Adaptive Neural Fixed-Time Control for Uncertain Nonlinear Systems

Abstract

The article addresses the adaptive fixed-time system control problem subject to unknown nonlinear functions. By employing backstepping technique, fixed-time control, and neural networks (NNs), an adaptive fixed-time control approach is developed. NNs are utilized to package the unknown nonlinearities in the controlled system. The hyperbolic tangent functions are firstly utilized to solve the singular problem, which may exist in the derivation of virtual controller. Meanwhile, the jitter phenomenon is avoided. The developed controller design scheme not only assures the signals’ boundedness within fixed-time interval, but also makes the tracking error converge. Finally, on the aid of the simulation results, the effective of the designed control technique is shown.