< Back to previous page
Vibration based condition monitoring of wind turbine gearboxes based on cyclostationary analysis
Book Contribution - Book Chapter Conference Contribution
Wind industry experiences a tremendous growth duringthe last few decades. As of the end of 2016, the worldwidetotal installed electricity generation capacity from wind poweramounted to 486,790 MW, presenting an increase of 12.5%compared to the previous year. Nowadays wind turbinemanufacturers tend to adopt new business models proposingtotal health monitoring services and solutions, using regularinspections or even embedding sensors and health monitoringsystems within each unit. Regularly planned or permanentmonitoring ensures a continuous power generation and reducemaintenance costs, prompting specific actions when necessary.The core of wind turbine drivetrain is usually a complicatedplanetary gearbox. One of the main gearbox componentswhich are commonly responsible for the machinery breakdownsare rolling element bearings. The failure signs of an earlybearing damage are usually weak compared to other sources ofexcitation (e.g. gears). Focusing towards the accurate and earlybearing fault detection, a plethora of signal processing methodshave been proposed including spectral analysis, synchronousaveraging and enveloping. Envelope analysis is based on theextraction of the envelope of the signal, after filtering arounda frequency band excited by impacts due to the bearing faults.Kurtogram has been proposed and widely used as an automaticmethodology for the selection of the filtering band, being onthe other hand sensible in outliers. Recently an emerging Address all correspondence to this author.interest has been focused on modelling rotating machinerysignals as cyclostationary, which is a particular class of nonstationary stochastic processes. Cyclic Spectral Correlation andCyclic Spectral Coherence have been presented as powerfultools for condition monitoring of rolling element bearings,exploiting their cyclostationary behaviour. In this work a newdiagnostic tool is introduced based on the integration of theCyclic Spectral Coherence along a frequency band that containsthe diagnostic information. A special procedure is proposedin order to automatically select the filtering band, maximizingthe corresponding fault indicators. The effectiveness of themethodology is validated using the National Renewable EnergyLaboratory (NREL) wind turbine gearbox vibration conditionmonitoring benchmarking dataset which includes various faultswith different levels of diagnostic complexity.
Book: Proceedings of the ASME Turbo Expo
Number of pages: 10