators in sign group recognition tasks. It establishes an intuitive evaluation index system to gauge drivers' cognitive response efficiency to directional signs with varying information levels. The research evaluates the comprehensive efficiency of drivers' cognitive response in scenarios with different sign group information, focusing on the visual communication effectiveness of underground cavern sign groups. Utilizing the BCC model in the DEA method, the study obtains comprehensive evaluation results, shedding light on drivers' cognitive response to different information levels of directional signs. It includes an analysis of experimental results, discusses DEA ineffectiveness, and offers suggestions based on underground cavern sign group design requirements.

The “Largo”, a unique type of urban boundary space exclusive to Macau, is a spatial product formed after the fusion of Chinese and Western cultures, with spatial functions and place meanings that are particularly distinctive. The Largo refers to the transitional space between buildings or roads, forming a boundary space between them. This paper quantitatively analyzes the characteristics of frontier spaces based on their different attributes using GIS and POI analysis methods. Furthermore, it examines the current situation quantitatively. Finally, an optimization strategy is proposed for strengthening rigid boundaries and implementing flexible designs for different types of frontier spaces. This provides a reference strategy for smart management of heritage buildings and smart urban renewal in Macau.

An adjusted mediation model was employed to investigate the mechanism through which risk perception influences carpooling intention. We delved into the interconnections among risk perception, attitudes towards carpooling, and the intention to carpool by employing correlation and mediation analyses. Furthermore, we utilized bootstrap tests to scrutinize how gender and the level of residential risk influence these associations. The main findings are as follows: Risk perception negatively influences carpooling attitude and intention, while carpooling attitude positively influences carpooling intention. The role of carpooling attitude serves as a full mediator in the relationship between risk perception and the intention to participate in carpooling. Gender does not exhibit a moderating effect in the mediation model, whereas residential risk level demonstrates a moderating effect. Based on these findings, recommendations are provided for epidemic prevention measures and platform management, offering insights for urban governance and carpooling optimization.

This paper presents a method for generating and simulating critical scenarios for Automatic Emergency Braking (AEB) testing in rainy and foggy weather conditions. The method utilizes a subgradient descent optimization algorithm and a six-neighborhood flood-fill algorithm to search and generate critical hazard scenarios, based on a scenario importance model and threshold setting. Simulation results demonstrate that under clear weather conditions, the AEB system successfully triggers and prevents collisions. However, in snowy weather conditions, although the AEB system triggers, it fails to prevent collisions due to late triggering and insufficient braking deceleration. The experimental results validate the effectiveness of the proposed method in effectively testing the performance of AEB systems under different weather conditions. This research is of significant importance for enhancing the safety performance of AEB systems.

This study tackles the inefficiencies of the prevalent electronic toll collection (ETC) systems on expressways, which are hindered by limited communication ranges and notable processing delays that significantly reduce actual transportation throughput compared to theoretical capacities. To address these issues, we have developed an innovative free-flow tolling system that integrates cutting-edge 5G communication and high-precision localization technologies, aiming to streamline the tolling process and boost efficiency.We constructed a comprehensive traffic flow model that factors in diverse vehicle classifications and distinct driving patterns. This model serves as the foundation for simulating the tolling operations and evaluating the performance of the proposed system. Employing a cellular automaton framework, our simulation study meticulously assesses the service level of the tolling system, enabling us to fine-tune the design of tolling channels and enhance traffic organization schemes.The findings indicate that the proposed system has the potential to significantly improve traffic flow, minimize congestion, and elevate the overall efficiency of expressway transportation networks. Our research provides robust theoretical support and practical insights for the deployment of future expressway free-flow tolling systems.

This paper takes the construction project of a road and bridge construction group in Gansu as an example, for its existing project cost prediction is subjective and lacks accuracy, and the project cost influencing factors are not considered comprehensively, etc., analyzes its cost composition and the project cost influencing factors in previous years, designs the process and method of cost prediction, establishes the cost prediction index system on the basis of which, and constructs a cost prediction model using the random forest algorithm to simulate the prediction of project cost on the basis of the accumulated historical data of the data service center. Data Service Center accumulated historical data as the basis of the project cost simulation prediction, the training samples and prediction test samples are fitted to test the effect of the model prediction, the results show that the regression model has a high prediction accuracy, and it can provide valuable reference for the budget management and bidding decision-making of highway construction cost.

