基于AWJPS-LOS算法的人员跟随路径规划

打开文本图片集

中图分类号：TP242 文献标志码：A 文章编号：2095-2945（2025）26-0011-04

Abstract：Toachievereal-timepersonnel follwingtasksinenvironments without prior maps，thispaper proposesan autonomousfolowingsystembasedonmulti-sensorfusion.Fortherobotsfront-endpathplanning，apathplanningalgorithm combiningtheimprovedJump Point Search （JPS）algorithmandtheLineof Sight（LOS）algorithmisproposed.Thisalgorithm significantlyimprovessearcheficiencybyeliminatingredundantnodesonthepathToaddresstherobot'sobstacleavoidance andpathsmoothingisues，aback-endtrajectoryoptimizationmethodisproposedThismethodconstructsasafetycoridorand appliesBeziercurvestogeneratesmothtrajectories.Theproposedalgorithmsweretestedonaphysicalplaformforindoor corridor follwingexperiments，andtheresultsdemonstratethatthepath planing methoddoesnotrelyonaglobalmapadcan effectively handle situations such as multiple pedestrian interferences and target loss.

Keywords：person following；path planning;obstacle avoidance； jump point search;trajectory optimization

近年来，随着感知手段与计算能力的快速发展，智能机器人技术取得了突飞猛进的发展，在人类生产、生活中得到了广泛应用。（剩余4604字）