Cooperative frequency control of multi-type energy storage systems based on improved consensus algorithm

• Considering applicability scenarios of diverse energy storage control technologies, fuzzy logic control is employed to determine weighting factors for hybrid control schemes; • Distributed grid control is implemented via multi-agent consensus algorithms; • A consensus-based state observer utilizing adjacent-bus frequency deviations accelerates grid frequency restoration. The high penetration of renewable energy sources, interfaced via power electronic converters, is displacing conventional synchronous generators. This transition reduces system inertia and exacerbates frequency stability challenges, thereby intensifying the demand for fast frequency regulation. Energy Storage Systems present a promising solution to mitigate these issues. This paper proposes a novel two-layer control strategy for grid frequency regulation utilizing a multiple energy storage system (MESS) that synergistically integrates Electrochemical Storage, Supercapacitors, and Flywheels . The strategy is grounded in a Multi-Agent System (MAS) consensus framework. In the primary control layer, a frequency response model is developed, accounting for the distinct dynamic characteristics of each storage type. Specifically, the complementary roles of Virtual Inertia Control (VIC) and Virtual Droop Control (VDC) are analyzed. A fuzzy logic controller is designed to dynamically allocate output power by determining the optimal weighting factors between VIC and VDC in real-time. For the secondary control layer, an improved MAS consensus protocol is introduced. It incorporates a state observer based on adjacent node frequency deviations and a nonlinear coupling term to accelerate convergence and improve robustness. Simulation studies conducted in MATLAB/Simulink validate the effectiveness of the proposed strategy. Results demonstrate a superior performance compared to existing advanced methods: under large disturbances, the maximum frequency deviation is reduced by 0.028 Hz, and the steady-state deviation is minimized by 0.011 Hz. The proposed approach significantly enhances the coordination capability of the MESS and substantially improves grid frequency stability.