果园二维激光雷达自主导航控制研究

打开文本图片集

中图分类号：S24 文献标识码：A 文章编号：2095-5553（2025）08-0090-07

Abstract：In modern agriculture，orchard plantprotection operations often face chalenges due to complex and unstructuredenvironments，necesitating thedevelopmentofautonomous navigationcapabilities fororchard vehicles.To address this need，anautonomous navigation system based on 2DLiDAR technology was developed.This system was designed to meetagronomicrequirementsof orchard planting，whileaccounting forkey environmental variables such as plantspacing，row spacing，and trunk diameter.Toenhance theaccuracyof navigation，noiseinthe2DLiDAR point cloud datawasefectivelyreduced bycustomized filtering thresholds.Feature extractionoftreetrunks was accomplished usingacombinationoftheEuclideanclusteringalgorithmandthethree-pointcolineargeometrytheoremtoaccurately identify centralfeature points fromthefiltered pointclouddata.Inadition，anavigation path was then generated using the least squares fiting methodtoensure high precisionpath navigation.Toenablereal-timeadaptability，a fuzzycontrol algorithmwasintegrated intothenavigation system，thereby alowing dynamicadjustmentsforautomatic drivingand path tracking.Experimentalresultsdemonstrated thesystem'seffectiveness，withtheaverageerorbetweenthefltered laser point cloud feature points and the actual tree positions measured at 8.96cm .During autonomous driving at a speed of 0.5m/s ，the vehicle maintained high navigation accuracy，with a maximum lateral deviation of 8.2cm and an average deviation of 6.81cm .These findings confirmed the system's robustness and practicality，showing that it significantly improved navigationprecision in thecomplex orchard environment and successfully enabled autonomous vehicle operatio Keywords：orchard;2DLiDAR;autonomous navigation;laser pointcloud data procesing;filtering and noisereduction

0 引言

近年来，随着工业化进程的加快和人口老龄化，农业劳动力不足的问题越来越严重。（剩余10368字）