Neural network inverse decoupling control of stator flux and torque for induction motor drives

This paper focuses on the development of a stator flux and torque decoupling mechanism based on artificial neural network (ANN) inverse system for induction motor (IM). Firstly, the existence of the inverse system is approved by inverse system theory. However, the analytic inverse model is hardly applied in the engineering fields because it is excessively depends on the IM parameters. Therefore, the method of synthesising neuro-network and analytic function is suggested in this paper, and an ANN-based inverse decoupling control scheme is constructed in order to eliminate the coupling between stator flux and torque. To accelerate the convergence speed of neuro-network and enhance its generalisation ability, the non-linear operations are realised by the analytic operation method and the corresponding results act as the inputs of network. In addition, simulation results are provided to validate the effectiveness of the proposed scheme.