Bearing health state evaluation method and system based on high-pass filtering

Abstract

The invention provides a bearing health state evaluation method and system based on high-pass filtering. The method comprises the following steps: acquiring a real-time vibration acceleration signal of a rolling bearing by using data acquisition equipment; the method comprises the following steps: performing fast Fourier transform to obtain a complex sequence containing time amplitude spectrum and phase spectrum information, performing high-frequency band-pass filtering processing in different frequency intervals by using high-frequency band-pass filters of different grades, and further performing inverse Fourier transform on multiple filtered complex sequences to obtain time domain waveform data; and respectively calculating envelope spectrums, calculating peak factors based on a set observation frequency range, selecting a maximum value from the peak factors, and determining the health state of the rolling bearing based on the maximum value in combination with a preset evaluation analysis rule. By adopting the scheme, the problems that the prior art depends on a full-cycle data model and hyper-parameter optimization and is complicated in evaluation operation are solved, the fault frequency characteristics of the envelope spectrum after high-frequency band-pass filtering are quantitatively described through the dimensionless indexes, the calculation complexity is low, and the health state evaluation timeliness is good.

Description

Technical Field

[0001] The present invention relates to the technical field of mechanical state monitoring and evaluation, and in particular to a bearing health state evaluation method and system based on high-pass filtering. Background Art

[0002] Rolling bearings are core components of industrial equipment. How to accurately and reliably evaluate the real-time health status of rolling bearings is the key to predictive maintenance of rotating machinery such as centrifugal pumps and fans. In recent years, with the rapid development of machine learning technology and the emergence of a large amount of industrial data, most related research in the field of rolling bearing operating status monitoring focuses on data-driven rolling bearing health assessment technology or spectrum analysis.

[0003] The data-driven rolling bearing health assessment technology has many restrictions in practical application, such as: comprehensive fault data or full life cycle data training model of the equipment or related equipment is required; and the evaluation and analysis calculation results are greatly affected by physical parameters such as equipment speed, bearing model, structural data, etc., and the model generalization is poor; in addition, the calculation process requires experts to rely on experience and knowledge to set hyperparameters, and rely on the trend of changes in historical data feature values ​​to judge the current health status of the bearing; affected by the scale and accuracy of historical data or intuitive factors set by personnel; the calculation results of this type of technology can only distinguish between normal and abnormal bearing operation, and cannot globally track the health status of the equipment, nor can it consider the impact of poor bearing lubrication on the health status of the bearing. The traditional spectrum analysis method also has the problem of relying on expert experience and knowledge to identify fault feature spectra, and the evaluation data is not accurate enough.

[0004] The information disclosed in the background technology section of the present invention is only intended to deepen the understanding of the general background technology of the present invention, and should not be regarded as acknowledging or suggesting in any form that the information constitutes the prior art known to those skilled in the art. Summary of the invention

[0005] To solve the above problems, the present invention provides a bearing health status assessment method based on high-pass filtering, which quantitatively describes the fault frequency characteristics of the envelope spectrum after high-frequency band-pass filtering through dimensionless indicators, has low computational complexity, and good timeliness of health status assessment, and overcomes the problems of the prior art that it relies on a full-cycle data model, relies on hyperparameter optimization, and has complicated assessment operations; preferably, in one embodiment, the method includes: a field signal acquisition step: using a data acquisition device to collect a real-time vibration acceleration signal of a rolling bearing on site;

[0006] The first transformation operation step: performing a fast Fourier transform on the collected vibration acceleration signal data to obtain a complex sequence corresponding to the vibration acceleration signal, wherein the complex sequence includes time amplitude spectrum and phase spectrum information;

[0007] Filtering processing step: based on the complex sequence, high-frequency bandpass filters of different levels are used to perform high-frequency bandpass filtering processing in different frequency intervals to obtain corresponding multiple filtered complex sequences;

[0008] The second transformation operation step: performing inverse Fourier transformation on the multiple filtered complex sequences respectively to obtain multiple corresponding time domain waveform data;

[0009] Envelope spectrum peak index calculation steps: calculate the corresponding envelope spectrum for each time domain waveform data, and calculate the peak factor of each envelope spectrum based on the set observation frequency range, and select the maximum value from the peak factors of each envelope spectrum as the bearing health status assessment index representing the bearing health status information;

[0010] Bearing status assessment step: determining the health status result of the target rolling bearing based on the value of the bearing health status assessment index combined with preset assessment and analysis rules.

