基于强化学习的多元混合储能聚合协调优化调度OA

To address the power fluctuations and economic dispatch challenges caused by the high penetration of renewable energy in park-level microgrids,this paper proposes a coordinated optimal dispatch method for multi-type hybrid energy storage aggregation based on deep reinforcement learning.The objective is to achieve dynamic and accurate allocation of heterogeneous energy storage resources and reduce the overall operating cost of the system.First,an aggregated system model incorporating flywheel energy storage,lithium-ion batteries,and vanadium redox flow batteries is established.Considering time-of-use electricity pricing,differentiated operation and maintenance costs,and a tiered carbon trading mechanism,an economic dispatch model is formulated with the minimization of the daily total operating cost as the objective.Subsequently,the deep deterministic policy gradient(DDPG)algorithm is introduced to solve the optimization problem.The results show that,the proposed method enables frequency-based cooperative operation among different storage technologies,where flywheel energy storage suppresses high-frequency power fluctuations,vanadium redox flow batteries perform long-duration peak shaving and valley filling,and lithium-ion batteries provide flexible regulation.Compared with the scenario without energy storage,the total daily operating cost of the system is significantly reduced,with the cost-saving rate eventually stabilizing at 23.9％.The results indicate that the proposed method effectively exploits the complementary characteristics of multiple energy storage technologies,improving the economic performance of the park while promoting low-carbon operation.