The Proceedings of the 11th International Conference on Traffic and Transportation Studies

Outside peak hours, urban train companies usually have to park a considerable amount of their rolling stock. In many urban rail networks, such parking takes also place on sidings outside any depot. Obviously, if according to the vehicle schedule a siding is occupied by a parked train, it is no longer available for disposition purposes. In particular, in the event of some unplanned track blockage, depending on the track topology, if some siding constitutes the only option to change from one track to the opposite track in order to continue as return trip, then such an available siding is of high operational value.We propose a procedure to assess significant parts of the operational value of an available siding. In particular, for each siding, we identify an amount of scheduled empty kilometers to park trains in stations that are farther away, that are likely to pay off during operations. This procedure is already in daily use in the planning department of the Berlin fast train company S-Bahn Berlin GmbH.

In the past, most of the research on car-following models in China focused on physical statistics, with relatively little collection and analysis of car-following data. Therefore, it is essential to establish a car-following model suitable for analyzing the car-following characteristics of Chinese drivers based on real driving data. This paper utilizes a vehicle-mounted monocular camera to capture traffic flow images on the highway and uses the YOLOv5 algorithm to detect vehicles. Through the homography transformation between the image plane in the front view and the spatial plane in the top view, the pixel coordinates are converted to 3D spatial coordinates. Then, the ranging model is established, and its effectiveness is verified using the length of the lane dividing lines. The factor analysis is used to determine the main influencing factors of car-following behavior. The appropriate linear car-following models are selected and its parameters are calibrated using multiple linear regression methods. This paper realizes the collection and subsequent processing of car-following behavior data under natural driving conditions, and transforms generalized car-following models into specific research theories for application in practical situations. This method of collecting car-following driving data has strong practicality and can provide effective data support for studying car-following models that accurately reflect the operation conditions of road traffic.

With the expansion of urban scale, traditional public transit struggles to fulfill long-distance travel demands (Rail transit is costly and fixed-route transit is time-consuming). This paper proposes a bimodal transit system with point-to-point transit (PPT) services and fixed-route transit (FRT) services. PPT is designed for passengers traveling long distances, efficiently linking two major transport hubs with few intermediate stops. We formulate the joint design problem as a mixed-integer program through a parsimonious continuum model. By minimizing the system cost, the sum of agency and user costs, we determine optimal design parameters using MATLAB built-in genetic algorithm. Through numerical experiments, we analyze how the performance of the bimodal transit system varies with demand, city size, and other factors. Our results demonstrate that the proposed system is more cost-efficient than the regular system in cities with high travel demand. The proposed model can be used for the sketchy design of such bimodal transit system and provide a theoretical foundation for the adjustment of public transit networks.

To explore the micro-level influencing factors of driving risks for private cars on freeway sections, the study integrates private car trajectory data matched with traffic accidents in both temporal and spatial dimensions. It systematically extracts indicators for driving risk assessment from four aspects: personnel, vehicle, road, and environment. The spatial–temporal range of 5 km upstream and downstream within 5 min before and after the occurrence of accidents was selected for data screening. An XGBoost-based driving risk identification algorithm was established to classify the data, and SHAP values were used for interpretability analysis of the indicators. The results indicate that the developed XGBoost algorithm achieved accuracy rate of 92.5% in identifying driving risks, with certain indicators showing strong interactions and inhibitory effects. The findings provide theoretical support and insights for traffic safety management and risk control.

To alleviate urban traffic pressure and reduce congestion costs, this paper constructs a multi-objective mixed integer programming model for optimizing the timetable and allocation of passenger and freight wagons in the case of operating cross-line trains. The model aims to minimize passenger waiting time and train operation costs, thereby optimizing the timetable and distribution of passenger and freight wagons to enhance the transport capacity. The paper conducts data experiments using the Beijing Subway Line Fangshan and Line 9 as a case study, and utilizes the optimization solver Gurobi for solution. Experimental results demonstrate that the timetable determined by the model can better meet the requirements and constraints. Compared with the scenario of not operating cross-line trains, operating cross-line trains on subway lines can significantly reduce average passenger waiting time, fully utilize redundant transport capacity, and reduce train operation costs.

The high-speed railway provides a safe, punctual, and comfortable mode of transportation underpinned by stable infrastructure, fast trains, and dependable control systems. Implementing an appropriate maintenance schedule is essential for ensuring operational status and reducing the malfunction events of these systems. This study proposes a reliability-centered maintenance (RCM) strategy for high-speed railway facilities maintenance scheduling optimization. A reliability model is developed to characterize the multi-stage deterioration process of facilities, and a maintenance optimization model is put forward to minimize the maintenance cost by incorporating reasonable critical reliability with multi-level maintenance tasks. Two baseline models that typify current maintenance strategies are also formulated to illustrate the enhanced effectiveness of the RCM approach. The numerical example demonstrates that the RCM can significantly reduce maintenance schedule costs while simultaneously prolonging the lifespan of the facilities, all within the bounds of the same reliability constraints.

Train operation efficiency is the basis for formulating and adjusting the train operation plan. Constructing an accurate train tracking operation model is the key to solving the low operation efficiency during the peak period. According to the actual operation scenario of the train, a cellular automata model based on spatial discretization and time continuity is constructed by using sparse matrix to store cellular state. The simulation results show that the train traction and braking performance parameters meet the requirements of the train speed limit. The simulation results of the train group operation curve under different departure intervals are consistent with the actual operation, which verifies the reliability and simulation accuracy of the model in this paper. On the lines with low speed limit and long station spacing, the adaptability of train speed control is stronger, and the increase of traction energy consumption with speed is smaller. The research conclusions can provide reference for urban rail transit optimization system and adjustment of operation strategy.

Studying the identification and resilience optimization of critical nodes in urban rail transit networks can prevent network interruptions and enhance the network’s ability to resist risks. This article adopts a TOPSIS model based on entropy weighting method to design a node importance evaluation approach that considers both local and global network structure information. Furthermore, under the constraint of financial budget and considering the comprehensive interests of different stakeholders in the network, a network resilience optimization model is constructed. Using Beijing’s subway as a case, when the network is subjected to sequential-deliberate attacks until collapse, the node coupling importance evaluation method proposed in this article reduces the number of attacked nodes by 4.39%, 2.59%, and 0.52% respectively, compared to Degree Centrality, Closeness Centrality, and Betweenness Centrality. This approach helps identify and maintain critical network nodes in different areas, ensuring the stability and safe operation of urban rail transit networks. Compared to the existing network, the optimized network exhibits a 3.36% improvement in flexibility, with an increase in the number of attacked nodes by 0.26% and 14.21% under random and sequential-deliberate attacks until collapse, demonstrating that the proposed elastic optimization method effectively improves the risk resilience of urban rail transit networks.

