Rotation frequency estimation method and detection device based on ridge probability distribution and local fluctuation

Abstract

The invention discloses an instantaneous rotational frequency estimation method based on ridge line probability distribution and local fluctuation characteristics and a detection device thereof. The method comprises the flowing steps that: a low-frequency region is separated from a vibration signal and a resonance frequency band is identified by using a fast spectral kurtosis method to realize frequency band separation, and time-frequency distribution characteristics of the signal low-frequency region and the resonance frequency band are enhanced based on an amplitude accumulation square algorithm; rotational frequency synchronization ridge lines of the signal low-frequency region and the resonance frequency band are searched respectively by using a peak searching algorithm and rotational frequency information is obtained by pre estimation, and then the resonance frequency band is synchronized; with a probability-distribution-based abnormal ridge line zone positioning method, a ridge line failure position is determined to improve a result of fusion of ridge lines obtained by extraction on low-frequency band and resonance envelope signals; and an abnormal zone fusion criterion based on local fluctuation characteristics is established and a standard deviation is introduced as a statistical indicator to indicate the fluctuation characteristic of the ridge line, and a fusion result is evaluated by using the standard deviation as an abnormal data segment fusion indicator to realize accurate rotational frequency estimation.

Description

technical field [0001] The invention belongs to the technical field of fault diagnosis of variable-speed rolling bearings, and in particular relates to an instantaneous rotation frequency estimation method and a detection device based on ridge line probability distribution and local fluctuation characteristics. Background technique [0002] Rolling bearings are the most commonly used general-purpose parts in various rotating machinery and one of the most vulnerable parts. About 30% of the failures of rotating machinery are caused by bearing failures. Therefore, it is extremely important to detect the health status of the bearing. The vibration of bearings under fault excitation often exists in the form of transient features. Effective extraction and research on them can accurately evaluate the operating conditions of bearings, which is the key to bearing fault diagnosis. In the actual operating environment, variable working conditions such as variable speed and variable lo...

AI Technical Summary

Problems solved by technology

It should be pointed out that although some new time-frequency analysis methods can improve the readability of signal time-frequency representation, they will bring problems such as complex algorithm calculation and poor timeliness.

Although time-frequency analysis methods such as short-time Fourier transform and wavelet transform can easily and quickly obtain the time-frequency distribution characteristics of the signal, due to the Heisenberg uncertainty principle and the weak energy of the target time-frequency ridge, the The time-frequency representation obtained by this type of time-frequency analysis method has poor aggregation, and algorithms such as peak search are not accurate enough to extract the conversion frequency information from the time-frequency representation results.

Images