Abstract：Research purposes: The railway line selection has a wide range of areas, multiple influencing factors, and strong multidisciplinary comprehensiveness. The research and development of intelligent line selection is difficult, and the efficiency and effectiveness of its application have not yet reached widespread promotion. The intelligent railway line selection system is developed, key and difficult points are identified to break through them step by step, new perspectives are proposed on path planning, local optimization, human-machine interaction, etc., in order to supplement and improve existing theoretical methods, and improve the efficiency and quality of line selection design. Research conclusions: (1) Conducting system research and development through three steps: channel search, scheme optimization, and human-machine collaboration, and fully reusing automatic vertical section design as the basic function, can significantly reduce research and development difficulty and improve research and development efficiency. (2) A method for automatically exploring terrain in large scenes is proposed to obtain high cost areas for bridges and tunnels, and to construct a triangulation cost graph based on the boundary feature lines of the areas, which can improve search adaptability and accuracy. (3) Based on Dijkstra's shortest path search algorithm, automatic path planning can greatly improve search efficiency and effectiveness. (4) Adaptive optimization of horizontal and vertical lines, real-time linkage of horizontal and vertical lines, and system fusion based GIS line selection can flexibly meet practical needs and enhance system applicability. (5) The research results can provide reference for the development of intelligent line selection systems for railways and highways.
孔国梁, 胡叙洪, 郑志霖, 李顶峰, 龙许友, 左峰, 袁伟. 基于代价图路径规划的铁路智能选线系统研发[J]. 铁道工程学报, 2024, 41(10): 18-23.
KONG Guoliang, HU Xuhong, ZHENG Zhilin, LI Dingfeng, LONG Xuyou, ZUO Feng, YUAN Wei. Research and Development of Railway Intelligent Line Selection System Based on Cost Graph Path Planning. Journal of Railway Engineering Society, 2024, 41(10): 18-23.
薛新功,李伟,蒲浩.铁路线路智能优化方法研究综述[J].铁道学报,2018(3):6-15.Xue Xingong, Li Wei, Pu Hao.Review on Intelligent Optimization Methods for Railway Alignment[J]. Journal of the China Railway Society, 2018 (3): 6-15.
[2]
朱颖,韩春华,许佑顶.铁路智能选线理论与方法[M].北京:中国铁道出版社,2018.Zhu Ying, Han Chunhua, Xu Youding.Theory and Method of Intelligent Railway Line Selection [M]. Beijing: China Railway Publishing House, 2018.
[3]
李伟,蒲浩,郑晓强.基于双向广义距离变换的复杂环境铁路线路优化[J].铁道学报,2017(2):90-98.Li Wei, Pu Hao, Zheng Xiaoqiang.Methodology for Railway Alignment Optimization in Complex Environment Based on Bidirectional Generalized Distance Transform[J]. Journal of the China Railway Society, 2017(2): 90-98.
[4]
李伟,蒲浩,赵海峰,等.基于分步编码改进遗传算法的铁路智能选线[J].西南交通大学学报,2013(10):831-838.Li Wei, Pu Hao, Zhao Haifeng, etc. Intelligent Railway Aignment Optimization Based on Stepwise Encoding Genetic Algorithm[J]. Journal of Southwest Jiaotong University, 2013(10): 831-838.
[5]
袁泉,曾文驱,李子涵,等.基于改进型D3QN深度强化学习的铁路智能选线方法[J].铁道科学与工程学报,2022(2):344-350.Yuan Quan, Zeng Wenqu, Li Zihan, etc. Intelligent Railway Location Design Approach Based on Enhanced D3QN Deep Reinforcement Learning[J]. Journal of Railway Science and Engineering, 2022(2): 344-350.
[6]
马青松,朱颖,高天赐,等.基于逆向强化学习的铁路线路方案优选研究[J].铁道建筑,2023(7):1-7.Ma Qingsong, Zhu Ying, Gao Tianci, etc. Research on Railway Route Scheme Optimization Selection Based on Inverse Reinforcement Learning[J]. Railway Engineering, 2023(7): 1-7.
[7]
孔国梁,李顶峰.铁路纵断面自动设计与优化方法研究[J].铁道工程学报,2012(12):40-44.Kong Guoliang, Li Dingfeng.Study on Methods for Autodesign and Optimization of Railway Profile[J]. Journal of Railway Engineering Society, 2012(12): 40-44.
[8]
韩元利. 基于地形演绎的纵断面自动拉坡设计方法[J].铁道工程学报,2014(10):37-43.Han Yuanli.Method on Automatic Vertical Design Based on Approach Integrating Terrain and Topography[J]. Journal of Railway Engineering Society, 2014(10): 37-43.
2024, Vol. 41 
版权所有 © 2014 《铁道工程学报》编辑部