具有时变约束的机械臂系统自适应控制器设计  

Adaptive control design for robot manipulators system with time-varying constraints

作　　者：范永青 郝美荣 FAN Yongqing;HAO Meirong(School of Automation,Xi’an University of Posts and Telecommunications,Xi’an 710121,China)

机构地区：[1]西安邮电大学自动化学院,陕西西安710121

出　　处：《西安邮电大学学报》2022年第1期102-110,共9页Journal of Xi’an University of Posts and Telecommunications

基　　金：国家自然科学基金项目(61903298,51875457,62003262);陕西省留学人员科技活动择优项目(35)。

摘　　要：对具有时变约束的不确定多自由度机械臂动态系统问题,设计出一种基于障碍李雅普诺夫函数(Barrier Lyapunov Function,BLF)的自适应径向基函数神经网络切换控制器。在传统的径向基函数神经网络(Radial Basis Function Neural Network,RBFNN)的基础上设计出一种带有时变参数的万能逼近器逼近机械臂系统中的不确定性,引入BLF进行稳定性分析,并设计一个滑模面使未知机械臂系统的状态在超出预先设计的定义域时,能够通过自适应律进入RBFNN的万能逼近域。仿真结果表明,该控制器可以保证系统在时变输出约束的条件下具有良好的跟踪性能,与其他的RBFNN控制器相比,该控制器的逼近精度可以自动在线更新,并将半全局稳定性扩展到全局稳定性。An adaptive radial basis function neural network switching controler based on barrier Lyapunov function is designed for an uncertain multi-freedom degree manipulator dynamic system,which has time-varying output constraints.Firstly,based on the classical radial basis function neural network,an universal approximator with time-varying parameters is designed to approximate the uncertainty in the manipulator system.Then the barrier Lyapunov function is introduced for stability analysis,and a sliding surface is designed so that when the state of the unknown manipulator system exceeds the pre-designed definition domain,it can enter the universal approximation domain of radial basis function neural network through the adaptive law.Simulation results show that the controler can ensure good tracking performance of the system under time-varying output constraints.Compared with other radial basis function neural network controllers,the approximation accuracy of the controller can be updated automatically on-line,and the semi-global stability can be extended to global stability.

关 键 词：机械臂 径向基函数神经网络系统 自适应控制 时变约束 全局一致最终有界 

分 类 号：TP13[自动化与计算机技术—控制理论与控制工程]