Real-time power forecasting method for photovoltaic power station based on SAGA-FCM-LSSVM model

Abstract

The invention relates to a method for real-time power prediction of photovoltaic power station based on a SAGA-FCM-LSSVM model, which includes collecting power generated in corresponding period of time of photovoltaic power station and corresponding meteorological parameters on meteorological station, and obtaining meteorological data; power parameter samples of the daily weather being pretreated;based on four statistical indexes and simulated annealing genetic algorithm, the fuzzy C-mean clustering algorithm clustering the samples from the first day of the history day to the day before the forecast day. According to the meteorological eigenvalue of each cluster sample set, the center point of each cluster meteorological eigenvalue is calculated, and the classification of the forecast date is judged by Euclidean distance. The least square support vector machine is trained by using the same kind of parameter samples as the predicted date, and the training model is obtained. The meteorological parameters and power values of the first 2 hours of the time to be predicted are input into the training model for real-time prediction of the power generation at each time of the time to be predicted. The invention can predict the output power value of the photovoltaic power station at each time in real time.

Description

technical field [0001] The invention relates to a real-time power prediction method of a photovoltaic power station based on a SAGA-FCM-LSSVM model. Background technique [0002] As an inexhaustible renewable energy, solar energy not only does not consume any earth resources, but also does not pollute the environment. Therefore, photovoltaic power generation has been widely concerned and applied under the attention of countries all over the world. However, the output of photovoltaic power generation is greatly affected by solar irradiance, temperature, humidity and other meteorological conditions. It will lead to a series of safety and stability problems in the power system. Therefore, with the large-scale application of photovoltaic arrays, accurate real-time prediction of photovoltaic power generation has become more and more important. [0003] In recent years, scholars have proposed various photovoltaic power prediction methods, which are mainly divided into two catego...

AI Technical Summary

Problems solved by technology

[0004] But these algorithms have some shortcomings: ANN model can be used for all classification and prediction problems, but it needs to specify various parameters related to network topology in the model, and it is easy to fall into local minimum

[0005] At present, there is no research on applying the SAGA-FCM-LSSVM algorithm to real-time power prediction of photovoltaic power plants in published literature and patents

Images