The stability of steel plate composite girder bridge is an important factor that must be considered in design. In this paper, the finite element method is used to analyze the overall stability and local stability of long-span steel plate composite girder bridge under single factor by selecting the cross beam spacing, height-span ratio, web height-thickness ratio and flange width-thickness ratio as the key design parameters. The orthogonal experimental method was used to analyze the multi factor combination of the bridge, and the importance of each design parameter was evaluated from the aspects of structural safety and economy. The results show that the safety factor of stability of steel plate girder bridge is positively correlated with the height-span ratio, and negatively correlated with the cross beam spacing, web height-thickness ratio and the flange width-thickness ratio. As far as the whole bridge is concerned, the cross beam spacing has the greatest influence on safety factor of stability and its value should be less than 15 m; the web height-to-thickness ratio has the greatest influence on the amount of steel, and its value should be between 100 and 120. The results of this paper will make this type of bridge structure more reliable and less costly.

Upgrades are continuously made to expressway traffic perception technology as part of the ongoing development of intelligent transportation, resulting in a large amount of data, including vehicle position, passing time, and speed. This real-time, accurate, and efficient source data is used to estimate short-term traffic flow on expressways. Aiming to enhance predictive accuracy and more accurately capture the intricate spatial and temporal features of traffic flow, this paper presents a convolutional long short-term memory neural network model based on attention mechanism to predict traffic flow. To extract spatial-temporal features rather than just single temporal features, convolutional long short-term memory neural networks combine the advantages of long short-term memory neural networks to extract temporal features and convolutional neural networks to extract spatial features. Attention mechanism is added to capture the influence of past feature states of time series data on traffic flow. On this basis, the error correction mechanism is introduced to further enhance the prediction accuracy of the model. The experimental results indicate that the proposed model outperforms other current techniques in terms of prediction accuracy.

At present, the increase of intercity railway trains has made it more convenient for passengers to travel. However, in the face of constantly adding trains, the work of traditional manual preparation of train working diagram has become very difficult. The rapid development of artificial intelligence technology makes it possible to optimize the train working diagram by intelligent means. In this paper, based on the existing intelligent adjustment strategy using the Sarsa algorithm, an optimization method for intercity railway train working diagram using the improved Sarsa algorithm is proposed. The proposed method adopts a differentiated optimization strategy for peak and off-peak hours to solve the tradeoff problem between train evenness and operation costs. The experimental results show that the proposed strategy can reduce passenger waiting time and train operation costs by 20.00% and 67.80%, respectively. Moreover, compared with the Sarsa algorithm for improving balance and reducing conflict severity in previous studies, the proposed strategy can not only reduce the train operation costs, but also reduce the waiting time of passengers by 12.12%.

As an important transportation packaging container, corrugated boxes cause pollution to the environment by the carbon emissions generated during their whole life cycle, and analyzing and calculating the carbon footprint of corrugated boxes has high economic value and social significance. This paper takes corrugated boxes as an example to study the carbon footprint. Taking a 5-ton van as a functional unit, based on the life cycle evaluation method, we focus on analyzing the carbon footprints of its production, storage, transportation, and waste recycling, and measure the carbon footprints of its whole life cycle using Gabi software. The study shows that, in the whole life cycle, the corrugated carton production and manufacturing stage has the largest carbon emissions, with a contribution rate of 64.55%, followed by the waste recycling stage accounting for 32.28% of the total carbon emissions. The warehousing and transportation stage accounts for a small proportion. This shows that measures should be taken to strengthen supervision and reduce carbon emissions in the production and recycling of corrugated boxes.

Tunnel fires represent a paramount threat within tunnels, encompassing multifaceted dangers beyond the immediate fire hazard. These threats extend to potential traffic congestion and associated casualties. A rapid and coordinated pedestrian evacuation strategy is pivotal in mitigating both human and property losses. However, conventional emergency plans often grapple with the complexities inherent in ever-evolving tunnel fire scenarios. This underscores the compelling need to introduce tunnel fire evacuation simulations as a proactive approach. Such simulations can authentically replicate fire scenarios and human behaviors, thereby providing a basis for informed emergency decision-making. This paper aims to comprehensively review the current state of tunnel emergency management simulation research, encompassing the evolution of fire numerical simulation techniques and pedestrian evacuation simulation technologies, as well as prospective avenues for future development. This endeavor seeks to enhance the capability to effectively address the challenges posed by tunnel fires.

This study divides the passenger travel path and constructs a mixed integer programming model for solving the cross-line operation plan of urban rail transit train. The optimization goal of model is to minimize the sum of passenger travel cost and enterprise operating cost. The model takes into account the constraints of passenger flow demand, train driving conditions and passenger service level, and is solved by using the enumeration method in short computation times. Through the analysis on the local urban rail transit network in a city, the model results are evaluated by the number of cross-line passenger flow, proportion of cross-line passenger flow, average transfer time of passengers, through rate for cross-line passengers and transfer coefficient. The results show that the operation of cross-line trains among different lines can effectively reduce the number of transfer and the travel time for cross-line passengers, and reduce the operating costs of enterprises.

Effective prestress under the anchorage refers to the prestress force retained by the prestress tendons under the anchorage opening of the working anchorage after the prestress tendons are tensioned and anchored. In order to improve the accuracy of the selection of the effective prestress under anchorage value of post-tensioned method small box girder, based on the measured data of effective prestress under anchorage of small box girder of a highway in Guangdong, four machine learning models, namely, multiple linear regression (MLR), adaptive boosting regression (AdaBoost), random forest (RF) and extreme gradient boosting (XGBoost), are used to comprehensively analyze the factors affecting the value. The results show that the mean square error (MSE) of the extreme gradient boosting model is 0.11, the mean absolute error (MAE) is 0.08, the Standard Deviation of the Residuals (SDR) is 0.33, and the goodness of fit (R-Squared) is 0.99. It means that the model predicts accurately, so it can be applied to the problem of accurately fetching the value of the effective prestress force under the anchorages of post-tensioned small box girders under certain conditions. In the XGBoost model, the degree of contribution of each factor to the model value is evaluated by introducing the Gini impurity (Gini) value, and the results show that the thickness of end web is the most important factor affecting the value of effective prestress under anchorages, accounting for 21.65%.

