Thermal Analysis of DC-link using Reduced Order Modeling Method

In the field of automotive component evaluation and verification, cost considerations play a pivotal role in the development of electrical devices. To address this, we propose to use virtual release techniques and integrating simulation through model order reduction methods. By doing so, we can efficiently reduce costs and testing durations. In this context, we demonstrate crucial steps involved in defining the entire process, specifying simulation requirements, and implementing this approach. Our focus is on building an inverter model capable of predicting transient temperature behavior of the DC-link with quantified accuracy, achieved through a combination of Finite Volume Method and Reduced Order Modeling methods. The techniques for verification and calibration are addressed. The process of validation using statistical methods is elaborated.