Archive

25 April 2022, Volume 22 Issue 2 Previous Issue    Next Issue

---

For Selected:

Download Citations EndNote Ris BibTeX

Toggle Thumbnails

Select

Overview of Life-oriented Travel Behavior Research ZHANG Jun-yi, LI Shuang-jin, MA Shuang 2022, 22(2): 1-16.  DOI: 10.16097/j.cnki.1009-6744.2022.02.001 Abstract ( )   PDF (1810KB) ( )   PDF(English version) (417KB) ( 194 )   The research on travel behavior should make a paradigm shift from traditional approaches to the life-oriented approach supporting the interdependence of life choices. The life-oriented travel behavior research argues that travel behavior not only results from various life choices, but life choices are also affected by travel behavior. Such arguments have not only theoretical foundations but also empirical evidence. Accordingly, policy decisions related to citizens' lives require communications and collaboration across sectors. Cross-sectoral actions need common languages. The lifeoriented approach presents a brand new theory to support cross-sectoral transportation planning and management. It can also serve as a new theory to support cross-sectoral urban and regional policy (or general public policy). The lifeoriented approach is developed as a truly scientific system, which can serve as a common language to support crosssectoral policy making. In this review paper, first, the life-oriented approach and its implications to deal with travel behavior research are described. Second, the research progresses of the following ten topics are overviewed: lifestyles, car dependence, household energy consumption, information and communication technology and life, health and life choices, tourism behavior, mobility of the elderly, risky behaviors of young people, responses to natural disasters, and mobility biography. These topics are closely related to travel behavior. Finally, travel behavior research challenges and prospects are discussed from the perspective of the life-oriented approach. References | Related Articles | Metrics

Select

Identifying Metropolitan Center Structure Based on Commuting Patterns LIU Xiao-bing , LI Feng-xiao , TIAN Xin-mei, YAN Xue-dong 2022, 22(2): 17-28.  DOI: 10.16097/j.cnki.1009-6744.2022.02.002 Abstract ( )   PDF (3554KB) ( )   With the increasing urban traffic problems, urban managers are seeking the way to guide the transformation of traffic patterns through urban spatial planning, and to balance the relationship between supply and demand and alleviate traffic problems. As an important step of urban spatial planning, however, center structure identification is still limited in analytical methods and data applications. Using the commuting data from Baidu location dataset, this paper established the grid-based DBSCAN(Density-Based Spatial Clustering of Applications with Noise) density clustering algorithm to identify the center layout of 35 major metropolitan areas in China. The paper also analyzed the center structures of metropolitan areas through developing five commuting theoretical models and determining corresponding quantitative indicators. A regression analysis was also performed to identify the factors that affect the commuting efficiency of the metropolitan areas. Results show that center layouts of different metropolitan areas are obviously different and most metropolitan areas exhibit an unbalanced polycentric structure. The distribution of the center structure has some regional characteristics. For example, the monocentric metropolitan center structure is mainly found in the central and western cities of China. The constrained diffusion and balanced polycentric structure are mostlydistributed in the east coastal cities. The city size and commuting time shows the most significant correlation. The workhousing balance also has a great impact on commuting time. The results of this study help to determine the effective strategies for resource allocation and commuting efficiency optimization for metropolitan areas which also provides useful references for the spatial planning and sustainable development of transportation in metropolitan areas. References | Related Articles | Metrics

Select

Evolution of Hinterland-port Accessibility Pattern and Its Spatial Spillover Effect LI Jie-mei , GAO Li, QI Jing-jie, YIN Qi 2022, 22(2): 29-36.  DOI: 10.16097/j.cnki.1009-6744.2022.02.003 Abstract ( )   PDF (2583KB) ( )   In order to explore the evolution characteristics and spillover effects of the hinterland-port accessibility, the hinterland-port accessibility measurement model is constructed. The spatial difference and spillover effect are analyzed by using Theil index and spatial Markov chain. The result shows that the high-value zones of hinterland-port accessibility extends along the main transportation channels and presents a cluster agglomeration phenomenon, and the high-value zones spread along one belt and one road. The spatial difference of hinterland-port accessibility shows a Vshaped characteristic. The spatial imbalance in the Eastern and Western regions is greater than that in the Central and Northeast. High accessibility cities have significant positive spatial spillover effects. When adjacent to high accessibility cities, the upward transfer probability of low accessibility cities is greater than that of adjacent to medium and low accessibility cities. Hinterland-port accessibility formed a spatial-temporal pattern dominated by the shortest travel time and supported by road network density. References | Related Articles | Metrics

Select

Job Accessibility Analysis Considering Travel Cost XU Qi, CHEN Yue , HUANG Jing-ru , GAO Shun-xiang , ZHANG Zhi-jian 2022, 22(2): 37-44.  DOI: 10.16097/j.cnki.1009-6744.2022.02.004 Abstract ( )   PDF (2833KB) ( )   PDF(English version) (879KB) ( 200 )   Good job accessibility contributes to the innovation of the job-living interface, which is a key issue in building sustainable cities. Existing accessibility studies have mostly considered travel costs limited to travel distance or travel time, without fully considering travel costs and their effects on different travel modes. Based on POI (Point of Interest) and route planning data from Internet maps and business data platforms, this paper obtains fine-grained employment and travel data and uses an improved two-step floating catchment area model to propose a job accessibility measure that takes into account travel costs, to study the job accessibility of both public transport and private cars and to evaluate the impact of adding travel costs on job accessibility. The case study in Beijing shows that the average travel cost of private cars changes from 54% to 6% higher than that of public transportation after considering travel costs. And job accessibility is sensitive to travel costs, and the interaction between commuting and travel costs cannot be fully captured by considering travel time only; the impact of travel cost is reflected in an overall average decrease of 7.3% and 4.8% for public transportation and private car accessibility, and without considering travel cost, the job accessibility of subdistricts along with the fifth to sixth ring subway will be underestimated; there is a boundary effect of the travel cost on accessibility, and the higher the threshold, the smaller the impact. This paper provides insights for planners and policymakers making job accessibility-oriented adjustment strategies for the balance between jobs and workers. References | Related Articles | Metrics

Select

Time Value Based Route and Port Selection for China and Africa Multimodal Transportation FENG Fen-ling , SUN Nan-jia 2022, 22(2): 45-53.  DOI: 10.16097/j.cnki.1009-6744.2022.02.005 Abstract ( )   PDF (2033KB) ( )   To distinguish and refine the route and port choice of different types of goods, this paper takes the time value of goods as a type of transportation cost. The time value cost of goods is defined as the capital occupation cost and the loss of goods value. Considering the liner's timetable, transportation quotes and the time value of goods, this paper proposes the route choice model of port selection and the way and mode of port transportation. The case study of the Changsha-Durban using the depth-first traversal algorithm show that the liner schedule, transportation price, and the time value of the cargo would affect the choice of route. In terms of freight amplitude modulation, for high value and high time sensitivity cargo, the shipper's port selection has strong robustness. For low value and low time sensitivity goods, the amplitude modulation of freight quotation can be controlled (1000 USD·TEU-1 ) to realize the change of shipper's choice of multiple ports, which has relative flexibility and controllability. References | Related Articles | Metrics

