Photovoltaic power generation power prediction method based on long short-term memory neural network

Abstract

The invention discloses a photovoltaic power generation power prediction method based on a long short-term memory neural network. The method comprises: obtaining historical data of photovoltaic power generation power, dividing the data into a training set and a test set by using a cross validation method, and performing normalization processing; substituting the data set into a long short-term memory neural network quantile regression model to obtain the predicted power of photovoltaic power generation; and finally, verifying a prediction result by adopting a root mean square error method. According to the method, the mode of the probability density function is used as a point prediction result, so that the prediction precision can be further improved, and the actual value of the prediction point has a relatively large probability to appear at the mode of the wind power probability density function; and the interval coverage rate meets the expected requirement, the interval average width is minimum, the quantile is minimum, and the superiority of the model in interval prediction compared with a traditional model is verified.

Description

technical field [0001] The invention relates to the field of photovoltaic power generation power prediction, in particular to a photovoltaic power probability prediction method based on a long short-term memory neural network. Background technique [0002] With the increasing demand for electricity in the future society, it can be predicted that the proportion of new energy application technology, especially photovoltaic power generation technology, in the power grid will continue to grow. Photovoltaic power generation has the characteristics of intermittent and random fluctuation. When large-scale photovoltaic power generation is integrated into the power grid, in addition to the inherent randomness and volatility of photovoltaic power generation, which will cause grid fluctuations, the power system itself is in the process of power consumption, power generation and power generation. The imbalance between the three transmission will also cause the grid to fluctuate to a cer...

AI Technical Summary

Problems solved by technology

[0003] At present, scholars at home and abroad have proposed a variety of models to predict photovoltaic power generation, such as the model using the particle swarm optimization algorithm to optimize the BP neural network, the prediction model based on the gray model and support vector machine, but these models have low prediction accuracy, Some shortcomings such as poor generalization and robustness make it difficult to judge the future based on its prediction results

Images