The Research Of Gearbox Fault Diagnosis Based On EMD And FastICA

Gearbox has the advantages of compact structure,strong bearing capacity,high transmission efficiency,high reliability,and it is an important device for mechanical power transmission system.Whether the working state is normal or not has an important influence on the movement and power transmission.Once an accident occurs,it will have a serious impact.Therefore,it is of great theoretical value and practical significance to research the fault diagnosis of Gearbox Based on EMD and FastICA algorithm.The research content of this paper is the application of a single channel blind source separation method based on EMD and FastICA algorithm in gearbox fault diagnosis.In theory,the vibration mechanism of the gear and the vibration mechanism of the rolling bearing are analyzed,and several typical faults of gear,bearings,shaft and box in gearbox are studied,including broken tooth,tooth wear,gear error,gear eccentricity,outer race fault,inner fault,fault and roller cage fault,shaft misalignment,shaft imbalance,resonance box,setting up a whole of gearbox fault understanding and laying the theoretical foundations for the fault diagnosis of gear box later.In the experimental aspect,on the basis of gearbox dynamic simulation system,artificially setting gearbox in different working conditions(normal,tooth wear,missing tooth fault and multiple fault),through the piezoelectric acceleration sensor and the laser speed sensor transmitting gearbox vibration signal to the HG8916 integrated data acquisition fault diagnosis system,using the time domain data acquisition module of HG8916 data acquisition module to collect the signal and observe the time domain characteristics of the signal,and then collecting the signal by the HG8916 export,converting the signal to TXT format.For a single channel signal,programming in MATLAB,using EMD program to decompose the signal into IMF components,according to modified singular value decomposition method estimating the source number,recombining virtual signals and then using the FastICA algorithm to process the signal,and extracting the singular value of the blind source separation matrix as the signal fault feature,then using BP neuralnetwork to classify and identify the fault diagnosis results.The experiment has achieved good results.Through the experiment we can draw the following conclusions: the single channel blind source separation method based on EMD and FastICA algorithm has the advantages of EMD algorithm and FastICA algorithm,and the principle and calculation is relatively simple,and the operation is convenient,there is no subjective factors,it is very suitable for the application of gearbox fault diagnosis.