Select

Pricing and Platform Revenue and Social Welfare Optimization for Online Car-hailing in Peak Period by Considering Passenger Independent Bargaining WANG Jian, WANG Hui , HU Xiao-wei , LI Yuan-yuan 2022, 22(2): 54-63.  DOI: 10.16097/j.cnki.1009-6744.2022.02.006 Abstract ( )   PDF (2193KB) ( )   Existing studies on dynamic pricing of online car- hailing mainly describe drivers' operating state from the perspective of drivers and platforms by the queuing theory, birth and death process, etc. However, the studies always pay little attention to the dynamic characteristics of market demand and ignore the independent bargaining of passengers. This paper formulates a dynamic matching model to describe the matching process between passengers and drivers in car-hailing markets by considering passenger independent bargaining. The model can capture the short-term fluctuations of car-hailing markets. The model uses demand and supply functions to describe the status of passengers and online car-hailing. Furthermore, this paper builds the platform profit optimal pricing model and the social welfare maximum pricing model. This paper puts forward the passenger bargaining influence factor and determines its function in different periods of market operation based on the existing data. We further incorporate the impact factor into the established models to revise the dynamic matching model and pricing model. This paper conducts numerical experiments to test the proposed model. The paper explores the effect of price changes on the market and analyzes theimpact of passenger bargaining on the dynamic car- hailing market. Results of numerical examples show that social welfare, platform profit, and matching amount increase first and then decrease with the increase of the multiple of the price change factor. Social welfare reaches the maximum when the multiple of the price change factor is 2.0. The meeting rate reaches its maximum when the multiple of the price change factor is 1.3. Comparative analysis shows that passenger bargaining will promote the market to balance supply and demand and increase platform profits and social welfare during peak hours of the online car-hailing market. References | Related Articles | Metrics

Select

Risk Decision of Railway "Go Global" Public-private Partnership Project Based on Variable Weight Extension Matter Element GENG Qing-qiao , JIA Yuan-hua , CHEN Jun-tuan 2022, 22(2): 64-71.  DOI: 10.16097/j.cnki.1009-6744.2022.02.007 Abstract ( )   PDF (1785KB) ( )   Aiming at the risk assessment and decision-making needs of railway "Go Global" public-private partnership (PPP) projects, this study proposes a risk decision-making approach based on the variable weight extension matter element and the cumulative prospect theory (CPT). The risk evaluation index system is constructed and the combined weight of each risk index is determined comprehensively through the classical domain, node domain matrix, and the quantitative value of the matter element level. Considering the psychological characteristics of decision makers such as loss avoidance and profit preference, the reference points are determined from the perspective of profit and loss by variable weight theory and proximity criterion. Moreover, the value function and probability weight function are developed based on Monte Carlo simulation. The corresponding changes of cumulative prospect value under different risk index are discussed through sensitivity analysis. A railway "Go Global" PPP project is taken as an example for the risk assessment decision making. The results show that the proposed approach can effectively identify the key risks in the construction and operation of railway "Go Global" PPP projects, distinguish the risk level closeness degree, and provide a reference for railway enterprises and government agencies to make risk assessment decision. References | Related Articles | Metrics

Select

Analysis and Overview of Influencing Factors on Autonomous Driving Takeover GUO Lie , XU Lin-li, QIN Zeng-ke, WANG Xu 2022, 22(2): 72-90.  DOI: 10.16097/j.cnki.1009-6744.2022.02.008 Abstract ( )   PDF (1965KB) ( )   Partial or conditional autonomous vehicles allow the driver to delegate driving tasks to the autonomous driving system. The driver must, however, continue to monitor the driving environment. In the event of an emergency or if the driving environment exceeds the operation design domain of the autonomous driving system, the driver must immediately take control of the vehicle. Human factors, traffic environment, and human-machine interaction are the main factors that influence the driving takeover process. This paper investigates the impact of driver's cognitive loadcharacteristics and other human factors on the takeover process and time budget. Disengaging from the driving task for a long time will cause the driver to experience passive fatigue or distraction, thereby reducing the driver's takeover performance. The appropriate non-driving related task can keep the driver's cognitive load and reduce the driver's passive fatigue level. With the application of connected technology, the early warning signal can be provided in two stages to improve the driver's takeover performance. The influence of the traffic environment such as traffic density and road conditions on the driver's perception, cognition, and decision- making was discussed. The transitions of control (ToC) management of connected vehicles at the transition area were investigated. In complex road traffic, the driver requires more time to recover the perception of the environment, and the driver is prone to large lateral deviation when taking over the vehicle in a curve road. In mixed traffic, to prevent concentrated ToC in the transition area, corresponding traffic management measures can be formulated to alleviate the mutual interference between vehicles. The advantages and disadvantages of visual, auditory, tactile, olfactory interaction and their combination modes as well as the method of ToC were summarized. Furthermore, the design of human-machine interaction systems underconnected environments and the ToC form was discussed. Each modality has its advantages and disadvantages. The combination of multiple types of modality can complement each other's advantages, and timely transfer the takeover information to the driver. The development of connected technology has increased the amount and types of available driving information. Connected information needs a better presentation strategy to ensure that the human-machine interface has higher usability and acceptability, and provides more accurate information for drivers. Meanwhile, the driver's state recognition technology can be also used to monitor the driver's state in real-time, and remind them to keep vigilant using the human-computer interaction system to improve takeover performance. Future researchers should pay more attention to the impact of non-driving tasks on cognitive characteristics. Combining with the driving environment and following the predictive algorithm helps to assist the driver in completing smooth ToC. With the continuous application of connected technologies, the performance of existing human-machine interaction systems needs to be improved gradually and further research on the management of TOC in the transition area is necessary. References | Related Articles | Metrics

