Energy Optimal Management of Microgrid With High Photovoltaic Penetration

Abstract

In order to further reduce carbon emissions, a large number of distributed photovoltaics (PVs) are connected to customer sider, which can form microgrids (MGs) with high PV penetration combined with energy storage system (ESS) adopting droop control. Due to the uncontrollability of PV output and frequent charging and discharging of ESS, the economic optimization of MG with high PV penetration is full of challenges, especially island state. Aiming at the lowest daily operating cost, the multi-factor collaborative energy optimization models are established for the grid-connected and islanded MG respectively. Then using particle swarm optimization (PSO) with inertial weight factor to find the optimal solutions of the models under stable operating constraints, the day-ahead energy optimal management strategy (EOMS) for the MG is obtained. In order to reduce the influence of PV and load prediction errors on the energy management accuracy, model predictive control (MPC) is applied to improve the day-ahead EOMS, and intraday rolling horizon energy optimal management strategy (RHEOMS) is obtained. The RHEOMS corrects the forecast errors by feeding back the PV and load current operating value continuously and rolling updating the EOMS control value. The economy and effectiveness of the proposed strategies are verified on a typical MG with high PV penetration.