In order to study the influence of the mechanical properties of the steel truss bridge in the rigid area, the 202 m simply supported through double-deck truss-arch system bridge is taken as the research object. Three finite element space models of joint hinged model, joint rigid model and rigid zone of joint are established. The influence of three kinds of joint rigidity on the internal force, combined stress and deflection of steel truss bridge is analyzed. The research shows that the rigidity of the joint has little effect on the axial force of the bar, but has a great influence on the shear force, bending moment and combined stress. The increase of the rigidity of the joint will cause the connecting bar to produce larger secondary bending moment and secondary stress. At the same time, the influence of joint rigidity on the deflection of member joints cannot be ignored. There-fore, in the design and construction of double-deck steel truss girder-arch system bridge, the influence of joint rigidity on the mechanical performance of the structure should be fully considered.

Severe weather events constitute a significant safety hazard for urban rail transit systems, emphasizing the importance of risk management. This paper introduces a resilience evaluation method for urban rail transit lines during adverse weather conditions, leveraging Bayesian networks. The proposed method encompasses a comprehensive resilience indicator system tailored to severe weather scenarios, encompassing topological characteristics, passenger organization, and equipment factors. A risk analysis framework utilizing a risk matrix is then applied to evaluate subsystem vulnerabilities under various weather conditions. Furthermore, a Bayesian network-based resilience evaluation approach is designed to integrate and analyze these diverse factors. To demonstrate the effectiveness of this approach, a case study is conducted using real-world data from the Changping Line of Beijing Subway, resulting in resilience scores for critical components like elevators, 400V, traction power supply, and on-board equipment, as well as an overall resilience value for the entire line. The results indicate that, on rainy days, the resilience levels of Changping Line's key indicators and the entire line are precisely predicted by the proposed method. Furthermore, the key indicators and the overall line are generally anticipated to exhibit high resilience, which closely aligns with actual operational performance. These findings not only establish the accuracy of the resilience evaluation method but also reinforce its practical relevance and applicability in real-world settings.

Amid the electric vehicle market's rapid growth, there's an urgent need for accessible, efficient charging services. Existing charging station planning research, while focusing on techno-economic efficiency, inadequately addresses user satisfaction. This study introduces a bi-level planning model that accounts for user satisfaction. The upper layer aims to minimize construction costs for investors and network losses, under constraints such as land prices and grid capacity, employing an improved particle swarm optimization for station siting and sizing. The lower layer measures user satisfaction via costs, minimizing time and travel expenses, with considerations for queue times and acceptable distances, utilizing Dijkstra's algorithm and queuing theory to calculate user charging costs. Interactions between these layers optimize both investor returns and user needs, validated through case studies.

In low demand areas, it has been proven that demand responsive transit (DRT) is more cost-effective compared to fixed route transit (FRT). The present study further examines the demand threshold of four public transit systems: DRT, FRT, fixed-route but non-fixed-point transit and dynamic stop transit, which are designated to feed urban rail transit. It constructs user cost models for these four transit systems, where we aim to find the system with the “best” user experience. The demand at which the user costs of two transit systems are equal is referred to as the critical demand. After deriving the analytical expression for the critical demand, a sensitivity analysis of the analytical solution is conducted. The results indicate that DRT is optimal for low demand density scenarios. If we allow more flexibility to transit services compared with FRT, fixed-route but non-fixed-point transit proves superior for higher demand. Additionally, the critical demands between DRT and FRT, as well as that between DRT and fixed-route but non-fixed-point transit, are influenced by the bus speed and passenger walking speed, where the impact has an upper bound; at the same time, the fleet size also affects the critical demands, where the impact has a lower bound.

With the rapid growth of railway passenger volume and heightened safety concerns from both operators and passengers, stricter demands have been placed on railway track maintenance and repair work. Current techniques for detecting track geometry faults largely rely on manual inspections or track inspection cars, which are often time-consuming and encounter limited occurrences. To address these shortcomings, this study integrates car-body vibration data with deep learning technology, aiming at enhancing the efficiency and accuracy of railway track maintenance work. By incorporating techniques such as sliding window sampling, data augmentation and attention mechanism, this method effectively resolves issues with imbalanced sample distribution and further enhances the model's classification performance. Experimental results demonstrate that this approach significantly improves the ability of deep learning models to detect and classify track geometry faults using car-body vibration data. The comprehensive application of car-body vibration data and deep learning technology provides an efficient, automated detection method for railway track maintenance, laying a data foundation for implementing preventive maintenance strategies.

This paper explores the method of domain specialization for Large Language Models in the field of railway track management, as well as their potential and practical effectiveness in knowledge-based question answering scenarios. Given the complexity of the railway track management work scenarios and the challenges associated with existing data acquisition and processing, this paper has proposed a domain-specialized large language model tailored for railway track management. Additionally, by integrating layout analysis, OCR image recognition technologies, and the self-instruct method, the efficiency and accuracy of data processing have been enhanced, and the training dataset has been expanded. Moreover, the paper employs knowledge retrieval augmentation techniques to mitigate the issue of hallucinations produced by the model, thereby improving the model's accuracy, relevance, and completeness in knowledge question answering. Furthermore, the paper has established a dataset and evaluation metrics specifically for question answering in the railway track management field and assessed them using a general Large Language Model. This research not only aids relevant personnel in their daily operations and enhances work efficiency but also provides valuable references for future applications of Large Language Models in similar domains.

With more and more flexible marshaling of EMUs and movement across lines train operation, in view of the fact that the movement across lines trains have been determined in advance when they operate on this line, and the trains on this line are considered to participate in it when the connection time allows, this paper aims at minimizing the operation cost, constructs a dynamic connection network model, which considers the constraints of reconnection and disconnection, and designs an alternating direction multiplier method (ADMM). By decomposing the original problem into a group of independent EMU connection problems, the sub-problem is effectively solved by the shortest path algorithm. A case study on Beijing-Baotou Railway, a high- EMU Routing Planning can reduce the use of three groups of EMUs, improve the use efficiency of EMUs and save operating costs.

The metro network is a significant part of the public transportation network, and measuring its robustness can help evaluate its performance when facing different operation incidents. Based on the shortest path between nodes, commonly used robustness parameters such as network efficiency neglect the variety of routes in the metro network. The perspective of node failure also ignores that metro trains can only turn back at stations with sidings, and a single node failure may lead to nearby normal nodes being out of service. In this paper, we define all possible routes chosen by metro passengers as effective routes, give two novel metro network robustness parameters, and a new perspective called section failure based on the sidings at specific metro stations. Taking the Chengdu metro as an example, we discuss its robustness when facing malicious continuous attacks. The result shows that the network efficiency parameter and the perspective of node failure underestimate the loss brought by metro operation incidents, and the Chengdu metro network is vulnerable when facing malicious attacks, especially based on node Laplacian gravity centrally. The paper gives a new method to analyze the robustness of metro networks and advises operators on protecting key metro stations to avoid large-scale failure.

