On-line short-term prediction method of photovoltaic power generation based on dynamic correlation of sky brightness information

Abstract

The invention relates to an online short-term prediction method of photovoltaic power generation based on dynamic association of sky brightness information. Aiming at the uncontrollability of photovoltaic power generation, a new online short-term prediction method of photovoltaic power generation is proposed, which takes into account the direct factors affecting photovoltaic power generation such as solar radiation, clouds, weather pollution and temperature, and changes the traditional model based on complex photovoltaic power generation mechanism. The forecasting idea is different from the classification and matching algorithm based on a large amount of historical data, and also different from the method based on low-precision cloud image collection and movement prediction. The invention directly uses a low-cost industrial wide-angle camera to obtain sky brightness information to characterize the comprehensive state of solar radiation, clouds and weather pollution at the photovoltaic array site. Using the characteristics of dynamic association of sky brightness information, and based on the actual output of photovoltaic arrays collected on site, online short-term prediction of photovoltaic power generation is realized.

Description

technical field [0001] The invention belongs to the field of photovoltaic power generation, relates to image processing technology, and specifically relates to an online short-term prediction method of photovoltaic power generation power based on dynamic correlation of sky brightness information. Background technique [0002] Solar photovoltaic power generation is considered as an important alternative to conventional power generation due to its renewability and zero pollution. However, due to uncontrollable meteorological conditions, photovoltaic power generation cannot provide stable and continuous power supply like traditional power generation methods. It is of great significance to improve the stable operation of the power system including photovoltaic power generation system. [0003] The current photovoltaic power prediction methods mainly include: (1) Construct a photovoltaic power generation model based on solar radiation and the working mechanism of solar cells. T...

AI Technical Summary

Problems solved by technology

The disadvantage of this method is that a large amount of historical data is required, and the objective definition of similar days does not exist, so it is difficult to guarantee the efficiency and accuracy of prediction

(3) Considering the important influence of clouds on solar radiation, there are currently methods to obtain weather conditions from high altitudes through satellite cloud images, but their temporal and spatial resolutions are low; there are also methods to obtain weather conditions based on ground-based cloud images, but ground-based all-sky The imager is expensive, and the image distortion acquired by the fisheye lens has a great impact on the prediction accuracy

Images