In the context of the development of new power systems dominated by new energy, this article proposes a new concept of using time-varying control methods to achieve real-time control of renewable energy stations. The core lies in introducing time-varying control methods, which can break through the limitations of traditional control methods in quickly and effectively controlling the intermittency and volatility of new energy generation, and solve the problems of inaccurate prediction, difficult prediction, and difficulty in responding to emergencies that exist in traditional control methods. The use of time-varying control can make real-time adjustments based on changes in system parameters, effectively avoiding the above-mentioned problems. Introducing it into the operation of the power system is forward-looking and feasible. Taking the wind farm system as the research object and maximizing the wind turbine output as the goal, this paper establishes a cost function model of wind turbine output, and then proposes a distributed time-varying optimal resource allocation algorithm based on the Lagrange multiplier method. Finally, by building a wind farm simulation model in simulation software, the proposed algorithm is proven to be able to achieve real-time optimal control of wind power generation and minimize the system cost.