Finite-time sliding mode fault-tolerant neural network control for nonstrict-feedback nonlinear systems

Abstract In the paper, finite-time tracking control for a class of nonstrict-feedback nonlinear systems subject to uncertain control gains and multiple actuator faults is explored. To circumvent the issue of “complexity explosion” arisen from the standard backstepping design, second-order command filter is introduced to estimate the virtual input and its derivative in every step. By means of a useful structural attribute of neural networks, the algebraic loop problem is excluded. Besides, the unknown control gains are handled via using Nussbaum functions. To attain robust control performance against external perturbations and chattering phenomenon, a novel sliding mode fault-tolerant controller is introduced to ensure finite-time stability of the controlled system. Furthermore, simulation cases are conducted to demonstrate the practical applications of the proposed approach to one-link manipulator and electromechanical system.