In recent years, the large-scale integration of wind turbines, characterized by strong uncertainty and weak support capability, has posed significant challenges to the frequency security of power systems. To enhance the stable operation capability of power systems with a high proportion of wind power, this paper proposes an optimal energy storage allocation strategy considering frequency security constraints. Firstly, the multi-agent frequency response expression for power systems with a high proportion of wind power is derived, and a dynamic frequency response model for the entire system is established. Secondly, aiming to minimize the annual total cost of the power system, an upper-level optimal energy storage capacity allocation model is formulated. With the objective of minimizing the day-ahead scheduling cost of the system, a lower-level typical daily optimal scheduling model considering frequency security constraints is constructed. An improved particle swarm optimization algorithm is adopted to solve this bi-level model. Finally, a case study is conducted based on a modified IEEE 39-bus system. The research results indicate that this strategy can ensure stable and adequate frequency regulation resources for the power system, effectively improving the economic performance of power system operation while satisfying frequency security constraints.