华中科技大学学报（自然科学版）2025，Vol.53Issue(11)：55-60,6.DOI:10.13245/j.hust.251113
基于障碍物状态分类的移动机械臂动态避障方法
Dynamic obstacle avoidance method for mobile manipulator based on obstacle state classification
摘要
Abstract
Aiming at the the defects in existing mobile manipulator control algorithms,such as poor adaptability and stability,in the face of high dynamic complex scene,a dynamic obstacle avoidance method for mobile manipulator based on obstacle state classification was proposed.Model predictive control(MPC)was used to predict the state and control input in the future period of time,the kinematics model of the mobile manipulator was coupled and the deviation between the end pose and the target pose was added to the cost function as a quadratic cost term.To ensure safety,track of surrounding obstacles was kept,and obstacle parameters were updated.Obstacle states were predicted and classified,and according to the motion state of the obstacles they were divided into static obstacles,uniform moving obstacles and irregular moving obstacles.A number of collision points were arranged at the prominent position of the mobile manipulator,and the distance between these collision points and obstacles in space was added to the cost function as a penalty term.Simulation results show that this method can effectively improve the security and stability of the mobile manipulator in high dynamic environment.关键词
移动机械臂/动态场景/运动控制/避障/路径规划Key words
mobile manipulator/dynamic scene/motion control/obstacle avoidance/path planning分类
信息技术与安全科学引用本文复制引用
ZHONG Hang,LUO Teng,ZHU Qing,WANG Yaonan..基于障碍物状态分类的移动机械臂动态避障方法[J].华中科技大学学报（自然科学版）,2025,53(11):55-60,6.基金项目
广东省基础与应用基础研究基金资助项目(2024A1515240062). （2024A1515240062）
