Wind Power Curve Fitting Method Based on Sparse Heteroscedastic Multivariate Regression

Abstract

The invention provides a wind power curve fitting method based on sparse heteroscedasticity multiple regression, comprising: adopting the fuzzy C-means algorithm to automatically detect abnormal points, obtaining data to remove abnormal points for the original wind power data; according to the acquired data Construct a sparse heteroscedastic multiple regression model; use the variational Bayesian method to optimize the constructed sparse heteroscedastic multiple regression model, and obtain the posterior distribution and parameter formulas of all parameters in the model; initialize the model parameters, according to the model The posterior distribution of all parameters and parameter formulas are used to obtain the estimated value of the parameters by iterative method. The wind power curve fitting method based on sparse heteroscedastic multiple regression provided by the present invention combines multiple spline basis functions, increases the nonlinear fitting ability of the model, and avoids the influence of redundant information on the final regression result .

Description

technical field [0001] The invention relates to the fields of new energy and statistics, in particular to a method for fitting wind power curves based on sparse heteroscedasticity multivariate regression. Background technique [0002] The development and utilization of new energy has become an important way to solve the world's energy shortage and environmental pollution problems. As a clean, environmentally friendly and inexhaustible renewable energy, wind energy has received more and more attention. An accurate wind power curve plays an important role in the wide application of wind energy. [0003] Usually, the fan manufacturers provide corresponding theoretical power curves for the fans they produce. These theoretical power curves are generally obtained at a fixed air density. However, climatic conditions change over time and geographically. Therefore, the performance of the same wind turbine varies in different seasons and in different wind fields. Therefore, it is...

AI Technical Summary

Problems solved by technology

[0007] The present invention provides a wind power curve fitting method based on sparse heteroscedasticity multiple regression, the purpose of which is to solve the two major defects of the existing power curve fitting, resulting in low power curve fitting accuracy and large errors question

Images