The increasing integration of photovoltaic (PV) power plants into power systems demands a high accuracy of yield prediction and measurement. With this aim, different global horizontal irradiance (GHI) estimations based on new-generation geostationary satellites have been recently proposed, providing a growing number of solutions and databases, mostly available online, in addition to the many ground-based irradiance data installations currently available. According to the specific literature, there is a lack of agreement in validation strategies for a bankable, satellite-derived irradiance dataset. Moreover, different irradiance data sources are compared in recent contributions based on a diversity of arbitrary metrics. Under this framework, this paper describes a characterization of metrics based on a principal component analysis (PCA) application to classify such metrics, aiming to provide non-redundant and complementary information. Therefore, different groups of metrics are identified by applying the PCA process, allowing us to compare, in a more extensive way, different irradiance data sources and exploring and identifying their differences. The methodology has been evaluated using satellite-based and ground-measured GHI data collected for one year in seven different Spanish locations, with a one-hour sample time. Data characterization, results, and a discussion about the suitability of the proposed methodology are also included in the paper.
Coherency Identification in Power Systems Through Principal Component Analysis