Imperialist competitive algorithm applied to the optimal integration of photovoltaic distributed generation units into a microgrid

Abstract

The present article aims to demonstrate that with an algorithm based on Imperialist Competitive Algorithm (ICA) the distributed generations of an electrical microgrid can be optimized through the connection or disconnection of small photovoltaic generations, having as its main objective to minimize electrical losses. Likewise, it aims to show that with ICA-based better results are obtained than even with the most used metaheuristic algorithms such as PSO. For this, a microgrid based on IEEE 37-bus feeder was modeled, where photovoltaic generations units were connected in 25 of its 37 buses. The modeling of the microgrid was done in OpenDSS. ICA-based and PSO algorithms were developed in MATLAB.Finally, results show that with ICA-based is possible to optimize electric scenarios. Then, although its reduction of electrical losses is as good as PSO one, ICA cloud shows greater accuracy and efficient due to the drastic decrease in computational resources required.