Select

Urban Smart Public Transport Studies: A Review and Prospect XU Meng , LIU Tao , ZHONG Shao-peng , JIANG Yu 2022, 22(2): 91-108.  DOI: 10.16097/j.cnki.1009-6744.2022.02.009 Abstract ( )   PDF (1711KB) ( )   PDF(English version) (470KB) ( 227 )   Based on the recent development of urban smart public transport and its status in China, this study systematically reviews four crucial research areas, including the analysis of passenger flow characteristics, operations management, network design and optimization, and system evaluation. The limitations of existing researches and future research directions in the four areas are discussed. Correspondingly, this study examines the emerging trends of urbansmart public transport from four aspects: intelligent recognition and prediction of passenger flow characteristics, urban smart public transport operations in complex scenarios, urban smart public transport network design and optimization, and urban smart public transport service evaluation. The primary research questions associated with each aspect are proposed. Particularly, this paper identifies the following four urgent research topics, including (i) mining passengers’ complete travel chain information and providing integrated urban smart public transport travel services; (ii) integrating and optimizing public transport infrastructure layout and operational efficiency from the perspective of land use and transportation integration; (iii) building a three-dimensional bus operation and evaluation system from the perspectives of the government, enterprises, and users, based on their multi-level development needs; and (iv) building an integrated urban smart public transport travel service platform. Furthermore, in viewing the current development of urban smart public transport, this paper summarizes its current status and points out the shortcomings of existing research and critical scientific issues. It is revealed that integrated applications of emerging and disruptive technologies, such as big data, cloud computing, autonomous driving, intelligent, connected, and new energy vehicles, offered new opportunities and challenges in providing multimodal urban smart transport services and promoting sustainable urban development. However, new technologies will not spontaneously improve public transport services and accelerate sustainable urban development. Finally, this study emphasizes that future research needs to strengthen the multi-disciplinary intersection, highlight the combination of industry, university, and research, and provide strong scientific support for the highquality development of urban smart urban public transport in China. References | Related Articles | Metrics

Select

Intelligent Construction Method of Vehicle Condition Based on Hybrid Constrained Autoencoder LIN Jian-xin, LIU Bo , ZHAO Xia , ZHANG Lei 2022, 22(2): 109-116.  DOI: 10.16097/j.cnki.1009-6744.2022.02.010 Abstract ( )   PDF (2256KB) ( )   In order to contract a representative vehicle driving cycle, real-world driving data recorded on 1 Hz of a motor vehicle in Fuzhou area for 20 days are collected, 14 characteristic parameters were selected based on the measured driving data to represent the kinematic fragment information. The principal component analysis and K-means clustering were applied for clustering the divided kinematic fragment, and the candidate fragments were chosen according to the distance of the clustering center and randomly combined to construct the condition set. Eleven characteristic parameters were selected to calculate the error of the construction driving cycle. We choose the driving state with the most minor error in the set as the construction driving condition, and propose the optimization of construction condition through hybrid constrained autoencoder, which reduces the average error from 2.97% to 2.39%. The hybrid constrained autoencoder model's analysis and validation show that the optimization strategy aligns with the actual situation and can effectively avoid the error uncertainty caused by the random selection. It verifies the rationality and accuracy of the driving condition's construction process and gives model parameter recommendations value. It has significant practical effects and significance for achieving the carbon emission prediction and emission control of vehicles under the carbon peak target. References | Related Articles | Metrics

Select

A Map Aided Visual-inertial Fusion Localization Method for Autonomous Vehicles CHENG Jun, ZHANG Li-yan, CHEN Qi-hong 2022, 22(2): 117-126.  DOI: 10.16097/j.cnki.1009-6744.2022.02.011 Abstract ( )   PDF (2848KB) ( )   Simultaneous Localization and Mapping (SLAM) method based on visual sensors is widely used for localization in the autonomous driving field. The traditional method uses the onboard camera to characterize the surrounding environment of the vehicle and estimate the vehicle locations. However, the accuracy and robustness are decreased when the vehicle moves fast. To solve this problem, this paper proposes a map-aided visual-inertial fusion localization method for autonomous driving. This method expands the map saving function on the basis of the ORBSLAM2 (Oriented FAST and Rotated BRIEF SLAM2) framework. The map building and positioning are divided into two independent modules. A map of environmental visual features was built and saved at a low driving speed. In the second run, the onboard computer loads the previously built- up map to realize high- precision and robust positioning performance. Since the graph-based optimization algorithm is adopted to integrate the inertial measurement unit (IMU) information in the map building stage, the errors of the visual map can be effectively corrected. The experimental results for scenes of the KITTI dataset and real-world scenes verified the good performance of the proposed method. The results indicate that the positioning error of the proposed method is respectively 2.59, 2.61, 2.73 m in the speed of 4, 8, 16 m/s. The frame lost rate (FLR) and path lost rate (PLR) are respectively 3.76% and 1.38%, 3.89% and1.69%, 4.27% and 1.84% for the three speed categories. Compared with the original ORB-SLAM2 framework, the positioning accuracy and robustness of the proposed method has been improved. References | Related Articles | Metrics

Select

Integrated Control of Trajectory Planning and Tracking for Vehicle Collision Avoidance WANG Guo-dong , LIU Li , MENG Yu , MA Zhi-ping, ZHENG Hao-qing , GU Qing, BAI Guo-xing 2022, 22(2): 127-136.  DOI: 10.16097/j.cnki.1009-6744.2022.02.012 Abstract ( )   PDF (2773KB) ( )   To improve the collision avoidance ability of intelligent vehicles in complex traffic environments, the path planning, speed planning, and tracking control are integrated into one optimization problem, and an integrated control method of trajectory planning and tracking for vehicle collision avoidance based on model predictive control (MPC) is proposed. Firstly, the collision avoidance scene in the real traffic environment is analyzed, and the collision avoidance control problem of intelligent vehicles is transformed into a multi-constraint optimization problem. Secondly, a 7DOF vehicle dynamic model and a UniTire model with combined slip conditions are established for MPC controller design. Thirdly, to solve the problem that, in the variable speed control problem, the traditional MPC based on the time-domain prediction model cannot accurately express the spatial and posture constraints of the vehicle over the prediction horizon, an MPC controller based on a spatial-domain prediction model is designed. Finally, based on the co-simulation platform of Matlab and CarSim, different collision avoidance scenarios are designed to verify the proposed method, and the existing integrated collision avoidance control method based on constant speed assumption is compared with the proposed method. The simulation results show that the proposed method can make full use of the vehiclemaneuverability, solve the problem of collision avoidance failure of existing integrated control methods in complex environments, and ensure the stability of the collision avoidance process and smoothness of trajectory. References | Related Articles | Metrics