[0011] Optionally, in one embodiment, in the field signal acquisition step, a vibration sensor whose distance from the target bearing meets set requirements is used to acquire real-time vibration acceleration signal data of the rolling bearing according to set sampling configuration parameters.

[0012] Furthermore, in one embodiment, the method further comprises:

[0013] Noise reduction optimization step: Before the bearing status evaluation step, the collected vibration acceleration signal is subjected to noise reduction processing, and the corresponding signal-to-noise ratio is calculated based on the vibration acceleration signal and the noise signal before noise reduction, and then the signal-to-noise ratio is applied to optimize the calculated bearing health status evaluation index, and the optimized bearing health status evaluation index is used as the final bearing health status evaluation index.

[0014] Preferably, in one embodiment, in the field signal acquisition step, the vibration sensor sampling frequency f s Not less than 20480Hz, sampling length n not less than 16384.

[0015] Specifically, in an optional embodiment, in the filtering processing step, high-frequency bandpass filters C1, C2, C3, and C4 are used based on the complex sequence to perform bandpass filtering at 0.9k~1.5kHz, 1.5k~5.5kHz, 3.5k~7.5kHz, and 5.5k~10kHz, respectively.

[0016] Furthermore, in one embodiment, in the envelope spectrum peak index calculation step, the Hilbert transform is used to calculate the corresponding envelope spectrum for each item of time domain waveform data.

[0017] Optionally, in one embodiment, in the envelope spectrum peak index calculation step, the peak factor of the 0-1 kHz spectral line range is calculated for each envelope spectrum. Among them, the peak factor of the envelope spectra {f1, f2, f3, f4} is calculated according to the following formula:

[0018] F1 = max(f1) / rms(f1)

[0019] F2 = max(f2) / rms(f2)

[0020] F3 = max(f3) / rms(f3)

[0021] F4 = max(f4) / rms(f4)

[0022] In the formula, f i is the envelope spectrum corresponding to the i-th item of time-domain waveform data, max() represents calculating the maximum value in the 0-1 kHz spectral line of the envelope spectrum, and F i is the peak factor of the 0-1 kHz spectral line range of the i-th envelope spectrum, and rms() represents calculating the root mean square value within the 0-1 kHz spectral line of the envelope spectrum.

[0023] In a preferred embodiment, in the bearing condition evaluation step, the health state result of the target rolling bearing is determined according to the following evaluation and analysis rules:

[0024] If the bearing health state evaluation index Hv ≤ 2, it is determined to be in a good state. At this stage, the bearing and its lubrication state are good, and continue to operate;

[0025] If 2 < Hv ≤ 20, it is determined to be in a sub-healthy state. At this stage, the bearing and its lubrication state are poor. Replace the lubricating oil and monitor the operation. If the Hv value does not improve after replacing the lubricating oil, replace the bearing;

[0026] If Hv > 20, it is determined to be in a risk state, and replace the bearing.

[0027] In one embodiment, in the noise reduction and optimization step, the fixed threshold wavelet method is used to denoise the vibration acceleration signal to obtain the denoised vibration acceleration signal and the noise signal.

[0028] Furthermore, in one embodiment, in the noise reduction and optimization step, the process of calculating the signal-to-noise ratio includes:

[0029] Calculate the power of the original vibration acceleration signal and the power of the noise signal before the noise reduction process respectively;

[0030] Then calculate the signal-to-noise ratio according to the following formula:

[0031] SNR = 10lg(Power signal / Powerniose )

[0032] In the formula, SNR represents the signal-to-noise ratio, Power signal Represents the power of the vibration acceleration signal before noise reduction, Power niose Represents the power of the noise signal.

[0033] Optionally, in one embodiment, in the noise reduction optimization step, the signal-to-noise ratio is applied to optimize the bearing health status evaluation index according to the following formula to obtain a final bearing health status evaluation index:

[0034] Hv'=Hv / SNR

[0035] Hv=max{F1,F2,F3,F4}

[0036] In the formula, Hv represents the bearing health status evaluation index before optimization, SNR represents the calculated signal-to-noise ratio, Hv' represents the bearing health status evaluation index after optimization, and F i is the peak factor of the 0-1kHz spectrum line range of the i-th envelope spectrum.

