一种基于改进混合A^(*)搜索方法的路径规划算法  

A Path Planning Algorithm Based on Improved Hybrid A^(*)Search Method

作　　者：谢涛 夏青元[1] XIE Tao;XIA Qingyuan(Nanjing University of Science&Technology,Nanjing 210094)

机构地区：[1]南京理工大学,南京210094

出　　处：《计算机与数字工程》2025年第2期422-426,共5页Computer & Digital Engineering

摘　　要：目前非结构化场景的路径规划存在效率低下的问题,论文提出了一种基于改进混合A~*的路径规划算法,在启发式函数中加入对航向角的动态约束,优化定长运动基元为长短混合的变长运动基元,运用共轭梯度法对初始路径进行平滑,最终获得了一条安全、平滑且易于车辆行驶的路径。经过仿真和真车试验,平均减少了31.11%的搜索节点损耗,降低了50.13%的搜索时间,显著提高了算法运行效率。At present,there is a problem of low efficiency in path planning in unstructured scenes.This paper proposes a path planning algorithm based on improved hybrid A^(*).Dynamic constraints on the heading angle are added to the heuristic function,and the fixed-length motion primitives are optimized as long and short.The mixed variable-length motion primitives use the conjugate gradient method to smooth the initial path,and finally obtain a safe,smooth and vehicle-friendly path.After simulation and real ve⁃hicle tests,the average loss of search nodes is reduced by 31.11%,and the search time is reduced by 50.13%,which significantly improves the operation efficiency of the algorithm.

关 键 词：无人车 混合A~*算法 启发函数 车辆运动学 路径规划 

分 类 号：TM743[电气工程—电力系统及自动化]