A compensation technique to aid the comparison of experimental and modeled data

The comparison of numerical model data against experimental mental data is common in EMC research and development. However, it is unlikely that the model will be an exact replica of the experimental situation, or vice versa. This is likely to lead to differences in the two sets of results, which complicate the comparison process. This paper presents an algorithm which allows compensation for complex variations between data sets, making the investigation of the level of (dis)agreement more focussed. The proposed technique allows better identification of features that are in disagreement between the results by correcting for artefacts arising, largely, from modelling assumptions. This paper describes the technique and demonstrates its benefits.