Data clustering‐based approach for optimal capacitor allocation in distribution systems including wind farms

Abstract

Optimal placement and sizing of shunt capacitors are important issues to improve voltage profile and reduce power losses in distribution networks. In this regard, dealing with stochastic nature of wind farms (WFs), the issue becomes more complex in wind-integrated radial distribution networks. This study aims at optimal siting and sizing of shunt capacitors considering wind uncertainty. Application of probabilistic approach causes to access more real and extensive information about the network. To do so, Monte Carlo simulation (MCS) is employed to assess stochastic variation of wind farm output. As the salient feature of this research, K -means-based data clustering approach is utilised, in which all data-points of the output power of WFs are bunched into desired clusters to ease the solving of simple MCS. The voltage profile and power losses are analysed to do the optimal siting and sizing of capacitors. In order to assess the effectiveness of the proposed method, results are compared with the simple MCS that proves the efficiency of the proposed method from the viewpoint of computation time.