With the continuous increase in the operating mileage of China's railways, especially high-speed railways, the role of improving the safety environment along the railway line in ensuring high-quality development of railways and the safety of people's lives and property has become more prominent. The aim of this study is to explore the impact of different air pressure differences behind high-speed trains passing through platforms and analyze the basis for setting safety markings to the railway department to ensure the safety of railway platforms based on proposed models. Two aspects will be explored in this paper, one is force analysis, the other is establishment of safety markings.

As a new building production mode, the construction method of prefabricated buildings is quite different from that of traditional buildings, and the construction safety management at each stage is more difficult. In order to reasonably analyze the relationship between the construction safety risks of prefabricated buildings, this paper identifies 5 types of 26 risk factors, and builds a multi-level hierarchical structure model of construction safety risks of prefabricated buildings based on Fuzzy-ISM. And MICMAC is used to calculate the driving force and dependence of each risk to classify the risk. It provides some reference for the identification, analysis, evaluation and early warning of the key risks in the construction management of prefabricated buildings.

To address the real-time applicability and reliability of traditional collision risk index calculation models. The quaternion ship domain model, safe encounter distance model and ship dynamic boundary model are introduced to build the safe navigation distance model. Select the distance to closest point of approach (DCPA), the time to reach the distance to closest point of approach (TCPA), ship spacing, the azimuth angle of the incoming ship relative to the ship, and the ship speed ratio between the two ships as collision risk index evaluation indicators. When evaluating the distance between DCPA and the ship, use the navigation safety distance model to divide the danger level into different numerical ranges, in order to achieve the goal of being suitable for the captain of the corresponding ship, and calculate in real-time based on the ship speed of the ship, Finally, the fuzzy set theory is used to establish the ship risk collision model. Simulation experiments were conducted in a situation where a single ship and multiple ships intersect. the results indicate that the established model can accurately calculate the collision risk index and trend of ships, remind drivers to pay more attention to more dangerous ships, and also assist the maritime supervision department in supervising and reminding the sailing ships.

Based on the LD 29-1 marginal oilfield development project, combined with the shallow sea gravity platform design scheme suggested in "863" project, numerical simulation and centrifuge test were carried out, and the mechanical deformation characteristic of offshore platform structure and foundation deformation law were analyzed in this paper. According to the numerical result, the gravity platform had obvious lateral deformation under the action of horizontal load, the deck structure warped and the vertical settlement of surface foundation was large. According to the centrifugal model test result of 80g acceleration, under horizontal load, when the gravity platform had no bucket skirt, sliding deformation occurred mainly between the caisson and the foundation, and when the bucket skirt was added, the rotation deformation around the bottom plate occurred, while the gravity platform was inclined. The deformation degree of foundation can be reduced and the overall stability of the platform structure can be improved well by setting bucket skirt at the bottom.

In order to understand the automobile information modeling based on sensor coupling, a research on automobile information modeling based on nonlinear multi-sensor coupling is proposed. In this paper, firstly, the coupling law of automobile motion in all directions is analyzed, the whole vehicle model of heavy-duty vehicle with three degrees of freedom is established, and the differential equation of system motion is deduced. Secondly, the nonlinear model is combined with the vertical point contact tire model to describe the three-dimensional nonlinear tire force, and the wheel speed is calculated in real time to calculate the slip rate. By numerical integration, the dynamic response of the vehicle during steering braking, uniform straight line and steering wheel angle jump is calculated, and compared with the traditional handling stability model, ride comfort model, virtual prototype model and test data, the effectiveness of the model is verified. Finally, by analyzing the influence of coupling on the vehicle response, it is shown that the coupling model can simultaneously study the ride comfort, handling stability and braking of the vehicle under complex working conditions.

Ship AIS data contains spatio-temporal information, and the spatio-temporal characteristics of ship behaviour can be obta ined by mining and analyzing a large amount of AIS data. The deep water channel of the Yangtze River estuary has a lar ge traffic flow of ships, limited navigation resources, and frequent congestion of navigation vessels, resulting in a chaotic navigation order. In view of this situation, this study collected AIS data from vessels in the lower part of the deep water channel on spring tide, medium tide and neap tide days combined with the tidal pattern of the day to perform spatio-temp oral clustering analysis. The conclusion was drawn that the distribution of vessel density in the deepwater channel was un even in time and the traffic-intensive period was the high tide phase. Based on this, proposals are made to extend the control time by 0.5 hours and to create a template for key time ship entry plans according to the current traffic situation of the deepwater channel, in order to improve the traffic order in the channel, enhance the navigation efficiency of the channel and ensure the safety of ship navigation.

