Wind speed forecasting device and method based on heteroscedastic noise twin LSSVR

Abstract

The invention discloses a wind speed forecasting method based on heteroscedastic noise twin LSSVR, and the method comprises the following steps: A, obtaining a wind speed data set D1, with the influence of heteroscedastic noise, of a to-be-forecasted region, and carrying out the calculation to obtain a loss function based on the characteristics of the heteroscedastic noise; b, deriving and solvinga dual problem based on the heteroscedastic noise characteristic twin least square support vector regression on the basis of the original problem of the heteroscedastic noise characteristic twin least square support vector regression; c, determining a penalty parameter and a kernel parameter of a twin least square support vector regression dual problem based on the heteroscedasticity noise characteristic, and selecting a proper kernel function; constructing an upper bound function and a lower bound function based on heteroscedasticity noise characteristic twin least square support vector regression, and finally constructing a decision function; and D, constructing a twinning least square support vector regression wind speed forecasting model based on the heteroscedasticity noise characteristics, and forecasting the wind speed. The defects in the prior art can be overcome, and the wind speed forecasting precision is improved.

Description

technical field [0001] The invention relates to the technical field of short-term wind speed forecasting, in particular to a wind speed forecasting device and method based on heteroscedastic noise twin LSSVR. Background technique [0002] As far as the linear system is concerned, since the Gauss era, the least squares technique has been used to fit the points on the plane to a straight line, and to fit the points in the high-dimensional space to a hyperplane. After more than 200 years of development, the classical least squares technique has become the most widely used technique for data processing in many fields. However, for the ill-posed problems in linear regression or nonlinear regression, the performance of least squares regression technology will become very bad. In view of this situation, many scholars have studied the improved model of least squares regression and proposed many new ones. regression algorithm. Twin least squares support vector regression (Twin Leas...

AI Technical Summary

Problems solved by technology

At this time, if TLSSVR, TLSSVR-GN and other regression techniques are used for forecasting, the

Images