Advanced approach for Static Part of LS Iron Loss Model
Résumé
Over the past two decades, the Loss-surface hysteresis model has been developed to provide an accurate, robust and uncomplicated estimate of iron losses in electromagnetic applications [1]. Its concept is to describe separately the static and dynamic behaviors of materials. Initially, we invented a technique for predicting the dynamic loss contribution and used an approach similar to that of Mayergoyz [2] for the static one. As time goes by, we need a simpler rate-independent model that requires fewer measurements for the parameter identification. As a result, in this research, we propose an improved static model considering simplicity as the most critical criterion while always ensuring the accuracy. One modeling approach of static hysteresis is to build reversal curves of materials which are generally not easy to handle. Instead, we represent mathematically the difference between these curves and the rising or decreasing curve of the major hysteresis loop, which varies monotonically following the induction level. The input data required are the experimental 1st magnetization curve, the major hysteresis loop and a few centered-cycles at lower levels. The number of parameters to identify is flexible but explicit and dependent on the accuracy demand of user. Besides, management of reversal points and history erasure allows the model to maintain the physical sense. In this paper, we will discuss in detail its principle, identification and validation procedure.