Adaptive And Sparsest Time-Frequency Analysis Method And Its Application To Gear Fault Diagnosis

Gears are widely used power transmission units in the machines.Once a gear malfunctions,it will directly affect the healthy operation of a mechanical device,even causing catastrophic accidents.Therefore,it is of significant value to conduct research on gear fault diagnosis.The core of gear fault diagnosis is to extract accurate fault feature information,which depends on the selection of suitable signal decomposition methods.Nowadays,the commonly adopted signal decomposition methods comprise Empirical Mode Decomposition(EMD)and Local Mean Decomposition(LMD).However,as to the forementioned methods,there are still many problems need to be resolved.Therefore,introducing new signal decomposition methods into the research field of gear fault diagnosis is of great importance.Adaptive and Sparsest Time-Frequency Analysis(ASTFA)is a novel method,which can transform the procedure of signal decomposition into the prolem of object optimization.ASTFA method regards the minimum number of mono-components obtained by decomposition procedure as the optimization target,thus realizing adaptive decomposition of signals in the process of object optimization.This dissertation has introduced ASTFA into the research area of gear fault diagnosis and combined ASTFA with Symmetric Difference Energy Operator(SDEO)and Envelope Normalization Demodulation(END),thereby analysing the gear fault signals,extracting fault information and further catching gear faults.This dissertation mainly aims to two research topics,namely theoretical research of ASTFA and applied research of ASTFA in the research filed of gear fault diagnosis.The primary research contents of the dissertation are listed as follows:(1)ASTFA method has been utilized to analyse simulated signals and the effectiveness of ASTFA has been confirmed by analysis results.Furthermore,the dissertation has also revealed two specific decomposition characteristics of ASTFA,namely the characteristic of target decomposition and the characteristic of separating high amplitude ratio multi-component signals.(2)Aimed at the problem that the decomposition capacity of ASTFA could be affected severely by improper values of initial phase function and bandwidth parameter,the dissertation has proposed Moth-Flame Optimization based Adaptive and Sparsest Time-Frequency Analysis(MFO-ASTFA).MFO-ASTFA method adopts Moth-Flame Optimization(MFO)algorithm to hunt for the optimal values of initial phase function and bandwidth parameter.Simulation analysis results have shown the effectiveness of MFO-ASTFA method.(3)Targeted at the modulation characteristic of fault signals of planetary gear sets,the dissertation has presented a fault diagnosis method for planetary gear sets based on MFO-ASTFA and SDEO.The presented method has been used to process fault simulation signals of planetary gear sets and the results have shown its effectiveness.Then the method has been applied to analyse actual vibration signals of local fault of a sun gear in a planetary gear set and the results have proven its practicability.(4)In consideration of the characteristic of amplitude modulation and frequency modulation of fault signals of fixed-axis gears,the dissertation has put forward a fault diagnosis method for fixed-axis gears based on MFO-ASTFA and END.The proposed method has been adopted to analyse the actual fault signals of a gear with root crack and a gear with a broken tooth,the analysis results have proven that the method can diagnose faults of gear root crack and gear tooth breakage accurately and effectively.