Data-driven Hierarchical Optimal Allocation of Battery Energy Storage System

The increasing penetration of distributed energy resources (DERs) may cause security and economic risks in the distributed network. In this paper, the optimal allocation of battery energy storage systems (BESS) is proposed to mitigate the risks in the radial distribution network by considering future uncertainties, such as the uncertainties of load and renewable energy. The interacting levels of the proposed hierarchical planning framework are (1) determination of the BESS location based on the calculated adjusted voltage violation risk; (2) obtaining the capacity of the BESS by solving an optimization problem assisted by supervised learning. In the previous works, the steady-state is evaluated by the DistFlow equations in the distribution system. In our paper, we have utilized a data-driven method to calculate the power flow and the voltage, thus leading to higher accuracy. Through case studies, the effectiveness of the proposed method is verified. Furthermore, the data-driven assisted optimization model reduces the computational burden to a large extent because massive state variables, the power flow constraints and voltage constraints are substituted.