Against the backdrop of the medical reform in China, the pharmaceutical distribution industry is stepping into the era of meager profit. Pharmaceutical companies are confronted with problems including superficial inventory classification, unreasonable procurement methods, etc. To address the above problems, this article proposes the drug inventory control strategy based on the classification of multi-criteria decision-making. We first use the weighted fuzzy C-means clustering method to classify the sampled drugs according to the established criteria. We then further subdivide different categories and determine corresponding inventory control strategies based on requirement analysis. Results suggest that this strategy enhances company inventory management and slashes costs.

This paper takes a specific company as an example to reveal the existing problems in its transportation management efficiency. Through on-site research and literature review, it is found that customer satisfaction rate is a key indicator for measuring transportation management efficiency. The use of SPSS software con-firms that there is a significant strong correlation between customer satisfaction rate, on-time delivery rate, and cargo damage rate. By applying the Analytic Hier-archy Process (AHP), a comprehensive evaluation index system for transporta-tion management efficiency is constructed in aspects of vehicles, personnel, oper-ational processes, and informationization. The final results show that the three factors with the highest weights are on-time delivery rate (C8), vehicle utilization efficiency (C3), and cargo damage rate (C7).

Aiming at the current problems of frequent traffic accidents, high subjectivity of risk warning features and low accuracy of road grade prediction, a random forest-matter-element risk warning model is proposed. The model selects the actual data set of the I880-N highway in the United States, defines 13 warning evaluation indicators from three aspects: traffic flow, speed and occupancy, and uses the random forest method to screen the warning features to achieve the purpose of removing dimensional disasters and weight distribution. The K-means method is introduced into the matter-element model to divide the risk level. The established model is applied to the accident-prone locations on the main line of the I880-N highway in the United States. The results show that the comprehensive risk goodness of the highway is 0.053, the warning level is level 4, and it has severe risk. This evaluation result is consistent with the actual situation, which verifies the applicability of the risk warning model and can provide new ideas for the development of road risk warning work.

This paper focuses on the energy consumption of urban rail transit trains with flexible train compositions. An energy-efficient speed profile model is proposed to calculate the motion state of various train types. Considering the traction energy consumption and regenerative braking energy, an energy consumption model is established to assess the total energy consumption under a given flexible train service plan and timetable. A set of numerical experiments based on real metro line data is conducted to analyze the energy consumption of different train operation scenarios. The results show that the energy consumption of the flexible train composition mode is reduced by 11.70% compared to the fixed train composition mode, highlighting its potential for enhancing energy efficiency and sustainability in urban rail transit systems.

With the extensive construction of modern highway bridges, incidents of bridge overturning due to vehicle overloading have become increasingly common, prompting a focus on the anti-overturning design of single-column highway bridges. Custom transport vehicles are typically required to travel along the centerline of the bridge at a uniform, low speed. However, in emergency situations where lane closures occur, custom transport vehicles cannot adhere to the centerline as specified. To investigate whether custom transport vehicles can traverse single-column pier bridges with lateral eccentric loads, this study conducted finite element simulations for five typical vehicle loads and 35 different load conditions. The stability against overturning of single-column pier bridges under the influence of lateral eccentric custom transport vehicle loads was analyzed. The results indicate the following: When the eccentric distance of the heavy vehicle exceeds 1m, the structural stability is negatively correlated with the vehicle load and lateral eccentric distance; Custom transport vehicles of types C-I and C-II can safely pass the bridge under all conditions; Custom transport vehicles of types C-III and C-V need to stay within a lateral eccentric distance of 5.32 m when crossing the single-column bridge; Custom transport vehicles of type C-IV must stay within a lateral eccentric distance of 3.97 m when crossing the single-column bridge; For the bridge studied in this paper, the eccentric load reaching characteristic state 1 can be considered as the limit condition for determining whether the bridge has the right of safe passage.

Line and slope adjustment is the last process of rail transit line design, through the design of slope adjustment to fully solve the problem of boundary incursions caused by construction deviations, thus ensuring the safe and good operation of trains in the future. The traditional method of line and slope adjustment relies on the experience of designers, which is a large amount of work and the design accuracy is not high enough. In this paper, the optimisation model of line and slope adjustment is constructed based on Grasshopper, which is currently focused on optimisation of the planar line position, and solved by genetic algorithm and simulated annealing algorithm. The above method is applied to a real rail transit shunting case, and the machine-selected scheme is better than the manual scheme, which verifies the validity of this paper’s method.

The fixed pricing strategy of high-speed railways has, for an extended period, presented a challenge to the enhancement of revenue. This research paper employs the theory of revenue management and applies it to the problem of dynamic pricing for multiple high-speed trains. It presents a solution to the issue by transforming it into a Markov Decision Process (MDP). Considering passenger choice behavior, a reinforcement learning environment for dynamic pricing of multiple trains is established with the objective of maximising the expected revenue. The Double Deep Q Network (DDQN) from deep reinforcement learning is employed to solve this issue. The efficacy of the algorithm is evaluated using the Beijing-Shanghai high-speed railway as a case study. Experimental results demonstrate that, compared to fixed pricing strategies and traditional dynamic pricing algorithms (Particle Swarm Optimization, PSO), the DDQN model increases total revenue by approximately 7.70% and 3.89%, respectively, indicating its practical application value.

Understanding users’ travel behaviors is an important subject in traffic science, which helps traffic management departments to formulate appropriate traffic control strategies. Travel mode identification is a critical aspect of analyzing users’ travel behaviors, aiming to precisely determine the travel modes of users and promote the sustainable development of urban intelligent transportation systems. In this article, we focus on identification methods for travel modes and propose a multi-scale convolutional neural network model based on channel attention mechanism. Our model efficiently extracts multi-scale sub-features and prioritizes key features. Additionally, to deeply extract effective information from the GPS data, we design 24 time-domain features and frequency-domain features and simplify them using principal component analysis and random forest algorithm. To verify the effectiveness of our method, we design five traditional machine learning models and two shallow deep learning models as comparative models, which are trained and tested on the Geolife dataset. Experimental results show that our approach is superior to the baseline models, achieving a maximum accuracy of 83.09%.

