Robust Control of Induction Motor Based on Fuzzy Neural Networks Generalized Inverse

Supported by National Nature Science Fund (60874014), the paper focuses on generalized linearizing reconstruction control of the AC induction motors. Fuzzy neural networks (FNN) generalized inverse system and robust control methods are applied on the induction motor based on dSPACE system to realize high performance speed regulation.Firstly, on the basis of introduction of inverse system theory, artificial neural networks (ANN), ANN inverse method, analysis of the deficiency of ANN inverse method is given. Then generalized inverse system theory, fuzzy neural networks (FNN), FNN generalized inverse method are introduced, confronted with the problem of model mismatch and disturbance, integrate internal model control(IMC), the methods, the steps, the design principles and the cautions are proposed to fabricate the FNN generalized inverse robust control system.Secondly, using adaptive neuro-fuzzy inference system (called FNN), the FNN generalized inverse system is constructed after analyzing the model of the induction motor and demonstrating its reversibility, by cascading the FNN generalized inverse system with the induction motor system, a generalized pseudo-linear system is completed. Then a robust controller is designed according to the generalized pseudo-linear system, and the specific method and steps of constructing the FNN generalized inverse robust controller are given. Based on MATLAB, simulation program, results and control performance are given.Finally, brief introduction of dSPACE is given, the designed FNN generalized inverse robust controller of the induction motor is validated on the self-designed dSPACE control platform, experiment step, results and control performance are given.Simulation and experiment results show that the reconstructed generalized inverse have open-loop stability and the FNN generalized inverse robust control system has the good static and dynamic performance and strong robustness to load torque load torque disturbance, parametric perturbation, un-modeled dynamics, which manifests robust control based on FNN generalized inverse method is an effective and applied method for the control of induction motor, and also proved the feasibility of integrated application of robust control and FNN generalized inverse method.