A strategy for induction motor stator flux estimation using neural networks

Given the importance of the stator magnetic flux in the control of speed and torque of induction machines, this paper aims at presenting an Artificial Neural Network technique for stator flux estimation. A special network configuration is proposed and simulation results are presented. The estimator results are promising in view of the new training strategy presented. An analysis is carried out which shows excellent performance within a spectral frequency range used in several applications.