Development of an artificial neural network based induction motor speed estimator

This paper derives two functional relationships between the stator quantities and speed from the induction motor dynamic equations. As singularities are present in these relationships, direct estimation of speed using artificial neural networks (ANNs) with these stator inputs is not possible. This paper proposes a scheme which employs two ANNs to recover motor speed. An added benefit of this scheme is that it can be used quite easily in the presence of an inverter without the need for a smoothing filter for the feedback quantities. This method is general, robust and implementable using commercially available ANN devices