Digital implementation of neural network inverse control for induction motor based on DSP

The induction motor is multi-variable, nonlinear and strong-coupled system. Due to parameters´ variation during operation of induction motor, the decoupling and linearization implemented by field oriented control and analytical inverse control is destroyed. For that, a novel linearization and decoupling method named as artificial neural network (ANN) inverse for induction motor control is proposed. With the combination of neural network and inverse system decoupling control method, the inverse model of induction motor is constructed by neural network and integrator. The neural network inverse system which has good robustness was obtained through reasonable and effective training. A pseudo-linear composite system was obtained by cascading the induction motor and neural network inverse system, and dynamic decoupling of induction motor was achieved. On the basis of dynamic decoupling for induction motor, A neural network inverse control scheme based on DSP for induction motor drive is presented. The hardware structure and software design of the neural network inverse control scheme are described in details. The experiment results show that the proposed scheme has excellent dynamic and static control performance.