Improved Neural Network-based High-frequency Behavioral Modeling of Induction Motors

High-frequency (HF) modeling of induction motors plays a key role in predicting motor terminal overvoltage and conducted emissions in a motor drive system. In this paper, a neural network based HF modeling method is proposed, which has the merits of high accuracy and good versatility. The proposed model of the induction motor consists of a three-phase equivalent circuit, which can be expanded by an increment of 7 elements per phase to ensure model accuracy. The motor physics knowledge, namely per phase impedances, are used in the artificial neural network to obtain the values of the circuit elements. The parameterization can be easily implemented using a common personal computer (PC). The case studies verify the effectiveness of the proposed HF modeling method.