Fixed-Time Prescribed Performance Neural Fault-Tolerant Control of Euler–Lagrange Systems Under Unknown Bounded Initial Conditions

摘要

This paper investigates the fixed-time prescribed performance tracking control problem for Euler-Lagrange systems with model uncertainties, external disturbances, and actuator faults. To the best of the authors’ knowledge, achieving prescribed transient and steady-state behaviors within a fixed time under unknown bounded initial conditions, while simultaneously ensuring effective compensation for system uncertainties and faults, still remains an open problem. To address these challenges concerning both performance-related and reliability-related constraints, we propose a novel adaptive neural fault-tolerant control with fixed-time prescribed performance (ANFTC-FPP). The performance-related constraints are handled through a unified framework that synergistically combines novel prescribed performance functions (PPFs) with barrier Lyapunov functions (BLFs). This integration relaxes initialization constraints and characterizes the relationship between initial conditions and transient performance, thereby considering overshoot for tracking errors within a fixed time while achieving specified steady-state accuracy, regardless of initial states. For reliability-related constraints, we establish a fixed-time compensation mechanism where model uncertainties are handled by extending radial basis function neural networks (RBFNNs) for uncertainty approximation, while actuator faults are addressed through adaptive fault-tolerant control (FTC). The semi-global practical fixed-time stability (SPFS) of all closed-loop signals is rigorously established through comprehensive Lyapunov stability analysis. The efficacy of the proposed control strategy is experimentally validated on a physical KINOVA robotic manipulator system through real-world implementation.