To address the challenges of accurately reconstructing passenger travel chains and predicting the spatiotemporal distribution of passenger flow in transportation hubs, this paper proposes a deep learning architecture based on mobility-driven spatial embedding and extended long short-term memory networks (MDSE-sLSTM). Initially, we conducted route choice behavior experiments in a virtual reality (VR) hub scenario to reconstruct passenger mobility chains. Then, using Skip-gram, we analyzed the semantic correlations between nodes representing functional areas to achieve a spatially embedded representation of the hub area, obtaining spatial correlation features. In the temporal dimension, an extended bidirectional long short-term memory network with multi-head attention mechanism was employed to jointly learn the time-varying patterns of passenger flow data. Validation on the passenger flow distribution dataset of Tianjin West Station shows that the MDSE-sLSTM model outperforms traditional prediction models. Ablation experiments further confirm that the method of extracting spatial features based on mobility-driven approaches significantly improves passenger flow prediction compared to extracting spatial features based on the actual physical network.

Stay cables of long-span cable-stayed bridges are susceptible to rain-wind induced vibration during windy and rainy conditions due to their low mass, high flexibility, and low damping. Such vibration may result in fatigue failure of the cable end anchorage devices, reducing the bridge’s durability and safety. To mitigate the damage caused by rain-wind induced vibration to bridge structures, it is essential to understand the vibration mechanism and vibration reduction measures of rain-wind induced vibration of stay cables. This paper reports a state-of-the-art review on rain-wind induced vibration of stay cables and summarizes the effects of wind speed and rainfall on rain-wind induced vibration. The possible mechanisms of rain-wind induced vibration is introduced. The application of existing vibration reduction measures is also discussed, including auxiliary cable vibration reduction, damper vibration reduction, and aerodynamic vibration reduction. The research results can offer valuable reference and guidance for the selection of vibration reduction measures for rain-wind induced vibration of stay cables.

Under the background of deepening the reform of railway freight, to improve the market competitiveness of railway freight, it is necessary to combine its own resources, fully consider the needs of customers, and provide quality services. Based on the service resources provided by railway, this study considers the importance of intangible resources in the process of freight service, and summarizes the four dimensions of organizational resources, information resources, transportation service resources and market resources, and builds a service resource index system from the perspective of customer participation. Through the extension analytic hierarchy process, the weight of each index is determined to optimize. The results show that the railway freight service needs to improve the transportation service capacity, optimize the organizational structure, strengthen the information transmission and personalized product provision.

In high-speed railway operations, disturbances from external environments and inherent control delays make it difficult to follow the optimal curve in the actual control process, which poses challenges for applying the proposed energy-efficient driving strategies in Automatic Train Operation (ATO) systems. To address issues for existing industrial speed tracking controllers, this paper proposed a PI controller with a feedforward compensation method based on optimization algorithms (OPFPI), incorporating the Smith predictor to mitigate the effects of control delays. A simulation model based on Simulink is developed to simulate the energy-saving and speed-tracking effects during actual train operations. Simulation results demonstrate that considering the inclusion of the feedforward compensator enables better tracking of the target trajectory and exhibits significant energy-saving benefits.

Forecasting interregional freight demand for large regions has been a challenging task due to intricate interactions between freight demand and multimodal transportation networks. With this, this paper endeavors to develop a high-fidelity integrated modeling approach to forecast freight demand within a large region. The proposed framework combines spatial economic modeling with a multimodal transportation supernetwork modeling approach to systematically model the quantitative relationship between freight demand, socioeconomic activities, and multimodal transportation networks. First, a “high-fidelity” spatial economic module is developed based on input-output tables, land use, and population/employment data. Parameters of the spatial economic model are calibrated using observed trip length distribution and traffic count data. Then a “high-resolution” multimodal transport supernetwork, consisting of a detailed highway, railway and waterway networks, is developed to assign freight demand to the multimodal transportation supernetwork. Finally, a case study is conducted using the proposed method for the Yangtze River Economic Belt, which demonstrates the utility of the proposed approach in systematically modeling and forecasting the freight demand, as well as in analyzing the impact of related economic, land-use, and transport policies in a high-fidelity fashion. The proposed approach supports improved decision-making in planning and operating regional multimodal transportation systems and related systems, such as economy, land-use and possibly environment.

There is a close correlation between airports and regional economic development, and ten-million-level airports drive the upgrading of the city’s industrial structure, improve the level of openness to the outside world, and become the engine of regional economic growth. Based on the panel data of 2010–2019 of 38 cities where ten-million-level airports are located, we constructed the fixed effect model and the mediation effect model, focusing on analyzing the influence of ten-million-level airports on regional economy and its functioning mechanism. The results of the empirical research show that ten-million-level airports have a significant effect on the regional economy, and every 1% increase in the workload of ten-million-level airports will drive the regional economy to grow by 0.59%. The positive effect of international trade is the main mechanism for ten-million-level airports to promote economic growth. Therefore, in the future, air transportation will play an increasingly important role in promoting international trade exchanges and driving regional economic growth.

With the increasing severity of urban traffic congestion, adaptive traffic signal control based on deep reinforcement learning has been widely studied. Q-value based Deep Q Network reinforcement learning algorithms have been introduced for various types of expansion due to its effective efficacy. However, most of the studies design rewards and algorithms for specific application scenarios, without considering the impact of sampling method in experience replay during DRL-based traffic signal control model training. Moreover, Traffic control is often a multi-objective optimization problem, which contains traffic-participant pedestrians and vehicles. Most studies design comprehensive rewards to solve multi-objective problem without considering the sample size for each target in multi-objective. To address these, this paper proposes a soft-constraint reward for demands of pedestrian crossing, and a distributed multi-head policy DRL-based traffic signal control method based on cluster sampling for pedestrians and vehicles, Cluster Sampling-Multi Head Dueling Deep Q Network. For multi-intersection traffic signal control, each agent at intersection contains two policy networks, collects the local and neighboring intersection states from environments and controls phase duration according to two policies respectively. The experiences produced in environments exploring are stored separately. For each policy network, experience will be clustered by rewards, and then sampled for training. It can make experience replay in line with reward distribution, which improves model stability and generalization. In experiments, Cluster Sampling-Multi Head Dueling Deep Q Network is compared with existing traffic signal control methods in the real traffic environment of xuancheng to verify its performance.

