Power grid frequency modulation control method based on combination of data driving and physical model driving

Abstract

The invention discloses a power grid frequency modulation control method based on the combination of data driving and physical model driving. The method comprises the steps of determining a state space set S and a control action set A of a power grid according to a historical frequency modulation sample of the power grid; clustering the elements of the state space set; taking a clustering result as a sample label of a conditional generative adversarial network; training the conditional generative adversarial network; generating a new sample having similar distribution with the historical frequency modulation sample; enhancing the historical frequency modulation sample by using the new sample; introducing a multi-layer perceptron (MLP) to establish a mapping model, using a Q learning controller to control a physical model of the power grid frequency modulation, taking a scheduling decision result of the mapping model as an initial value of the physical model, and outputting an optimal solution of a power grid frequency modulation strategy, i.e., the power grid frequency deviation at a certain moment and the corresponding power regulation quantity, and performing frequency modulationon the power grid. According to the present invention, the generative adversarial network is introduced for data enhancement, and the efficiency of an initial iteration process of an existing model-driven power grid frequency modulation strategy is improved.

Description

technical field [0001] The invention relates to the technical field of power system frequency modulation control, in particular to a power grid frequency modulation control method based on the combination of data driving and physical model driving. Background technique [0002] With the passage of time, a large number of power grid frequency modulation control methods have been used in actual power grids and have accumulated a large number of historical decision-making schemes. These schemes have not only been verified by engineering, but also have been corrected by dispatchers based on actual conditions, and have high engineering application value. . However, the development level of big data technology was relatively low before, and it did not have the ability to deal with massive historical decision-making solutions, so the data-driven frequency modulation method was relatively lacking. In recent years, with the rapid development of artificial intelligence technology and...

AI Technical Summary

Problems solved by technology

[0003] The traditional power grid frequency modulation control method is basically based on model-driven, which has high requirements on the quality of mathematical models and algorithms. The initial value is generally selected randomly, resulting in a long time-consuming algorithm at the initial stage, which is prone to local convergence, and does not make full use of historical data. Decision-making schemes result in a waste of data resources. When encountering problems that have already been processed, it is time-consuming and labor-intensive to perform complete calculations again.

Images