Research On Condition Monitoring And Fault Diagnosis Of Hydraulic Rudder Rotating Machine Of Aircraft

The thesis is mainly about the research on condition monitoring and fault diagnosis ofhydraulic controlling rudder machine of aircraft. The research is done to find a more mature andreliable method for condition monitoring and fault diagnosis of the machine.In this thesis, twofaults are set for researching, damage of turning shaft and turning prevented fault of bearing.Signals of vibration are sampled to analyse and diagnose faults.Wavelet Analysis and Empirical Mode Decomposition are two pop methods for signalanalysis at present. In this thesis, the two methods are used for the issue of condition monitoringand fault diagnosis of hydraulic controlling rudder machine. At first to validate the two methods’feasibility, emulation signals are made and the fault features are picked-up from them. Fault signalsamples are rare ones in praticality for a complex system such as an aircraft. How to improve thefault identification from the rare samples is also an important issue. Support VecorMachine(SVM）is a good measure for rare samples’ identification. In this research, SVM is usedto identify fault features which are picked-up by Wavelet Packet and EMD. From theidentification, the diagnosis precision of the two methods is got. The higher precision one of thetwo methods is selected to use in the physical testing platform. From the research, a method basedon Wavelet Packet for fault diagnosis of hydraulic controlling rudder machine is found. UsingWavelet and EMD to de-noise the signal is another research issue in the thesis.