CN-122020107-A - Fault signal phase synchronization method and device based on data fusion algorithm

CN122020107ACN 122020107 ACN122020107 ACN 122020107ACN-122020107-A

Abstract

The invention relates to the technical field of data analysis, in particular to a fault signal phase synchronization method and device based on a data fusion algorithm. The method and the device comprise the steps of preprocessing collected data signals of bearing operation, fusing the preprocessed data signals, calculating reliable phase offset estimation of each operation period and position, processing the fused data signals through a waveform matching and phase correction algorithm, eliminating phase offset caused by driving speed and load change, obtaining data signals after phase synchronization, extracting bearing fault characteristics based on the data signals after phase synchronization, and calculating a phase consistency coefficient. According to the technical scheme, through multi-sensor data fusion and an advanced phase synchronization algorithm, the identifiability of fault signal characteristics is improved, and an important technical means is provided for fault diagnosis of rotary machinery.

Inventors

REN ZENGLE
WANG YUAN
YANG TIANYU
SHI YIFAN
WANG PING

Assignees

中国科学院深圳先进技术研究院

Dates

Publication Date: 20260512
Application Date: 20241112

Claims (10)

1. The fault signal phase synchronization method based on the data fusion algorithm is characterized by comprising the following steps of: s101, preprocessing collected data signals of bearing operation; s102, fusing the preprocessed data signals, and calculating reliable phase offset estimation of each operation period and position; S103, processing the fused data signals through a waveform matching and phase correction algorithm, eliminating phase offset caused by driving speed and load change, and obtaining data signals after phase synchronization; And S104, extracting bearing fault characteristics based on the data signals after phase synchronization, and calculating a phase consistency coefficient.
2. The method for phase synchronization of fault signals based on a data fusion algorithm according to claim 1, wherein the method further comprises, prior to step S101: and a plurality of sensors are used for collecting vibration signals of the bearing, and detailed data signals of bearing operation are obtained.
3. The method for synchronizing the phase of fault signals based on a data fusion algorithm according to claim 1, wherein the step S101 includes performing super-resolution sampling on the collected original vibration signals of the bearing operation to improve the time and spatial resolution of the signals.
4. The method for phase synchronization of fault signals based on a data fusion algorithm according to claim 1, wherein step S102 comprises: Constructing a synchronous information matrix containing all operation periods and channel data, and recording relative phase shift and waveform similarity data; And fusing the data signals of the plurality of sensors, and calculating the reliable phase offset estimation of each operation period and position to form a synchronous model.
5. The method for synchronizing the phase of fault signals based on a data fusion algorithm according to claim 4, wherein the windows are moved along the signal time axis in a set step size, and the waveform similarity of each window position is calculated step by step.
6. The method for phase synchronizing fault signals based on a data fusion algorithm according to claim 4, wherein the synchronization information matrix is a high-dimensional matrix, wherein each element specifically records a relative phase deviation under a specific period and channel and an acceleration signal of a corresponding channel.
7. The method for phase synchronization of fault signals based on a data fusion algorithm according to claim 1, wherein in step S102, an internal error between channels in each period is estimated, and data of different channels are fused to form a phase synchronization model.
8. The method for phase synchronization of fault signals based on a data fusion algorithm according to claim 1, wherein in step S103, the signals are subjected to final phase correction by using the calculated relative phase error and channel phase error, and phase synchronization of different channel data and phase synchronization of different running period data are ensured by two interpolation.
9. A fault signal phase synchronization device based on a data fusion algorithm, comprising: The preprocessing unit is used for preprocessing the collected data signals of bearing operation; The data fusion unit is used for fusing the preprocessed data signals and calculating the reliable phase offset estimation of each operation period and position; The phase synchronization unit is used for processing the fused data signals through a waveform matching and phase correction algorithm, eliminating phase offset caused by driving speed and load change and obtaining phase synchronized data signals; And the phase consistency unit is used for extracting bearing fault characteristics of the data signals based on phase synchronization and calculating a phase consistency coefficient.
10. The data fusion algorithm-based fault signal phase synchronization device of claim 9, wherein the device further comprises: And the data acquisition unit is used for acquiring vibration signals of the bearing by using a plurality of sensors and acquiring detailed data signals of bearing operation.

