Liu Xia, Zhao Yang
School of Electrical Engineering and Electronic Information, Xihua University, Chengdu 610039, China.
Rev Sci Instrum. 2024 Oct 1;95(10). doi: 10.1063/5.0226681.
This paper focuses on a robotic system with actuator fault and presents a sliding mode active fault-tolerant control method based on coupled position error. The actuator fault is first detected by a fault alerter with a predetermined threshold. After the fault is successfully detected, the fault is estimated by a fault observer. Coupling the position error with the weighted position error of different joints, a non-singular fast terminal sliding mode surface is constructed, and a super-twisted algorithm is introduced to design a coupled error super-twisted sliding mode controller (CESSMC). Furthermore, the fault estimation is incorporated with the CESSMC to accomplish the coupled error super-twisted sliding mode active fault-tolerant control. The stability and the finite-time convergence of the system are theoretically proved. Simulations and experiments are conducted to verify the effectiveness of the proposed method. The proposed method can achieve the position error convergence in finite time and reduce chattering in the control input. With the tight coupling between position error and weighted position error, the fault compensation effect and trajectory tracking accuracy can be improved.
本文聚焦于一个存在执行器故障的机器人系统，并提出了一种基于耦合位置误差的滑模主动容错控制方法。执行器故障首先由具有预定阈值的故障警报器进行检测。在故障被成功检测后，通过故障观测器对故障进行估计。将位置误差与不同关节的加权位置误差相耦合，构建了一个非奇异快速终端滑模面，并引入超扭曲算法设计了一种耦合误差超扭曲滑模控制器（CESSMC）。此外，将故障估计与CESSMC相结合，以实现耦合误差超扭曲滑模主动容错控制。从理论上证明了系统的稳定性和有限时间收敛性。进行了仿真和实验以验证所提方法的有效性。所提方法能够在有限时间内实现位置误差收敛，并减少控制输入中的抖振。通过位置误差与加权位置误差之间的紧密耦合，可以提高故障补偿效果和轨迹跟踪精度。
