Automatic learning of pulse current shape for torque ripple minimisation in switched reluctance machines

In servo control applications or when smooth control is required at low speeds, torque ripple reduction becomes the main issue for switched reluctance machines. In this paper, the design and experimental evaluation of a novel technique of adjusting the machine currents to minimize its torque ripple is shown. In the proposed technique, a compensating signal, which is based upon a self-tuning neuro-fuzzy system, is added to the PI speed-controller to minimize automatically the ripple. Experimental results are presented to show how the current is modulated reducing torque ripple for different motor speeds and load values.