Abstract: As the penetration rate of renewable energy increases gradually, the inertia support capacity of the power system decreases correspondingly, and scientifically increasing the frequency regulation performance of the power grid becomes an urgent problem to be solved.Therefore, a multi-objective optimization strategy for virtual power plant to participate in automatic generation control（AGC） frequency regulation is proposed.The virtual power plant aggregates electric vehicles and energy storage to respond to the low-frequency and high-frequency components of the deviation between thermal power units and the AGC signal, and establishes a multi-objective optimization model by taking the maximum comprehensive revenue, the optimal energy storage power balance, and the best frequency regulation effect as the objective function and considering the relevant constraints.The optimization results are obtained by using the NSGA-III algorithm and fuzzy set theory.Simulation results show that the strategy proposed in this paper can effectively tap the frequency regulation potential of adjustable resources and improve the grid frequency regulation performance.