Inverse System Decoupling Control for Induction Motor Based on Neural Network On-Line Learning

Abstract

The induction motor is a MIMO, nonlinear and high coupling system. The reversibility of the induction motor is testified. Consequently, a pseudo-linear system is completed by constructing a neural network inverse (NNI) system and combining it with the motor system. The inverse can transform the MIMO nonlinear system into two SISO linear subsystems (i.e., rotor speed and flux subsystems). In order to approach the inversion exactly in operation of the motor, the control method online learning based on NNI system is proposed, in which connection value can be amended continuously on-line to make the NN adapt to the changes of environment to strengthen its robustness. Experiment results have shown that NN can be adjusted in the control process. The good applicability of NN along with the strong stability and robustness of the system can be achieved by using the proposed method.