Data Clustering Method for Probabilistic Power Flow in Microgrids

Abstract

Microgrids are paving the way for the integration of renewable energy-based distributed resources. Operators must deal with uncertainties linked to renewable generation and electric load fluctuations. One of the reliable tools for steady-state analysis of microgrids is probabilistic power flow (PPF). In this chapter, the concept of PPF is introduced via a literature review. Then, the detailed power flow formulation is presented for microgrids with or without reconfigurability characteristics. In the next part, the K-means algorithm is presented, and it is explained how this algorithm, combined with the LAPO algorithm, can help to model data clustering-based PPF for microgrid steady-state analysis. Moreover, it describes 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. Last but not least, four different case studies are simulated, and the results are visualized and discussed to simplify the learning process.