Accurate traffic accident clearance times prediction can help road managers make effective decisions and reduce property damage. This paper aims to develop a framework for traffic accident clearance time prediction and find the best prediction model. We propose a multi-model prediction framework for traffic accident severity. This framework consists of three parts: preprocessing of imbalanced data, variable selection and establishment of hybrid models: RF-SVM, RFBPNN, and RF-BN. Four highways in Shandong Province's traffic accident data are used as a case study in this paper. Based on the data used in this paper and previous literature exploration, three mixed models are constructed. Comparing the outcomes, we discover that the RF-SVM model has the highest prediction accuracy, up to 0.98, for the oversampled data set. This framework can be used to forecast the clearance time for traffic accidents, allowing for prompt emergency response and a reduction in fatalities and property damage.

In order to enhance ship detection accuracy and improve processing speed, an enhanced ship detection algorithm based on the YOLOv5 algorithm is introduced in this paper. Firstly, by adopting the Bidirectional Weighted Feature Pyramid Network, the model achieves higher accuracy in ship detection. Secondly, all the regular convolutions in the model are replaced with Ghost convolutions to achieve model lightweighting. The experimental results show that the average precision has been improved to 83.5%, exhibiting a 2.6 percentage point increase compared to the original model. The improved algorithm reduces the model's parameter size and computational complexity while maintaining high precision in ship detection.

Inspection is an important means of managing the safety of water projects and an important task for ensuring the safe operation of dams. In response to issues such as scattered inspection data, severe information islands, insufficient visualization, inadequate information management, and poor timeliness in traditional manual patrol inspections, this paper has developed an intelligent safety monitoring and inspection system based on micro inertial navigation technology. The system integrates "inertial navigation + internet of things + digital map" technology, which can accurately locate the patrol route based on indoor inertial navigation positioning or outdoor GNSS positioning. Through multimedia methods such as intelligent terminals taking photos and videos, the system records inspection data and generates inspection results based on the patrol route and inspection data. This effectively improves the intelligence level of safety monitoring and patrol inspection work. Engineering applications have shown that this system is worth promoting in other projects.

As the significance of truck transportation in the modern economy continues to grow, the issue of fatigue driving in tanker trucks has garnered significant attention. Therefore, this study proposes a multimodal fatigue driving detection method. It involves conducting driving experiments with tanker trucks, collecting driving operation data, electrocardiogram data, and eye-tracking data. After data preprocessing, a multimodal driving dataset is generated. Data mining techniques are used to extract 42 driving feature values, and then, through correlation analysis, 27 feature values are selected for fatigue state detection. Subsequently, the K-means method is employed to classify driving data into four fatigue levels, and a random forest algorithm is used for fatigue state recognition. Experimental results demonstrate that the proposed fatigue detection algorithm achieves a precision rate of 92.6%, effectively identifying different fatigue driving states. This approach provides insights and theoretical support for subsequent driver risk assessment and targeted traffic management.

Freeway merging areas are accident-prone roadways due to the frequent vehicle lane-changing interactions. It is necessary to ensure traffic safety on this roadway. And driver behavior is crucial to traffic safety. Therefore, it contributes to traffic safety to analyze and anticipate driving behavior in merging areas. To analyze the characteristics of driver behavior in merging areas, three scenarios were created based on Unity 3D. A total of 280 records were collected for the three different scenarios. A total of 120 people were invited to attend the test. The participants were clustered into three different types of driving styles using the clustering voting method. General drivers have the highest frequency of rapid acceleration and deceleration. Aggressive drivers have the fastest speed and the shortest minimum following distance, which makes it vulnerable to crashes, especially on highly accident-prone roads. Cautious driving style drivers have low speed, which can lead to traffic congestion and may also cause other vehicles to be forced to accelerate or decelerate sharply. Therefore, we should pay attention to these characteristics and make different responses to different drivers arriving at the merging area, based on these characteristics to reduce traffic crashes

The traditional logistics model is difficult to adapt to personalised and frequent express demand, and urban development and road network restrictions increase the difficulty of path planning and cost control. This study is dedicated to optimising logistics transport, proposing intelligent path planning and dynamic demand forecasting. By accurately predicting the number of urban express delivery, resources are deployed in advance to avoid insufficient or excess capacity and improve the quality of express delivery service. Finding the optimal cost solution under path constraints improves logistics efficiency and achieves more economical and sustainable logistics transport. Apply ARIMA model combined with grid search algorithm to forecast the number of express delivery with actual data.

