Optimal Energy Management Among Multiple Households with Integrated Shared Energy Storage System (ESS)

The integration of artificial intelligence with home energy management systems (HEMS) due to the development of advanced metering infrastructure is a promising scheme to improve the usage of renewable energy in a residential application. In the paper, energy management among multiple co-operative households with PV-Storage integrated generation system in a home micro-grid in the presence of short-term prediction of power generation and consumption is studied. In such a home microgrid system, the central energy storage system (C.ESS) is considered that is connected with multiple household and PV panels. The key parameters that are responsible for optimum scheduling of C.ESS are forecasted PV power generation, forecasted household energy consumption, dynamic state of charge (SOC), and base level of energy consumption. In this paper, firstly, the prediction of short-term generation and consumption based on the long short-term memory (LSTM) algorithm is done. Then, this forecasted data is used as the constraint to the control algorithm for optimum scheduling. Therefore, the amount of power that will be supplied from C.ESS is also determined for properly utilizing the stored energy. The simulation results of the proposed scheme show the robustness and effectiveness in the home microgrid environment.