Ultrashort-period solar radiation predication method and ultrashort-period solar radiation predication device based on hybrid model

Abstract

The invention discloses an ultrashort-period solar radiation predication method and an ultrashort-period solar radiation predication device based on a hybrid model. The method comprises the steps of periodically acquiring solar radiation observation samples which are acquired by an observation state; obtaining a clearness index time sequence according to the solar radiation observation samples and a solar-radiation-clearness-index relation; performing wavelet transformation processing on the clearness index time sequence, and obtaining a transformed clearness radiation sequence; respectively inputting the clearness radiation sequence into a preset linear predication model, a first support vector machine model and a neural network model, obtaining three input sequences; superposing the three input sequences and inputting the superposed input sequence into a second support vector machine model; and introducing the output result of the second support vector machine model as the predicated clearness index time sequence into the solar-radiation-clearness-index relation, thereby obtaining a predicated solar radiation value. The ultrashort-period solar radiation predication method and the ultrashort-period solar radiation predication device realize integral preset precision reduction by the predication error of a single predication model in a manner of hybrid predication of a plurality of predication models.

Description

technical field [0001] The invention relates to the technical field of photovoltaic power generation, in particular to a hybrid model-based ultra-short-term solar radiation prediction method and a device thereof. Background technique [0002] Affected by the country's policy incentives on photovoltaic power generation in recent years, large-scale photovoltaic power plants have been connected to the grid one after another. Due to the randomness and volatility of the output power of photovoltaic power generation, it has an impact on the safety and stability of the power grid and economic operation. Therefore, it is necessary to accurately predict the output power of photovoltaic power plants, so as to make overall arrangements for the coordination and cooperation of conventional power sources and photovoltaic power generation, so that photovoltaic resources can be fully utilized. [0003] The output power of the photovoltaic power generation system is largely determined by the...

AI Technical Summary

Problems solved by technology

Among them, for the linear model, when the prediction time is long, the correlation between the data is weakened, and the linear model will be difficult to achieve accurate prediction; while the selection of the kernel function of the support vector machine has a great impact on the running results, and there is an optimal neural network. It is difficult to determine the network structure, and both of them may have over-fitting phenomenon during model training

Images