基于EDDPG的农业巡检机器人路径规划研究

打开文本图片集

中图分类号：S951 文献标识码：A 文章编号：2095-5553（2026）04-0187-08

Abstract：To addressthe issues of lowpath planning eficiency，long path length，and insuficient dynamic obstacle avoidancecapabilityinagricultural inspectionrobotsoperating incomplexenvironments，animprovedDep Deterministic Policy Gradient（EDDPG）-based path planing method isproposed.This method incorporates Screw Theory to enhance path planning efficiency and integrates the Artificial Potential Field method with the A* algorithm to construct a reward function，therebyimprovingdynamicobstacleavoidanceandreducing theoptimal path length.Experimentalresults demonstrate that，in atwo-dimensionalenvironment，the improved DRLalgorithmachieves the shortest path length and the best path smoothness compared to baseline algorithms. Under obstacle rates of 20% ， 40% and 60% ，itspath planning success ratesreach O.936，0.931and 0.918，respectively，all outperforming the compared methods. Furthermore，a higher frequencyof Screw Theory application leads toan increased path planning succssrate.Compared to other path planning models，the improved DRL algorithm achieves a maximum reduction of 27.23% in the shortest path length and decreases the average planning time by up to 24.16s

Keywords：agricultural automation；inspection robot；path planning；deepreinforcement learning；inteligent algorithm

0 引言

随着现代农业的发展，智能巡检机器人成为提高农业生产效率、环保和智能化的重要工具。（剩余12749字）