Neural network based torque ripple minimisation in a switched reluctance motor

Abstract

This paper presents an artificial neural network (ANN) solution to torque ripple reduction in a switched reluctance motor. Magnetic saturation together with salient stator and rotor poles give rise to a highly nonlinear torque/current/angle characteristic. The approach in this paper allows the neural network to be used to its full potential, that is, learning the nonlinear flux linkage characteristic while also incorporating a priori analytical knowledge of the torque production mechanism of the machine. This combination of neuro-learning and analytical insight results in a greatly simplified controller. Simulation results are presented to illustrate the performance of the proposed technique. Experimental results based on a floating point DSP processor are included.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>