Utilizing Hopfield neural networks in the analysis of reluctance motors

Reluctance motors are currently being used widely in different applications. Sometimes, the rotor inherent saliency may introduce some difficulty in pursuing an analytical solution to the motor electromagnetic field problem. In this paper, Hopfield artificial neural networks are used to minimize the air-gap magnetic energy function. Thus, a numerical electromagnetic field solution is obtained automatically. Performance of the motor may then be computed from the obtained field solution. Simulations for a motor having typical dimensions are presented in the paper. It is found that the results of these simulations are in full agreement with reported results as well as well known theoretical aspects.