Enhanced Control of Magnetic Levitation Systems Using Type-2 Fuzzy Elman Neural Network Cerebellar Model Articulation Controller
Trong-Hien Chiem, Van‐Phong Vu, Do Duc Tri, Tien-Loc Le · 2024
This paper endeavors to introduce a novel design concept: a type-2 fuzzy Elman neural network cerebellar model articulation controller tailored for uncertain nonlinear systems, with a primary focus on enhancing stability and accuracy in the control of magnetic levitation systems. The proposed controller merges the functionalities of a type-2 fuzzy Elman neural network system and a cerebellar model articulation controller. To automatically generate the network structure, a self- organizing algorithm is employed. Additionally, adaptation laws, derived from the gradient descent method, facilitate the online updating of network parameters. To uphold system stability, a Lyapunov stability function is integrated into the design. Finally, numerical simulation results concerning trajectory tracking control of magnetic levitation systems are presented to underscore the effectiveness and practicality of the proposed control methodology.