Addressing the issue of data security sharing within the railway sector, this study explores the construction of a secure railway data sharing system based on blockchain and privacy computing, integrated with the architecture of railway data service platforms and the railway data privacy computation framework. The system is specifically designed for operational data storage in the data sharing module of railway data service platforms. It incorporates blockchain technology to manage, monitor, and trace every data operation by various stakeholders involved in the data sharing process. The aim is to enhance the security and controllability of the data sharing process, facilitate internal and external data sharing within the railway sector, break down data silos, and transform the value of data.

Nowadays, high-speed railway has become an important mode of transportation. In real life, train operation is often disturbed by various factors, which may lead to unexpected delay in train running time. Therefore, train timetable should be designed with robustness to improve the ability to cope with disturbances. The research on optimal layout of slack time of high-speed railway timetable can provide important support and guarantee for the safety, reliability and efficiency of high-speed railway operation. Firstly, based on the research of many experts at home and abroad, the present situation of train slack time research is summarized, and the shortcomings of current research are put forward. In this paper, a light robustness model is established. Taking the operation performance of Beijing-Shanghai high-speed railway within one month as an example, trains in Tengzhou-Bengbu South section of Beijing-Shanghai high-speed Railway are selected, and the optimized slack time setting scheme and related train timetable are solved by using the Gurobi solver, which provides a new idea for the optimization and adjustment of train timetable. Finally, by setting the disturbance conditions, it is proved that the scheme of optimizing the slack time using the light robustness model can improve the flexibility of the train timetable.

Effective air traffic management (ATM) is crucial for ensuring the safety, efficiency, and sustainability of air travel. Traditional ATM systems, while capable of monitoring aircraft, often lack the capability to optimize aircraft trajectories, resulting in inefficient airspace usage and increased environmental impact. This paper introduces a novel approach utilizing the column generation algorithm for optimizing four-dimensional (4D) trajectories in ATM. Unlike previous methods, this approach incorporates an optimization method that simultaneously integrates multi-aircraft 4D trajectories with considerations for fuel efficiency, timing delays, and adjustments in aircraft altitude and speed. The column generation framework decomposes the 4D trajectory optimization problem into a master problem and subproblems. The master problem selects suitable 4D trajectories for aircraft, and subproblems generate new 4D trajectories. Computational results indicate that the framework gives high-quality solutions with very small provable optimality gaps in short run time, confirming the method’s effectiveness in enhancing real-time decision-making in ATM systems.

Public transport plays a crucial role in economic and social integration by facilitating mobility. However, disruptions frequently impact public transport services, resulting in delays and cancellations that significantly hinder mobility and annoy passengers. This paper focuses on improving mobility resilience through service management, in the stage of line planning where a set of lines for operating services is developed. We propose a line planning model (in an integer linear programming) for public transport networks to determine the offered lines and their frequencies, with consideration of multiple service types such as train, tram, and bus. The objective comprises two terms: passenger convenience and service resilience, with respect to the given operating budget and static demand. We quantify service resilience from three aspects: connectivity, balance, and diversity. A case study is performed on a small 6-node network to examine the effectiveness of the proposed method.

To research the travel choice behaviour of travellers in a mobile-connected environment, this study starts by examining the travel decision mechanism. Using a Bilevel Programming model, the influence of utility perception on travel utility is analysed. Additionally, the study considers and quantifies the impact of travellers’ travel habits on their choice of travel mode, constructing a Nested Logit model that includes a dynamic utility change perception function. The feasibility and validity of the model are verified through case studies, and the influence of travel habits on travel decisions within the model is further analysed. The results show that during the morning peak hour, mobile connectivity has the greatest impact on the likelihood of choosing metro travel, followed by cab, and the smallest impact on bus travel. There are also differences in the impact effects depending on the destination points. Moreover, travel habits inhibit changes in travel modes in a mobile-connected environment. The larger the travel utility perception threshold, the greater the probability that travellers will maintain their original travel mode, making it more difficult to switch to a mode with higher travel utility. As the travel utility perception threshold increases, the dependence on the original travel mode has an increasing influence on travel decisions.

To improve the efficiency of urban transportation systems and reduce congestion, this article focuses on the problem of dynamic traffic flow assignment with uncertain multimodal transportation demand. The objective is to minimize travel costs for commuters. We constructed a multimodal dynamic traffic flow allocation model with road capacity and solved it by combining variational inequalities with the MSA algorithm. The example analysis shows that the traffic volume allocation of different travel modes conforms to the Logit probability selection function, and the increase in the road network information perception coefficient makes travel mode selection more deterministic. This provides a new perspective for urban transportation planning and management.

Interchange is an important link in the urban transport system, but the problem of inefficient travelling of urban residents has not been improved. In this paper, we take the metro-transit composite network in Putuo District, Shanghai as the research object, use Space-L model to construct the metro-transit composite network with walking as the transfer mode and walking plus shared bicycle as the transfer mode respectively, study the basic topological properties of the metro-transit network and compare the network characteristics of the composite network with different transfer modes. The results show that the average degree value, network density, local efficiency and global efficiency of the composite network with two modes of transfer are higher than that of the composite network with walking as a single mode of transfer, and it also proves that the two modes of transfer cooperate with each other, so that the whole public transport system can play a greater efficiency.

Under the background of carbon peaking and carbon neutrality, the low-carbon transformation of multimodal transport has attracted much attention.In view of the exploration of transportation route optimization by the transportation industry, this paper comprehensively considers transportation cost, carbon emission and other factors. At the same time, the carbon tax policy is added to establish a low-carbon multimodal transport path optimization model with the total cost of goods transportation and the lowest carbon emissions as the goal, and combined with the case through the genetic algorithm to calculate the optimal low carbon multimodal transport combination mode and path. The research results show that this model can optimize the transportation structure, productively decrease transportation costs and carbon emissions, and reasonable carbon tax policy can provide freight enterprises with transportation mode and path decision guidance, to achieve efficient low carbon transportation, and further facilitate the sustainable progress of society.

Due to the special nature of the dangerous goods warehouse itself, for safety reasons, the monitoring requirements for the dangerous goods storage environment are more stringent than those for ordinary warehouses. However, the internal environment of the dangerous goods warehouse is highly hazardous to human beings, prone to personal injury and death accidents, and not suitable for personnel on-site task operations. This paper intends to combine the characteristics of UAV space flight flexibility, easy to carry ancillary equipment, as well as the wireless transmission of data from wireless sensor networks, low power consumption and other characteristics, the application of UAV technology, Internet of Things technology, research and design a set of UAV cargo information collection and environmental monitoring system solutions suitable for dangerous goods warehouses, to carry out real-time recording and monitoring of a variety of environmental parameters for the dangerous goods in the storage of the warehouse, and realize the Intelligent collection of dangerous goods information and real-time monitoring and management of environmental parameters in dangerous goods warehouses, thereby improving the safety of dangerous goods storage, and at the same time improving the efficiency of storage inventory and inspection and the intelligent level of environmental monitoring.

