An improved virtual synchronous generator power control strategy considering time-varying characteristics of SOC

Abstract

Virtual synchronous generator (VSG) is proposed to solve the lack of inertia of the power system. However, as the core of the VSG providing inertia, the dynamic characteristics of energy storages tend to be ignored in the VSG. In this paper, the dynamic characteristics of energy storages are first considered into the VSG control. Based on this concept, an improved VSG power control strategy considering time-varying characteristics of state of charge (SOC) is proposed in this paper. Firstly, the dynamic characteristics are analyzed in terms of external and output characteristics. Then, Radial Basis Function (RBF) neural network is used to learn data features and handle nonlinear relationships between port voltage and output power of energy storages. The dynamic response of the VSG is improved since the dynamic indexes of the power are optimized during transients. Finally, simulation and experimental results validate the efficacy of the proposed power control strategy.Index Terms—Energy storage, VSG, SOC, RBF neural network, Power control.