Load Frequency Control Strategy of Hybrid Power Generation System: a Deep Reinforcement Learning-Based Approach

To solve the problem of frequency modulation performance degradation caused by large-scale renewable energy access to the power grid, this paper proposes a data-driven load frequency coordinated optimization control method for hybrid energy system consisted of wind, thermal power and energy storage. Firstly, this paper establishes a mathematical model of the multi-area hybrid energy system through mechanism analysis. Secondly, a reward function with control performance standard (CPS), wind power casting and dynamic performance index is established. The load frequency control problem is transformed into a maximum reward function problem, and the deep deterministic policy gradient (DDPG) algorithm is introduced to solve this problem. Through pre-learning and online application, the optimal adaptive coordinated control strategy can be obtained under acturl output of wind turbine. Finally, the performance of the proposed method in improving the performance of load frequency control (LFC) is verified by adding continuous stepped disturbance and actual wind speed disturbance. Simulation results show that when the power system is disturbed, the introduction of energy storage equipment and the proposed method can not only suppress fluctuations effectively, but also shorten the adjustment time required by LFC and increase the proportion of wind power consumption. © 2022, Electrical Technology Press Co. Ltd. All right reserved.