Flight Regime Mapping for Aircraft Engine Fault Diagnosis

One of the issues that impair the performance of aircraft engine fault diagnosis is the flight regime. When an aircraft travels from one point to another in flight regime, engine performance parameters that are used for fault diagnosing change and such changes mask the parameter changes caused by engine faults, thus make the engine fault diagnosis much more difficult. Properly addressing the flight regime issue is the key in achieving good aircraft engine fault diagnosis. Currently, the flight regime issue is typically addressed by flight regime partitioning. That is, the flight regime is partitioned into several smaller regions and each of the regions is assigned a classifier that is appropriate just for that region. There are several drawbacks associated with this approach. The fundamental one is that it requires the design and implementation of a large number of classifiers, which result in a significant increase of the costs and complexity. In this paper, a novel approach – flight regime mapping is introduced for tackling the flight regime issue in aircraft engine fault diagnosis. The proposed flight regime mapping essentially compensates for flight regime induced parameter changes, thus accentuates the engine condition related changes, by mapping the engine parameter values from the actual flight regime to sea level static equivalent. The mapping enables classifiers that are designed for the sea level static condition to work over the entire flight regime without using multiple classifiers for different regions of flight regime. More importantly, the mapping is able to improve the performance of aircraft engine fault diagnosis. Empirical studies are conducted to demonstrate the effectiveness of the flight regime mapping approach in tackling the flight regime issue in the design of aircraft engine fault diagnostic systems.