宽幅机器人作业平台最优路径跟踪控制系统研究  

Optimal path tracking control system for wide-span robot operation platform

作　　者：李蓉萱 闫洪峰[1,2] 李法镰 赵传波 李璐 刘成李 邓育荣 LI Rongxuan;YAN Hongfeng;LI Falian;ZHAO Chuanbo;LI Lu;LIU Chengli;DENG Yurong(Chinese Academy of Agricultural Mechanization Sciences Co.,Ltd.,State Key Laboratory of Agricultural Equipment Technology,Beijing 100083,China;Beijing Jinlun Kuntian Special Machine Co.,Ltd.,Beijing 100083,China;Dalian University of Technology,Dalian Liaoning 116024,China)

机构地区：[1]中国农业机械化科学研究院集团有限公司,农业装备技术全国重点实验室,北京100083 [2]北京金轮坤天特种机械有限公司,北京100083 [3]大连理工大学,辽宁大连116024

出　　处：《农业工程》2025年第3期26-33,共8页AGRICULTURAL ENGINEERING

基　　金：中国机械工业集团有限公司科研专项(ZDZX2022-1)。

摘　　要：为提升宽幅机器人作业平台在田间作业的精度并减少由于土壤土质不均和农机几何形状不对称等因素导致的行驶横向偏移,提出一种最优路径跟踪控制系统。该系统采用无迹卡尔曼滤波(UKF),融合北斗卫星定位和机器视觉技术,监测和控制车辆行驶状态参数。介绍非线性模型预测控制(NMPC)策略,建立车辆运动学模型,与自适应模型预测控制(AMPC)方法相比较,Simulink仿真结果表明,对于路径跟踪系统,NMPC有更快的响应和超调速度。An optimal path tracking control system was proposed to improve accuracy of wide-span robot operation platform in field and reduce lateral deviation of running caused by uneven soil quality and asymmetric geometric shape of agricultural machinery.System adopted unscented kalman filter(UKF),integrated information of Beidou satellite positioning and machine vision,and observed vehicle running state parameters.In addition,a nonlinear model predictive control(NMPC)strategy was introduced to establish a vehicle kinematic model,which was compatible with adaptive model predictive control(AMPC).Simulink simulation results showed that NMPC has faster response and overshoot speed for path tracking system.

关 键 词：宽幅机器人作业平台 多传感器数据融合 无迹卡尔曼滤波 路径跟踪 模型预测控制 

分 类 号：S24[农业科学—农业电气化与自动化]