Skip to main content
APEC 2026 Main banner

Thermal Management

D29.1 - Advanced Online DC-Link Capacitor Thermal Model Incorporating Nonlinearities with Robust Accuracy Under Extreme Load and Environmental Variations

Thursday, March 26, 2026
12:00 PM - 1:45 PM CT
Location: Hemisfair Ballroom
This work proposes an advanced lumped parameter thermal network (LPTN) for accurate temperature estimation of DC-link capacitors, enabling robust overload operation in automotive inverters. The model incorporates a multi-node architecture, an adaptive link to the ambient environment, and temperature dependent ESR to account for system nonlinearities. Validated against test bench data, the model achieves a prediction accuracy within 10 K for critical hotspots, even during extreme overload scenarios.