Comparison of the ANN with SVRM Method on Determining the Magnetic Characteristics of the E-Core Transverse Flux Machine

Abstract

The E-core Transverse Flux Machine (ETFM) is combined of transverse flux and reluctance principle. An accurate ETFM model needs a good knowledge of the magnetic characteristics to know its electrical and mechanical behaviors. This paper proposed Support Vector Regression Machine (SVRM) and Artificial Neural Network (ANN) methods for determining of the magnetic characteristics of the ETFM. The data for the training and testing is obtained by experimental measurement methods. To reveal the accuracy of the nonlinear learning methods, the SVRM performance is compared with ANN's.