A State Space Model for Vibration Based Prognostics

Eric Bechhoefer, Steve Clarck, and David He
Submission Type: 
Full Paper
phmc_10_040.pdf640.22 KBOctober 4, 2010 - 6:53am

Operation and maintenance of offshore wind farms will be more difficult and expensive than equivalent onshore wind farms. Accessibility for routine servicing and maintenance will be a concern: there may be times when the offshore wind farm is inaccessible due to sea, wind and visibility conditions. Additionally, maintenance tasks are more expensive than onshore due to: distance of the offshore wind farm from shore, site exposure, and the need for specialized lifting equipment to install and change out major components.
As a result, the requirement for remote monitoring and condition based maintenance techniques becomes more important to maintain appropriate turbine availability levels. The development of a prognostics health management (PHM) capability will allow a strategy that balances risk of running the turbine versus lost revenue. Prognostics would give an estimate of the remaining useful life of a component under various loads, thus avoiding component failure.
We present a state space model for predicting the remaining useful life of a component based on vibration signatures. The model dynamics are explained and analysis is performed to evaluate the nature of fault signature distribution, and an indicator of prognostic confidence is proposed. The model is then validated under real world conditions

Publication Control Number: 
Submission Keywords: 
State Space
extended Kalman filter
Submitted by: 

follow us

PHM Society on Facebook Follow PHM Society on Twitter PHM Society on LinkedIn PHM Society RSS News Feed