[0037] Based on other aspects of the method described in any one or more of the above embodiments, the present invention further provides a storage medium storing program codes that can implement the method described in any one or more of the above embodiments.

[0038] Based on the application of the method described in any one or more of the above embodiments, the present invention further provides a bearing health status assessment system based on high-pass filtering, which executes the method described in any one or more of the above embodiments. In a preferred embodiment, the system includes:

[0039] A field signal acquisition module, which is configured to acquire real-time vibration acceleration signals of rolling bearings on-site based on data acquisition equipment;

[0040] A first transformation operation module is configured to perform a fast Fourier transform on the collected vibration acceleration signal data to obtain a complex sequence corresponding to the vibration acceleration signal, wherein the complex sequence includes time amplitude spectrum and phase spectrum information;

[0041] A filtering processing module, which is configured to perform high-frequency bandpass filtering processing in different frequency intervals using high-frequency bandpass filters of different levels based on the complex sequence to obtain corresponding multiple filtered complex sequences;

[0042] A second transformation operation module is configured to perform inverse Fourier transformation on the multiple filtered complex sequences respectively to obtain multiple corresponding time domain waveform data;

[0043] An envelope spectrum peak index calculation module is configured to calculate the corresponding envelope spectrum for each time domain waveform data, and calculate the peak factor of each envelope spectrum based on a set observation frequency range, and select the maximum value from the peak factors of each envelope spectrum as a bearing health status assessment indicator representing the bearing health status information;

[0044] Bearing status assessment step: determining the health status result of the target rolling bearing based on the value of the bearing health status assessment index combined with preset assessment and analysis rules.

[0045] Furthermore, in one embodiment, the system further comprises:

[0046] The noise reduction optimization module is configured to perform noise reduction processing on the collected vibration acceleration signal before bearing status evaluation, calculate the corresponding signal-to-noise ratio according to the vibration acceleration signal and the noise signal before noise reduction, and then use the signal-to-noise ratio to optimize the calculated bearing health status evaluation index, and use the optimized bearing health status evaluation index as the final bearing health status evaluation index.

[0047] Compared with the closest prior art, the present invention also has the following beneficial effects:

[0048] The present invention provides a bearing health status assessment method and system based on high-pass filtering. The method uses a data acquisition device to collect the real-time vibration acceleration signal of the rolling bearing; performs a fast Fourier transform to obtain a complex sequence containing time-amplitude spectrum and phase spectrum information, uses high-frequency bandpass filters of different levels to perform high-frequency bandpass filtering in different frequency intervals, and then performs inverse Fourier transform on multiple filtered complex sequences to obtain corresponding time domain waveform data, calculates the envelope spectrum and calculates the peak factor based on the set observation frequency range, selects the maximum value from the peak factors of each envelope spectrum, and determines the health status of the rolling bearing based on the combination of the preset assessment and analysis rules. With this scheme, the fault frequency characteristics of the envelope spectrum after high-frequency bandpass filtering are quantitatively described by dimensionless indicators, the calculation complexity is low, the timeliness of the health status assessment is good, and there is no need to collect historical bearing vibration monitoring data to build a bearing diagnosis database. The calculated bearing health status data is not affected by physical parameters such as bearing speed and bearing model.

[0049] Furthermore, the bearing health status assessment is realized by adopting the scheme of the present invention, without the need for a full life cycle data training model, and without the limitation of less fault data in industrial production. In addition, the present invention does not require optimization of hyperparameters, and does not require extraction of multi-dimensional time domain, frequency domain, and time-frequency domain feature values.

[0050] In addition, the present invention comprehensively considers the impact of poor bearing lubrication on the calculated bearing health status value. The calculated bearing health status value can well standardize the bearing and its lubrication health status, and the calculation logic has strong engineering online application performance. It can be widely used in rolling bearing status monitoring of rotating equipment in different fields such as petroleum, chemical industry, nuclear power, etc., and has good practicality.

