A Self-training Numerical Method to Calculate the Magnetic Characteristics for Switched Reluctance Motor Drives

摘要

Based on the two-dimensional least squares algorithm, this paper presents a novel numerical method to compute the magnetic characteristics for switched reluctance motor drives. In this algorithm, the coefficients of the orthogonal polynomials, which are ased to describe the magnetic characteristics, are determined by the two-dimensional least squares technique. Furthermore, these coefficients can be determined off line and trained on line, it is shows that the computed and experimental results agree welt The proposed algorithm is very helpful for torque prediction, simulation studies and sensorless control of switched reluctance motor drives.