Optimal Design of Energy Storage System Assisted AGC Frequency Regulation Based on DDPG Algorithm

Abstract

In recent years, battery energy storage system (BESS) participating in power system frequency regulation gradually enter people’s view, because it has the characteristics of rapid response to load changes, so they can assist in the output of the active power required for secondary frequency regulation to achieve rapid frequency stabilization. In this paper, a proportional-integral-differential (PID) controller based on the deep deterministic policy gradient (DDPG) algorithm is designed to precisely control the frequency modulation power output of a battery energy storage system. The final simulation results show that the DDPG-PID controller can effectively reduce the number of frequency oscillations, especially under continuous load disturbance, and its control capability is excellent. In this study, a standard two-region interconnected power system load frequency control (LFC) model and a Thevenin equivalent circuit model of the battery energy storage system are first built. the DDPG agent optimizes the controller PID parameters of the LFC model according to the load variations. The frequency regulation performance of the battery storage system controlled by the DDPG-PID controller is superior when tested and compared with the conventional PID control.