Research Papers

Modeling and Prediction of Gearbox Faults With Data-Mining Algorithms

[+] Author and Article Information
Andrew Kusiak

e-mail: andrew-kusiak@uiowa.edu
Department of Mechanical
and Industrial Engineering,
The University of Iowa,
Iowa City, IA 52242-1527

1Corresponding author.

Contributed by the Solar Energy Division of ASME for publication in the JOURNAL OF SOLAR ENERGY ENGINEERING. Manuscript received January 21, 2012; final manuscript received September 20, 2012; published online March 26, 2013. Assoc. Editor: Christian Masson.

J. Sol. Energy Eng 135(3), 031007 (Mar 26, 2013) (11 pages) Paper No: SOL-12-1023; doi: 10.1115/1.4023516 History: Received January 21, 2012; Revised September 20, 2012

A data-driven approach for analyzing faults in wind turbine gearbox is developed and tested. More specifically, faults in a ring gear are predicted in advance. Time-domain statistical metrics, such as jerk, root mean square (RMS), crest factor (CF), and kurtosis, are investigated to identify faulty components of a wind turbine. The components identified are validated with the fast Fourier transformation (FFT) of vibration data. Fifty neural networks (NNs) with different parameter settings are trained to obtain the best performing model. Models based on original vibration data, and transformed jerk data are constructed. The jerk model based on multisensor data outperforms the other models and therefore is used for testing and validation of previously unseen data. Short-term predictions of up to 15 time intervals, each representing 0.1 s, are performed. The prediction accuracy varies from 91.68% to 94.78%.

Copyright © 2013 by ASME
Your Session has timed out. Please sign back in to continue.



Grahic Jump Location
Fig. 1

Schematic of gearbox used in current study (courtesy of NREL)

Grahic Jump Location
Fig. 2

Sensor locations across the unit (courtesy of NREL)

Grahic Jump Location
Fig. 3

Maximum jerk across 12 vibration sensors

Grahic Jump Location
Fig. 4

Analysis of sensor AN3 and AN4 data: (a) root mean square, (b) crest factor, (c) kurtosis, (d) combined (COM1), and (e) combined (COM2)

Grahic Jump Location
Fig. 5

Analysis of the torque signal: (a) first min, (b) second min, and (c) fourth min

Grahic Jump Location
Fig. 6

Baseline spectrum of a healthy gearbox

Grahic Jump Location
Fig. 7

Power spectrum of vibrations across the ring gear: (a) first min and (b) tenth min

Grahic Jump Location
Fig. 8

Trend of vibration amplitude across the ring gear over the test run (10 min)

Grahic Jump Location
Fig. 9

Run chart of the test results obtained using NN based models: (a) scenario 1, (b) scenario 2, (c) scenario 3, and (d) scenario 4

Grahic Jump Location
Fig. 10

The values of MAE and MRE for different time stamps

Grahic Jump Location
Fig. 11

Test results predicting jerk in ring gear (a) t + 1, (b) t + 9, and (c) t + 15




Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In