Research Papers: Gas Turbines: Controls, Diagnostics, and Instrumentation

Constrained Sparse Estimation for Improved Fault Isolation

[+] Author and Article Information
S. Borguet

O. Léonard

 University of Liège, Turbomachinery Group, Campus du Sart-Tilman, B52/3, 4000 Liège, Belgiumo.leonard@ulg.ac.be

J. Eng. Gas Turbines Power 133(12), 121602 (Sep 01, 2011) (8 pages) doi:10.1115/1.4004013 History: Received April 08, 2011; Revised April 09, 2011; Published September 01, 2011; Online September 01, 2011

Least-squares-based methods are very popular in the jet engine community for health monitoring purpose. Their isolation capability can be improved by using a prior knowledge on the health parameters that better matches the expected pattern of the solution i.e., a sparse one as accidental faults impact at most one or two component(s) simultaneously. On the other hand, complimentary information about the feasible values of the health parameters can be derived in the form of constraints. The present contribution investigates the effect of the addition of such constraints on the performance of the sparse estimation tool. Due to its quadratic programming formulation, the constraints are integrated in a straightforward manner. Results obtained on a variety of fault conditions simulated with a commercial turbofan model show that the inclusion of constraints further enhance the isolation capability of the sparse estimator. In particular, the constraints help resolve a confusion issue between high pressure compressor and variable stator vanes faults.

Copyright © 2011 by American Society of Mechanical Engineers
Your Session has timed out. Please sign back in to continue.



Grahic Jump Location
Figure 1

Penalty set by the regularization term in the traditional and sparse least-squares (LS) estimators

Grahic Jump Location
Figure 2

Feasible domain for the health parameters of a turbomachinery component - Part of the third quadrant for a compressor (left), part of the fourth quadrant for a turbine (right)

Grahic Jump Location
Figure 3

Turbofan layout with station numbering and health parameters location

Grahic Jump Location
Figure 4

Signatures of SW26 and VSV on the sensors

Grahic Jump Location
Figure 5

Comparison of the unconstrained and constrained sparse estimate of a hpc fault

Grahic Jump Location
Figure 6

Misclassified LPC faults (left) and HPC faults (right)




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