Applications of neural networks to the real-time prediction of metal temperatures in gas turbine engine components

Michael Widrich, Alok Sinha, Eva Suarez, Brice Cassenti

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    1 Scopus citations

    Abstract

    Several methods for predicting metal temperatures in a turbine engine are presented. Proper Orthogonal Decomposition (POD) is used to determine the system modes from temperature data sets from an engine mission. The coefficients of the system POD modes are used to identify the system dynamics. The linear state space model in conjunction with a multi-layer feedforward neural network is shown to produce superior prediction values for untrained temperature data when compared to those values produced by the state space model alone.

    Original languageEnglish (US)
    Title of host publicationProceedings of the ASME Turbo Expo 2006 - Power for Land, Sea, and Air
    Pages561-569
    Number of pages9
    Volume2
    DOIs
    StatePublished - 2006
    Event2006 ASME 51st Turbo Expo - Barcelona, Spain
    Duration: May 6 2006May 11 2006

    Other

    Other2006 ASME 51st Turbo Expo
    CountrySpain
    CityBarcelona
    Period5/6/065/11/06

    All Science Journal Classification (ASJC) codes

    • Engineering(all)

    Fingerprint Dive into the research topics of 'Applications of neural networks to the real-time prediction of metal temperatures in gas turbine engine components'. Together they form a unique fingerprint.

  • Cite this

    Widrich, M., Sinha, A., Suarez, E., & Cassenti, B. (2006). Applications of neural networks to the real-time prediction of metal temperatures in gas turbine engine components. In Proceedings of the ASME Turbo Expo 2006 - Power for Land, Sea, and Air (Vol. 2, pp. 561-569) https://doi.org/10.1115/GT2006-90317