基于MK一RRT算法的果园机器人多目标点路径规划研究

打开文本图片集

中图分类号：S24；TP242 文献标识码：A 文章编号：2095-5553（2025）12-0240-08

Abstract：Toaddresstheissuesof high path costand por executabilityincontinuous movementbetween multiple target points incomplex orchard environments，this paper proposes a KinematicConstraint-based MultiobjectiveRapidlyexploring Random Tree（MK—RRT）algorithm.The algorithm first preprocesses the map using grid-based connected component analysisaccording tointer-target distance，headinganglevariation，andenvironmentalcomplexitySimultaneously，apath cost prediction model between regions is constructed to solve the optimal traversal sequence of target points.Considering the kinematicconstraints of orchard mobilerobots，a sampling method incorporating motionconstraints is designed，along witha path smoothing optimization strategy to generate an optimal smoth pathconnecting multiple targets.This approach establishesatheoreticalmodelformulti-targetpathplanningthatintegratesthecomplexityofrealorchardenvironmentsand therobot's intrinsic constraints，providing reliable technical support for continuous movement among operational points. Experimental results demonstrate that in diverseorchard scenarios，compared tothe discrete brainstorming algorithm，the proposed method reduces average path length by 12.3% and decreases maximum turning angle by 90.3% .These findings confirmthatMK—RRTnotonlyoptimizes path length but alsoenhances smoothnessand executability，providing theoretical and technical references for full-field multi-target continuous path planning of orchard robots.

Keywords： orchard mobile robots；multiple objective points； kinematic constraints；path planning；RRT

0 引言

水果是我国第三大种植产业，种植面积大，产量高1。（剩余11059字）