A practical framework for managing uncertainties in power device model evaluation

All models are only abstractions of the real systems and their performances are greatly dependent on the assumptions. As different models with different assumptions are introduced, the circuit designers face evaluation problems. In this paper we adopted uncertainty reasoning and the framework of decision-making under incomplete knowledge to solve the problems. The framework produces a range of the model-fit index according to the circuit designers' subjective or a priori specification on the importance of the performance criteria. Case studies were performed for practical applications of the framework. In the case study, it could be observed that the outcome of the model-fit index is more influenced by the order of the criteria than the way they are ranked.