0

Full Content is available to subscribers

Subscribe/Learn More  >

Benchmarking Diagnostic Algorithms

[+] Author Affiliations
Lokesh Kumar Sambasivan, Joydeb Mukherjee

Honeywell Technology Solutions Lab, Bangalore, India

Dinkar Mylaraswamy

Honeywell Labs, Minneapolis, MN

Paper No. GT2007-28194, pp. 857-864; 8 pages
doi:10.1115/GT2007-28194
From:
  • ASME Turbo Expo 2007: Power for Land, Sea, and Air
  • Volume 1: Turbo Expo 2007
  • Montreal, Canada, May 14–17, 2007
  • Conference Sponsors: International Gas Turbine Institute
  • ISBN: 0-7918-4790-X | eISBN: 0-7918-3796-3
  • Copyright © 2007 by ASME

abstract

The problem of fault diagnosis has gained considerable significance in a cost conscious aerospace industry. This has resulted in the development of novel methods as well as novel approaches. Following a cost benefit analysis, wherein a diagnostic algorithm has to buy its way on an aircraft, the practitioner is faced with the difficult problem of choosing the best algorithm from among many candidates. Despite the appeal of multiple algorithm solutions, practical applications are limited by engineering resources—the practitioner is forced to pick few among many. Specifically, this paper addresses this important engineering decision making step. Our approach to evaluating diagnostic algorithms is based on two key elements—non recurring engineering cost and recurring engineering cost. Corresponding to each of these criterions, we define metrics. Since development data is a major cost element, non recurring engineering cost is derived using a metric that measures how well an algorithm has used this data. Recurring cost is measured with respect to the algorithm’s robustness and hence the cost associated with sustaining it. Further, we outline procedures for calculating these metrics, making minimal assumptions regarding algorithm internals; allowing the practitioner to evaluate both in-house as well as third party algorithms. The utility of this benchmarking procedure is illustrated using two sets of examples. One of them is a standard vowel recognition problem, while the second one is related to gas turbine diagnosis. For each of these problems, we evaluate a series of candidate algorithms and illustrate the utility of the proposed approach to filter out weak ones. Concluding sections discuss the use of these procedures for exiting technical feasibility and entering engineering feasibility on the technology readiness level (TRL).

Copyright © 2007 by ASME
Topics: Algorithms

Figures

Tables

Interactive Graphics

Video

Country-Specific Mortality and Growth Failure in Infancy and Yound Children and Association With Material Stature

Use interactive graphics and maps to view and sort country-specific infant and early dhildhood mortality and growth failure data and their association with maternal

NOTE:
Citing articles are presented as examples only. In non-demo SCM6 implementation, integration with CrossRef’s "Cited By" API will populate this tab (http://www.crossref.org/citedby.html).

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 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