一种复杂场景下便捷通行路径规划新方法  

A new method of convenient path planning in complex scenarios

作　　者：李益沛[1] 宋鹏波 赵嘉琦 章玉龙 LI Yipei;SONG Pengbo;ZHAO Jiaqi;ZHANG Yulong(State Grid Shanxi Electric Power Company,Taiyuan 030025,China;State Grid Shanxi Electric Power Company Information Communication Branch,Taiyuan 030025,China;Xi'an Jiaotong-liverpool University,Suzhou 215028,Jiangsu China;Anhui Jiyuan Software Co.,Ltd.,SGITG,Hefei 230088,China)

机构地区：[1]国网山西省电力公司,山西太原030025 [2]国网山西省电力公司信息通信分公司,山西太原030025 [3]西交利物浦大学,江苏苏州215028 [4]国网信通产业集团安徽继远软件有限公司,安徽合肥230088

出　　处：《粘接》2025年第5期147-150,共4页Adhesion

摘　　要：目前常规的通行路径规划算法主要通过构建深度学习模型,在此基础上对最短路径进行寻优处理。但由于缺乏对复杂场景的细致建模,导致路径规划效果不佳。对此,提出复杂场景下便捷通行路径规划优化算法。首先采用车载传感器对障碍物的关键点信息进行提取,构建出全局坐标系。然后以场地的可通行程度为标准,构建出栅格地图,实现场景建模。并将路径规划问题转换为最短行驶时间求解问题,结合路径重复情况以及行驶速度,对目标函数进行约束。最后通过引入启发式函数,对目标函数规划结果的平滑性进行优化,从而输出最优规划结果。测试结果表明,采用提出的方法对最优路径进行规划时,算法的平均路径长度较短,具备较为理想的路径规划效果。At present,the conventional traffic path planning algorithm mainly optimizes the shortest path by constructing a deep learning model.However,due to the lack of detailed modeling of complex scenes,the path planning effect is not good.In this regard,an optimization algorithm for convenient passage path planning in complex scenarios was proposed.Firstly,the vehicle-mounted sensor was used to extract the key point information of the obstacle and construct a global coordinate system.Then,based on the passability of the site,a raster map was constructed to realize scene modeling.The path planning problem was transformed into the shortest travel time solving problem,and the objective function was constrained by combining the path repetition and travel speed.Finally,the heuristic function was introduced to optimize the smoothness of the objective function planning results,so as to output the optimal planning results.The test results showed that when the proposed method was used to plan the optimal path,the average path length of the algorithm was shorter,and the path planning effect is more ideal.

关 键 词：复杂场景 路径规划 最短路径 目标函数 通行 

分 类 号：TP391.92[自动化与计算机技术—计算机应用技术]