[0051] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the description, claims and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0053] Figure 1 It is a flow chart of a bearing health status assessment method based on high-pass filtering provided in one embodiment of the present invention;

[0054] Figure 2 It is a schematic flow chart of the operating principle of a bearing health status assessment method based on high-pass filtering provided in another embodiment of the present invention;

[0055] Figure 3 It is an overview diagram of a centrifugal pump in an implementation case of a bearing health status assessment method based on high-pass filtering provided in an embodiment of the present invention;

[0056] Figure 4 It is a schematic diagram of a curve of a vibration acceleration signal of a centrifugal pump bearing in a bearing health status assessment method based on high-pass filtering provided in an embodiment of the present invention;

[0057] Figure 5 It is a schematic diagram of a health status data distribution curve of a centrifugal pump bearing in a bearing health status assessment method based on high-pass filtering provided in an embodiment of the present invention;

[0058] Figure 6 It is a schematic diagram of a frequency distribution curve of bearing fault characteristics at different points of an envelope spectrum in a bearing health status assessment method based on high-pass filtering provided in an embodiment of the present invention;

[0059] Figure 7 It is a structural schematic diagram of a bearing health status assessment system based on high-pass filtering provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0060] The following will describe the implementation methods of the present invention in detail in conjunction with the accompanying drawings and embodiments, so that the implementers of the present invention can fully understand how the present invention applies technical means to solve technical problems and achieve the implementation process of technical effects and implement the present invention specifically according to the above implementation process. It should be noted that as long as there is no conflict, the various embodiments and various features of the embodiments in the present invention can be combined with each other, and the technical solutions formed are all within the protection scope of the present invention.

[0061] Although the flowcharts describe the operations as sequential processes, many of the operations may be performed in parallel, concurrently, or simultaneously. The order of the operations may be rearranged. A process may be terminated when its operations are completed, but may also have additional steps not included in the accompanying drawings. A process may correspond to a method, function, procedure, subroutine, subprogram, etc.

[0062] Computer devices include user devices and network devices. Among them, user devices or clients include but are not limited to computers, smart phones, PDAs, etc.; network devices include but are not limited to a single network server, a server group consisting of multiple network servers, or a cloud based on cloud computing consisting of a large number of computers or network servers. Computer devices can be operated alone to implement the present invention, or they can be connected to the network and implement the present invention through interactive operations with other computer devices in the network. The network where the computer device is located includes but is not limited to the Internet, wide area network, metropolitan area network, local area network, VPN network, etc.

[0063] The terms "first", "second", etc. may be used herein to describe various units, but these units should not be limited by these terms, and these terms are used only to distinguish one unit from another unit. The term "and / or" used herein includes any and all combinations of one or more of the listed associated items. When a unit is referred to as being "connected" or "coupled" to another unit, it can be directly connected or coupled to the other unit, or there can be intermediate units.

[0064] The terms used herein are only for describing specific embodiments and are not intended to limit exemplary embodiments. Unless the context clearly indicates otherwise, the singular forms "one", "one" and "item" used herein are also intended to include plural numbers. It should also be understood that the terms "include" and / or "comprise" used herein specify the existence of stated features, integers, steps, operations, units and / or components, without excluding the existence or addition of one or more other features, integers, steps, operations, units, components and / or combinations thereof.

[0065] In the field of machinery, rolling bearings are the core components of industrial equipment. How to accurately and reliably evaluate the real-time operating health status of rolling bearings is the key to predictive maintenance of rotating machinery such as centrifugal pumps and fans. In recent years, with the rapid development of machine learning technology and the emergence of a large amount of industrial data, most related research in the field of rolling bearing operating status monitoring focuses on data-driven rolling bearing health assessment technology or spectrum analysis.

[0066] The data-driven rolling bearing health assessment technology has many restrictions in practical application, such as: comprehensive fault data or full life cycle data training model of the equipment or related equipment is required; and the evaluation and analysis calculation results are greatly affected by physical parameters such as equipment speed, bearing model, structural data, etc., and the model generalization is poor; in addition, the calculation process requires experts to rely on experience and knowledge to set hyperparameters, and rely on the trend of changes in historical data feature values ​​to judge the current health status of the bearing; affected by the scale and accuracy of historical data or intuitive factors set by personnel; the calculation results of this type of technology can only distinguish between normal and abnormal bearing operation, and cannot globally track the health status of the equipment, nor can it consider the impact of poor bearing lubrication on the health status of the bearing. The traditional spectrum analysis method also has the problem of relying on expert experience and knowledge to identify fault feature spectra, and the evaluation data is not accurate enough.

