Artificial Intelligence-Aided Thermal Model Considering Cross-Coupling Effects

This letter proposes an artificial intelligence-aided thermal model for power electronic devices/systems considering thermal cross-coupling effects. Since multiple heat sources can be applied simultaneously in the thermal system, the proposed method is able to characterize model parameters more conveniently compared to existing methods where only single heat source is allowed at a time. By employing simultaneous cooling curves, linear-to-logarithmic data re-sampling, and differentiated power losses, the proposed artificial neural network-based thermal model can be trained with better data richness and diversity while using fewer measurements. Finally, experimental verifications are conducted to validate the model capabilities.

I. Introduction

Thermal modeling is essential to reliability analysis and thermal management of power electronics [1]. With an increasing request for higher power density and lower parasitics in power electronic systems, a more compact power device packaging with multichips is a trend. One of the challenges that come with it is the thermal cross-coupling (TCC) effects among different semiconductor chips. Junction temperature of a chip is not only affected by its own power losses but also that of neighboring devices. Conventional one-dimensional (1-D) lumped models (e.g., Cauer and Foster models [2]) neglect the TCC effects, and might result in misleading results for multichip modules. Finite-element method can provide detailed thermal results, but it is restrained to a limited time span. Beyond multichip components, a power electronic system consisting of multiple components also has the TCC effects [3], [4].