Adaptive Neural Network Tracking Control for Switched Uncertain Non-Linear Systems with Actuator Failures and Time-Varying Delays

Qinghua Hou,Jiuxiang Dong,Changjiang Xi
DOI: https://doi.org/10.1049/iet-cta.2018.6188
2019-01-01
Abstract:This study focuses on the problem of adaptive neural network (NN) tracking control for a class of strict-feedback non-linear switched systems under arbitrary switching. The considered systems are with unknown external disturbances, time-varying delays, and actuator failures. NNs are used to approximate unknown functions and Lyapunov-Karsovskii functionals are utilised to compensate for the time-varying delays. Different from the existing results, piecewise switched adaptive laws are proposed for each subsystem, which can reduce conservativeness caused by using common adaptive laws for all subsystems. Besides, prescribed performance bound (PPB) technique is developed to further improve the transient performance of the systems, especially when actuator failures occur and system switchings take place. Finally, under the framework of Lyapunov theory, an adaptive NN reliable tracking control method is proposed. It is proved that all the signals in the closed-loop system are semi-globally uniformly ultimately bounded and the tracking error remains within the PPBs. A simulation study is given to illustrate the effectiveness of the authors' results.
What problem does this paper attempt to address?