Traffic is a complex system with great randomness and uncertainty. It is affected by a combination of many factors, including social, environmental and road factors. Therefore, the results of highway traffic forecasts are often unstable. In response to the above problems, a forecasting model based on artificial neural networks is proposed. The model is applied to traffic forecasting on the Xixian section of the Da-Guang Expressway in China, proving its stability and effectiveness. The application of artificial neural networks to predict the traffic volume of highways can greatly improve the efficiency and accuracy of traffic volume prediction. This is of great significance for solving the problem of road traffic congestion and improving the planning of highway network and regional development planning.

In order to overcome the disadvantage that the acceleration integral method isn’t suitable for low-speed detection, the track inclination integral method is proposed by installing one-dimensional laser displacement sensor and inclinometer on the frame. The advantages and disadvantages of the acceleration integral method and the track inclination integral method are studied from the transfer function and the signal-noise ratio. Combining the advantages of the above two algorithms, the complementary filtering algorithm is proposed, and the space compensation filter and the complementary filter are designed. Finally, the estimation accuracy of the three algorithms is verified by the simulation data. The results show that the complementary filtering method can effectively make up for the integral drift of the acceleration integral method and the shortcomings of the track inclination integral method to the short-wave irregularity attenuation. At the speed of 20km/h and above, the maximum estimated error of the complementary filtering method is 1.07mm, which meets the requirements of traditional track detection system, indicating that the complementary filtering method has good adaptability at low speed.

Driving maneuver recognition is one of the important problems in the field of intelligent transportation and is the basis of driving behavior research. The data collected from the accelerometer, gyroscope and GPS of the smartphone during natural driving were processed to obtain 14 types of driving maneuvers, including lane changing, obstacle avoidance, overtaking, 45° turning, 90° turning, 180° turning, noise, and idle (where noise refers to the noise data received while using the smartphone), and establish a driving maneuver sample dataset, where the distribution of sample categories is extremely imbalanced, the idle class with the most samples accounts for 97.006%, while the 180° right turn class with the least samples only accounts for 0.002%. So this paper proposes a driving maneuver recognition model based on XGBoost (eXtreme Gradient Boosting) algorithm, which utilizes the XGBoost algorithm to extract and analyze features from the sample dataset without artificially balancing the number of samples in each category, and the test results show that the recognition accuracy is 0.997, and the macro averages of precision, recall, and F1 score are 0.982, 0.956, and 0.968, respectively, which are significantly better than that of Random Forest, Adaboost and LightGBM, and overall better than that of CNN (Convolutional Neural Network) and LSTM (Long Short Term Memory). Therefore, XGBoost has excellent driving maneuver recognition capability, which can solve the problem of low recognition rate due to the extreme imbalance in the number of samples of each category in the multi-category driving maneuver recognition task.

This research employs the Prophet model, a time series forecasting tool, to tackle the intricacies and fluctuations in bus passenger flow. Through the utilization of historical ridership data, a Prophet model was formulated, accounting for seasonal changes and potential holiday impacts. Empirical results demonstrate the remarkable performance of the Prophet model in predicting bus passenger flow, showcasing high preciseness and robustness. The model not only captures cyclic ridership variations but also adjusts to seasonal dynamics and the influence of special events. In-depth analysis uncovers the model's predictive capabilities across various time scales, furnishing robust decision support for bus operators. The significance of this research is underscored by its introduction of the Prophet model into the of bus passenger flow prediction, presenting a novel and efficient tool for the analysis and prediction of short-term urban transit passenger flow. This has the potential to elevate the sustainability and service standards of public transit systems.

Traditional target signalized intersections assume that the arrival of vehicles on the shared lane for left turns and straight traffic is Poisson's, meaning that the home of vehicles is independent and arrives one by one. But according to the actual road network, it is found that multiple vehicles always arrive at the signalized intersection together. Therefore, in response to the permissible phase of left turn traffic flow at intersections, this paper establishes a left turn traffic queueing model with a Markov arrival process and vacation, and uses matrix analysis techniques to evaluate the traffic performance of left turn vehicles at intersections. This model considers many factors, including the demand for direct traffic and the demand for vehicles within shared lanes. By conducting numerical experiments on these factors, we aim to obtain the queueing process of vehicles at signalized intersections that dynamically change with the cycle under different factors. Combine the obtained experimental results with the intersections in the actual road network to optimize the left turn queueing problem and reduce traffic congestion.