Description

Fault signal phase synchronization method and device based on data fusion algorithm Technical Field The invention relates to the technical field of data analysis, in particular to a fault signal phase synchronization method and device based on a data fusion algorithm. Background Due to its periodic and repetitive movement, the rotary machine structure is particularly prone to problems such as structural wear and cracking under high load conditions over time. These problems are particularly pronounced in bearings, especially in precision equipment such as machine tools and engines. The operational reliability of the bearing structure is critical to production safety, output efficiency and personal safety of the operator. Failure to detect and handle faults in time can lead to equipment downtime, production breaks, and serious safety hazards. Statistics show that more than 30% of rotating machine failures are related to rolling bearing structural problems, which are mainly caused by manufacturing errors, internal lubrication failure and coupling operating force defects. Existing studies utilize a variety of sensors to monitor the operating state of the bearing, including thermal imaging, noise, stator current, vibration, and multisensor fusion. Vibration signals are generated on all components that generate relative motion, such as balls and cages, and contain information on the operating state of each bearing component, and are therefore an important source of information for structural failure analysis. However, the conventional signal feature extraction method is vulnerable to external noise and unstable speed of the driving motor, and thus an advanced data processing method is required to correct raw data and enhance the identifiability of fault features. Time Synchronization Analysis (TSA) exhibits comb filtering characteristics that enable truncated average denoising of the target signal to extract quasi-periodic features from the complex signal. Although this approach has been proposed for over 30 years, TSA has remained widely used in fault diagnosis of multi-stage gear systems and rotating bearing structures in the past decade. Many researchers have been working on improving or applying TSA methods to solve various problems. The learner applies TSA as a data preprocessing method to the maximum second order cyclostationary blind deconvolution (CYCBD) iterative algorithm. TSA is also used to pre-process raw data when bearing failure time series predictions are made using a nonlinear autoregressive exogenous (NARX) model. Although the TSA method performs well in fault diagnosis, the synchronization effect of TSA depends on the time scale calibration provided by the tachometer due to fluctuations in speed and load during operation of the device. Most researchers have installed tachometers in experiments to provide targeted truncation and averaging of the measured vibration signal during analysis. However, in many engineering scenarios, it is not feasible to install a tachometer, which results in the inability to obtain frequency information of interest, particularly high frequency information after TSA processing. Due to continuous fluctuation of speed and load, phase errors generated by direct signal truncation can accumulate, and finally the fault analysis result of the whole signal is affected. Currently, for phase synchronization improvement in TSA algorithms, most researchers use a single index such as pearson correlation coefficient, kurtosis, or cross-power spectrum to quantify similarity and perform phase compensation. Although these methods are helpful in terms of the actual alignment effect, there is room for improvement in application in the actual engineering setting. Disclosure of Invention The embodiment of the invention provides a fault signal phase synchronization method and device based on a data fusion algorithm, which are used for at least solving the phase deviation problem of the fault signal of the existing rotating structure. According to an embodiment of the present invention, there is provided a fault signal phase synchronization method based on a data fusion algorithm, including the steps of: s101, preprocessing collected data signals of bearing operation; s102, fusing the preprocessed data signals, and calculating reliable phase offset estimation of each operation period and position; S103, processing the fused data signals through a waveform matching and phase correction algorithm, eliminating phase offset caused by driving speed and load change, and obtaining data signals after phase synchronization; And S104, extracting bearing fault characteristics based on the data signals after phase synchronization, and calculating a phase consistency coefficient. Further, the method further comprises, before step S101: and a plurality of sensors are used for collecting vibration signals of the bearing, and detailed data signals of bearing operation are obtained. Further, step S101 inclu