Intelligent Dispatch of Wind-Thermal Hybrid System Based on Deep Reinforcement Learning Considering Flexible Ramping Capacity Provided by Wind Power

With the continuous improvement of the penetration rate of renewable energy, the flexible ramping capacity of the power system is facing enormous challenges. Relying only on conventional units to provide flexible ramping capacity may result in insufficient ramping capability of units, which will make it difficult to maintain the balance of power supply and demand, and then affect the reliability and economy of power system operation. This paper proposes an intelligent dispatching method for wind-thermal hybrid system based on deep reinforcement learning, considering the flexible ramping capacity provided by wind power. Firstly, the dispatching model considering the flexible ramping capacity provided by wind power is constructed. Then, the dispatching problem is transformed into reinforcement learning task. Based on the deep deterministic policy gradient (DDPG) approach, an intelligent dispatching method of wind-thermal hybrid system considering the flexible ramping capacity provided by wind power is proposed. Finally, the effectiveness of the proposed method is verified with the test data on PMJ 5 bus system. The results show that the proposed method can effectively realize the dispatching of wind-thermal hybrid system with low scheduling cost and can be applied to real-time dispatching decision with short decision time.