The evaluation index system for the selection of railway container hub nodes is screened and constructed from the three aspects of regional economic development level, node capacity scale and transportation network structure conditions. On the basis of 18 railway container hubs, 4 cities, Jinan, Lianyungang, Nanning and Hefei, are added, and finally 22 cities are taken as the alternative cities, and entropy value method and entropy weight TOPSIS method are utilized to comprehensively evaluate the alternative city nodes respectively. Using the comprehensive evaluation results to transform the traditional gravity model, the improved gravity model applicable to hub selection is obtained. After comprehensively analyzing the evaluation results of the indicators and the total gravitational value of the cities calculated by the improved gravitational model, and combining with the geographical factors, Jinan, Wuhan, Chongqing, Shanghai, Guangzhou, Tianjin and Hefei are finally selected as the hub cities of the railway container transport network.

The study of charging choice for electric vehicle holds significant importance for the promotion of new energy vehicles, the realization of energy saving and emission reduction. Based on one month's global positioning system data of electric vehicles, this study segments the daily usage behaviors of electric vehicles. It identifies the starting and ending points of trips and extracts the instantaneous state information of vehicles at trip endpoints. This information is used as independent variables for electric vehicle charging choice, with the presence of charging at trip endpoints as the dependent variable. A binary Logit model is employed to establish a model analyzing the charging choice behavior at the endpoints of electric vehicle trips. The model's parameters are then estimated and tested. The results indicate that the state of charge, cumulative mileage, and time significantly impact the charging choice of electric vehicle at trip endpoints.

The station passing capacity is the key performance indicator to evaluate the planning and operation of railway stations. This paper mainly studies the calculation method about the effective passing capacity of railway stations that incorporates the train operation process into the calculation, which is of great significance for the theoretical research of station passing capacity calculation. At first, this paper introduces the concept of train operation chain, and then introduces the construction of double-layer alternative set based on train operation chains. On this basis, the effective passing capacity model of railway station is constructed. A case study based on the actual data of Jingping Logistics Park verifies the effectiveness of the model and the station technical operation scheduling is drawn based on the effective passing capacity. In the end, this paper analyzes the bottleneck of the logistics park's capacity and gives suggestions for capacity improvement.

The swift formulation of dispatch adjustments amidst disturbance scenarios poses a critical challenge in the daily operations of high-speed railways. Accurate prediction of train delays, along with their potential impact, is essential for effective dispatching and command operations. Therefore, a hybrid method named Density clustering - Bayesian optimization - Memory network (DBM) hybrid method is proposed, merging data mining and machine learning to increase prediction efficiency and accuracy. Leveraging techniques like Hierarchical Density-Based Spatial Clustering of Applications with Noise, the DBM hybrid method uncovers train delay evolution patterns. Within each pattern, Bayesian Long short-term memory is used to predict train delays. This DBM hybrid method integrates density clustering and Bayesian optimization parameter of the memory network, offering a forward-looking approach to dispatch adjustments. Utilizing real operation data from the Beijing-Guangzhou high-speed railway, the results demonstrate a prediction accuracy of 93.475% with a permissible error of 1 min. This improvement in accuracy and efficiency has significant practical implications.

In the field of autonomous driving, automatic parking stands out for its ability to enhance convenience, alleviate traffic congestion, and improve parking efficiency. Traditional parking technologies, relying on complex sensor integration and precise path planning, often fall short in unpredictable and intricate situations. Deep Reinforcement Learning (DRL) has emerged as a promising approach to address these challenges, yet its application in parking still faces limitations, such as a narrow focus on conventional parking methods and a lack of capturing the full unpredictability of real-world parking conditions. This research seeks to advance the field by utilizing the Highway_env repository to create realistic parking scenarios that include reverse, parallel, and advanced diagonal parking scenarios. These scenarios vary from open to obstructed environments, increasing the realism and complexity of the parking task. Additionally, this study integrates Convolutional Neural Networks (CNNs) with DRL to enhance parking precision in diverse settings. Through detailed analysis, the study reveals that while both Soft Actor-Critic (SAC) and Twin Delayed Deep Deterministic Policy Gradient (TD3) algorithms demonstrate high success rate in simple scenarios (up to 0.99 success rate), TD3 displays superior performance in complex environments by exhibiting greater adaptability and successfully acquiring safer parking strategies, achieving success rate levels between 0.8 and 0.93. This highlights TD3’s exceptional adaptability and resilience in tackling a wide range of parking challenges, establishing it as a key solution capable of navigating the complexities of automatic parking.

The high-speed railway system often encounters inevitable disturbances due to adverse weather and equipment failures. In orderssss to achieve integrated train timetable and vehicles rescheduling under disturbance and restore normal operation of the high-speed railway system as soon as possible, an integer linear programming (ILP) model based on a time-space-state network is proposed. This model can avoid train cancellations, add additional stops for trains and adjust the train sequence through overtaking. It allows for flexible changes when the types of vehicle used by train are the same. In order to solve the model, the model predictive control (MPC) framework is used to decompose large-scale problems into several small-scale problems, while ensuring the real-time nature of the solution process. Within each time period, the Lagrangian relaxation algorithm is used to solve small-scale problems. Verify the effectiveness of the model and algorithm using actual data from the Beijing Zhangjiakou high-speed railway. Compared with commercial solvers, algorithms can reduce solving time while ensuring solution quality. Compared to fixed vehicle circulation, flexible circulation can effectively reduce the total delay of trains.

The development of China’s “eight vertical and eight horizontal” railway network has led to the transition of China’s railway from the “big construction” stage to the “big operation” stage. To improve the efficiency of railway transportation organization and meet the growing travel needs of the people, how to comprehensively consider the issues of train line planning optimization and timetable optimization has become a key factor. To this end, this paper first puts forward the simple algorithm of the train line planning optimization, that is, first determine the train running section, then calculate the number of trains and determine the stopping plan of each train. Then, taking the minimum total train travel time as the objective function, the optimization model of the high-speed railway timetable is constructed. The model adds train proportionality constraints to ensure the proportionality of the timetable with a small number of trains. Finally, taking the Rizhao-Qufu section of RLHSR as an example, the optimization research of the operation schedule and timetable is carried out. A total of 19 pairs of trains are running on the Rizhao-Qufu section of RLHSR, including 17 pairs from the Rizhao West-Qufu East section and 2 pairs from the Linyi north-Qufu East section. The total travel time of trains on the RLHSR is 2760 min.

