空间机器人的可变增益滑模控制方法

A variable gain sliding mode control method for space robots

摘要:

针对空间机器人抓捕目标的轨迹跟踪过程,提出了一种基于可变增益的滑模控制算法,并利用MATLAB/Simscape对空间机器人的理论动力学模型和控制算法进行了仿真校验。首先,基于拉格朗日法构建了空间机器人的动力学模型,并给出了系统参数不确定时的动力学方程。设计了一种滑模控制器,在控制器中引入了可变增益,该增益可根据系统的不确定性和系统的运动状态进行调整,从而使控制器具有很强的鲁棒性。基于Simscape Multibody软件模块,对空间机器人的理论动力学模型进行了验证。在此基础上,针对多关节运动和系统模型不确定两种情形,利用可变增益滑模控制算法进行了空间机器人的轨迹跟踪仿真实验。实验结果表明,与传统的计算力矩法相比,本文提出的控制算法呈现出更高的运动控制精度和误差收敛速度。

Abstract:

A variable gain sliding mode control algorithm is proposed for a space robot to track the path aimed at capturing a target. MATLAB/Simscape is used to verify the dynamic model of the space robot and the control algorithm. First, Lagrange method is adopted to establish the dynamic model, and subsequently a model considering the uncertainty of the system is developed. A sliding mode controller which introduces a variable gain is developed. The gain can adapt to the uncertainty of the system and its motion thereby facilitating the controller to be robust. Further, the theoretical model of the space robot is verified using Simscape Multibody. In addition, for the scenarios of single joint motion, multiple joints motion and uncertain system, the proposed control method is applied to the robot for tracking a desired path. Numerical results have demonstrated that in comparison with computed control method, the new method presents higher control accuracy and faster convergence speed.