Troubleshooting of wind turbine gearboxes

Abstract Qinghai Province is rich in wind power resources, actively building wind power generation clean energy highland, large installed capacity, and in the rapid expansion, how to achieve the safe and stable operation of the wind power base is particularly critical. Wind turbine gearboxes operate 24 hours a day under poor working conditions, and key components are prone to failure, accounting for the highest proportion of maintenance costs, so it is of great significance to research intelligent assessment methods and key technologies for the health status of wind turbine gearboxes. A dynamic model of wind-machine-network coupling in wind turbine transmission chain is established to investigate the failure mechanism and dynamic evolution of transmission chain gearboxes under variable wind load. A time-frequency energy concentration algorithm based on a synchronous extraction operator is found to explore the mapping relationship between fault sources and signs. Using a multi-source signal fusion strategy and migration learning method for order spectral analysis, a system fault characteristic parameter identification model is established to implement online monitoring and fault prediction of the wind turbine operation condition, and to improve its operation reliability.