Mixed-signal separation method

Abstract

The invention discloses a mixed signal separation method, which comprises the following steps: obtaining a fault signal of a gearbox; The spectrum characteristic parameters are obtained by spectrum analysis of the fault signal. A stationary modulation basis function set is generated according to the spectrum characteristic parameters, the preset fluctuation range and the fluctuation step length. Correlation operation is performed on the stationary modulation basis function set and the fault signal, and the stationary modulation component which is most correlated with the fault signal is extracted. Subtracting the stationary modulation component extracted in the previous step from the fault signal, and recording the difference between the stationary modulation component and the stationary modulation component as a residual term; Comparing the difference between the fault signal and the remaining term, and executing the next step if the difference is smaller than the preset value; Otherwise, the remaining items are taken as fault signals, and the above two steps are repeatedly executed; Adding all the extracted stationary modulation components to obtain a stationary modulation signal; Wavelet transform is used to extract the impulse modulation signal from the remaining terms. The invention is capable of separating the stationary modulated signal mixed together from the shock modulated signal.

Description

technical field [0001] The invention relates to the technical field of gearbox fault diagnosis, in particular to a mixed signal separation method. Background technique [0002] A gearbox is an important part of a mechanical device used to transmit motion and power. Once the gearbox fails, the consequences will be very serious. When faults such as pitting, broken teeth, and fatigue spalling occur in the gearbox, periodic pulse impact force will be generated, resulting in the shock modulation phenomenon of the vibration signal. The modulated vibration signal includes the shock modulation signal. In the frequency spectrum The above shows that there are evenly spaced modulation sidebands on both sides of the meshing frequency or natural frequency. Therefore, most of the current diagnosis methods for gearbox faults are to collect the modulated vibration signals in the gearbox and identify the shock modulation signals. [0003] However, due to the influence of factors such as co...

AI Technical Summary

Problems solved by technology

[0003] However, due to the influence of factors such as cooperation between components and non-completely rigid supports during the operation of some gearboxes, the vibration signal includes another modulation in addition to the shock modulation, called smooth modulation, and the smooth modulation signal It will show similar spectral characteristics to the shock modulation signal on the spectrum, which will interfere with the extraction and identification of the shock modulation signal, making the existing gearbox fault diagnosis methods unable to effectively and accurately extract and identify the shock modulation signal, resulting in gear The result of box fault diagnosis is not accurate enough

Images