A performance detection and health assessment method for wind turbine gearbox based on temperature parameters

Abstract

The invention discloses a performance detection and health assessment method of a wind turbine gearbox based on temperature parameters. Through the analysis of the wind turbine state detection data, combined with the characteristics of the wind turbine operation, the model established by combining KECA and GRNN is used to analyze the gearbox temperature and gear The state detection and fault prediction of the gearbox bearing temperature are carried out, and according to the predicted residual, the evaluation decision is made on the operation state of the wind turbine gearbox through the threshold value discrimination and the idea of ​​statistical process control. At the same time, the health status of the gearbox is obtained by comprehensive fuzzy discrimination. Applying this method to an actual wind turbine case with small sample data can achieve accurate prediction and evaluation of the health status of wind turbines. The method of the invention is applied to an actual wind turbine with small sample data, and can realize accurate prediction and evaluation of the health state of the wind turbine.

Description

technical field [0001] The invention is a performance detection and health assessment method applied to wind turbine gearboxes, especially considering the characteristics of a large number of diverse and nonlinear SCADA parameters of wind turbines, and improving the reliability of the gearbox health assessment method; it belongs to data-driven state detection, Fault early warning and health assessment technology field. Background technique [0002] With the continuous increase of human energy requirements and the rapid development of the power industry, the wind power industry has become the main development trend of clean energy due to its relatively low cost, abundant wind resources, and green energy. The maintenance cost after the completion of the wind farm directly determines the benefit of the wind farm. The wind power project is a project with a long investment time, about 7 years, and the profit period is also very long, even more than 10 years. Long-term operation ...

AI Technical Summary

Problems solved by technology

[0003] Considering the large variety and nonlinear characteristics of wind turbine SCADA parameters, there is a lack of general guidelines for the selection and processing of the input parameters of the prediction model

Moreover, the choice of predicted state parameters is single, and the health analysis of the gearbox based on a single state lacks reliability. Therefore, a state assessment method for wind turbine gearboxes based on state detection and prediction models is proposed.

Images