Wind power prediction method based on forecast wind speed correction and multi-model fusion

Abstract

The invention discloses a wind power prediction method based on forecast wind speed correction and multi-model fusion. The method comprises the steps that data are collected and preprocessed to obtain sample data; respectively calling historical numerical weather forecast data and historical actually measured meteorological data as an input sample and an output sample, and building and training a wind speed correction model; calling factory wind speed-power curve comparison table data of the wind power plant, and building and training a naive wind power prediction model; calling historical actually measured meteorological data and the naive wind power prediction model to form a new sample, and building and training a wind power prediction model based on forecast wind speed correction and multi-model fusion; and calling the numerical weather forecast data, the wind speed correction model and the naive wind power prediction model to form an input sample of the wind power prediction model, and outputting a final wind power prediction value through the wind power prediction model. According to the invention, wind power prediction is carried out based on fusion of the wind speed correction model, the physical model and the statistical model, and the accuracy of wind power prediction is further improved.

Description

technical field [0001] This application relates to the technical field of wind power forecasting, in particular to a wind power forecasting method based on forecasted wind speed correction and multi-model fusion. Background technique [0002] At present, the mainstream short-term wind power prediction methods are mainly divided into two categories. One is the traditional physical model method. etc. are used as model input to predict wind power. The limitations of the physical model algorithm are mainly reflected in the following: it is necessary to comprehensively consider the influence of wind farm topography, surface roughness, and wind turbine wake effects, and many physical parameters are required. Modeling It is more difficult; the factory power curve of the fan has low accuracy because it does not consider the influence of wind direction, humidity, and air pressure, and generally cannot be directly used for prediction; the other is the statistical model method, which i...

AI Technical Summary

Problems solved by technology

[0002] At present, the mainstream short-term wind power prediction methods are mainly divided into two categories. One is the traditional physical model method. etc. are used as model input to predict wind power. The limitations of the physical model algorithm are mainly reflected in the following: it is necessary to comprehensively consider the influence of wind farm topography, surface roughness, and wind turbine wake effects, and many physical parameters are required. Modeling It is more difficult; the factory power curve of the fan has low accuracy because it does not consider the influence of wind direction, humidity, and air pressure, and generally cannot be directly used for prediction; the other is the statistical model method, which is generally based on measured meteorological data and numerical weather data. The forecast and the actual output of wind turbines are used as the basis of modeling data, and the data are learned through artificial intelligence models such as artificial neural networks, support vector machines, and decision trees, and the characteristics and laws of wind power are represented by model parameters. For wind power prediction, the limitations of statistical model algorithms are mainly reflected in the following: certain historical data are required, and a high degree of coupling between model input and wind power is required, otherwise a good model cannot be established

[0003] Both traditional physical models and statistical models show great limitations when applied to wind power forecasting alone, and due to the influence of numerical weather forecast errors, especially the influence of wind speed forecast errors, the existing wind power forecasting algorithms low predictive accuracy

Images