CN-121977841-A - Bearing fault diagnosis method based on boundary rearrangement Chirplet transformation

CN121977841ACN 121977841 ACN121977841 ACN 121977841ACN-121977841-A

Abstract

The invention discloses a bearing fault diagnosis method based on boundary rearrangement Chirplet transformation, which comprises the following steps of S1 collecting bearing fault signals and rotating speed signals by utilizing an acceleration sensor and an encoder, S2 constructing generalized matching Chirplet transformation to process the bearing fault vibration signals and obtain corresponding time-frequency representation results, S3 searching local amplitude maximum values of time-frequency representation along the time direction and defining a new time rearrangement strategy, S4 designing a compression boundary algorithm based on energy distribution characteristics to reduce environmental noise influence, S5 redistributing energy in a compression boundary according to the time rearrangement strategy, calculating time intervals of two adjacent fault impacts extracted based on the obtained time-frequency representation results, and S6 comparing the time intervals with the time intervals of the fault impacts actually extracted to further judge the occurrence position of the bearing fault. The invention improves the energy aggregation of time-frequency representation, has stronger noise robustness, and provides powerful guarantee for extracting the fault characteristics of the rotary mechanical bearing.

Inventors

ZHAO DEZUN
Du Shuaicong

Assignees

北京工业大学

Dates

Publication Date: 20260505
Application Date: 20251202

Claims (6)

1. The bearing fault diagnosis method based on boundary rearrangement Chirplet transformation is characterized by comprising the following steps of: step S1, setting sampling frequency and sampling time, and collecting bearing fault vibration signals and rotating speed signals by using an acceleration sensor and an encoder; s2, introducing a time matching operator, constructing generalized matching Chirplet, converting and processing a bearing fault vibration signal, and obtaining a corresponding time-frequency representation result; step S3, searching a local amplitude maximum value of the time-frequency representation along the time direction and defining a new time rearrangement strategy; s4, designing a compression boundary algorithm based on the energy distribution characteristics to reduce the influence of environmental noise; S5, redistributing energy in a compression boundary according to a time rearrangement strategy to obtain boundary rearrangement Chirplet transformation, and then calculating the time interval of two adjacent fault impacts extracted based on the obtained time-frequency representation result; and S6, calculating the time interval of theoretical fault impact by using the related parameters and the rotating speed of the experimental bearing, comparing the time interval of theoretical fault impact with the time interval of actually extracted fault impact, and judging the occurrence position of the bearing fault.
2. The method for diagnosing bearing faults based on the boundary rearrangement Chirplet transform according to claim 1, wherein in the step S2, a time matching operator is introduced to construct a generalized matching Chirplet transform: Time matching operator: , Based on the time matching operator, constructing generalized matching Chirplet transformation: And (2) and ; Wherein, the The frequency domain form of the bearing fault signal is represented and obtained through Fourier transformation; a variable of the time is represented and, A variable of the frequency is represented and, Representing imaginary units; represent the first Undetermined coefficients of the frequency interval; Is the first A frequency node; is a gaussian window function in the frequency domain; Representative of Is a complex conjugate of (a) and (b).
3. The method for diagnosing bearing failure based on the boundary rearrangement Chirplet transform according to claim 1, wherein in the step S3, the local amplitude maximum value of the time-frequency representation is searched along the time direction and a new time rearrangement strategy is defined, and the specific expression is as follows: ; Wherein, the Representation of Is used for the mold length of the mold, The energy diffusion section in the time direction is shown.
4. The method for diagnosing bearing faults based on the boundary rearrangement Chirplet transform according to claim 1, wherein in the step S4, the compression boundary algorithm designed based on the energy distribution characteristics can be expressed as: , ; Wherein, the And A left boundary and a right boundary representing a compression boundary, respectively; Representing the window support interval length.
5. The method for diagnosing a bearing failure based on the boundary rearrangement Chirplet transform according to claim 1, wherein in the step S5, the energy in the compression boundary is redistributed according to the time rearrangement strategy, and the expression of the boundary rearrangement Chirplet transform is obtained as follows: ; Wherein, the Is a dirac function.
6. The method for diagnosing bearing faults based on the boundary rearrangement Chirplet transformation according to claim 1, wherein in the step S6, the theoretical fault characteristic frequency of the experimental bearing can be calculated as: , , ; Wherein, the Is the number of rolling elements; , And Respectively representing the diameter of the rolling element, the pitch diameter and the contact angle; The theoretical time interval is the reciprocal of the characteristic frequency of the theoretical fault, and is obtained by the following formula: , , 。

