Validation of electric transportation system simulation models using multiple performance indexes based on scale and slope evaluation

Validation of a simulation model is among other things the evaluation of accuracy against experimental data; quantitative judgment is based on performance indexes that measure various kinds of distances between simulation and data vectors: amplitude, slope, peaks, etc. Different indexes may privilege different characteristics always giving a somewhat biased judgment. For this reason three performance indexes that proved to be effective, robust and reliable are used on a complete test case, which allows verifying which system characteristics may have the most significant impact on validation results. The used performance indexes are Theil, Modified Pendry and FSV (the latter reported in the IEEE Std. 1597.2). The overall judgment of the simulator is “very good”, ranging between good” and “excellent” for various positions and configurations, with the three indexes in substantial agreement, and some minor differences.