Study on improvement of SE core loss model based on Taylor Expansion under variable temperature conditions

Magnetic component loss has become a key factor restricting the further development of power converter products. Currently, research on core materials under varying temperature conditions still has certain limitations, and the prediction accuracy of core loss under temperature variation needs further improvement. To address this issue, this paper analyzes existing core loss models under varying temperature conditions, where current models assume a linear relationship between core loss and temperature, which does not align with reality. Therefore, this paper improves the SE core loss model based on Taylor expansion, assuming that the coefficients in the SE equation for core loss are high-order differentiable functions of temperature. Parameters are solved through nonlinear least squares fitting, ultimately optimizing the SE equation. In repeated core loss prediction comparisons under varying temperature conditions for multiple core materials, the proposed improved model reduces the prediction error by an average of over 85% compared to the SE equation and Model 2, demonstrating the effectiveness and robustness of the proposed method.