Vehicle computing tasks, which need high computational resources and be sensitive to delay, is widely concerned in the field of Mobile Edge Computing (MEC). At present, considerable progress has been made in the research of Vehicle-to-Vehicle (V2V) task offloading. However, in order to further improve the efficiency of task offloading, the base station should also be considered to have reserved computational resources. Determining the level of resources reservation is challengeable, as setting it too low will have bad efficiency in high-task scenarios, while setting it too high will lead to a waste of resources. Hence, we introduce the Edge-server Resources Reservation framework (ERRF), in which we study the impact of traffic flow on the reserved computational resources, introduce an algorithm to calculate a balanced reserved computational resource, and design a pre-processing algorithm for task offloading on this basis. We conduct the simulation and compare it with two V2V task offloading algorithms. Our results demonstrate the effectiveness of our framework, particularly in terms of average latency and task success rate across different scenarios.

A real-time physiological index anomaly detection method for subway drivers has been proposed. We use smart wristbands to collect indicator data and determine whether the current physiological indicator values of subway drivers are in an abnormal state. We use historical data of three human physiological indicators, heart rate, body temperature, and respiration, and generate confidence intervals for various indicators of drivers based on the principle of statistical normal distribution, to establish a model for judging abnormal physiological indicators of drivers. Python programming is applied to simulate historical data, establish confidence intervals, collect indicator data, and calculate results for judgment. The simulation results can determine the current state of the driver through measurement values, and the historical data generated by the measurement is in a continuously updated state. The confidence intervals of each indicator can be continuously adjusted, which is more applicable than other methods. Moreover, monitoring indicators can be added on the basis of technology and equipment.

In recent years, with the improvement in living standards and economic development, the number of cars in the city has been increasing year by year. This has put pressure on the carrying capacity of the roads and led to traffic congestion. However, it is difficult for traditional time series models to predict future traffic flows accurately. In this paper, we take the traffic data of the coastal urban area as our research object, combining spatial and temporal factors of the city. We observed the excellent performance of the improved model in complex environments by the attention temporal graph convolutional network(A3T-GCN) model. The root mean square error(RMSE) value is 33.305, the mean absolute error (MAE) value is 21.257, and the variance(var) is 90.3%. This study provides a practical algorithmic foundation for intelligent transport. The A3T-GCN algorithm, which is based on the self-attention mechanism, has a significant advantage in traffic flow prediction, especially when there is limited data

To study the differences in driver visual behavior characteristics between bridge sections and roadbed sections of highways, the Smart Eye Pro5.7 non-invasive eye tracker was used to record the driver's eye movement data. Eye movement indicators such as pupil area, X-angle, Y-angle, and percentage of fixation time were used to analyze the data. The results show that compared to general roadbed sections, the pupil area of drivers on bridge sections is larger, indicating that drivers on bridge sections with open spaces on both sides are prone to feeling insecure, and drivers pay more attention to the central divider. The driver pays more attention to the center left position, with the central anti-glare board as the visual focus. On the roadbed section, drivers pay more attention to the center right position and pay less attention to the central divider. Drivers on bridge sections pay more attention to the left side, while those on roadbed sections pay more attention to the middle right side.

The identification of accident-prone locations on roads remains a crucial area of focus in road safety research. Among the various approaches used, the cumulative frequency curve method stands out as one of the most widely employed techniques for this purpose. However, it is important to address the limitations of the traditional cumulative frequency curve method and propose enhancements. In this article, we present an improved version of the method that builds upon the foundation of the traditional approach. Our enhanced method introduces the concept of equivalent accident numbers, which enables us to gauge the severity of different accidents more effectively. By utilizing the sliding window method, we divide road sections into manageable segments and construct cumulative frequency curves based on equivalent accident numbers per kilometer. To determine critical values for identifying accident-prone and potential accident-prone locations, we examine the equivalent accident numbers at specific cumulative frequencies, namely 80% and 95%. These critical values, denoted as N80 and N95, serve as benchmarks for our analysis. An empirical validation was conducted using historical traffic accident data from the Guizhou Section of the G75 Lanzhou-Haikou Expressway. The results of the analysis using the traditional cumulative frequency curve method revealed accident-prone sections accounting for 9.67% of the total length and 40.17% of the accident counts. However, our improved method identified accident-prone sections representing only 3.78% of the total length but capturing 20.06% of the accident counts. Furthermore, we successfully identified potential accident-prone locations spanning 11.03% of the total length and accounting for 33.95% of the accident counts. Compared to the traditional algorithm, our enhanced approach demonstrates superior accuracy and precision by pinpointing more accident-prone locations and potential accident-prone locations. These findings hold significant implications for the graded management of road safety, ultimately leading to improved road safety outcomes.

