受约束空间机器人降阶自适应神经网络滑模控制

The Reduced Order to the Adaptive Neural Network Sliding Mode Control of Constrained Space Robot

摘要:

为了实现受约束空间机器人的高精度控制，提出了一种基于U-K(Udwadia-Kalaba)方程的降阶自适应神经网络滑模控制算法。首先，基于U-K方程，同时考虑受约束空间机器人末端的理想约束力与非理想约束力，推导得到详细的动力学方程。接着，考虑到非理想约束力（主要是受约束空间机器人末端受到的切向力）具有不确定性且单独采用滑模控制会出现抖振现象，提出了自适应神经网络滑模控制算法，实现各关节角度、角速度以及非理想约束力的高精度跟踪。然后，针对系统受约束模型，对动力学方程和滑模控制器进行了降阶求解，减少了变量并简化了计算过程。最后，为了验证所提算法的正确性与合理性，以2自由度受约束空间机器人为例进行了仿真验证。仿真结果表明：受约束空间机器人的各关节角度、角速度以及非理想约束力的跟踪误差均低于10-4量级。

Abstract:

In order to achieve high precision control of the constrained space robot, a reduced order adaptive neural network sliding mode control algorithm based on U-K (Udwadia-Kalaba) equation is proposed. Firstly, On the basis of the U-K equation and considering the ideal and non-ideal constrained forces at the terminal of the constrained space robot, the detailed dynamic equations are derived. Secondly, considering the uncertainty of the non-ideal constrained force (mainly the tangential force on the terminal of the constrained space robot) and the chattering phenomenon when using sliding mode control alone, the adaptive neural network sliding mode control algorithm is proposed to realize the high-precision tracking of each joint angle, angle speed and the non-ideal constrained force. Thirdly, for the constrained model of the system, the dynamic equation and the sliding mode controller are reduced to decrease the variables and simplify the calculation process. Finally, in order to verify the correctness and rationality of the proposed algorithm, the 2-DOF constrained space robot is taken as the simulated object. The simulation results show that the tracking errors of joint angles, angle speed and the non-ideal constrained force are less than 10-4 order of magnitude.