Iterative Learning-Based High-Performance Current Controller for Switched Reluctance Motors

Abstract

Switched reluctance motors (SRMs) are being considered for variable speed drive applications due to their simple construction and fault-tolerant power-electronic converter configuration. However, inherent torque ripple and the consequent vibration and acoustic noise act against their cause. Most researchers have proposed a cascaded torque control structure for its well-known advantages. In a cascaded control structure, accurate torque control requires accurate current tracking by the inner current controller. As SRM operates in magnetic saturation, the system is highly nonlinear from the control point of view. Developing an accurate current tracking controller for such a nonlinear system is a big challenge. Additionally, the controller should be robust to model inaccuracy, as SRM modeling is very tedious and prone to error. In this paper, we have reviewed various current controllers reported in the literature and discussed their merits and demerits. Subsequently, we have proposed and implemented a novel high-performance current controller based on iterative learning, which shows improved current tracking without the need for an accurate model. Experimental results provided for a 1-hp, 8/6-pole SRM, demonstrate the effectiveness of our proposed scheme.