In this paper, the issue of iterative learning fault diagnosis (ILFD) and fault tolerant control (FTC) is studied for stochastic repetitive systems with Brownian motion. Different from existing fault diagnosis (FD) methods, a state/fault simultaneous estimation observer based on iterative learning method is designed. The convergence condition of the ILFD algorithm is given. By employing the fault estimation information, the FTC algorithm is proposed to compensate for the fault effect on the system and to keep the stochastic input-to-state stability of the control system. Finally, the simulation results of an induction motor system and a single-link robotic flexible manipulator system are given to show that the proposed method is validated.

本文研究了具有布朗运动的随机重复系统的迭代学习故障诊断（ILFD）和容错控制（FTC）问题。与现有的故障诊断（FD）方法不同，设计了一种基于迭代学习方法的状态/故障同时估计观测器。给出了ILFD算法的收敛条件。通过利用故障估计信息，提出了FTC算法来补偿故障对系统的影响，并保持控制系统的随机输入到状态的稳定性。最后，给出了感应电机系统和单连杆机器人柔性机械臂系统的仿真结果，表明所提方法是有效的。