Fourier series based characterization of switched reluctance motor using runtime data

High performance applications of SR motors strictly depend on accurate modeling (or characterization), as well as an appropriate torque control strategy. Most of torque control strategies need torque feedback for realization. Nonlinearity between electrical and mechanical terminals makes simple analytical modeling impossible by means of measured quantities. Therefore, estimation of the motor torque is an important issue to work on. There have been some attempts on motor modeling. However, they have drawbacks due to low accuracy or complexity in realtime implementation. This research concentrates on characterization of an SR motor using runtime data and promises to overcome the problems of recent modeling methods. The algorithm is based on Fourier series and presents accurate and simple static torque computation. The method also enables to derive an estimator for instant torque estimation during realtime operation of the SR motor, which is very important for high performance control.