Peak-current estimation using simplified current-rise model of switched reluctance generator operating in single-pulse mode

Abstract

This paper proposes a simplified current-rise model to estimate the steady state peak-current of the switched reluctance generator for variety of operational conditions such as speed, dc-link voltage, turn-off position and turn-off current level. This model requires machine magnetization characteristics which are obtained from an experimental setup of an 8/6 pole machine in this study. The current rise is estimated in a given generation interval using the turn-off current level and position as the information. This interval can be subdivided into multiple segments for better accuracy, however, this increment in number of sub-interval is beneficial up to certain extent. Interestingly, with the increased number of segments, the proposed model captures the dynamics of the current rise in the generation interval with good accuracy, except near to unaligned position. The proposed model is benchmarked against an existing non-linear model which uses energy-based approach. The model with 1, 5, and 10 segments give maximum error of 20%, 7.1%, and 6.6%, respectively, in the estimation of peak-current. Thus, the proposed model with 5 or 10 segments can be used as a fast alternative to dynamic simulation models for steady state peak-current estimation.