Xpe Tube Welding Machine,Xpe Inslulation Pipe Making Machine,Pe Insulation Pipe Welding Machine,Xpe Foam Pipe Bonding Machine Laizhou Jinlida Machinery Co.,Ltd , https://www.jinlidafoammachine.com
Extraction of feature vector of gear wear fault based on continuous wavelet transform. By analyzing the relationship between the wavelet energy spectrum and the degree of wear of the vibration signal under different wear conditions of the gear, the integral value of the wavelet energy spectrum of the vibration signal of the gear between the scales 2 and 18 can be used to indicate the wear condition of the signal. This method can extract the distribution of the energy spectrum of the signal over the feature scale interval without losing the total energy spectrum of the signal over the feature interval. Such a processing method can make the characterization information of the signal wear condition more comprehensive. Because the Euclidean distance is a direct measure of the degree of similarity between vectors, the similarity of the data of the same operating state is high, the distance between the vectors should be small, and the distance between the feature vectors of different operating states should be large.
The Euclidean distance between the feature vectors of the same kind of wear condition signal is small, and the Euclidean distance between the characteristic vectors of different wear conditions is larger. Therefore, extracting the signal feature vector by extracting the energy spectrum of the signal over the feature scale interval can not only characterize the wear condition of the gear well, but also transform the wear condition into a feature vector representation, which enriches the representation information of the sample state. The energy spectrum of the continuous wavelet transform of the vibration signal of the Shandong filter plate at the scale shows the characteristics of the signal in a relatively intuitive manner. Through the continuous wavelet transform, the wear information of the gear can be characterized as the energy spectrum of the gear vibration signal on the scale; the integral value of the energy spectrum of the gear vibration signal can be used to represent the wear condition of the gear. The eigenvector composed of the energy spectrum of the gearbox vibration signal on the scale can be used as a feature vector to characterize the wear condition of the gear. The Euclidean distance between the eigenvectors shows that the eigenvectors extracted by this method can distinguish the signals of the running-in period, the pre-wearing period, the middle-wearing period, the late wear stage and the ultimate wear period. This method provides a new method for in-depth study of dynamic wear of machinery and its components and its online monitoring.