Adaptive locally weighted learning tracking control for a class of unknown strict-feedback nonlinear systems via differentiable higher order kernels

Abstract

This study addresses the problem of adaptive locally weighted learning (LWL) tracking control for a class of n-order unknown strict-feedback nonlinear systems (SFNSs). Without involving a priori information on the system dynamics, an adaptive tracking control algorithm is designed by fusing LWL and the technique of backstepping, in which the LWL is employed to identify the unknown nonlinear functions. Particularly, by introducing a novel weighting function with sufficient differentiability, the obstacle of the integration of LWL and backstepping to control SFNSs is successfully circumvented. The developed adaptive LWL tracking control scheme ensures that all closed-loop signals are bounded. Finally, simulation results are performed to testify the effectiveness of the proposed LWL tracking control approach.