Investigation on SOC Estimation Algorithms for VRFB

Abstract

Increasing the use of renewable energy based distributed generation (DG) embedded with energy storage systems (ESS) and smart grids are the recent development trend in power and energy systems. Considering the nature of power fluctuation in the DG systems, certain ESS are necessary in realizing optimal energy management and control of power systems. Of the known batteries, the all vanadium redox flow battery (VRFB) is a chemical energy storage device with many merits, e.g., high application flexibility, high efficiency, re-scalability, fast response, long life, and low maintenance requirements. In practice, the real-time estimation of battery’s state of charge (SOC) plays a very important role in operating smart grid with DG systems. In this paper, a novel SOC estimation method based on neural networks (NN) and the electrochemical impedance spectroscopy (EIS) analysis is proposed for the VRFB. Basic principles of VRFB and existing SOC estimation methods are firstly reviewed, followed by a set of test results demonstrating the feasibility and effectiveness of the proposed NN based on-line detecting algorithm.