Optimal VAR compensation with DG penetration using an enhanced Gravitational Search Algorithm

The paper presents an original approach to the VAR compensation problem in distribution networks with Dispersed Generation (DG) penetration, based on an enhanced form of the Gravitational Search Algorithm (GSA). The enhanced GSA proposed in this paper is based on three concepts: elitism, hyper-mutations and a new exploration strategy called Dynamic Supervision (DS). The performances of the proposed method were assessed twofold: on a set of benchmark function similar to those used by the standard GSA and on a test distribution system with radial, tree-like structure, and DG sources, considered in the framework of the VAR compensation problem. The results of the proposed method were compared to those produced by a standard GSA for the first test function, and to those generated by an alternative metaheuristic method.