Select

Traffic Performance Identification Method Based on Adaptive Congestion Index CHEN Ding , ZHOU Shui-ting, CHEN Yun , SU Min-xian 2022, 22(2): 137-144.  DOI: 10.16097/j.cnki.1009-6744.2022.02.013 Abstract ( )   PDF (2391KB) ( )   The congestion index is a useful tool for objectively assessing the performance of urban traffic. The current congestion index model does not take into account the spatial-temporal characteristics of traffic free-flow speed sufficiently. The probability density segmentation method is introduced into the spatial-temporal matching of vehicle positioning and road network in this study to divide the vehicle speed into distance and time segments. A congestion index with adaptive adjustment capability is proposed for the classification and identification of urban traffic performance. Using Xiamen's holiday floating vehicle positioning data as an example, the applicability of the congestion index model in complex roads and urban traffic is investigated, and the change law of urban traffic performance is revealed. The results show that the congestion index model can effectively distinguish the idle and busy periods of the roads, and adjust the traffic speed adaptively according to the traffic speed in the corresponding period. For complex interchanges and dedicated collinear roads, the calculation results of the congestion index are increased by 6.5% and 3.2% compared with the traditional methods. Xiamen's traffic is characterized by a single evening peak. The average congestion index is between 1.0 and 3.0, and the peak time for congestion is 18:10 to 21:00. The spatial distribution is concentrated in the hub of the main road in and out of Xiamen Island and the East-West main road in the island. The result plays a positive role in quickly establishing traffic performance identification methods for different cities and the needs of special periods. References | Related Articles | Metrics

Select

Mixed-coordinated Decision-making Method for Arterial Signals Based on Reinforcement Learning MA Dong-fang , CHEN Xi , WU Xiao-dong , JIN Sheng 2022, 22(2): 145-153.  DOI: 10.16097/j.cnki.1009-6744.2022.02.014 Abstract ( )   PDF (2769KB) ( )   Traffic congestion has become a common social problem in many large and medium- sized cities. Signal control, as one of the important measures to alleviate congestion, has attracted great attention. Signal optimization methods can be divided into two types: model-driven and data-driven. With the development of traffic big data, datadriven methods based on reinforcement learning have become an emerging development direction. However, the existing data-driven researches mainly focus on algorithm design while lacking the discussion of agent design. Meanwhile, the distribution strategy is mostly used in multi-intersection coordinated problems, which ignores the communication between agents and cannot guarantee the overall optimization. Therefore, this paper proposes a multiagent cooperative decision-making optimization method for arterial signals. First, given the diversity, heterogeneity, and data imbalance of the traffic state, a single-agent model with a memory palace is designed, in which the state space and reward function are optimized. Secondly, the advantages of distributed and centralized learning are integrated to design an interaction method. Based on the distributed control, a central agent is set up to evaluate behaviors of local agents and provide additional rewards to adjust the model of local agents to realize coordinated control. Finally, a simulation platform is built to conduct the test and algorithm comparison. The results show that compared with independent control and distributed coordination, the proposed method reduces the stop times on the arterial road by 14.8% and 13.6%, respectively, and has a better control effect without affecting the branch road traffic flow. References | Related Articles | Metrics

Select

Trajectory Optimization of Connected Vehicles at Isolated Intersection in Mixed Traffic Environment LIU Chun-yu , LIU Yong-hong , LUO Xia , ZHU Ying 2022, 22(2): 154-162.  DOI: 10.16097/j.cnki.1009-6744.2022.02.015 Abstract ( )   PDF (2180KB) ( )   Signalized intersection has an important impact on urban road capacity and vehicle fuel consumption. This paper presents a trajectory optimization method for connected vehicles in mixed traffic flow. , The vehicle travel time estimation model was developed considering the traffic signal timing plan. The trajectory planning and connected vehicle control were carried out with the goal of minimizing fuel consumption and maximizing traffic operation efficiency. The Gaussian Pseudo-Spectral method was used to discretize the vehicle trajectory rolling optimization model. The Simulation of Urban Mobility (SUMO) simulation platform was used to verify the model results. The simulation results show that the connected vehicle can optimize control variables to affect the state of the human-driven vehicle, and reduce the queuing of traffic flow and the stop-and-go phenomenon. The proposed vehicle trajectory optimization method is important to reduce the overall fuel consumption of the fleet, improve the average speed of the fleet and shorten the average travel time. References | Related Articles | Metrics

Select

Vehicle Routing Problem of Intercity Transportation Platform for Less-than-truck-load Cargo WANG Ning, ZHANG Jia-rui, ZHAO Jiao 2022, 22(2): 163-177.  DOI: 10.16097/j.cnki.1009-6744.2022.02.016 Abstract ( )   PDF (3408KB) ( )   Under the subsidy schemes of intercity transportation platforms for less-than-truck-load cargo, we established a vehicle routing optimization model considering goods-vehicles matching and three-dimensional loading constraints. The model minimized the sum of platform subsidy cost, vehicle operation cost, and fuel cost. To solve this model, we designed a hybrid quantum particle swarm optimization algorithm to determine the optimal cargo matching, vehicle path, cargo loading and unloading, and platform subsidy. The experimental results show that the gap between the solution obtained by the hybrid quantum particle swarm optimization algorithm and the optimal solution obtained by CPLEX software is 3.31% on average in small-scale cases. By introducing the fitness function value as the weight in solving the optimal middle position, the solution in the large-scale examples is 0.91% higher than the traditional quantum particle swarm optimization algorithm. By analyzing the characteristics of the optimal solution, the improved hybrid quantum particle swarm optimization algorithm is combined with a heuristic algorithm, and the solution quality is improved by 4.05% . Through the comparative experiments of subsidy modes, it is found that in a reasonable planning cycle, the increase of cargo owner time subsidy and no-load subsidy can effectively improve the platform profit and vehicle utilization while maintaining the total cost basically unchanged. References | Related Articles | Metrics

Select

Optimal Wage Level for Ride-sourcing Drivers Under Uncertainty of Travel Demand WANG Jing-peng , ZHU Rui , WANG Peng-fei 2022, 22(2): 178-185.  DOI: 10.16097/j.cnki.1009-6744.2022.02.017 Abstract ( )   PDF (1627KB) ( )   In the ride-sourcing system, the demand uncertainty of passengers will affect the setting of the wage for ridesourcing drivers, which in turn changes the platform's profit. This paper studies the optimal wage level for ridesourcing drivers under demand uncertainty. According to the mathematical description of the problem, we assume the travel demand follows a random distribution, and the number of drivers follows the labor supply theory. To describe the matching relationship between passengers and drivers, the perfect matching function widely used in the ride-sourcing market is adopted. In this way, a model framework for solving the platform's profit-maximization problem is established. We find that there is the optimal wage level that maximizes the platform's profit, and the optimal wage level, the number of drivers is positively related with the maximum travel demand, while the optimal wage level is negatively correlated with the platform's influence. Finally, our model gives an optimal solution that maximizes the expected platform's profit, compared with the average model, the optimal solution has a certain robustness and effectively protects the platform's profit. The results revealed in this paper show that the optimal wage model can be effectively applied to the ride-sourcing system under demand uncertainty References | Related Articles | Metrics

