Fuzzy clustering based fault diagnosis for aircraft engine health management

Fault diagnosis plays a crucial role in aircraft health management for modern military and commercial aircrafts. Accurate detection and diagnosis of impending aircraft faults can lay the foundation to reduce maintenance turnaround times, operational costs and improve flight safety. Modern aircrafts are capable of generating massive amount of in-flight data and maintenance reports, which makes the task of developing a robust fault diagnosis scheme greatly challenging. Using flight parameters such as exhaust gas temperature (EGT), fuel flow (FF), engine fan speeds (N1 and N2), total air temperature (TAT) decisions can be made on current and future health of aircraft engines. In this paper such flight parameters are used as the basis to develop a diagnostic scheme which can identify a fault and relate this information with the ground reports and maintenance data to allow the maintainer decide necessary maintenance procedures. The baseline values for the in-flight parameters are used as a reference for this evaluation. Any deviation from the baseline values can be considered as a system fault and has to be addressed by the maintenance crew. The data used for this analysis is obtained from flight data recorders. The final decision on a fault being accurately detected is taken by the ground maintenance crew or engineers. Once the fault has been accurately detected and identified fault isolation manuals (FIM) are used to identify necessary maintenance actions required to repair the system or sub-system under fault. A robust fault diagnosis scheme combined with the maintenance actions can give the maintainer enhanced foresight in aircraft system health thus reducing unnecessary maintenance actions.