In recent years, the state has issued a series of policies to encourage the priority development of public transport. But the expanding size of cities and growing population pose a major challenge for public transportation. In this paper, we study the urban bus system with the non-Markovian higher-order network model considering the path-dependent characteristics for the first time. An empirical study on the bus transport of Xi'an(BT-X) system was conducted. The theoretical higher-order bus transport network(THBTN) is established, and the improved weighted k-core decomposition method based on the road grade, subway connection, the number of POI within the service area, and the population density is proposed. It is used to divide the THBTN into the core layer, bridge layer and periphery layer. We analyse the importance of bus stops. Twenty two important bus stops in the BT-X system are discovered and ranked. On this basis, the problem is found in the bus network of Xi'an: poor interconnectivity between important hub stations. Therefore, forty four routes are found to be required to interchange at intermediate stops, which gives a route optimization plan for BT-X system. The research method of this paper can be applied to the optimization of bus routes in other cities, with a view to solving the interchange problem of urban bus systems in China.

In response to the trend of developing aircraft with heavier loads and higher tire pressures, a full-scale nine-block pavement panel three-dimensional finite element analysis model was established using ANSYS software to study the stress and deformation of cement concrete pavement under different loads applied at different positions. The results showed that the thickness of the pavement panel significantly affected the stress at the bottom of the panel. By considering the reliability, the ultimate limit state of the pavement was analyzed, and it was found that for aircraft with a tire pressure greater than 1.6 MPa, the use of the aircraft type should be strictly controlled when the pavement thickness is 24 cm. Finally, the ACNPCN method was used to study the overloading situation of the pavement, and the requirements for overloading use were established. By consulting the chart, the airport management department can intuitively understand the requirements for pavement thickness for different aircraft tire pressures and the range of tire pressure that the pavement thickness can accommodate, making it easier to control the overloading use of aircraft at the airport.

As an important part of the highway, the weaving section undertakes the merging and the diverging of the traffic flow, which is an important traffic bottleneck that deserves to be studied.[1] This paper focuses on analyzing and evaluating the traffic capacity of this bottleneck. The vehicle trajectory dataset chosen for this study originates from the Next Generation Simulation (NGSIM) initiative conducted by the Federal Highway Administration (FHWA). The study area spanned roughly 640 meters in length and comprised of five primary lanes throughout the section, as. Additionally, an auxiliary lane was present along a portion of the corridor, running between the on-ramp at Ventura Boulevard and the off-ramp at Cahuenga Boulevard. The parameters for evaluation are capacity, average queue length, average delay, average time in the queue, and average excess accumulation. To improve the traffic, some strategies will be given. VISSIM simulation is used to see the effectiveness of these solutions.

According to the IMO International Maritime Solid Bulk Cargo Rules, solid bulk cargo, including iron ore concentrate powder, nickel ore powder, etc., is classified as Group A cargo, which is easily fluidized cargo. This type of goods is in the form of solid particles and often contains moisture. Under the combined effects of harsh environmental conditions and ship movement during sea transportation, the gradual increase in moisture content of such goods can cause solid particle cargo to transform into a liquid phase, which means fluidization occurs. The fluidized cargo can move within the cargo hold, greatly affecting the stability of the ship. In severe cases, it can cause the ship to capsize and cause huge personnel and property damage. Therefore, in order to ensure the safety of transporting such goods by sea, it is necessary to conduct research on goods that are prone to flow.A 3D cargo hold model with equal proportions was established, and the model was reduced by ten times during simulation. Due to the harsh marine environment, where strong winds and waves cause ships to move, the main movement that causes fluidization of cargo is roll motion. Therefore, the LIGGHTS software is used to set up a cargo hold model for a roll motion with a period of 10 seconds and an amplitude of 20 ° to simulate the marine environment and analyze the movement of granular cargo in the cargo hold.

With the rapid development of transportation industry, various pollution and transportation loss costs are increasing. Green, safe and efficient has become the development direction of future collection and distribution mode. In this paper, a linear integer programming model is established to minimize transportation costs, air pollution costs, noise pollution costs, traffic accident costs and freight damage costs. Taking Yangshan Port as an example, the validity of the model is verified, and it is found that the proportion of railway and water transportation can be increased by taking external costs into account.

In this paper, a simulation-based method for calibrating the fundamental diagram of traffic network is proposed, which has the advantage of finding the optimal performance for performance-decaying road parameters in a limit number of iterations, considering the need for accurate operational performance decaying traffic network simulation based on limited traffic data to serve traffic managers. A two-layer optimization model is constructed, with the lower layer being the stochastic dynamic user equalization (SDUE) based on the LTM model, which is used to simulate the acquisition of the OD and the traffic flow of each road segment; the upper layer objective function is set to minimize the variance of the average true and simulated vehicle counts of each road, which is approximated and replaced by the resolvable metamodel, and is solved by nested into the trust-region(TR) method. A small simulation network is constructed for parameter calibration validation under operational performance decaying, and the calibration errors of each OD level are within 2% at 10% and 20% decaying levels, the calibration deviation on the complex network is also within 2% in 20 iterations, which demonstrating the applicability of the algorithm.