Select

Travel Mode Choice Analysis with Shared Mobility in Context of COVID-19 ZHANG Xiao-yu , SHAO Chun-fu , WANG Bo-bin , HUANG Shi-chen 2022, 22(2): 186-196.  DOI: 10.16097/j.cnki.1009-6744.2022.02.018 Abstract ( )   PDF (2095KB) ( )   To analyze the impact of COVID-19 on the travel mode choice behavior with diverse shared mobility services, this study designed the stated preference (SP) questionnaire for the multi-modal transportation system which include conventional travel modes, ride hailing, ride sharing, car sharing, and bike sharing. The mixed Logit models with panel data were proposed to investigate the travel mode choices before and during COVID-19. The influence differences of explanatory variables are compared, and the joint effects of perceived pandemic severity and mode choice inertia are examined. Based on the elasticity analysis, the mode choice preferences are predicted corresponding to different management policies under COVID-19 pandemic. The results indicate that the perception to pandemic severity has significant impacts on the ridership of ride sharing and car sharing, and the mode choice inertia obviously affects the usage of ride hailing, car sharing, and bike sharing. When the perceived pandemic severity reduces to 30%~ 50%, the strategy of increasing parking charge to 1.6~3.0 times would reduce the usage of private car to pre-pandemic condition, and the car sharing with lower close contact risk could become a main substitute. When the perceived pandemic severity is higher than 60%, the strategy of increasing the travel safety of ride sharing to 1.4~3.6 times would improve the ridership. References | Related Articles | Metrics

Select

Optimization Methods of Combined Passenger and Freight Transportation Based on Flexible Train Composition Mode QI Jian-guo , ZHOU Hou-sheng, YANG Li-xing, ZHOU Ya-ru, ZHANG Chun-tian 2022, 22(2): 197-205.  DOI: 10.16097/j.cnki.1009-6744.2022.02.019 Abstract ( )   PDF (2433KB) ( )   To make full use of the resource of rail transit, this paper investigates the optimization methods for combined passenger and freight transportation based on the flexible train composition mode. The proposed method takes the selection of train types and the departure headway of consecutive trains as the main decision variables. A mixed-integer linear programming model is developed to minimize the waiting time of passengers and operation cost for metro companies. The competitive relationship between the freight and passenger demands are taken into consideration in the model. The numerical experiments based on Beijing Metro Batong Line are conducted to verify the effectiveness of the proposed methods, which are all computed by CPLEX solver code by Visual Basic (VB) programming language. The computational results show that compared to the operating mode with single type of train for passenger transportation, the proposed method reduced the operation cost by 41.86% with only 1.1 minutes increase of average passenger waiting time. The method can bring a better benefit and achieve a balance between the service quality and operation costs. References | Related Articles | Metrics

Select

Joint Optimization of Urban Rail Transit and Local Bus Transit: Continuous Approximation Approach LI Xin , DAI Zhang , LI Huai-yue, HU Jia 2022, 22(2): 206-213.  DOI: 10.16097/j.cnki.1009-6744.2022.02.020 Abstract ( )   PDF (1963KB) ( )   This paper presents a bi-level mixed-integer program model to jointly optimize the urban rail and bus transit on a grid network. The model takes a variety of route types of passengers into consideration and solves the headway of the urban rail and bus and the line spacing and stop spacing of the bus system. The upper model is a continuous model aiming at minimizing the total system cost of the patrons and the transit agency. It is used to find the tradeoff relationship between passengers and operators, and deduce each cost in detail. The sequential quadratic programming algorithm is used and the convex algorithm is used to envelop the nonconvex problem. The lower model is a route assignment problem according to probability allocation, and the method of successive averages (MSA) algorithm is embedded to weighted allocate the flow of each route. Considering the non-convexity of the analytical form of the problem, the heuristic method is used to solve it. Finally, this paper takes an actual network in Jianye District of Nanjing as an example to verify the effectiveness of the model. Taking the peak period as an example, the average travel time of passengers in this study is reduced from 41.6 minutes to 33.0 minutes, which reduces by 20.6% through the model optimization. The proposed optimization model in this paper can provide a reference and basis for the planning of public transport networks in grid network cities. References | Related Articles | Metrics

Select

Optimization Model and Algorithm of Balanced Loading Layout of Railway Container Mixed Goods ZHANG Ying-gui , YAO Ying-hua, GAO Quan, LEI Ding-you 2022, 22(2): 214-222.  DOI: 10.16097/j.cnki.1009-6744.2022.02.021 Abstract ( )   PDF (1590KB) ( )   Railway container loading layout can be taken as the problem of maximizing the utilization of container space and load capacity under the restriction of balanced loading for mixed goods. Aiming at this type of problem, this paper proposes the constraint quantitative method for load balance of the container gravity center and the allowable bending moment of the concentrated-weight after loading. The loading layout optimization model is then developed to maximize the container utilization rate for mixed goods. In addition, the study proposes method to classify goods into block units based on mixed goods classifications and goods structure index, which is different from directly group goods into goods blocks. The method that can select and place a set of goods block units, and the rule that can update available space are also designed. The optimization algorithm is proposed to solve the model. The results from numerical example show that the proposed method can keep the average container space utilization and load capacity not less than 87%. There are 92.8% probability that the container gravity center is balanced after loading, and there are 97.87% probability that the allowable moment of concentrated-weight meet the requirement. The method can effectively match the size of the goods blocks with available spaces and objectively reflect the utilization of container space and meet the requirements of balanced loading, which provides support for railway container goods loading and layout optimization. References | Related Articles | Metrics

Select

One-way Pedestrian-bicycle Mixed Flow Model and Self-organization Phenomenon HAO Yan-xi, LIU Rong-yang, HU Hua , FANG Yong, LIU Zhi-gang 2022, 22(2): 223-229.  DOI: 10.16097/j.cnki.1009-6744.2022.02.022 Abstract ( )   PDF (1923KB) ( )   Pedestrian and bicycle mixed traffic is a common non-motorized vehicle flow, but the current research on its mixed dynamics is still limited. Based on the social force model, this paper proposes a pedestrian-bicycle mixed floworiented mixed flow social force model (MFSF) to analyze the movement mechanism of the pedestrian-bicycle mixed flow. The observation data of the mixed flow of pedestrians and bicycles on the road is calibrated to the model, and a simulation platform is constructed. The simulation platform is used to reproduce the self-organization phenomenon of pedestrians and bicyclists on one-way roads. From the comparison and verification of the simulation results and the measured data, the model can accurately describe the motion characteristics of the mixed flow. The simulation results show that the MFSF model can reproduce the self- organization phenomenon of the mixed flow of pedestrians and bicycles. When the ratio of pedestrians to bicyclists is the same, the lower the density of pedestrians and bicyclists, the smaller the change in lane width for bicyclists. When the ratio of pedestrians to bicycles changes, the greater the ratio of bicyclists, the wider the road occupied by bicyclists, and the more unstable the lane width of bicyclists. If the pedestrian input is significantly reduced, the lane width for bicyclists is significantly increased. References | Related Articles | Metrics