[0067] For example, patent document CN114216681A discloses a method and device for determining the health status of rolling bearings of EMUs. The method first constructs a bearing acoustic feature database through historical bearing acoustic monitoring data and bearing maintenance information. Then, the bearing acoustic feature database is used to train the bearing fault diagnosis model to obtain relevant characteristic parameters, and the bearing health index is determined based on the relevant characteristic parameters to obtain a bearing health index database. Finally, the acoustic signal of the bearing to be evaluated is input into the above model to obtain the bearing health index to be evaluated, and compared with the bearing health index database to determine the bearing health status; during the operation of this scheme, it is necessary to collect historical bearing vibration monitoring data to construct a bearing diagnosis database, and the calculated bearing health status value is affected by physical parameters such as bearing speed and bearing model.

[0068] Patent document CN105528504A discloses a rolling bearing health status assessment method based on CFOA-MKHSVM. The method first extracts time domain and frequency domain statistical features and time-frequency features based on wavelet packets from the vibration signal of the equipment throughout its life cycle. Then, the multi-core hypersphere support vector machine model is constructed using the above features, and the fruit fly algorithm is combined with chaos theory to optimize the model hyperparameters. Finally, the vibration signal to be evaluated is input into the hypersphere support vector machine model, and the normalized difference coefficient evaluation index is established to obtain the rolling bearing health status assessment curve. This method uses a hypersphere support vector machine to identify and perform multi-core convex combination optimization to evaluate the degree of rolling bearing performance degradation in response to the uneven distribution of data and data heterogeneity of each state of the rolling bearing. This solution requires a full life cycle data training model, and is inevitably limited by the small amount of fault data in industrial production when applied. In addition, the hyperparameters need to be optimized during the calculation process, and multi-dimensional time domain, frequency domain, and time-frequency domain eigenvalues ​​need to be extracted. The calculation complexity is too high and the practicality is limited.

[0069] To solve the above problems, the present invention provides a bearing health status assessment method and system based on high-pass filtering. The present invention is based on the actual application needs of engineering and provides a bearing health status monitoring and analysis strategy based on high-frequency bandpass filtering. The evaluation value obtained by analysis represents the bearing health status and bearing lubrication status, and is not affected by the physical parameters of the bearing. It can provide favorable technical support for enterprises to carry out predictive maintenance of rolling bearings.

[0070] Existing theoretical research and actual tests show that the characteristic frequencies of early bearing failures and poor bearing lubrication appear in the high-frequency region, among which the characteristic frequencies of poor bearing lubrication failures are roughly between 0.9 and 1.5 kHz. The vibration frequencies of general mechanical vibration failures such as misalignment and imbalance are usually low, while the failures of the inner ring, outer ring, rolling element, etc. of the bearing itself are generally high, and the earlier the failure occurs, the higher the characteristic frequency of the failure. In the later stage of the bearing failure, the high-frequency bearing measurement value decreases, the noise level increases, and the fault frequency and its sidebands become peak-shaped (commonly known as "haystacks").

[0071] Based on the above analysis and the performance of vibration acceleration sensors actually applied in industrial fields, the present invention proposes a bearing condition assessment method based on high-frequency filtering. This method focuses on the vibration information within a set frequency range and filters out most of the frequency components that are not related to the bearing health and its lubrication status, thereby specifically calculating the characteristic values ​​that only characterize the bearing health status.

[0072] In addition, the present invention characterizes the characteristics of the fault characteristic frequency in the envelope spectrum after filtering by calculating the peak factor of the observed spectrum line below 1k of the envelope spectrum after filtering. The abnormality of the fault characteristic frequency in the envelope spectrum can be dimensionlessly characterized, and this value is independent of physical parameters such as bearing speed and model. Reliable evaluation data of the bearing health status that is not affected by the physical parameters of the bearing is obtained, and the calculation results can guide enterprises to carry out corresponding predictive maintenance work, providing technical support for the integrity management of mechanical equipment.

[0073] Next, the detailed process of the method of the embodiment of the present invention is described in detail based on the accompanying drawings. The steps shown in the flowchart of the accompanying drawings can be executed in a computer system including a set of computer executable instructions. Although the logical order of each step is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than here.

[0074] Embodiment 1:

[0075] Figure 1 The flow chart of the bearing health status assessment method based on high-pass filtering provided in the first embodiment of the present invention is shown. Figure 1 It can be seen that the method includes the following steps.

