Optimizing Chip Area in Power Module Design: Comparison of Traditional and AI Surrogate Models for Thermal Resistance Calculation
Konferenz: PCIM Conference 2025 - International Exhibition and Conference for Power Electronics, Intelligent Motion, Renewable Energy and Energy Management
06.05.2025 - 08.05.2025 in Nürnberg, Germany
doi:10.30420/566541240
Tagungsband: PCIM Conference 2025
Seiten: Sprache: EnglischTyp: PDF
Persönliche VDE-Mitglieder erhalten auf diesen Artikel 10% Rabatt
Autoren:
Patani, Dhruvil; Shri S J, Kaviya; Schneidewind, Eike Alexander; Dr. Orlik, Thomas; Dr. Avakian, Artjom; Dr. Tashakor, Nima
Inhalt:
Optimizing chip area in power module design is crucial for balancing cost, performance and thermal management. This study introduces a methodology for chip area scaling based on junction temperature, calculated using power loss and thermal resistance values. Three different approaches for thermal resistance calculation are investigated: the hyperbolic scaling method, improved analytical method and AI-based surrogate model. The accuracy of each method is validated against FEM simulations. Results indicate that the AI-based surrogate model outperforms the other two methods, reaching ~99.93% accuracy with reduced computational time. The findings highlight the trade-offs between computational efficiency and accuracy, demonstrating that AI-driven models provide an effective alternative to FEM for rapid design iterations.