Fault-Tolerant Control for Uncertain Nonstrict-Feedback Stochastic Nonlinear Systems With Output Constraints
针对存在执行器故障和输出约束的非严格反馈随机非线性系统，提出一种基于神经网络和障碍李雅普诺夫函数的自适应容错控制策略，无需故障先验信息即可在线补偿故障，保证跟踪性能和输出约束。
This article investigates the problem of fault-tolerant control (FTC) for nonstrict-feedback (NSF) stochastic nonlinear systems subject to actuator faults and output constraints. Most of the existing backstepping-based FTC techniques have been developed for strict-feedback systems, which may cause algebraic loop problems when applied directly to NSF systems. In this work, unknown nonlinear functions are approximated by neural networks (NNs). Based on the properties of NN basis functions and a barrier Lyapunov function, a novel adaptive FTC strategy for an NSF system is proposed to achieve tracking control without violating the output constraint. Adaptive fault accommodation terms are constructed to compensate for the faults without a priori information to achieve online fault tolerance. The designed controller guarantees the fault-tolerant tracking performance while ensuring that all signals are bounded in probability and that the output is within the specified constraint. The performance of the proposed FTC strategy is illustrated by simulation studies.