A Rough Set-based gas turbine fault classification approach using enhanced fault signatures

摘要

Gas turbine engine health management has become more and more important because of its ability to optimize the total gas turbine operation. Gas path fault classification is one of the most important techniques in gas turbine engine health management. In this article, a Rough Set-based gas turbine fault classification approach is introduced to enhance gas turbine engine health management by taking its advantages in selecting appropriate measurements for fault classification and dealing with uncertainties caused by measurement noise. In the approach, a Rough Set-based knowledge discovery tool is used to find the knowledge hidden in fault samples, and transfer the knowledge into rules representing the logical relationship between the faults and the fault signatures. Such rules can then be used by the Rough Set diagnostic approach to classify faults. Enhanced fault signatures, represented by the measurement deviations and their ranking pattern in terms of their magnitude, are used to make the diagnostic approach more effective. The Rough Set-based diagnostic approach was applied to a model two-spool turbofan gas turbine engine for the classification of single- and dual-component faults. The results show that such Rough Set-based diagnostic approach is able to classify complex-component faults accurately in the presence of measurement noise.