Smart Grid Operation Using Hybrid AI Systems

Abstract

Battery storage system is an essential component for the depletion of the microgrid, and of the smart grid on a larger scale. It acts as an interface between the highly intermittent renewable energy resource such as solar photovoltaic or wind energy system, the load, such as residential homes and businesses, electric vehicle fleets and the electric power grid. In principle, battery storage is intended to store excess electricity production from the renewable resource. The energy stored is then typically discharged in the evening when demand is higher, or sold to the grid when the available electricity is higher than the demand. However, the implementation of such a system becomes more complicated when taking into account the stochastic nature of the renewables, the variations of the load demand and the electric vehicle availability. In this project, a fuzzy logic management system, optimized through genetic algorithm process, is proposed. It seeks to obtain a fully functional microgrid that minimizes electricity consumption from the utility grid, while offering ancillary services to the grid such as peak shaving. These objectives are accomplished by maximizing the energy input from the renewables and optimizing the utilization of the energy storage system. Finally, the microgrid’s control main objective is to maintain a constant voltage at the DC bus under all operating conditions. Initial simulations and experimental data show some promising results.