| Summary: | Validation of power system simulation models is essentially a similarity analysis problem based on multivariate time series. With the development of the internet of things (IoT) technology in the power system, the interoperability and integration of devices in the practical project are improved, and the cross interaction in the simulation process becomes more complex correspondingly. It is critical to explore the inherent correlation from the high dimensional data to evaluate the credibility and to locate the error of the simulation model. Thus, a model validation method based on factor analysis and the Prony method is proposed in this paper. Firstly, the multivariate time series of the simulation model and the practical/acknowledged system are replaced by a low number of common factors with physical meanings by factor analysis. Secondly, the modified adaptive Prony method is applied to extract the features of each common factor to ensure the best fitting of the non-stationary signal. Then the complete similarity evaluation model of the simulation system is established based on energy proportion, information entropy, and variance of the contribution rate. Finally, the error location is identified in the evaluation process based on the physical meaning of extracted features. The feasibility and effectiveness of the proposed method are verified by an application in the simulation model of a power electronics system developed in PSASP.
|
|---|
