Bearing state diagnostic method and apparatus, storage medium and electronic device

Abstract

The invention relates to a bearing state diagnostic method and apparatus, a storage medium and an electronic device. The method comprises the steps: acquiring a vibration signal of a bearing; performing local mean decomposition on the vibration signal; extracting a set of dimensionless parameters from the vibration signal subjected to the local mean decomposition as a feature vector of the vibration signal; inputting the feature vector into a preset radial basis function neural network for processing to obtain a set of actual output values; determining the state type of the bearing according to a range to which each value of the set of actual output values belongs. Since the local mean decomposition method is applicable to non-Gaussian, non-stationary and non-linear signals such as the vibration signals of wind generating sets, dimensionless parameters that are not affected by factors such as bearing load and rotational speed are adopted as the feature vector of state diagnosis, analysis is performed in combination with the radial basis function neural network, the diagnosis results can be made more accurate, and the use in practical applications is more convenient and faster.

Description

technical field [0001] The present disclosure relates to the field of bearing detection, and in particular, to a bearing state diagnosis method, device, storage medium and electronic equipment. Background technique [0002] Bearings are widely used in various fields, and bearing structures are required in many mechanical structures, for example, fixed bearings are required in wind turbines. The detection and diagnosis of the health status of the wind power generating set can be judged mainly based on the state of the bearings in the generating set. Usually, the state of the bearing in the unit is obtained by detecting and analyzing the vibration signal of the bearing, so that the state of the wind power generating set can be judged based on the state of the bearing. [0003] For the processing and analysis of vibration signals of bearings, time domain analysis methods are mostly used in the initial stage of analysis and research. However, since the natural characteristics ...

AI Technical Summary

Problems solved by technology

For example, the corresponding wavelet base must be selected in the analysis process, but the wavelet base is not a dynamic change, but a fixed function, so its adaptability is still poor

When the wavelet base is fixed, they cannot be changed during the whole analysis process, which will easily cause the analysis effect of some frequency bands in the vibration signal to be poor

[0006] Therefore, in view of the disadvantages of wavelet analysis, some researchers proposed to use Empirical Mode Decomposition (EMD) to decompose the signal to solve the problem that the signal analysis effect of certain frequency bands in wavelet analysis is not good, but this This method has relatively large interference for a specific signal, and under certain conditions, this method may cause serious consequences such as endpoints and aliasing effects

Images