Feasibility Conditions-Free Prescribed Performance Decentralized Fault-Tolerant Neural Control of Constrained Large-Scale Systems
针对具有非对称时变全状态约束的不确定非线性大规模系统，提出了一种基于命令滤波的分散预设性能自适应容错补偿控制策略，无需虚拟控制信号的可行性条件，并补偿了无限时变执行器故障的影响。
This article investigates the command filter-based decentralized prescribed performance adaptive fault-tolerant compensation control strategy for uncertain nonlinear large-scale systems subject to asymmetric time-varying full-state constraints. Via integrating the performance function with command filter-based backstepping technique, the prescribed performance control problem is addressed, under which the complexity of controller design is reduced. Under the prescribed performance control framework, the nonlinear transformed function is constructed so as to ensure that the asymmetric time-varying full-state constraints free from feasibility conditions imposed on virtual control signals are not violated. Besides, the effect of infinite number of time-varying actuator faults is compensated with the aid of projection adaption design. Furthermore, based on the piecewise Lyapunov function analysis, it is rigorously testified that entire involved signals are bounded, desired constraints are not breached and tracking errors within the predefined domains. Finally, the effective performances of the developed control algorithm are confirmed by some simulation results.