A Data-driven Optimal Sizing and Control Methodology for Hybrid Storage System

In this paper, a novel data-driven optimal sizing and control methodology is proposed. The historical operation data of grid is used to calculate the operating characteristic parameters of the local power grid. Based on the characteristic parameters, a typical grid operating curve is fitted by K-Means clustering method. Additionally, an improved filter-based hybrid energy storage system control strategy is presented. Combining the proposed strategy and the typical grid operating curve, the sizing of high energy density storage device is provided while the performance requirement of high power density storage device is calculated. The simulation results based on an East China system proved that the presented methodology improve the renewable energy consumption and the grid peak-shaving ability with the lowest cost of construction.