Researches on vehicle routing for logistics enterprises often neglects the impact of vehicle speed on operational costs. In order to address this issue, a genetic algorithm is designed to optimize the BP neural network model for vehicle speed prediction, which utilizes big data to predict the vehicle speed on different segments of the distribution network. Taking into account the demand for goods at distribution outlets, a vehicle routing model is constructed. The decision variables of the model include the vehicle delivery routes and the corresponding vehicle types. The objective is to minimize the total cost of distribution for the enterprise. A case study validates the effectiveness of the model by using a distribution network with 21 distribution outlets as nodes. We implemented the genetic algorithm program in the Matlab software to solve the case study. The results demonstrate that the optimized solution can effectively reduce distribution costs for logistics enterprises.

This paper proposes an emergency obstacle avoidance algorithm based on Dubins analytical method for high-speed intelligent driving lane change scenarios, which solves the optimal path under high-speed emergency obstacle avoidance through optimal control theory and Dubins analytical algorithm, which has less calculation time than existing algorithms when solving high-speed obstacle avoidance problems, and shorter distance under the condition of ensuring that the trajectory can be executed, which effectively solves the obstacle avoidance problem in this scenario. Then, a series of lane change trajectory simulation tests are carried out in this scenario, which verifies the robustness of the algorithm under actual working conditions.

Train Automatic Stop Control (TASC) system based on reinforcement learning(RL) is a crucial technology to improve self-driving performance for automatic train operation(ATO). Although significant progress has been achieved, existing RL-TASC systems still face challenges related to sparse rewards and lower control precision. In this paper, we design a train automatic stop strategy based on the Deep Q-Network (DQN) algorithm with both global rewards and single-step rewards to enhance the precision and training stability of RL-TASC systems. Initially, a single-point motion equation for the train was established and the state-action space was defined. Subsequently, a global reward function was designed based on stopping error. To address the challenges posed by sparse global rewards, which led to poor training effects, and imprecise stopping performance, a method for setting a single-step reward function based on expected velocity was proposed. Finally, through simulation experiments, the results showed that the improved control method could maintain the train stopping error within 0.05 meters. Comparing to the pre-improved model, the average value of stopping error decreased by 76.56% and the standard deviation of stopping error decreased by 67.74%, confirming the effectiveness of the improved method. This paper provides a high-precision and highly robust method for automatic stopping control in high-speed train ATO systems.

Road Traffic Accidents (RTAs) are a serious safety issue, especially in fast-growing cities, and have become one of the leading causes of death worldwide. This study takes Addis Ababa, Ethiopia, as a case study for the period from 2017 to 2020 and uses advanced interpretable machine learning techniques to analyse the key features that influence road safety. The results highlight the superior performance of the Random Forest model. Interestingly, findings indicate that a large number of accidents occurred under normal road and weather conditions, highlighting the significant influence of driver characteristics. This study provides relevant authorities with effective strategies to significantly reduce mortality in persistent RTAs.

The proposal of China's dual-carbon strategy has put forward new requirements for the operation of Suburban railway trains, which should minimize carbon emissions during operation under the premise of satisfying the dynamics of Suburban railway passenger flows, and further realize the low-carbon greening of urban (suburban) travel. In this paper, we consider the dynamics of passenger flows from the uneven arrival of passengers during the peak period, and establish a multiobjective optimization model of Suburban train stopping scheme with the objectives of minimizing the broad travel cost of passengers, minimizing the carbon emission cost of enterprises, and minimizing the operating cost of enterprises by integrating the three aspects of passengers, enterprises and environment. The NSGA-II algorithm is designed by combining the characteristics of urban (suburban) passenger flow and passenger mode choice. After analyzing the results of the algorithm, the stopping scheme determined by the model can effectively reduce travel time, carbon emissions and operating costs.

