Abstract: Based on the virtual synchronous generator controlled grid-forming energy storage system, which serves as a critical supporting technology for power grids with high penetration of renewable energy, maintaining its grid-connected stability is a core challenge for ensuring system security. This paper focuses on a typical scenario involving a vanadium redox battery- based grid-forming energy storage system connected to the grid, and conducts systematic research on the stability mechanism of electrochemical energy storage systems. First, a loss circuit model of the vanadium redox flow battery is constructed, and on this basis, a small-signal model of the system suitable for transient stability analysis is established. Next, the oscillation modes of the system are analyzed using the eigenvalue method. The results indicate that the system exhibits high-frequency oscillations and sub-synchronous oscillations. Furthermore, root locus analysis is employed to elucidate the influence of parameters such as moment of inertia, damping, and the series-parallel configuration of the battery stack on the system’s oscillation modes. Finally, an oscillation suppression strategy based on a multi-objective weighted architecture using genetic algorithms and adaptive weight adjustment is proposed. The feasibility of this approach is verified through MATLAB/Simulink simulations, providing a theoretical basis for parameter tuning and structural configuration of battery energy storage systems.