Operational Modal Blind Identification And Fault Diagnosis Of Gearbox

Gearbox fault diagnosis methods based on vibration signals ignores the inherentcharacteristics of system. When it is applied to the non-uniform conditions it has someshortcomings for fault diagnosis. Modal parameters include inherent frequency, modal shapeand damping ratio, which reflects the dynamic characteristics of the mechanical systems.When something is wrong with the parts of gearbox, the dynamic characteristics must bechanged corresponding to the fault. So the Research on fault diagnosis based on modalanalysis has important theoretical meaning and practical value.Independent component analysis is a tool to analyze the signal, which is at the sourceand transport conditions are unknown condition, according to the statistical properties of thesource signal is separated from the mixed signals in the observed signal only from out of theseparate components.Exciting signals no needed, the operational modal analysis is a kind ofblind identification method of the system modal parameters in fact. It has the same ideas withthe independent component analysis method for the blind source signal identification. Theresearch on operational modal parameters blind identification method based on ICA canexpand the application field of the ICA technique.Operational modal parameters of gearbox are blind identified based on ICA technologyusing vibration signals according to theory modal of gear, gearbox and transmission axes inthis paper. The identification results are applied to gearbox fault diagnosis to recognize thefault effectively. In order to verify the validity of the ICA-based blind identification algorithm,a simulation based on virtual prototype is conducted in the system dynamics analysis softwareAdams. The data from simulation results are analyzed using ICA method and compared withthe theoretical analysis. The effectiveness of referred algorithm in this paper are proved bycomparing with the results of polymax in Test.Lab software.