A method for predicting the power generation of photovoltaic power plants

Abstract

The invention discloses a method for predicting the power generation power of a photovoltaic power station, comprising: using the historical meteorological data of the photovoltaic power station to extract six meteorological features in daily units, and establishing a meteorological feature database; Carry out clustering to realize the classification of weather types, and classify the power data and meteorological data of each day; according to the category marking, a SVR sub-model is established for the power data and meteorological data in each category; The weather characteristics of the target day, use SVM to identify the weather type of the target day, select the corresponding SVR sub-model; use the real-time monitoring data of the target day to establish an ARIMA model, and use the rolling forecast method to realize real-time prediction of radiation intensity and temperature; radiation intensity and The predicted value of temperature is input into the selected SVR sub-model to obtain the power prediction result of the photovoltaic power station. The invention improves the prediction accuracy of the generating power of the photovoltaic power station.

Description

technical field [0001] The invention relates to a method for predicting the generated power of a photovoltaic power station, which belongs to the technical field of electric power system automation. Background technique [0002] Photovoltaic power generation, as one of the main technical means of solar energy utilization, has developed rapidly at home and abroad in recent years. By the end of 2015, the cumulative global photovoltaic installed capacity reached 227GW. The accumulative installed capacity of photovoltaic power generation in China has reached 43.18GW, of which the photovoltaic power station is 37.12GW, becoming the country with the largest installed capacity of photovoltaic power generation in the world. At the same time, China's new installed capacity in 2015 was 15.13GW, which is also one of the fastest growing countries in the world. The large-scale development and application of photovoltaic power plants have brought serious impacts on the stability of the ...

AI Technical Summary

Problems solved by technology

[0004] There are two main defects in the current photovoltaic power prediction algorithm. First, the accuracy of photovoltaic power prediction cannot be guaranteed. Relatively obvious volatility, and different weather conditions, the degree of volatility is not the same, the randomness of fluctuations seriously affects the accuracy of photovoltaic power forecasting; second, the time scale of forecasting is long, and it is difficult to achieve real-time forecasting

Images