Inverse Neuro-Fuzzy Control of Nonlinear Systems

Abstract

This article, we addressed the use of Artificial Intelligence techniques in the field of control engineering. Our work focuses on the use of neuro-fuzzy networks, specifically ANFIS (Adaptive Neuro-Fuzzy Inference System), for the identification of inverse models required for implementing the control laws of a nonlinear dynamic system. In the first approach, the identified model is used as an open-loop controller with the system (Direct Inverse Control) for regulation purposes. In the second approach, Internal Model Control (IMC) is applied to improve the performance of the neuro-fuzzy model when the system is subject to a constant disturbance. The final section presents an application of these control structures to a nonlinear system. The results are validated through simulations carried out in the MATLAB environment.