A Sensor Fusion Based Temperature Estimation Model for Oil-Cooled Windings

This article introduces an innovative method to monitor the average winding temperature in real-time during the operation of an oil-cooled electric motor, particularly suitable for high-performance automotive applications. It is based on the sensor fusion of two distinct dynamic models: one considering the measurement of local temperature through a thermal sensor positioned on the end-winding of the motor, and the other utilizing a first-order model to predict the average winding temperature. The integration of these models significantly enhances the accuracy in estimating the average winding temperature, particularly for short-duty and high current densities.