<![CDATA[This article explores the problem of adaptive control for nonlinear dynamic systems operating under uncertainty. It presents a model reference adaptive control (MRAC) method that integrates a neuro-fuzzy network with B-spline basis functions. The proposed approach allows effective approximation of nonlinear behaviors and ensures high control accuracy despite external disturbances and structural uncertainties within the system. The paper compares the performance of conventional linear MRAC with the neuro-fuzzy controller. Simulation results demonstrate that the neuro-fuzzy MRAC achieves superior stability and accuracy in closed-loop control. Additionally, the study examines the system’s local stability under specific conditions of the learning rate. To address the challenge of computational complexity, a decomposition strategy dividing the controller into smaller sub-models is introduced, effectively mitigating the “curse of dimensionality.” The findings support the applicability of neuro-fuzzy controllers for the intelligent control of a wide range of nonlinear systems.]]>
Copyrights © 2026
