Determining Signatures for Energy Mix Produced by Photovoltaic Systems and Wind Turbines

The article contains an analysis of power generation by a photovoltaic system with a peak power of 3 MWp and a wind turbine with a power of 3.45 MW. The acquired time series of generated power was analyzed using traditional and modern analytical methods. The power generated by these two Renewable Energy Sources was characterized separately and then by their mix. In this article, the power signature was defined as the power generated by the photovoltaic system and the wind turbine in the state space over a period of one month. The state space was extracted from the results of cluster analysis. The experiment with clustering was carried out into 10 classes. The K-Means clustering algorithm was used to determine the clusters in a variant without prior labeling of classes with the method of learning without the participation of the teacher. In this way, the trajectories of the power generation process from two Renewable Energy Sources were determined in the 10-state space. Knowing which class each data record belongs to, the frequencies of staying in each state were determined. The computational algorithm presented in the article may have great practical application in balancing the power grid powered by energy produced from renewable sources.