A novel fault diagnosis system for aircraft based on adaboost and five subsystems with different pattern recognition methods


With the development of aerotechnics and requirements of aircraft performance, the system of aircraft becomes multi-functional and more complex. Such a complex system, it is difficult to fault diagnosis by traditional methods, which are performed by expertise, by observing how the system works (generally through unit tests on each component of the complex system). It costs more time and more human resources. Through modern aircrafts generally have a good robust performance and the fault tolerant system can neutralize some faults, it arouses more hidden danger. The small symptoms of faults are covered and it becomes more difficult to detect them by maintenance staff. If they can not be found in time, maybe it leads to some terrible accidents. Consequently, aircraft needs to embed a fault diagnosis system to achieve self-diagnosis and the system needs to be intelligent to analysis different condition of aircraft, give an alarm to pilot in necessary and send a report to engineer and maintenance staff.

in 2012 International Conference on Machine Learning and Cybernetics