Thermal model of permanent-magnet synchronous motor based on support vector machine and fuzzy modeling

This paper presents a thermal model of a certain permanent-magnet synchronous motor over its entire operating range. An experimental setup was designed and a series of experiments have been conducted to obtain the thermal power of the motor on various speeds and torques according to the input-output measurement principle. A regression model of the thermal power was proposed based on the support vector machine algorithm, and a fuzzy inference model for the thermal power calculation was also established based on the T-S fuzzy inference system with the assistant of the operating characteristics of the permanent-magnet synchronous motor. Analysis and comparison of the thermal power results via the two models was developed, both of the two thermal models proposed could predict the thermal power of the PMSM with a high accuracy.