Economical Energy Storage Systems Scheduling Based on Load Forecasting Using Deep Learning

Energy storage system is a key device for load-leveling which can shift the load from on-peak time to offpeak time in time-of-use. Customers of the behind-the-meter energy storage system can schedule charging/discharging of energy storage system for electricity cost saving at peak-load. In order to maximize the reduction of electricity cost, smart charging and discharging algorithms based on accurate load forecasting are needed. This paper proposes an energy storage system scheduling algorithm based on water filling optimization followed by short-term load forecasting by using long short-term memory neural network. The proposed method is expected to reduce electricity cost for customers in behind-the-meter by scheduling charging and discharging of an energy storage system. For practical implementation, the satisfaction index of the optimization and the daily electricity cost are compared according to the change of scheduling intervals. Finally, case studies are conducted to confirm the effectiveness of the proposed method.