Clustering Time Series over Electrical Networks

Daniil Vankov, Ivan Zorin, David Pozo

Abstract: The growing number of renewable energy sources in electrical networks introduces new uncertainties in the electrical network nodes. Reducing the size of electrical networks helps to understand their structure better as well as to plan capacity updates more effectively. The ways of reducing representation of an electrical networks is not a trivial task. In this paper, we consider different methods of clustering of nodal time series data renewable power networks. We propose a clustering method for spatial and temporal data size reduction with local renewable energy as a main driver. The proposed methods are applied to an illustrative 9-bus, 118-bus case studies, and the RE-Europe dataset network.