Select

Bivariate Traffic Conflict Extreme Value Model of Truck Collision Prediction on Two-lane Mountain Highway JI Xiao-feng , GENG Zhao-shi 2022, 22(2): 230-238.  DOI: 10.16097/j.cnki.1009-6744.2022.02.023 Abstract ( )   PDF (2774KB) ( )   To predict collisions between trucks and conflicting vehicles on two-lane mountain highway, this paper proposes a bivariate traffic conflict extreme value (BTCEV) model. The high-precision trajectory data of trucks and interactive vehicles were extracted based on drone video. The traffic conflict indicators suitable for different trajectories were selected and the extreme value theory was also considered in the model. The Post Encroachment Time (PET) and Time to Collision (TTC) were incorporated into a unified framework to realize the collision prediction of trucks and conflicting vehicles on two-lane mountain highway. The prediction performance of the BTCEV model was verified by the example of the two-lane mountain highway that has a high truck incidence rate in Yunnan Province. The result shows that PET is 0.382 s and TTC is 4.471 s are the thresholds for serious conflicts of trucks on two-lane mountain highway. The annual accident probability of truck on two-lane mountain highway predicted by the BTCEV model is 5.84%, and the prediction accuracy is as high as 98.92%. The prediction accuracy is respectively 167.33% and 10.80% higher than that of the PET model and the TTC model. Compared with the univariate model, the confidence interval estimated by the bivariate model is narrower and the prediction accuracy is higher. The proposed method extends the truck collision prediction method on two-lane mountain highway from single variable to double variables, and has broad application prospects in the truck safety analysis on two-lane mountain highway. References | Related Articles | Metrics

Select

Bi-level Optimization Model in Transportation Evacuation Network Based on Rational Inattention ZHAO Chuan-lin , HE Shao-song, SUN Shu-min, WANG Yu-han 2022, 22(2): 239-246.  DOI: 10.16097/j.cnki.1009-6744.2022.02.024 Abstract ( )   PDF (1999KB) ( )   Many studies on traffic network evacuation have been conducted, but none have examined traffic network evacuation using rational inattention theory. With the randomness of evacuation network traffic state and the limited capacity of traveler information processing, this paper internalizes the cost of traffic state information and develops a bilevel evacuation network optimization model based on rational inattention theory. The upper model minimizes the total evacuation time of the system, with the decision variable being whether a link is unidirectional. The lower model is the user equilibrium model for rational inattention travelers. The DPSO-MSA hybrid heuristic algorithm is designed. The upper model is solved using the particle swarm optimization algorithm, and the resulting one-way strategy is passed down to the lower model, which is solved using the MSA algorithm. An example is provided to demonstrate the model's effectiveness. It is found that the optimal one-way strategy outperforms both the non-one-way strategy and the whole one-way strategy. The designed algorithm can quickly identify the key links of the evacuation network. For the whole evacuation system, more information obtained by the traveler may not lead to better performance. The results can provide a reference for the formulation of evacuation strategy References | Related Articles | Metrics

Select

Interval Optimization Model of Intersection Signal Timing Under Uncertain Traffic Demand CHEN Xiao-hong , LAN Qiu-yu 2022, 22(2): 247-256.  DOI: 10.16097/j.cnki.1009-6744.2022.02.025 Abstract ( )   PDF (1699KB) ( )   To deal with the demand uncertainty of the mixed traffic flow, a nonlinear interval optimization model of traffic signal control is established by taking the traffic flow interval and saturated flow rate interval as the input parameters based on the interval optimization theory in this paper. Firstly, the traffic volume interval is constructed by using the 5-minute acquisition segment in the peak hours, and the saturated flow rate interval is estimated by modifying the traditional Highway Capacity Manual 2010(HCM2010). Secondly, taking the intersection service level as the performance, this model is transformed into a deterministic mathematical model by using the interval order relation and possibility degree model and solved by using a genetic algorithm. Finally, two-phase and three-phase intersections in which the phases have significant difference grades in rush hours in Beijing are adopted in numerical cases by VISSIM simulation software. The results show that the interval optimization method of signal timing is feasible and effective. The interval optimization model with the saturated flow rate interval is more suitable for two-phase intersection when the saturated flow rate of the key signal phase has large fluctuation. Furthermore, the vehicle average delay decreases by 35.9%, and the intersection capacity increases by 14.9%, compared with the Webster model. References | Related Articles | Metrics

Select

Capacity and Delay of Right-turn Vehicles at Signalized Intersections Under Influence of Pedestrian Two-stage Crossing LI Bing, WANG Zheng-hui, MA Ming-wei, YANG Hong-yu, FENG Yue 2022, 22(2): 257-267.  DOI: 10.16097/j.cnki.1009-6744.2022.02.026 Abstract ( )   PDF (2496KB) ( )   This paper analyzes the impact of pedestrian two-stage crossing on right-turn vehicles at signalized intersections. Based on the gap acceptance theory and random distribution, this paper develops the capacity and delay models for right-turn vehicles considering the conflicts with one-way and two-way pedestrian flows. Taking a typical four-phase signal release mode as an example, the study also evaluates the right-turn movement capacity and delay when pedestrian use two-stage crossing or one time crossing. The occupancy of pedestrian crossing time and the arrival distribution of pedestrians are used as the comparison indexes. The results show that compared to the pedestrian one time crossing, the average capacity of right-turn movement is reduced by 16.68% and the average delay is increased by 21% after installation of pedestrian two- stage crossing. Therefore, when the traffic demand for right-turn vehicles is heavy, it is necessary to consider the traffic operation of both pedestrians and right-turn vehicles to determine whether to use pedestrian two-stage crossing. This would avoid right-turn vehicles exceeding the tolerance limit and increasing the probability of conflicting with pedestrians. References | Related Articles | Metrics

