A Quantitative Analysis of Electromagnetic Simulation Model Credibility

The evaluation of an electromagnetic simulation model plays a key role in the field of computational electromagnetics. However, the current validation methods cannot provide quantitative validation results when the simulation model is used to a specific target; therefore, a quantitative analysis needs to be done carefully for a model's credibility. This letter proposes an approach for solving this problem that infers the model's credibility from the calculation results for certain benchmark targets that cover the characteristics of a specific target. Because this inference process has uncertain performance, we introduce an uncertainty reasoning theory to analyze the credibility. A method for the quantitative analysis of an electromagnetic simulation model is created that contains five credibility evaluation factors and constructs two inference rules. Based on this method, the Dempster-Shafer evidence combination rule is used to synthesize integrated belief intervals. Two examples are given to show the feasibility and effectiveness of the proposed approach.