Rotor flux estimation using reactive power and artificial neural networks in indirect field oriented control for induction motor drives

In the indirect field-oriented control of induction motors, the proximity of the estimation of rotor flux plays an important role. Stator resistance affects the flux estimation when the rotor flux is derived from the stator circuit variables. The change in stator resistance at low frequencies leads to errors in the flux estimation and affects the performance of the control structure. In this study, the induction motor flux estimation was performed by using the reactive power of induction motor and Artificial Neural Networks (ANN). The feed-forward training algorithm was used to train the Artificial Neural Network for simulations. It was compared with MRAS (Model Reference Adaptive System) method to show the performance and efficiency of the results. Simulations are implemented in MATLAB / Simulink platform.