Select

Optimization Model and Algorithm for Supply of War Storage Materials on Distant Islands and Reefs Based on Sea Air Cooperation LIU Zhong-bo , WANG Qi-xian, ZHENG Hong-xing 2022, 22(2): 268-279.  DOI: 10.16097/j.cnki.1009-6744.2022.02.027 Abstract ( )   PDF (2128KB) ( )   Aiming at the wartime supply of war storage materials in offshore islands and reefs, this paper develops a two-stage optimization model (2E-MLRP). The shortest total system time and the lowest material support cost are taken as the goals and the proposed model considers the advantages of sea air cooperative transportation. The structural characteristics of the solution are applied to improve the congestion comparison operator and elite retention strategy to form a targeted improved NSGA-II algorithm. The results of the example analysis show that the optimal scheme obtained by the proposed method meets the requirements of "cost performance". Compared with the whole shipping model, the total system time of sea air coordination is reduced by 53.15%, while the cost increase is only 22.27%. The delivery time of the first batch of materials is also reduced by 2.95 days. The comparison results show that the Spacing Metric index value and Maximum Spread index value obtained by the improved algorithm are better than the traditional NSGA-II. The Pareto solution set obtained by the improved algorithm shows better distribution. Compared with other schemes, the transportation scheme has significantly reduced cost and narrow increase in transportation time, which can ensure the efficient utilization of transportation equipment, and the number of demand islands in a single route is reasonable. The results meet the requirements of war storage material transportation and provide a reference for the formulation of war storage material supply scheme for far island reefs in wartime. References | Related Articles | Metrics

Select

Delay Propagation Characteristics of Multilevel Handlings Hypernetwork at Container Terminals XU Bo-wei, WANG Ling-ling , LI Jun-jun 2022, 22(2): 280-289.  DOI: 10.16097/j.cnki.1009-6744.2022.02.028 Abstract ( )   PDF (3044KB) ( )   The multilevel handlings delay occurs frequently in uncertain environments at automated container terminals (referred to as automated terminals). The delay will propagate among the multiple handlings over time, resulting a "chain" effect that will have a significant impact on the overall efficiency of the port. To improve the ability of enterprises to cope with handlings delays, a multilevel handlings hypernetwork is constructed based on the theory of hypernetwork and propagation dynamics. Combined with the topology characteristics of the small-world network theory, the propagation speed, propagation breadth, and propagation capacity of delay in multilevel handling hypernetworks are discussed and numerically simulated from three perspectives: characteristic path length, clustering coefficient, and SIS simulation. The results show that the shorter characteristic path length and larger clustering coefficient will make the delay propagated more quickly and widely, the fault nodes with smaller clustering coefficients will be restrained by clustering characteristics. The network's propagation capacity is directly proportional to the initial fault nodes' number and inversely proportional to the repair probability. This study is beneficial for enterprise managers to actively respond to the delay caused by uncertain events and improve the operation efficiency of ports. References | Related Articles | Metrics

Select

A Capacity Estimation Approach for Waterway Traffic Under LNG Carriers Navigation Mode LIAO Shi-guan , WENG Jin-xian, HU Shen-ping 2022, 22(2): 290-297.  DOI: 10.16097/j.cnki.1009-6744.2022.02.029 Abstract ( )   PDF (2232KB) ( )   An empirical method in the ship domain is proposed to quantify the scale of the moving safety zone for LNG carriers and extended to estimate the traffic capacity under LNG carriers navigation mode. Based on the Automatic Identification System (AIS) data, a case study of the waterway near Wuhaogou LNG terminal was applied to testify the applicability of the proposed model in estimating the traffic capacity. Results show that the scale error of the moving safety zone is between 15 to 28 meters, indicating a higher accuracy. When LNG carriers sail into the waterway, the traffic capacity in the same and reverse directions is 110~250 ship · h-1 and 120~230 ship · h-1 , respectively. When LNG carriers sail out of the waterway, the traffic capacity in the same and reverse directions is 140~240 ship · h- 1 and 90~ 230 ships· h-1 , respectively. To maximize the navigational efficiency, LNG carriers are suggested to enter the waterway one hour before the high tide during the daytime at a speed of at least 11 knots but without exceeding the permissible maximum speed. To guarantee navigational safety, LNG carriers are advised to depart the waterway one hour after the high tide during the daytime and maintain the speed below 9.5 knots. References | Related Articles | Metrics

Select

Connection Mechanism of Small and Medium-sized Airports Based on Link Prediction ZHANG Pei-wen , DU Fu-min , WANG Yu 2022, 22(2): 298-304.  DOI: 10.16097/j.cnki.1009-6744.2022.02.030 Abstract ( )   PDF (1552KB) ( )   This paper uses the link prediction method to analyze the route network of small and medium-sized airports and establishes four prediction proximity indexes considering network exogenous attributes and endogenous factors. The prediction effects of each index are compared and three coupling proximity prediction algorithms are designed. The algorithm with the highest prediction accuracy is selected to predict the new routes of small and medium- sized airports for the future condition. The prediction results are then verified. The results show that the prediction accuracy of the four endogenous factor indexes is higher than that of the exogenous attribute indexes except takeoff and landing sorties, among which the prediction accuracy of the local path (LP) index is the highest. Mining the internal structure of the network is more effective to predict the network connection, and the prediction effect of the coupling algorithm is better than that of a single index. Comparing the prediction results based on the coupling algorithm with the actual planed new routes in the future, the coupling algorithm actually predicted more than one third of the new routes in the future for small and medium-sized airport routes. The predicted routes are mainly concentrated in the eastern region, and most small and medium-sized airports will still choose to connect with central cities. The results are consistent with the small and medium-sized airport connections in real situation. References | Related Articles | Metrics

Select

Air Traffic Complexity Model Based on Aircraft Self-separation Operation WANG Hong-yong , GUO Yu-peng 2022, 22(2): 305-312.  DOI: 10.16097/j.cnki.1009-6744.2022.02.031 Abstract ( )   PDF (2388KB) ( )   This paper focuses on the air traffic situation assessment under aircraft self-separation operation mode. The air traffic complexity assessment method is proposed based on distributed air traffic management (ATM) system, and the application is simulated in aircraft conflict detection and autonomous flight path planning. First, the complexity of air traffic was defined based on the free route airspace and the autonomous operation mode of aircraft. Based on quantizing the space-time position of aircraft through three-dimensional airspace grid model, the complexity calculation model is developed to reflect the real-time complexity influence of aircraft position, course and speed on each grid of airspace. Then, based on the actual sector structure (ZSSSAR01) and flight level, the free route and the fixed route were simulated to compare the difference of thel space-time distribution of air traffic complexity. The study also investigates the correlation between air traffic complexity and aircraft conflict index (i.e. conflict rate and conflict ratio) and determines the airspace complexity threshold based on the free route airspace operation mode. At last, this paper preliminarily explores the aircraft autonomous flight path adjustment method based on airspace complexity threshold and evaluates the operational effect. The result show that the maximum complexity of airspace can be significantly reduced by free route airspace operation mode (average decrease is 119% ). There is a strong correlation between air traffic complexity calculated by the model and aircraft conflict index in airspace (correlation coefficient greater than 0.90). The path adjustment strategy based on complexity and complexity threshold is applicable to some extent. References | Related Articles | Metrics