Description

Bearing fault diagnosis method based on boundary rearrangement Chirplet transformation Technical Field The invention belongs to the field of rotating machinery bearing fault diagnosis, and particularly relates to a bearing fault diagnosis method based on boundary rearrangement Chirplet transformation. Background Rotary machines are widely used in industry, aviation, agriculture, etc., wherein rolling bearings are used as key basic components, the operating state of which directly affects the stability and reliability of the device. The bearing defect usually generates a periodic pulse vibration signal, and the early warning and diagnosis of bearing faults can be realized by extracting and analyzing impact components in the vibration signal, so that serious faults and accidents are avoided. The pulse signal has broadband characteristics, and the traditional time domain or frequency domain method cannot effectively represent key information in the signal. The time-frequency analysis technology is widely applied to the field of bearing fault diagnosis because of the capability of simultaneously representing the time and frequency characteristics of pulse signals. In order to improve the performance of the traditional time-frequency analysis method, scholars introduce group delay as a key parameter, and a series of post-processing methods are provided, such as time redistribution multi-synchronous compression transformation, generalized horizontal synchronous compression transformation, transient extraction transformation and the like. However, the time-frequency analysis method based on group delay definition causes a problem of non-reassigned points, which reduces the readability of the time-frequency representation. In addition, the methods are based on short-time Fourier transform results, and the time-frequency resolution of the method still has room for improvement. Therefore, scholars introduce frequency modulation terms based on the traditional time-frequency analysis method and propose Chirplet transformation. As research continues, various improved methods based on Chirplet transform are sequentially proposed, such as frequency domain polynomial Chirplet transform, base match Chirplet transform, and horizontal reassignment frequency domain Chirplet transform. The frequency domain polynomial Chirplet transform can effectively process nonlinear group delay signals by introducing a group delay operator to match the group delay curve. The base match Chirplet transform allows the dispersion to be exactly matched to the group delay trajectory by constructing a new phase kernel function. However, the time-frequency energy aggregation of the two methods needs to be further improved. The horizontal reassignment frequency domain Chirplet transform constructs a group delay selection criterion based on the geometric characteristics of the window function magnitudes, further improving the energy aggregation of the time-frequency representation by preserving only the time-frequency coefficients on the group delay trajectories. The method cannot accurately estimate the group delay track in a strong noise environment, and the noise robustness needs to be further improved. In summary, although the conventional time-frequency analysis method has been widely used for bearing fault diagnosis, it is still required to further improve both energy concentration and noise robustness. Disclosure of Invention Aiming at the defects of the prior art, the invention provides a bearing fault diagnosis method based on boundary rearrangement Chirplet transformation, which can solve the problems of serious energy diffusion and poor noise robustness of the existing time-frequency analysis method and remarkably improve the readability of time-frequency representation and the reliability of impact positioning. In order to achieve the purpose, the technical scheme adopted by the invention is that the bearing fault diagnosis method based on boundary rearrangement Chirplet transformation comprises the following steps: step S1, setting sampling frequency and sampling time, and collecting bearing fault vibration signals and rotating speed signals by using an acceleration sensor and an encoder; s2, introducing a time matching operator, constructing generalized matching Chirplet, converting and processing a bearing fault vibration signal, and obtaining a corresponding time-frequency representation result; step S3, searching a local amplitude maximum value of the time-frequency representation along the time direction and defining a new time rearrangement strategy; s4, designing a compression boundary algorithm based on the energy distribution characteristics to reduce the influence of environmental noise; S5, redistributing energy in a compression boundary according to a time rearrangement strategy to obtain boundary rearrangement Chirplet transformation, and then calculating the time interval of two adjacent fault impacts extracted based on the