Data Clustering Method for Probabilistic Power Flow in Microgrids

Abstract

Microgrids, as small-scale power systems, are paving the way for the integration of renewable energy-based distributed resources. As a result, microgrid operators have to deal with uncertainties linked to renewable generation as well as electric load fluctuations. One of the reliable tools for the steady-state analysis of microgrids is probabilistic power flow. In this book chapter, at first, the concept of PPF is introduced briefly via a literature review. Then, the detailed power flow formulation is presented for microgrids with or without reconfigurability characteristics. Moreover, it is explained how to take advantage of different probability density functions, such as Beta, Gaussian, and Weibull distributions to model uncertainties regarding solar photovoltaic generation, electric demand, and wind power generation, respectively. In the next part, the K-means algorithm is presented, and it is explained how this algorithm in combination with the LAPO algorithm can help us to model data clustering-based PPF for microgrid steady-state analysis. Last, but not least, four different case studies are simulated, and the results are visualized and discussed to simplify the learning process.