Energy Management of Microgrid Considering Different Battery SOC

With wind and solar energy are integrated into power grid especially in microgrid (MG), the randomness nature of renewable energy power generation and load demand bring great difficulties to optimize scheduling of MG. In this work, a hybrid prediction method is put forward to address this problem. Firstly, to improve forecasting accuracy, variational mode decomposition (VMD) is adopted to analyze the input power time-series data. Secondly, each decomposed mode is sent to self-adaptive evolutionary extreme learning machine (SaE-ELM) separately. The final prediction result is obtained by rebuilding the result of every mode. In order to reduce the volatility of renewable energy generate electricity and load, an energy management strategy is proposed. Different from traditional energy management method, the battery SOC is considered when calculating the energy storage cost. Finally, a non-dominated sorting genetic algorithm II (NSGA-II) is applied to solve the multi-objective optimization problem. The simulation results reveal that the proposed energy management strategy is superior to other reported methods in terms of economic and environmental benefits.