Method, system and device for economic dispatching decision-making of electric power system and medium

Abstract

The invention discloses a power system economic dispatching decision method, system and device and a medium, and the method comprises the steps: obtaining the first operation state data of a power system, and constructing a first operation state diagram according to the first operation state data; constructing a graph reinforcement learning model integrated with the graph convolutional neural network; according to a reward mechanism, training the graph reinforcement learning model by using the first operation state graph; and obtaining second operation state data of the power system, constructing a second operation state diagram according to the second operation state data, inputting the second operation state diagram into the trained diagram reinforcement learning model, and outputting an economic dispatching decision of the power system. According to the method, the operation state of the power system is converted into the graph data in the non-Euclidean space based on the graph theory, the graph convolutional network can effectively aggregate node data in the operation state graph, the graph neural network is introduced to be combined with reinforcement learning, a new economic dispatching decision scheme of the power system is formed, and the method can be widely applied to the power system.

Description

technical field [0001] The invention relates to an electric power system, in particular to a method, system, device and medium for economic dispatching decision-making of the electric power system. Background technique [0002] With the acceleration of the construction of new power systems, the power grid is facing randomness and uncertainty brought about by large-scale distributed power generation. Dispatching operation is the core business of the power system. How to achieve the balance between the power generation and load of the system on the premise of ensuring the safe operation of the system urgently needs better algorithms to improve the quality of decision-making. How to deal with the random decision-making problem of the complex power grid connected with massive distributed new energy is a difficult problem faced by the current power system. [0003] There are mainly two kinds of existing economic dispatching decision-making methods, one is mathematical analysis m...

AI Technical Summary

Problems solved by technology

[0003] There are mainly two kinds of existing economic dispatching decision-making methods, one is mathematical analysis method; as a variety of new energy sources have different characteristics, on the one hand, it brings challenges to system modeling, on the other hand, it makes the economic dispatching model unique Typical characteristics of non-convex nonlinearity, it becomes difficult to solve

Moreover, mathematical analysis methods often rely on forecast data, and the accuracy of future load forecasting will directly affect the quality of the solution.

Another solution method is based on heuristic algorithms such as genetic algorithm and particle swarm algorithm; such algorithms can often obtain better solutions, but it is difficult to obtain the global optimum, and because the algorithm relies on exploration, the solution time is long and the effect is not good. Stablize

Images