使用高斯过程回归的四轮机器人终端滑模控制
首发时间:2023-05-12
摘要:滑模控制(sliding mode control, SMC)是一种非线性控制,可以在控制过程中根据系统当前的误差状态,迫使系统向设计的滑模面运动并最终完成收敛。终端滑模(terminal sliding mode control, TSMC)是滑模控制中的一种,通过将滑模面设计为非线性函数使得系统误差可以在指定时间内收敛,得到更好地控制效果。滑模控制可以有效地对不确定性进行处理,具有很强的鲁棒性,因此被广泛应用于机器人控制中。但是,滑模控制器往往存在着抖振的现象,需要在控制中进行改善以适用于实际系统。本文针对四轮移动机器人的路径跟踪任务,设计了终端滑模控制器。并通过高斯过程回归对控制器中包含的不确定性进行处理,从而在减小抖振的同时提高跟踪性能。通过理论分析证明了系统的稳定性。数值仿真的结果表明,与传统的终端滑模相比,设计的使用高斯过程回归的终端滑模控制器能够获得更好的控制效果,同时保持控制量抖振较小,在可实际执行的范围内。
关键词: 自动控制技术 四轮机器人 高斯过程回归 终端滑模控制
For information in English, please click here
Terminal Sliding mode control of four-wheel robot with Gaussian process regression
Abstract:Sliding mode control (SMC) is a kind of nonlinear control, which can force the system to move to the designed sliding mode surface and finally complete convergence according to the current error state of the system in the control process. Terminal sliding mode control (TSMC) is one of sliding mode control. By designing the sliding mode surface as a nonlinear function, the system error can converge within a specified time, and better control effect can be obtained. Sliding mode control can deal with uncertainty effectively and has strong robustness, so it is widely used in robot control. However, the sliding mode controller often has chattering phenomenon, which needs to be improved in the control to apply to the actual system. In this paper, a terminal sliding mode controller is designed for the path tracking task of a four-wheel mobile robot. And the uncertainty contained in the controller is processed through Gaussian process regression, thereby reducing chattering and improving tracking performance. The stability of the system has been proven through theoretical analysis. Numerical simulation results show that compared with the traditional terminal sliding mode controller, the designed terminal sliding mode controller using Gaussian process regression can achieve better control effect, while keeping the control amount chattering small and within the range of practical implementation.
Keywords: Automatic control technology Four-wheel robot Gaussian process regression Terminal sliding mode control
基金:
引用
No.****
同行评议
勘误表
使用高斯过程回归的四轮机器人终端滑模控制
评论
全部评论