Fault Simulation And Diagnosis For Marine Transmission Gear And SSS Clutch

Transmission system of marine gas turbine plant is hub of power transmission, gear andSSS clutch are the two of the marine gas turbine transmission system, once these parts failed,the operation of the power plant will be influenced, even more can’t work, and huge losses arebrought. It has great significance in the fault diagnose relevance research for core componentof marine transmission system. This paper taking the gear and relay SSS clutch as theresearch object, fault simulation and feature extraction and fault diagnosis are launched. Thespecific research contents are as follows:(1) Failure mechanisms analysis of gear and SSS clutch. Characteristic forms of severalgear failure are analysed combined to failure mechanism of gear. The SSS clutch faultmechanism and characteristics are analyzed According to the SSS clutch structure andworking principle.(2) Failure simulation of gear and SSS clutch. Dynamic model of gear is established, themeshing stiffness of normal, crack, broken of gear teeth are simulated, vibration accelerationsignal of gear are simulated on the basis of the meshing stiffness. Dynamic model of the maincomponents of SSS clutch are established, simulated for SSS clutch, the response of clutchengagement process in normal and the ratchet-and-pawl and relay gear and follower gearfaults are abtained.(3) The failure feature research for gear and SSS clutch. Failure feature of vibrationsignal for gear is extracted by the method of vibration signal, wavelet decomposition andHilbert demodulation are used to analyse the fault signal of gear and extract failure feature infrequency domain, and is compare to Fourier transform. The results revealed that the methodof wavelet decomposition and Hilbert demodulation is better to effectively extracted faultfeature than the nomal Fourier transform. Failure feature of relay SSS clutch is analyzed bythe displacement method. The displacement curve of the middleware and relay mechanism ofthe SSS clutch process are researched, normal mesh displacement zone and componentsfunction area of the SSS clutch are divided, and then failure process are compared to normalprocess, fault feature is found.(4) The fault diagnosis method for gear and SSS clutch. Gear is fault diagnosed by theBP neural network. Firstly the time domain and the frequency domain feature which extracted by wavelet decomposition are used to eigenvector, which is normalized. Finally network istrained and diagnosed, diagnosis results revealed that the BP neural network can welldiagnose to gear. The relay SSS clutch’s fault diagnosis method of the expert system isresearched. The fault of relay SSS clutch is diagnosed by knowledge base and diagnosis rulesof expert system for it, Diagnosis results revealed that the expert system can effectivelydiagnose to SSS clutch.