Intelligent electric power dispatching system based on reinforcement learning

The invention relates to an electric power system dispatching system based on reinforcement learning, and aims to improve the efficiency and stability of an electric power system. According to the method, by establishing interaction between a reinforcement learning agent and an environment model, an environment state and a scheduling action of a power system are associated, and the environment state and the scheduling action are converted into a Q value through a neural network model. Meanwhile, according to the method, various limiting conditions in the power system are considered, and a constraint optimization method is adopted to process the limiting conditions. Experimental results show that the method has better adaptivity, intelligence and real-time performance, can better adapt to the dynamic change and uncertainty of the power system, and can effectively improve the efficiency and stability of the power system. 本发明涉及一种基于强化学习的电力系统调度系统，旨在提高电力系统的效率和稳定性。该方法通过建立强化学习智能体和环境模型的交互，将电力系统的环境状态和调度动作联系起来，并通过神经网络模型