Data-driven dynamic modeling method and storage device for photovoltaic power plants

Abstract

The data-driven photovoltaic power station dynamic modeling method and storage device provided by the present invention, the method includes: S10, based on the operation monitoring information of the photovoltaic power station, establishing a database under the normal grid-connected operation state of the photovoltaic power station; S20, establishing the photovoltaic power station The relational equation of input variables and output variables; the input variables include: external light intensity and temperature, and the output variables include: the output voltage and power of the grid-connected point when the photovoltaic power plant is in normal operation; S30, the data in the database is divided into Training set and test set; S40, the convolutional neural network model is trained through the training set, and the trained convolutional neural network model has a dynamic model of the operating characteristics of the photovoltaic power station; S50, the test set is input into the dynamic model, and the voltage and power prediction to obtain prediction results; the invention has the beneficial effect of high reliability and accuracy, and is applicable to the field of photovoltaic power stations.

Description

technical field [0001] The invention relates to the technical field of photovoltaic power plants, in particular to a data-driven photovoltaic power plant dynamic modeling method and a storage device. Background technique [0002] In the process of building a new power system with new energy as the main body, new energy has brought many difficulties to the planning and regulation of the power grid due to factors such as uncertain output and large power fluctuations; in order to achieve effective management of new energy power generation, improve the new Reliability and stability of power system operation, a large number of modeling methods and prediction models for new energy power generation have been proposed in the field of electrical engineering. [0003] Generally, in new energy power generation, the proportion of photovoltaic power plants is gradually increasing. Photovoltaic power generation is a technology that directly converts light energy into electrical energy by ...

AI Technical Summary

Problems solved by technology

[0006] 2. Equivalent modeling of photovoltaic power generation inverter: The starting point of this method is to model the control and working conduction state of the inverter. Generally, the method of switching function modeling or parameter identification is used for modeling, and the control model is established. The relationship equation between the loop control parameters (PI parameters) and the working circuit of the inverter, and the working characteristics of the inverter are expressed by the switching function equation. When power generation is affected by changes in the external environment, its power, voltage, etc. will undergo transient fluctuations, and the working state of the inverter will also change. At this time, the equivalent switching function equation of the inverter is no longer applicable to this situation. There are limitations in engineering applications

[0008] In summary, the photovoltaic power generation system is a complex nonlinear system, and the traditional equivalent modeling method generally linearizes it to obtain a modeling method and model suitable for specific research purposes; however, with the continuous development of new power systems , the power grid's demand for new energy is gradually increasing, and the impact of new energy on the power grid is gradually expanding and deepening. Therefore, traditional modeling methods and models are no longer suitable for engineering research and application

Images