Asymptotic tracking and dynamic regulation of SISO non‐linear system based on discrete multi‐dimensional Taylor network

Abstract

For non-linear control, it is important to secure a generally structured controller that promises wide application and desirable performance. This study deals with the problem of asymptotic tracking and dynamic regulation of single-input single output (SISO) non-linear systems via output feedbacks by the discrete multi-dimensional Taylor network (MTN) controller, a novel controller with fixed structure and sampled-data control mechanism. For verification of its validity, differential geometry and polynomial approximation are adopted. Using the emulation technique and regional pole assignment, the asymptotic tracking and dynamic regulation without online optimisation of the system by discrete MTN controller is tested. With the dynamic change of error signals, the dynamic regulation by given index is realised. As a convex optimisation problem, the controller parameters can be acquired by parametric learning. Based on the delta operator model, the procedure of the controller design is given in detail. Simulation results confirm the feasibility and effectiveness of the proposed approach.