skip to main content

SciTech ConnectSciTech Connect

Title: Gear-box fault detection using time-frequency based methods

Gear-box fault monitoring and detection is important for optimization of power generation and availability of wind turbines. The current industrial approach is to use condition monitoring systems, which runs in parallel with the wind turbine control system, using expensive additional sensors. An alternative would be to use the existing measurements which are normally available for the wind turbine control system. The usage of these sensors instead would cut down the cost of the wind turbine by not using additional sensors. One of these available measurements is the generator speed, in which changes in the gear-box resonance frequency can be detected. Two different time-frequency based approaches are presented in this paper. One is a filter based approach and the other is based on a Karhunen-Loeve basis. Both of them detects the gear-box fault with an acceptable detection delay.
Publication Date:
OSTI Identifier:
Report Number(s):
DOE Contract Number:
Resource Type:
Journal Article
Resource Relation:
Journal Name: Annual Reviews in Control, 40:50-58
Research Org:
Pacific Northwest National Laboratory (PNNL), Richland, WA (US)
Sponsoring Org:
Country of Publication:
United States