The Speed Estimation Of Gear Transmission System Based On Meshing Vibration And Its Application

Rotating machinery is an important class of power machinery, and accounted fora large proportion of mechanical equipment. The safe and reliable operation ofrotating machinery is of great significance to enterprises and the national economy.The rotating machine characteristic analysis is an important part of rotating machineryfault diagnosis and condition monitoring. Order analysis is effectively way to extractvibration characteristics with reference to shaft rotational speed. To achieve constantangle sampling data, Order analysis need to get constant angular increment of time,that is Order Tracking. Traditional Order tracking method need key phase device, noteasy to install hardware limits its application. The speed estimation based on gearmeshing vibration is worthy of further study because gear meshing vibration is closelyrelated to shaft rotates. With the aim of speed estimation of gear transmission, withthe target of gear meshing vibration of gear transmission, the paper proposes themethod of speed estimation from the gear meshing vibration. The theoretical researchand application research are studied in depth, respectively.On the basis of the method of correlation filtering, speed information is estimatedthrough establishing the most similar Morlet wavelet to the meshing vibration. TheMorlet wavelet center frequency parameters can be seen as the instantaneous gearmesh frequency, and then divided by the number of teeth is the shaft rotationfrequency.For traditional Speed Estimation based on time-frequency distribution hasproblems of low accuracy and weak noise immunity, a new speed estimation methodbased on time-frequency fusion distribution is proposed.Time-frequency fusiondistribution is of the fusion of WVD and wavelet scalogram in accordance with the fusion algorithm, which not only eliminate WVD cross-term problem, but alsogreatlyretain the WVD high frequency resolution. Speed information contained in themeshing vibration can be extracted by using peak search and least-squares fit.In order to test the effectiveness of the speed estimation methods, the each speedestimation methods are verified by experiments, the test includes the gear fault andthe rolling bearing fault of the gear transmission system. In this paper, through theresearch on Laplace wavelet correlation filtering and time-frequency fusiondistribution, it is confirmed that both methods were effective in speed estimation formeshing vibration in gear transmission, which is of certain theoretical and practicalvalue for fault diagnosis of rotating machinery.