Abstract: In energy storage power systems, previous research methods often only consider setting the objective function at a single time scale, resulting in significant energy waste. Therefore, this article proposes a distributed control method for frequency regulation of energy storage power systems based on multiple time scales. This article considers the influence of multiple factors to construct a frequency response model, which is equivalent to a single machine system model, and extracts frequency recovery time and maximum frequency deviation as key features. Define sensitivity coefficients based on multiple time scales and calculate the maximum power support that energy storage can provide; introducing real-time adjustment time scale, selecting distributed power output adjustment and energy storage device charging and discharging adjustment as control variables, thus constructing an objective function that considers adjustment cost and frequency stability. Design a distributed controller using the frequency regulation objective function, dynamically optimize the controller parameters, and formulate a distributed control strategy by solving the frequency fluctuation control effect and comprehensively considering multiple factors to achieve the optimal effect of frequency regulation in energy storage power systems. The experimental results show that in frequency fluctuation control, the design method significantly reduces the difference between the maximum and minimum values of the system frequency by 8.2 Hz compared to uncontrolled conditions. At 12 o'clock, the load transfer rate reaches 35%, which is better than other methods. This indicates that it can effectively cope with power fluctuations in the power system and reduce energy waste.