电子测量技术ELECTRONICMEASUREMENTTECHNOLOGY第45卷第23期2022年12月DOI:10.19651/j.cnki.emt.2209909改进RRT结合B样条的移动机器人路径规划研究*陈都侯明张学东(北京信息科技大学自动化学院北京100192)摘要:针对快速扩展随机树(RRT)算法在移动机器人路径规划中的时间过长、路径冗长曲折且平滑度较差等问题,提出一种新的采样节点生成和路径节点选取的改进型RRT算法。该算法以融合采样点概率和目标点引力的方式,再结合动态变随机步长的策略,加快了向目标点的扩展,降低了规划时间;最后为完成路径优化,对利用正向寻优和二次选取过后的路径节点,采用B样条函数进行处理,综合改善了路径的长度及其平滑性。在仿真实验中分别与传统RRT算法,P-RRT算法进行比较,实验结果表明:所提算法在路径长度,规划时间以及路径节点个数均有一定优化,且路径平滑性得到了有效提升。关键词:RRT算法;移动机器人;路径优化;二次选取;B样条中图分类号:TP24文献标识码:A国家标准学科分类代码:510.80ResearchonpathplanningofmobilerobotbasedonimprovedRRTandB-splineChenDuHouMingZhangXuedong(SchoolofAutomation,BeijingInformationScienceandTechnologyUniversity,Beijing100192,China)Abstract:Toaddresstheproblemsoftherapidly-exploringrandomtrees(RRT)algorithminmobilerobotpathplanningincludingtimecost,longandtortuouspath,andpoorsmoothness,animprovedRRTalgorithmofnewsamplingnodegenerationandpathnodeselectionisproposed.Thealgorithmcombinestheprobabilityofsamplingpointsandthegravitationofthetargetpoint,anddynamicallychangestherandomstepsize,thusacceleratingtheexpansiontothetargetpointandreducingtheplanningtime.Finally,tocompletethepathoptimization,thepathnodesobtainedaftertheforwardoptimizationandthesecondaryselectionareprocessedusingtheB-splinefunction,whichcomprehensivelyimprovesthepathintermsoflengthandsmoothness.TheproposedalgorithmhasbeencomparedwithtraditionalRRTalgorithmandP-RRTalgorithminthesimulationexperiments,whoseresultsshowthattheproposedalgorithmhasimprovedthepathlength,planningtimeandthenumberofpathnodestoacertainamount,andeffectivelythepathsmoothness.Keywords:RRTalgorithm;mobilerobot;pathoptimization;quadraticselection;B-spline收稿日期:2022-05-11*基金项目:国家自然科...
