Energy management system of low voltage dc microgrid using mixed-integer nonlinear programing and a global optimization technique

Abstract

Integration of a group of distributed generators (DGs) and loads in the form of a microgrid (MG) enhances the supply reliability, reduces the amount of greenhouse gasses (GHG) emission, and reduces the generation cost. However, this integration requires a refined energy management system (EMS). The MG-EMS dispatches the power between DGs to optimize either of the generation cost, the amount of GHG emissions, or both simultaneously. Minimum cost scenario needs an accurate generation cost model for each DG. In this paper, generation cost model includes the no-load cost and nonlinear behavior of losses inside the DG. In addition, the start/stop cost of DGs and power losses within the MG are also considered. The proposed optimization problem becomes a non-convex mixed integer nonlinear (MINL) problem which is hard to solve. The paper suggests using branch and reduce optimization navigator (BARON) algorithm, a global optimization technique. Minimum cost and minimum emission scenarios are single objective problems, while operation at reasonable emission and cost is a multi-objective scenario. Results of the single objective scenarios are exploited to determine feasible constraints to optimize the multi-objective scenario using constraints approach. Eventually, the capability of an energy storage system (ESS) to compensate generation shortage and minimize the emission is also investigated. The study is carried out on a DC microgrid comprised of photovoltaic (PV), fuel cell (FC), microturbine (MT), diesel generator (DE), battery ESS, and Egyptian grid load profiles over four seasons of the year.