Abstract: To address the frequency random fluctuation caused by high-penetration wind power integrated with power system, a fast frequency regulation strategy was proposed based on moment generating function. Firstly, through collecting the local frequency data at connection points of each wind turbine in wind farm, and obtaining the frequency data by monitoring the dispatch center, the frequency deviation signal at current moment was obtained. Secondly, the probability density function (PDF) of frequency deviation was calculated with kernel density estimation, so as to establish a random distribution model of frequency deviation based on the moment generating function, and performance indicators for frequency control system. Thirdly, an analytical solution for frequency control was obtained after optimization solution, and the stability of frequency control system was analyzed. After which, simulation experiments were conducted on a domestic 1 100 MW wind-thermal hybrid power system. Results show that, under both of load and wind speed disturbance conditions, the proposed fast frequency control strategy can maintain the quasi-steady-state frequency deviation of the points of common coupling and each wind turbine within ±0.02 Hz, while the PDF of frequency deviation exhibits a sharp peak and remains a concentrated distribution around zero. Compared with the traditional droop control, the deviation range of quasi-steady-state frequency is reduced by over 80%, and the nadir of frequency is increased by about 0.05 Hz. Relevant results verify the effectiveness of the proposed strategy in mitigating the frequency fluctuation caused by wind speed and load disturbances.