A signal adaptive de-envelope method and computing equipment based on optimal signal-to-noise ratio

Abstract

The present invention discloses a signal self adaption envelope solving method based on optimal signal to noise ratio. The method comprises: obtaining signals to be processed comprising fault information and obtained through sampling a preset sampling frequency when a rotary machine performs fault monitoring; constructing a tree-shaped search structure, wherein the frequency band corresponding to each node in the tree-shaped search structure is obtained through calculation according to the preset sampling frequency and the number of each layer of the nodes; performing filtering processing of the signals to be processed according to the frequency band of each node, and obtaining the filtering signals corresponding to each node; performing envelope demodulation processing of the filtering signals, and obtaining the envelope solving signals corresponding to each node and an envelope spectrum thereof; calculating the signal to noise ratio of each envelope solving signal according to the envelope spectrum, and obtaining the signal to noise ratio corresponding to each node; and selecting the envelope spectrum corresponding to the maximum signal to noise ratio as an optimal envelope spectrum to facilitate analysis of the fault frequency and harmonic wave of the signals to be processed. The present invention discloses a calculation device configured to execute the method.

Description

technical field [0001] The invention relates to the technical field of mechanical fault diagnosis, in particular to a signal adaptive de-enveloping method and computing equipment based on an optimal signal-to-noise ratio. Background technique [0002] Rotating machinery is currently widely used in various industrial sites, whether it is gears or bearings, the quality of its operation will directly affect the performance of the entire equipment. Therefore, early fault detection of related equipment is of great significance. According to theoretical analysis, when there is a defect on the working surface of the rolling bearing, a series of broadband impacts will be generated at a certain passing frequency due to the impact of the defect. At the same time, these broadband shocks will excite the entire bearing system to produce corresponding shock attenuation responses, and these shock attenuation responses contain relevant fault information corresponding to the bearings. [0...

AI Technical Summary

Problems solved by technology

[0004] However, the traditional envelope demodulation method has obvious limitations. In the calculation process, we need to determine the filter center frequency and filter bandwidth, and the selection of these parameters requires prior knowledge, and the resonance of the signal in different equipment environments The bands are very different, and the selection of these parameters will ultimately have a great impact on the analysis results.

In the traditional method, one is to use the resonant frequency of the sensor as the center frequency for filtering, but different installation methods and installation environments of the sensor will greatly affect its actual resonant frequency, and the connection method of the sensor and monitoring components will affect the signal. Transfer effect, the resonant frequency of the sensor used in practice is also difficult to obtain accurately

The other is to use the spectral curve peak or spectral kurtosis method to determine the center frequency and bandwidth, but at the position of the spectral curve peak of the signal, sometimes there will be a large number of noise signals, resulting in a low signal-to-noise ratio of the demodulated signal; Using the spectral kurtosis method will introduce two new parameters in the short-time Fourier transform (STFT) process, the window position and the window length. The selection of related parameters is difficult and the amount of calculation is also large.

In addition, in the current manual diagnosis process, diagnostic engineers also need to artificially select different envelope frequency bands to observe the envelope effect, and further give a diagnostic conclusion, which seriously affects the signal processing efficiency

Images