基于代价图路径规划的铁路智能选线系统研发  

Research and Development of Railway Intelligent Line Selection System Based on Cost Graph Path Planning

作　　者：孔国梁 胡叙洪 郑志霖 李顶峰 龙许友 左峰 袁伟 KONG Guoliang;HU Xuhong;ZHENG Zhilin;LI Dingfeng;LONG Xuyou;ZUO Feng;YUAN Wei(China Railway Design Corporation,Tianjin 300308,China)

机构地区：[1]中国铁路设计集团有限公司,天津300308

出　　处：《铁道工程学报》2024年第10期18-23,共6页Journal of Railway Engineering Society

基　　金：中国国家铁路集团有限公司科技研究开发计划重大课题(K2022G001);中国铁路设计集团有限公司科技开发重点课题(2022A02238009)。

摘　　要：研究目的:铁路选线区域广、影响因素多、多专业综合性强,智能选线研发难度大,应用的效率效果尚未达到广泛推广。研发铁路智能选线系统,找准重难点分步突破,对路径规划、局部优化、人机交互等提出新的视角,以期对现有理论方法进行补充和完善,提高选线设计效率和质量。研究结论:(1)按照通道搜索、方案优化、人机协同三个步骤开展系统研发,并将纵断面自动设计作为基础功能充分复用,可显著降低研发难度,提高研发效率;(2)提出大场景地形自动探查获取桥隧高代价面域、基于面域边界特征线的三角网代价图构建方法,可提高搜索适应性和精确性;(3)基于Dijkstra最短路径搜索算法开展路径自动规划,可极大提高搜索效率和效果;(4)平纵横自适应优化、平纵横实时联动、系统融合式GIS选线可灵活应对实际需求,增强系统适用性;(5)本研究成果可为铁路、公路的智能选线系统研发提供参考和借鉴。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.

关 键 词：铁路智能选线 代价图 通道搜索 差异演化 平纵横联动 GIS选线 

分 类 号：U212.3[交通运输工程—道路与铁道工程]