Abrupt fault diagnosis of aero-engine based on affinity propagation clustering

Aiming at aero-engine faults, an abrupt fault diagnosis method based on affinity propagation clustering was proposed. Abrupt fault historical monitoring data were used to establish faults database. Through affinity propagation clustering, all the exemplars of abrupt faults in the database were found and the affinity propagation clustering was applied once again to find the exemplar of the new collected data. The fault type was then identified by matching the center with the centers obtained from the faults database. The method was used in the aero-engine abrupt fault diagnosis. The simulation and experiment results show that the method is feasible to diagnose abrupt fault, and compared with other methods, it needs shorter time consuming and produces lower error.