Information
Edited & Published: Editorial Office of Journal of Dalian Jiaotong University
Publication Period: Bimonthly
Publisher: Dalian, Liaoning Province, China
Language: Chinese
Superintendent: Liaoning Provincial Department of Education
Sponsored by: Dalian Jiaotong University
Editor in Chief: Chen Bingzhi
Executive Editor in Chief: Guo Ruijun
China National unified serial number: CN 21-1550/U
International Standard Serial Number: ISSN 1673-9590
Periodical website: https://dltd.cbpt.cnki.net
Related Links
Page Views
A Survey of UAV Aerial Object Detection for Intelligent Transportation
JIA Shijie;ZHANG Huidi;Having three-dimensional(3D) moving ability and an air high-angle view, Unmanned Aerial Vehicles(UAVs) have appeared as an unreplaceable complement to ground sensing technologies in the area of Intelligent Transportation Systems(ITS). Even so, the aerial image data brings very obvious difficulties to high-accuracy object detection, which include the dense arrangement of tiny targets, size change, visual angle deformation, and environment domain transformation. This review focuses on UAV-based object detection technology in intelligent transportation scenes, this paper systematically reviews its development path, analyzes key technical problems and progress, and sums up application results, as follows in detail:(1) It follows the development course of UAV object detection technologies, from traditional manual feature engineering approaches to deep learning-based frameworks—including two-stage, single-stage, and Transformer-based detection structures. Typical data collections which are custom made for ITS situations are carried out systematic classification, and core technical directions are analyzed deeply, including network frame optimization, training method promotion, light-weight arrangement technology, and RGB-infrared multi-model combination methods.(2) It makes overall induction of the present applying situation and performance promotion of UAV object detection in core intelligent transportation scenes, for example traffic flow monitoring and management, initiative traffic safety prevention and control, intelligent identification and law execution of traffic breaking rules, and dynamic checking of transportation infrastructure foundations.(3)It further makes more precise the restrictions of current studies, pointing out key difficult points such as the non-consistency between task definition and assessment targets, uneven data covering among different situations, and the absence of universal capability under complicated environment conditions.
Research on the Integrated Optimization of Locomotive Routing Plan and Train Formation Plan Based on Fuzzy Programming
WANG Zhenyu;GUO Ruijun;LI Xiang;SHEN Yaoming;LU Dishen;LIN Boliang;The locomotive routing plan and train formation plan are at the same tactical level, and the locomotive routing plan has a certain influence on the train formation plan. This paper provides a mathematical formulation of the locomotive routing plan, based on its characteristics and components. Furthermore, the influence mechanism of the locomotive routing plan on the train formation plan is explored, followed by the development of an integrated optimization model for both plans. Additionally, considering the elastic reclassification capability of railway yards, the membership function in fuzzy programming is introduced to further improve the model. A case study based on a railway network is conducted at the end of the paper. The results show that the proposed model can effectively optimize both the locomotive routing plan and the train formation plan, further validating its correctness and effectiveness.
Research on the Limit Values of Longitudinal Slope and Slope Length of Expressways Based on Electric Vehicle Performance
CAI Xiaoxi;HU Lintao;PEI Yulong;Current freeway longitudinal slope design standards are primarily based on the performance characteristics of internal combustion engine vehicles and fail to adequately capture the energy consumption patterns and speed attenuation behavior of battery electric vehicles during slope driving. This study establishes a co-simulation model using CarSim and Matlab/Simulink to quantitatively analyze the energy consumption and speed reduction characteristics of electric vehicles under various combinations of slope gradient, slope length, and design speed. Based on the simulation results, a longitudinal slope design threshold method is proposed that simultaneously considers energy efficiency and speed attenuation. Taking the equilibrium speed corresponding to the slope with the highest motor efficiency as the reference, an acceptable limit of speed reduction is set to ensure driving safety while maximizing energy conversion efficiency of the powertrain. The proposed method provides technical guidance for the planning, design, and reconstruction of freeways primarily serving electric vehicles, particularly for dedicated passenger car freeways and specialized routes such as airport expressways.
Pedestrian Origin-Destination(OD) Identification Method Based on Wi-Fi Probe Requests Under Randomized MAC Address Mechanisms
ZHU Chengjuan;ZHAO Xingji;CHEN Aitong;To address the privacy protection and data invalidation issues faced by traditional Origin-Destination(OD) crowd flow monitoring methods based on real MAC addresses, this paper proposes an anonymous OD crowd flow identification method based on Wi-Fi probe request information elements. By analyzing stable features in fields such as HT Capabilities and VHT Capabilities within probe requests, we design an enhanced DBSCAN clustering algorithm to accurately distinguish randomized MAC devices. To address cross-region trajectory matching, we introduce an OD association strategy leveraging information element group filtering and distance matrix computation. Experimental results in a public facility environment demonstrate a device clustering accuracy of 97.87% and an OD matching accuracy of 81.8%, validating the method's effectiveness. Compared to existing techniques, our approach achieves high precision and low computational complexity while preserving user privacy, offering novel insights for public safety, transportation planning, and emergency management.
Layout and Size Optimization of Connecting Rivets for Composite Bearing Parts of EMU Car Body
XIE Suming;HU Zhiyuan;WANG Yanxiang;ZHANG Qian;Using topology and size optimization technology, the strength of connecting rivets of composite car body of EMU is studied. According to the bearing characteristics of composite materials and rivets, the strength analysis model of composite car body of an EMU is established. According to BS EN 12663-2014 and TB/T 3451—2016 standards, the working load of the car body is determined, and the strength of the car body is analyzed under the working load. The redundant area of rivet connection strength is determined as the side wall and the hood and underframe area; taking the rivets in these two regions as the design variables, the minimum weighted flexibility as the objective function, and the volume fraction as the constraint condition, the rivet layout optimization is carried out to obtain the new layout of the rivets. The rivet size of the new layout is taken as the design variable, the minimum mass is taken as the objective function, and the axial force and transverse shear force of the rivet are taken as the constraint conditions to optimize the rivet size. After optimization, the total number of rivets in the connection area between the side wall and the underframe is reduced by 19.91 %, and the diameter of rivets is reduced from 4.8 mm to 4.0 mm. The total number of rivets in the connection area between the head cover and the chassis is reduced by 26.60 %, and the maximum diameter is reduced by 50 %. After optimization, the stiffness and mode of the car body meet the design requirements, and the maximum failure factor of the rivet is 0.629, which meets the strength design requirements of rivet.
WeChat official
