Trade-Off Between Accuracy and Computational Time for Magnetics Thermal Model Based on Artificial Neural Networks (ANN)
Daniel Santamargarita,Guillermo Salinas,David Molinero,Emilio Bueno,Miroslav Vasic
DOI: https://doi.org/10.1109/jestpe.2022.3203934
IF: 5.462
2022-01-01
IEEE Journal of Emerging and Selected Topics in Power Electronics
Abstract:This paper analyzes the possibility to employ Artificial Neural Networks (ANN) for thermal modeling of magnetic components, with special emphasis on Medium Frequency Transformers (MFTs). The objective is to obtain a MFT thermal model that can be easily used in multivariable optimizations and automated design of power electronics converters, when it is necessary to rapidly explore the complete design space, optimizing the system from the point of view of the efficiency and power density, without losing the accuracy. This can be accomplished using low computational cost ANNs that are trained using the results obtained from thermal Finite Element Method (FEM) simulations. Nevertheless, depending on the approximations that are often applied to speed up the simulations (without modifying the mesh) and, consequently, to reduce the dataset generation time, the precision of the simulations and ANN based models vary significantly, especially in the case when the forced air cooling is used. For example, we demonstrate that different approaches can accelerate the dataset generation (460 simulations) by up to 40 hours, but would introduce an error of up to 40 °C. Therefore, a trade-off of the different alternatives to generate the ANN dataset is analyzed in this paper. The proposed design and optimization approach have been verified using the experimental platform, a Medium Frequency Transformer (MFT), designed and optimized for a 7.5 kW Dual Active Bridge (DAB), designed by brute force optimization using the ANN trained with FEM simulations. The experimental results obtained in different operating points show errors lower than 2.6 °C for the estimated average surface temperature. Additionally, the FEM obtained results are compared with the results obtained by state-of-the-art simplified analytical models and simplified FEM simulations, clearly showing the necessity of highly accurate FEM simulations to generate the training dataset. Finally, we discuss the drawback of the proposed ANN approach in the cases when the ANN based model should be scaled, trying to design a MFT with characteristics outside the range that was used for ANN training. We show that one possible solution for this issue may be ANN re-training, allowing to reduce the error from 50 °C to 5 °C by using only 40 new data points.
engineering, electrical & electronic