Select

Calculation Method of Total Carbon Emission and Efficiency of Logistics Enterprises JIANG Xiao-hong, CHEN Sha, ZHANG Yi 2022, 22(2): 313-321.  DOI: 10.16097/j.cnki.1009-6744.2022.02.032 Abstract ( )   PDF (1388KB) ( )   PDF(English version) (409KB) ( 163 )   The carbon emission coefficient method is used to measure the carbon emissions of logistics enterprises' transportation and storage. The input and output indicators of carbon emission efficiency are selected, and the Superefficiency Slack-based Measurement Model is adopted to obtain the three index values of technical efficiency, pure technical efficiency, and scale efficiency to evaluate the carbon emission efficiency of logistics enterprises. The empirical analysis is performed using SF Express's operating data from 2013 to 2020. The analysis results show that the technical efficiency from 2013 to 2016 and 2020 are all greater than 1, indicating that resource allocation has reached the optimal state. The technical efficiency values from 2017 to 2019 are all less than 1, mainly because of the low scale efficiency value. SF Express's transportation business volume has soared since 2017, and the scale efficiency has been gradually adjusted to find a relatively reasonable balance point by 2020, and the technical efficiency has been improved in 2020. The pure technical efficiency values from 2013 to 2020 are greater than 1. In recent years, SF has continued to increase its capital investment in technology to seek technological improvements. These results show that the proposed method is feasible, and the super-SBM model can effectively evaluate the carbon emission efficiency of logistics enterprises. Finally, we put forward the improvement measures for logistics enterprises to improve carbon emission efficiency from the aspects of innovative production technology, and energy saving and emission reduction technology. References | Related Articles | Metrics

Select

Relational Analysis Between Bus Drivers' Violation Type and Traffic Accident ZHU Tong , QIN Dan, DONG Ao-ran, DU Yu-meng, WEI Wen 2022, 22(2): 322-329.  DOI: 10.16097/j.cnki.1009-6744.2022.02.033 Abstract ( )   PDF (2108KB) ( )   Bus drivers' violating behavior is one of the important causes of traffic crashes. Previous studies have considered the relationship between driving violations and crash frequency, but failed to distinguish the types of violations and consider the heterogeneity. To reveal the differentiated impact of different violations and then provide a theoretical basis for behavior control measures, using the data of 4532 bus-involved violations and crashes in a city, the bus drivers' crash frequency in the accidents was focused on. Besides, 17 violating behaviors and demography factors were taken as explanatory variables. Based on the data characteristics and heterogeneity analysis, the random parameter zero-inflated Poisson model with heterogeneity in means and variances is established. The results indicated that different types of violating behaviors showed a significant difference in crash frequency. When drivers had several specific violating behaviors, the crash frequency was significantly increased. Specifically, not driving on bus lanes and other behaviors were associated with high crash frequency. The parameters of variables such as not wearing a seat belt, not yielding to pedestrians, and updating the expired license timely have random characteristics, indicating that its effect has the attribute of random fluctuation. In addition, the random parameter of not yielding to pedestrians shows that heterogeneity not observed in the data may appear in the mean and variance of the parameters. When the times of drivers pressing the line and not traveling on the bus lanes increased, the impact on the crash frequency would be more obvious. References | Related Articles | Metrics

Select

Driver Characteristics Clustering Under Impact of Varying Traffic Operation Conditions ZHANG Jian-bo , SUN Jian-ping , XU Chun-ling , GUO Jing-xia, WEN Hui-min , SONG Guo-hua 2022, 22(2): 330-336.  DOI: 10.16097/j.cnki.1009-6744.2022.02.034 Abstract ( )   PDF (2180KB) ( )   To improve the stability and reliability of driver characteristics clustering, this paper proposes an improved driver clustering method considering the impact of varying traffic conditions using motor vehicle trajectory data. First, the analysis of vehicle trajectories indicates that different road types and average speeds significantly affect the aggregate characteristics of driving behavior. Considering the road types and average speed, a slicing and classification method of vehicle trajectory is designed to steadily extract the driving behavior characteristics under typical traffic conditions. Then, the driver characteristics are clustered using the Gaussian Mixture Model (GMM). Data analysis demonstrates that road types and average speed can significantly affect driving behavior characteristics. The clustering results show that the improved clustering method shows better performance on intra-class aggregation and inter-class separation, and this method can improve the applicability and reliability of driver clustering. References | Related Articles | Metrics

Select

Analysis on Contributing Factors Influencing Severity of E-Bicycle and Motor Vehicle Crashes MA Jing-feng, REN Gang , LI Hao-jie, CAO Qi, DU Jian-wei 2022, 22(2): 337-348.  DOI: 10.16097/j.cnki.1009-6744.2022.02.035 Abstract ( )   PDF (2202KB) ( )   It is critical to quantify the differential influence of different factors on the severity of e-bicycle and motor vehicle crashes in order to mitigate the losses caused by these crashes. Based on the data of 10304 e-bicycle and motor vehicle crashes in Shangyu in 2018, the severity distribution and spatiotemporal characteristics of the crashes were analyzed. As the dependent variable, the crash injury severity levels were coded into three classifications: uninjured, minor injury, and serious injury crashes. A total of 17 contributing factors were selected from six perspectives, i.e., time, space, road, environment, cyclist, and vehicle type. The multinomial Logit model, the ordered Logit model, the generalized ordered Logit model, and the partial proportional odds model were applied to explore the potential influencing factors by comparing the goodness of fit. The optimal model, i.e., the partial proportional odds model, as well as the marginal effects was carried out to quantitatively analyze the significance of the contributing factors in the crashes. The results evidenced that, the day type, the speed limit of lanes, intersection angle of traffic flow, and temperature have no significant influence on the crash severity, while the other 13 factors have significant influence. The factors, including crash time, lighting condition, rider gender, rider age and vehicle type, do not comply with the parallel-lines assumption in the crashes. The top four factors with the greatest impacts were e-bicycle type, vehicle type, rider age, and gender. The maximum absolute marginal values all exceeded 61% . Crash area, road type, andlighting condition had moderate impacts (20% ~30% ), crash time and wind level had less impact (10% ~20% ), and season, crash location, non-motorized traffic interference, and the weather had minor effects (≤8% ). Based on the differences in the influencing factors, effective improvement suggestions had been put forward for the traffic management department. References | Related Articles | Metrics