Modeling and adaptive control of permanent magnet synchronous motors using multilayer neural networks

Abstract

In this study, identification and adaptive tracking control of a permanent magnet synchronous motor (PMSM)-widely used in high-performance drives because of high torque, power density, and power factor-is implemented using Multilayer Neural Networks (MNN). A problem generally encountered in industrial drives is the identification of nonlinear, unknown motor-load dynamics. The purpose of this study is to identify nonlinear motor-load dynamics using MNN, and to control the motor with neural networks so that it can follow an arbitrarily selected reference speed or position. Hence, MNNs are used for the emulation and control of PMSM. The proposed algorithm is simulated using the MATLAB program, and the simulation results show that MNN is more effective than conventional adaptive control methods in nonlinear system modeling and control.