Conceptual Development of Strain-Gage Signal Interpretation System for Jet Engine Applications
Abstract
This study sought to define the hardware/software configurations of an expert system capable of monitoring and interpreting strain-gage signals from jet engine compressor stages and immediately taking necessary evasive actions to ensure rig safety. First, a comprehensive strain-gage data bank was generated that contains all major categories of aeromechanical phenomena and noise signals from fan and compressor stages of various engines. These data were analyzed in both time and frequency domains to identify significant signal features for various types of aeromechanic phenomena and noise. These individual signal features were combined with spatial (interblade and interstage), temporal and other syntactic features to form a complete feature set for each aeromechanical phenomenon or noise type. These feature sets were then used in a three level classifier system to determine the probability values of each possible aeromechanical phenomenon and noise type. The first classification uses only the individual strain gage feature to determine the category probability for each phenomena and noise type. The second level uses the spatial features to refine the category pseudo-probability values and yield the event probability values. The third level combines the temporal features with the event probability values to yield the final probabilities based on which decisions are made on the type of actions to be taken to ensure rig safety. Baseline software/hardware system specifications have been defined to acquire strain-gage data, calculate global features, then use these features to determine the predominant aeromechanical phenomena. Up to 100 strain-gage channels are to be incorporated into the system on a modular basis.
Document Details
- Document Type
- Technical Report
- Publication Date
- Oct 01, 1985
- Accession Number
- ADA165489
Entities
People
- Deborah C. Haas
- Philip E. Zwicke
- Ray M. Chi
- William C. Mcclurg
Organizations
- United Technologies Corporation