Aiming at the problems of high distribution cost and low customer satisfaction in the current cold chain distribution link, this paper constructs a cold chain distribution path optimization model with the objective of minimizing the comprehensive distribution cost according to the characteristics of cold chain distribution, and under the constraint of overall customer satisfaction, and by comprehensively considering the transportation cost, refrigeration cost, timewindow penalty cost and cargo damage cost. Then, the improved particle swarm algorithm is designed and solved the model. Finally, an example of a cold chain distribution center is analyzed, and a distribution plan that meets customer satisfaction and cost optimization is obtained, which verifies the effectiveness of this paper's model and the improved algorithm; at the same time, the applicability of this paper's model and the improved algorithm is verified by solving the standard case of the Solomon dataset.

In order to analyze the dynamic performance differences of light rail transit (LRT) under different marshalling, this paper establishes the nonlinear system dynamics models of three-module marshalling, six-module marshalling and twelvemodule marshalling low-floor LRTs, and then analyzes the dynamic performance of the vehicle under straight lines and curves. The results show that in the straight line, the vehicle of the three modules is in the range of 20km/h ~ 70km/h, and the difference of the stability index and the safety index is very small, and all meet the standard requirements. Under the curve line, the difference between the three modular marshalling types is very small in the range of 20km/h ~ 60km/h, and the twelve modular marshalling vehicles are slightly larger than the other two types. From the perspective of the critical speed of the vehicle, the critical speed difference between the three-module marshalling vehicle and the other two marshalling vehicles is small, only 5km/h. Under the calculated line condition and speed, there is no significant difference in vehicle dynamic performance.

To solve the problem of emergency material support for geological disaster rescue, a collaborative mode of transport vehicles and drones is applied for material distribution. Vehicles-drones collaborative path optimization model is established with the goal of minimizing the total distribution time. The K-means clustering analysis method is used to cluster the emergency material demand points in the disaster stricken area, in order to divide the distribution areas for the vehicle-drone groups, Genetic algorithm is used to code the task planning of the two transportation tools, and the damage repair operation in the adaptive large neighborhood search algorithm is embedded in the solution to complete the path optimization. Based on standard calculation examples, the results show that Vehicles-drones collaboration can reflect time advantages and meet the requirements of timeliness, accuracy, and sustainability in the delivery of emergency rescue tasks at the end.

With the gradual improvement of intelligent transportation system, unmanned driving, as a representative product of the intelligent development of automobiles, has attracted much attention for its broad application prospects and influence. Technically speaking, path planning is the key to the realization of unmanned driving, and it is also one of the research hotspots in the field of unmanned driving technology. Aiming at the problems of complicated calculation, low efficiency and unsmooth path in the path planning process of unmanned cars, this paper will focus on the global trajectory planning algorithm, take Rapidly-exploring Random Tree (RRT) algorithm as the research object, and put forward corresponding optimization strategies, so as to improve the overall performance of the vehicle path planning algorithm. Practice has proved that the improved RRT algorithm has obvious advantages in sampling time, the number of points used, the number of sharp points, the path cost and other indicators after MATLAB simulation test. Compared with the commonly used original RRT algorithm and RRT* algorithm, it can solve the optimal path of unmanned cars more quickly.

Aiming at the problem that the traditional grey wolf algorithm has low convergence efficiency and to fall into local optimum when solving the underwater glider path planning problem, this paper uses an optimization algorithm with an adaptive head wolf to improve the exploration ability of the grey wolf individual by increasing the head wolf, and applies it to the path planning of underwater gliders. It is experimentally verified that the improved grey wolf algorithm saves 43.877% of time on underwater glider path planning.

The optimization of the railway freight network is crucial for achieving sustainable development. And the reform in railway carrier liquidation has injected new vitality into the railway freight market by significantly enhancing the quantitative accuracy of redundant payment elements within the network. This paper introduces a railway freight network optimization model, utilizing the current railway carrier liquidation mechanism in China as a basis. A global search algorithm, based on Dijkstra, is designed for efficient solving. Based on the real-world case, when planning and selecting freight transportation routes, the railway carriers must consider not only the transportation distance but also carefully weigh the revenue and expenditure elements associated with the freight transportation clearing mechanism.

Based on the underwater pipeline laying project in port, this article analyzed the movement characteristics of pipeline laying ship and limiting operational condition under the influence of wave in port by using numerical simulation method. The results show that the amplitude of the movement response of the pipeline laying ship increases with the increase of wave height. Under the actions of waves in different wave directions, the movement response of the pipeline laying ship varies greatly. The heave and roll are the most significant in the wave condition with a direction of 90°, and the pitch is the most significant when the wave direction is 180°. The limiting operational conditions of the pipeline laying ship vary on the impacts of waves in different directions. The limiting operational wave height is the smallest when the wave direction is 90° while the limiting operational wave height is the largest when the wave direction is 0°.