农业工程学报2025，Vol.41Issue(18)：28-41,14.DOI:10.11975/j.issn.1002-6819.202412242
丘陵山地果园纯电驱动除草机器人研制
Design and development of a fully electric weeding robot for hilly and mountainous orchards
摘要
Abstract
Weed management has been confined to small spaces and obstructive branches in densely planted orchards,particularly in the hilly and mountainous regions.The conventional mowing equipment cannot fully perform the intra-row weeding or navigate the tight transitions.This study aimed to develop an intelligent,fully electric-driven robot on both inter-row and intra-row weeding operations with obstacle avoidance.Specifically,the modular hardware and control architecture was integrated for the spatial and terrain constraints of the closed-canopy orchard environments.Four core systems were developed:A dual-motor tracked chassis to enhance the terrain adaptability;an electric push-rod mechanism for the adjustable cutting height in response to undulating terrain;a torsion-spring passive avoidance for the intra-row weeding blades;and an isolated direct current to direct current(DC-DC)converter system for the stable and safe power distribution across high-and low-voltage subsystems.A fuzzy proportional-integral-derivative(PID)controller was implemented for the chassis drive system,in order to improve the accuracy of the motion control under unstructured field conditions.Furthermore,an improved version of the Sparrow Search Algorithm(SSA)was proposed to optimize the parameters of the controllers.This optimization was incorporated with the chaotic population initialization,adaptive dynamic step adjustment,and reverse learning strategies,in order to prevent the premature local optima for the convergence performance.Simulation tests demonstrated that the improved fuzzy PID controller exhibited significantly enhanced tracking performance and robustness.Compared with both standard SSA-tuned fuzzy and conventional PID controllers,the improved controller reduced the steady-state error and overshoot,when subjected to the step inputs,indicating superior response stability and dynamic adaptability.Field experiments were conducted to validate the performance of the robots under full-load operations in a closed-canopy hilly orchard.The better performance was achieved,with an average working speed of 0.781 1 m per second,and an average turning trajectory diameter of 984 mm.Reliable operation was also maintained on the slopes with the gradients up to 16.8°.The heading deviation remained within±3° during navigation.In terms of agronomic effectiveness,the inter-row weeding rate reached an average of 91.97%.The success rate of obstacle avoidance reached 95.58%,indicating better performance in safely maneuvering around tree trunks and irregular obstacles.The consistency coefficient of the stubble height exceeded 85%,indicating the uniform cutting height.The cutting width utilization rate surpassed 90%for the high efficiency.All evaluated metrics fully met the requirements of the original design,indicating technical feasibility and functional robustness.The high maneuverability,terrain adaptability,and precision weeding were realized in the hilly,spatially constrained orchard environments.An optimized fuzzy PID controller and the metaheuristic tuning algorithm were integrated to enhance control performance and autonomous decision-making.This finding can offer valuable theoretical and technical support for the future development of electric-driven weeding robots targeting closed-canopy orchards.A great contribution can also be gained to advance intelligent orchard machinery in sustainable agriculture.关键词
农业机械/机械化/株间除草/割茬高度调节/模糊PID控制/性能试验Key words
agricultural machinery/mechanization/inter-row weeding/stubble height adjustment/fuzzy PID control/performance testing分类
农业科技引用本文复制引用
黄薛凯,施印炎,汪小旵,林乐彬,王延鑫,徐浪涛..丘陵山地果园纯电驱动除草机器人研制[J].农业工程学报,2025,41(18):28-41,14.基金项目
国家重点研发计划项目(2023YFD2000304 （）
2023YFD2000305) （）
甘肃省农机研发制造推广应用一体化试点项目(6-1) （6-1）
