基于模型预测控制的机器人姿态控制策略研究  

Research on Robot Attitude Control Strategy Based on Model Predictive Control

作　　者：姚建均 张宜坤 柯运 钱琛 王超 YAO Jianjun;ZHANG Yikun;KE Yun;QIAN Chen;WANG Chao(College of Mechanical and Electrical Engineering,Harbin Engineering University,Harbin 150000)

机构地区：[1]哈尔滨工程大学机电工程学院,哈尔滨150000

出　　处：《机械工程学报》2024年第19期88-100,共13页Journal of Mechanical Engineering

摘　　要：针对机器人复杂灵活的任务,需要机器人具有运动平滑、大范围定位与高精度跟踪的能力。为了解决离线规划导致的低系统实时性,与低精度运动学数值逆解导致机器人抖动的问题,提出一种基于模型预测控制的机器人姿态控制策略。首先,建立机器人正运动学模型,将其表达为状态空间模型;其次,考虑机器人极限关节位置与运动平滑性两种约束条件,构造其为紧、凸集合;然后,将机器人求解运动学逆解的问题转换为求解带约束最优控制问题,并得出最优控制量;最后,理论证明了控制器的稳定性;仿真表明,相对于经典控制策略,提出的控制策略具有更好的大范围定位能力;设计基于六自由度机器人的大范围定位实验与复杂轨迹跟踪实验,证实了提出的控制策略具有良好的跟踪与定位能力。For the complex and flexible tasks of robots,robots are required to have the ability of smooth motion,large range positioning and high precision tracking.In order to solve the problem that low real-time performance of offline trajectory planning and low precision numerical solution of inverse kinematics lead to robot jitter,a robot attitude control strategy based on model predictive control is proposed.Firstly,the forward kinematics model of the robot is established and expressed as a state space model.Secondly,considering the two constraints of the ultimate joint position and motion smoothness of the robot,and it is constructed as a compact and convex set.Then,the inverse kinematics of the robot is transformed into the constrained optimal control problem,and the optimal control action is obtained.Finally,the stability of the controller is proved theoretically.Simulation results show that compared with the classical control strategy,the proposed control strategy has better positioning ability in a large range.Apart from this,large range positioning experiments and complex trajectory tracking experiments based on six degree of freedom robot are designed,which confirm the good tracking and positioning ability of the proposed control strategy.

关 键 词：机器人姿态控制 模型预测控制 约束系统 轨迹跟踪 

分 类 号：TG242[金属学及工艺—铸造]