For a significant period in the future, vehicles with different levels of automation will share road resources, with connected vehicles capable of traveling in closely spaced platoons. In this study, a mixed traffic flow model considering connected vehicle platooning is proposed based on a Markov chain model. With numerical experiments, the effects of vehicle penetration rate and maximum platoon size on fundamental diagrams and lane capacity are investigated. The results indicate that the increasement of penetration rate of connected vehicles can significantly promote the capacity. In addition, there is a threshold value for the maximum platoon size. When the maximum platoon size is smaller than the threshold value, lane capacity will increase with the increasement of maximum platoon size. While if the maximum platoon size is larger than the threshold value, the capacity will decrease as the increase of the maximum platoon size.

Intercity traffic networks support the integrated and synergistic development of city clusters. However, there is a notable gap between outdated traffic construction and growing intercity travel demand. This paper constructs a comprehensive intercity traffic network for the Guanzhong city cluster. By projecting population, GDP, and employment at district level, the intercity travel demand is quantitatively predicted by applying the improved four-stage method. The applicability of intercity traffic network is analyzed, and strategies are proposed to solve the problem of imbalanced supply and demand of intercity traffic. The study indicates that (1) with the economic and social development, the total scale of intercity passenger traffic has grown significantly in the planning year. (2) There is a 66% high level of intercity traffic within the provinces, and intercity travel between provinces mainly originates from border cities. (3) The road network saturation in the core area has increased significantly, leading to increasing “centripetal” transit traffic pressure. A series of parameter calibration methods for intercity traffic models are developed to improving the decision-making level of Guanzhong city cluster traffic planning.

China's urban rail transit is gradually transitioning from independent single line operation to a multiline network operation mode. The rolling stocks is the most important mobile transportation resource. Compared with the existing single line operation mode, the network operation organization mode can effectively improve the efficiency of train base operation and reduce operating costs. This article takes the shared use of multiple lines in a rolling stocks depot as the background, and takes the routes of two shared rolling stocks depots as an example to establish a model for planning the use of stocks scheduling for multiple lines in rolling stocks depots. Select the maximum and minimum ant colony algorithm for solving, calculate the rolling stocks scheduling for independent mode and shared mode of the vehicle depot, and compare the obtained rolling stocks scheduling. The results indicate a decline in the number of locatives and vehicles below 100 minutes and above 400 minutes, exceeding all equilibrium levels. The plan of using multiple shared locomotives and vehicles can effectively reduce the number of locomotives and vehicles used, improve rolling efficiency, but it will lower the balance level of locomotives and vehicles.

Pantograph is a key equipment of the electrical system in rail transit vehicles, which is easy to wear during operation. The monitoring of pantograph slider wear is beneficial to the maintenance of electrical system, which is of great significance to the safety and stability of rail transit. In this paper, an image-based strategy for extracting the wear edge of the pantograph slider is proposed. A bi-directional cascade network structure is utilized to detect the edge of pantograph slider. Reciprocal Dice coefficient loss is introduced to enhance the edge detection results. With the design of a post-processing module, the complete framework for extracting the wear edge of pantograph slider is established. This method is validated on the pantograph image data from Beijing Metro Line 6. Experimental results show that the proposed method has better performance than the original bi-directional cascade network with the number of model parameters significantly reduced. ODS, OIS and IMP increased by 1.2%, 1.3% and 2.54%, respectively with the detection speed improved greatly, and the model efficiency is approaching that of a lightweight network.

Autonomous vehicles have already been introduced on city roads, and they are poised to become a significant component of the transportation system. Buses, being a critical part of public transport, are particularly suitable for the advancement of higher levels of autonomous driving technology due to their fixed routes and scheduled times. This paper investigates the stopping behavior of autonomous buses on future urban roads under mixed traffic conditions. Specifically, it examines the lane-changing and following behavior of autonomous buses when entering and exiting stops. The study proposes a game theory-based control strategy for bus entry and exit, modeling and simulating driving decisions using game theory principles. The model is solved and simulation is conducted to verify the feasibility and advantages of the proposed strategy. This strategy can provide enhanced decision support for the control of autonomous buses in mixed traffic flow.

This study focuses on the over-crowded situation after disruptions, contributing to fill the gap that applying virtual coupling technology in rescheduling problems in the metro lines. Specifically, both stop-skipping instructions and virtual coupling/decoupling instructions are considered to be sent to the trains in real-time, and a mixed-integer nonlinear programming (MINLP) model is built based on these two kinds of instructions. The proposed model is solved by GUROBI after linearization, and computational results based on a real-world metro line show that by applying virtual coupling technology, the total waiting time of passengers could significantly reduce by nearly 50%. In addition, it is verified that combining virtual coupling technology with the stop-skipping strategy could have an advantage compared with only applying the virtual coupling technology or the stop-skipping strategy.

This paper studies the site selection problem for aviation emergency logistics centers. Based on the different task requirements faced by aviation emergency logistics centers, they are divided into two categories: first level aviation emergency logistics centers and secondary aviation emergency logistics centers. For first level aviation emergency logistics centers, a comprehensive evaluation model using the TOPSIS method based on entropy weight is applied for site selection. For secondary aviation emergency logistics centers, a multi-objective model aimed at minimizing costs and maximizing coverage area is established, and the NSGA-II algorithm is used for site selection and solution. Finally, using the Northwest region of China as a case study background, the feasibility and practicality of the models are discussed. The optimal site selection schemes for aviation emergency logistics centers when the number is 10, 13, and 15 are calculated, and the universality and feasibility of the two models proposed in this paper are analyzed.

Risk assessment plays a crucial role in ensuring safety and mitigating potential losses in highway tunnel construction. In response to the challenges commonly encountered in this domain, this paper presents a comprehensive risk assessment index system along with evaluation method to establish a risk assessment model based on the t-s fuzzy neural network. First, the WBS-RBS method was used to identify the risk assessment indices and enhanced by taking soil erosion factor into account. Then, the quantitative criteria of safety risk and risk assessment indices of highway tunnel construction were determined. And the principal component analysis (PCA) method was used to refine and process the quantified indices. As a result, a highway tunnel construction risk assessment model based on the T-S fuzzy neural network was developed and validated using specific construction cases in China. The findings demonstrate the model's validity and reliability, making it a useful reference for similar projects and future research.