Coupling Principal Component Analysis and Kalman Filtering Algorithms for On-Line Aircraft Engine Diagnostics

Engine health monitoring has been an area of intensive research for more than three decades. Numerous methods have been developed with the goal of determining a faithful picture of the engine condition. It becomes largely admitted that a practical implementation of a monitoring tool will rely on an adequate fusion of some techniques. In this framework, the present contribution proposes an original approach for coupling two diagnosis tools in order to enhance the capability of an engine health monitoring system. One tool is based on a Principal Component Analysis scheme and the other one on a Kalman filter technique. The three methodologies are compared and the benefit of the fused tool is demonstrated on simulated fault cases which can be expected on a current turbofan layout. Keywords—Condition monitoring, jet engine, Kalman filter, Principal Component Analysis, Estimation, Information Fusion