Research on Monitoring Method of Wind Turbine Gearbox Operation Status based on SSA-Informer

Yuzhi Wang

Abstract: In view of the complexity and redundancy of the operating data brought by the harsh environment of the wind turbine, and the difference of the operating state of the wind turbine in the harsh environment. In this paper, singular spectrum analysis is used to decompose the normal operation data with high temperature correlation of the gearbox, and the serial signals of different time scales are obtained. Then, the generation power of the wind turbine is predicted by Informer model, which provides fast and convenient monitoring for the wind turbine gearbox.

Keywords: wind power, forecasting; deep learning, ensemble learning