[0076] On-site signal acquisition steps: Use data acquisition equipment to collect real-time vibration acceleration signals of rolling bearings on-site;

[0077] The first transformation operation step: performing a fast Fourier transform on the collected vibration acceleration signal data to obtain a complex sequence corresponding to the vibration acceleration signal, wherein the complex sequence includes time amplitude spectrum and phase spectrum information;

[0078] Filtering processing step: based on the complex sequence, high-frequency bandpass filters of different levels are used to perform high-frequency bandpass filtering processing in different frequency intervals to obtain corresponding multiple filtered complex sequences;

[0079] The second transformation operation step: performing inverse Fourier transformation on the multiple filtered complex sequences respectively to obtain multiple corresponding time domain waveform data;

[0080] Envelope spectrum peak index calculation steps: calculate the corresponding envelope spectrum for each time domain waveform data, and calculate the peak factor of each envelope spectrum based on the set observation frequency range, and select the maximum value from the peak factors of each envelope spectrum as the bearing health status assessment index representing the bearing health status information;

[0081] Bearing status assessment step: determining the health status result of the target rolling bearing based on the value of the bearing health status assessment index combined with preset assessment and analysis rules.

[0082] The embodiment of the present invention proposes a bearing state assessment method based on high-frequency filtering, which specifically studies the vibration information of 0.9k~1.5kHz, 1.5k~5.5kHz, 3.5k~7.5kHz, and 5.5k~10kHz, filters out most of the frequency components that are irrelevant to the bearing health and its lubrication status, and thus specifically calculates the characteristic value that only characterizes the bearing health status. The present invention characterizes the characteristics of the fault characteristic frequency in the envelope spectrum after filtering by calculating the peak factor of the observed spectrum line below 1k in the envelope spectrum after filtering, and can dimensionlessly characterize the abnormal degree of the fault characteristic frequency in the envelope spectrum, and this value is independent of physical parameters such as bearing speed and model; thus, reliable assessment data of the bearing health status that is not affected by the physical parameters of the bearing is obtained.

[0083] In a preferred embodiment, in the process of collecting the real-time vibration acceleration signal of the rolling bearing through the field signal collection step, a vibration sensor whose distance from the target bearing meets the set requirements is used to collect the real-time vibration acceleration signal data of the rolling bearing according to the set sampling configuration parameters.

[0084] In practical applications, the real-time vibration acceleration signal X0{x1,x2,x3,…x n}, the installation position of the vibration sensor should comply with relevant standards such as ISO20816-1, API STD 670-2014, etc., to effectively collect accurate real-time vibration acceleration signals of rolling bearings. Preferably, the sampling frequency of the vibration sensor is f s Not less than 20480Hz, sampling length n is not less than 16384. Sampling length refers to the length (number of points) of the collected signal during the collection of a single vibration acceleration signal.

[0085] Furthermore, a first transformation operation step is performed to perform a fast Fourier transform on the collected vibration acceleration signal data to obtain a complex sequence corresponding to the vibration acceleration signal, wherein the complex sequence includes time-amplitude spectrum and phase spectrum information.

[0086] In this step, preferably, the collected bearing vibration acceleration signal X0 is subjected to a fast Fourier transform (FFT) according to the following formula to obtain a complex sequence Y0{a1+b1i,a2+b2i,a3+b3i,…,a n +b n i}.

[0087] Y0=fft(X0);

[0088] In the formula, a n +b n i represents a unit complex number, based on which, the vibration acceleration signal X0 with a length of n is subjected to fast Fourier transform (fft) to obtain a complex number sequence with a length of n.

[0089] Next, by executing the filtering processing step, based on the complex sequence, high-frequency bandpass filtering processing is performed in different frequency intervals using high-frequency bandpass filters of different levels to obtain corresponding multiple filtered complex sequences.

[0090] Preferably, in one embodiment, in the filtering processing step, high-frequency bandpass filters C1, C2, C3, and C4 are used to perform bandpass filtering at 0.9k to 1.5kHz, 1.5k to 5.5kHz, 3.5k to 7.5kHz, and 5.5k to 10kHz, respectively, based on the complex sequence.

[0091] Among them, in an optional embodiment, in the filtering process step, the frequency outside the passband is set to zero, and the frequency within the passband remains unchanged.

