Estimation of PWM-Induced Iron Loss in IPM Machines Incorporating the Impact of Flux Ripple Waveshape and Nonlinear Magnetic Characteristics

Abstract

This article presents an improved estimation model to predict the additional iron loss due to pulsewidth modulation (PWM) switching in interior permanent magnet (IPM) machines. The input current including the predicted PWM-induced current ripple is fed to two-dimensional (2-D) finite-element analysis to calculate the PWM-induced flux ripple distribution in the analyzed machine. A special frequency decomposition method is proposed to separate the predicted flux ripple into 1x and 2x PWM frequency components. Combining these results with the modified dynamic Jiles-Atherton model, the PWM-induced iron loss is calculated incorporating the impact of the actual flux ripple waveshape and the lamination steel's nonlinear magnetic characteristics. Experiments have been conducted to evaluate the prediction accuracy and investigate the impact of several factors on PWM-induced core losses (i.e., both iron and magnet loss) in IPM machines. These results demonstrate that the model offers promising estimation accuracy over a wide range of machine and inverter operating conditions.