Optimization of Shared Energy Storage Capacity Based on GRU Neural Network Prediction

Abstract

With the rising proportion of renewable energy in the electrical power systems, the importance of energy storage as a two-way energy device that can provide rapid response has become increasingly prominent. In this paper, the neural network is used to analyze the historical output data of each equipment in the microgrid, provide a more accurate prediction curve of processing level, and determine the upper limit of energy storage in the next day. Thus, the excess energy storage capacity can be arranged in other business models in advance to obtain benefits. Then a dynamic shared energy storage lease model is proposed to reduce the shared energy storage capacity of microgrid as much as possible. On this basis, a two-level optimization model of shared energy storage capacity allocation based on multi energy unit output is proposed. The upper model is to maximize the benefit of shared energy storage, and the lower model is to minimize the total operating cost. Finally, an example is analyzed on the MATLAB platform to check the feasibility and correctness of the model.