Photovoltaic power station super-short-term power prediction device

Abstract

The invention discloses a photovoltaic power station super-short-term power prediction device which comprises a meteorological data receiving device used for receiving predicted meteorological data of a photovoltaic power station in the future, a power station monitoring device used for measuring the real-time meteorological data and real-time power of the photovoltaic power station, a historical power generation server used for recording and collecting historical generated power and corresponding real-time meteorological data of the photovoltaic power station, a measured meteorological change system used for analyzing change information of the real-time meteorological data, a power prediction system used for processing the data information of the meteorological data receiving device and the historical power generation server to get a predicted value of power, and a power prediction correction system used for correcting the predicted value of power of the photovoltaic power station. According to the invention, real-time change information of meteorological elements is introduced into the power prediction correction system, and the predicted power of the photovoltaic power station is corrected according to the laws of real-time meteorological change on the prediction day. Thus, the accuracy of photovoltaic power station super-short-term power prediction is improved.

Description

technical field [0001] The invention relates to an ultra-short-time power prediction device for a photovoltaic power station. Background technique [0002] Photovoltaic power station refers to a photovoltaic power generation system that is connected to the grid and transmits power to the grid. As of the end of 2014, the cumulative installed capacity of photovoltaic power stations in my country exceeded 28GW, with a growth rate of over 60%. In 2014, photovoltaic power generation was about 25 billion KWh, a year-on-year increase of more than 200. %. [0003] With the continuous expansion of the scale of photovoltaic power plants in my country, improving the prediction accuracy of photovoltaic power generation will benefit the safe and stable operation of the power grid. The current photovoltaic power station power prediction method mainly adopts the direct prediction method. The direct prediction method uses information such as historical power generation, meteorological measu...

AI Technical Summary

Problems solved by technology

[0004] At present, the output power of the photovoltaic power generation system can be predicted by establishing a prediction model through the above-mentioned multiple linear regression algorithm, artificial neural network (ANN) algorithm, support vector machine (SVM) algorithm, and gray theory algorithm. The resource utilization efficiency of photovoltaic power generation, but these prediction methods are only based on the same historical cycle and the same day type. The disadvantage of the forecast method is that the forecast day is only attributed to a certain day type, such as sunny, and the weather changes and power generation power of photovoltaic power plants with the same historical cycle and the same day type are considered to be similar, but the real-time analysis of various meteorological elements on the forecast day Insufficient processing of change information, for example, when the day type is sunny, sudden cloudy weather is difficult to achieve ultra-short-term accurate prediction

Images