The Use of NSGA - II for Optimal Placement and Management of Renewable Energy Sources When Considering Network Uncertainty and Fault Current Limiters

Abstract

Due to the abundant benefits of renewable energy sources (RESs), their participation in distribution networks is booming. However, they could have adverse effects on the protection coordination schemes. This paper proposes a nondominated sorting genetic algorithm (NSGA-II) that is a multiobjective optimization procedure to obtain the best locations and sizes of renewable energy sources (RESs) with fault current limiters (FCLs), reducing the short-circuit level of buses. The support vector regression, a supervised time series prediction approach in machine learning, is introduced to consider the uncertainty of load demands, network bid changes, and the generated powers of some RESs based on probabilistic states. The efficiency of the proposed procedure is established on the IEEE 33-bus test network.