Magnet temperature estimation based on a novel frequency determination algorithm for the five‐phase PMa‐SynRM

Abstract

This study presents a novel frequency determination algorithm for the magnet temperature estimation of a five-phase permanent magnet assisted synchronous reluctance motor (PMa-SynRM) using the signal injection method. The signal injection method is a convenient and non-invasive approach to determine the magnet temperature in permanent magnet motors by exploiting the temperature dependency of magnet resistance. However, its implementation as a magnet temperature estimation tool warrants an adaptive frequency determination method in order to ensure its effective employability under more pragmatic dynamic operations. The proposed method in this study proposes an analysis of the frequency spectrum to isolate the fundamental and fault frequencies and their respective harmonics to determine a range of available frequencies that may be employed in this method. The available frequencies are further concentrated based on the signal-to-noise ratio constraints. As an additional part of the contribution of this study, an innovative rotating hardware set-up is developed which would transmit the magnet temperature data wirelessly to verify the accuracy of the proposed temperature estimation algorithm.