Three-phase Induction Motor Speed Estimator Using Genetic Algorithm Based on Elman Recurrent Neural Network

In the present work, modeling of a three-phase induction motor is achieved by Recurrent Neural Network (RNN), the three-phase data of input voltage, current, and speed are collected digitally via suitable interface card type National Instrument (NI-PCI-6023E DAQ). These data are stored in a file to be used for training the Recurrent Neural Network which is carried out by the Genetic Algorithm (GA) and it has yielded a good result with an overall error of 0.1 5962%. In addition, the trained Recurrent Neural Network (RNN) is used to substitute the missing speed data due to any reasons of the practical set-up.