基于Taylor Series-LQR复合策略的双关节机械臂时滞补偿控制  

Taylor Series-LQR Compound Strategy Based Time-delay Compensation Control for Dual-Joint Manipulators

作　　者：陈士安[1] 刘金裕 CHEN Shi’an;LIU Jinyu(College of Automobile and Transportation Engineering,Jiangsu University,Zhenjiang Jiangsu 212013,China)

机构地区：[1]江苏大学汽车与交通工程学院,江苏镇江212013

出　　处：《电子器件》2023年第3期752-757,共6页Chinese Journal of Electron Devices

基　　金：国家自然科学基金项目(52072158)。

摘　　要：为解决时滞导致的颤振问题,提出了一种基于Taylor Series-LQR(Linear Quadratic Regulator)复合策略的双关节机械臂时滞补偿控制方法。首先,利用两个一阶泰勒级数分别构建基于两机械臂当前理想控制力预测下一时滞时刻的理想控制力的预测方程;然后,利用这两个预测方程对描述双关节机械臂运动的状态方程进行扩展;最后,根据LQR控制策略设计时滞补偿预测控制器求取预测控制力。应用结果显示,随着时滞的变大,使用无补偿措施的LQR控制器的双关节机械臂的颤振逐渐增大,并最终走向运动发散,而使用Taylor Series-LQR时滞补偿预测控制器的双关节机械臂在较大时滞范围内均能稳定跟踪期望轨迹而不产生颤振,说明Taylor Series-LQR复合策略具有良好的双关节机械臂时滞补偿效果。To solve the problem of flutter caused by time delay, a time delay compensation control method for dual-joint manipulators based on Taylor Series-Linear Quadratic Regulator(LQR)compound strategy is proposed. Firstly, two first-order Taylor series are utilized to construct prediction equations using the current ideal control forces of the two manipulators to calculate the predictive control force at the next time delay. Then, these two prediction equations are used to augment the system state equation describing the dual-joint manipulators motion. Finally, a LQR based time-delay compensation predictive controller is designed to obtain the predictive control force. The application results show that, as the time delay increases, flutter of the double-joint manipulators regulated by the LQR controller without compensation measurement gradually deteriorates, and eventually leads to divergence. When using the Taylor Series-LQR controller, the dual-joint manipulators can track the desired trajectory stably without chattering within a large time delay range. It indicates that the Taylor Series-LQR compound strategy has a good time-delay compensation effect for the dual-joint manipulators.

关 键 词：双关节机械臂 LQR 泰勒级数 时滞补偿 

分 类 号：TP241[自动化与计算机技术—检测技术与自动化装置] TP301.6[自动化与计算机技术—控制科学与工程]