[0092] In practical applications, a high-frequency bandpass filter {C1, C2, C3, C4} is used to perform high-frequency bandpass filtering on Y0. Preferably, as shown in the following formula, high-frequency bandpass filters C1, C2, C3, and C4 are used to perform bandpass filtering at 0.9k to 1.5kHz, 1.5k to 5.5kHz, 3.5k to 7.5kHz, and 5.5k to 10kHz, respectively, and the frequencies outside the passband are set to zero, and the frequencies within the passband remain unchanged, and the four complex sequences {Y1, Y2, Y3, Y4} after filtering are obtained respectively.

[0093] Y1=Y0*(C1)

[0094] Y2=Y0*(C2)

[0095] Y3=Y0*(C3)

[0096] Y4=Y0*(C4)

[0097] In the formula, Y0 represents the complex sequence obtained by fast Fourier transforming the collected bearing vibration acceleration signal X0. Except for the frequency interval corresponding to the C1 interval, the frequency intervals of other C2, C3, and C4 intervals can be slightly adjusted according to needs and actual conditions.

[0098] Furthermore, a second transformation operation step is performed to perform inverse Fourier transform on the multiple filtered complex sequences respectively to obtain multiple corresponding time domain waveform data.

[0099] In an optional embodiment, the complex sequence {Y1, Y2, Y3, Y4} is subjected to inverse Fourier transform (ifft) to obtain four time domain waveform data {X1, X2, X3, X4} after filtering in different frequency bands.

[0100] X1=ifft(Y1)

[0101] X2=ifft(Y2)

[0102] X3=ifft(Y3)

[0103] X4=ifft(Y4)

[0104] Where, X i Represents the time domain waveform data after the i-th frequency band is filtered.

[0105] Next, the envelope spectrum peak index calculation step is executed to calculate the corresponding envelope spectrum for each time domain waveform data, and the peak factor of each envelope spectrum is calculated based on the set observation frequency range. The maximum value is selected from the peak factors of each envelope spectrum as the bearing health status evaluation indicator that characterizes the bearing health status information.

[0106] Preferably, in one embodiment, in the envelope spectrum peak index calculation step, the Hilbert transform is used to calculate the corresponding envelope spectrum for each item of time domain waveform data.

[0107] In actual application, after using Hilbert transform to calculate the envelope spectrum {f1, f2, f3, f4} of the time domain waveform data {X1, X2, X3, X4}, the set formula is used to calculate the peak factor of the envelope spectrum {f1, f2, f3, f4}. Preferably, the envelope spectrum observation frequency is 0 to 1 kHz, that is, only the peak factor of the 0 to 1 kHz spectrum line is calculated.

[0108] Therefore, in an optional embodiment, in the envelope spectrum peak index calculation step, the peak factor of the 0-1kHz spectrum line range is calculated for each envelope spectrum, wherein the peak factor of the envelope spectrum {f1, f2, f3, f4} is calculated according to the following formula:

[0109] F1=max(f1) / rms(f1)

[0110] F2=max(f2) / rms(f2)

[0111] F3=max(f3) / rms(f3)

[0112] F4=max(f4) / rms(f4)

[0113] In the formula, f i is the envelope spectrum corresponding to the i-th time domain waveform data, max() means calculating the maximum value of the envelope spectrum in the 0-1kHz spectrum line, F i is the peak factor of the 0-1kHz spectrum line range of the i-th envelope spectrum, and rms() represents the root mean square value within the 0-1kHz spectrum line of the envelope spectrum.

[0114] Further, select the maximum value from the peak factors of each envelope spectrum according to the following formula as the bearing health state evaluation index characterizing the bearing health state information:

[0115] Hv = max{F1, F2, F3, F4}

[0116] In the formula, Hv represents the bearing health state evaluation index.

[0117] In the embodiment of the present invention, further determine the bearing health state according to the bearing health state value Hv calculated in S8, execute the bearing state evaluation step, and determine the health state result of the target rolling bearing based on the value of the bearing health state evaluation index in combination with the preset evaluation and analysis rules.

[0118] Preferably, the bearing health state value Hv is divided into the following three states:

[0119] 1. Hv ≤ 2, good state. In this stage, the bearing and its lubrication state are good, and it is recommended to continue running;

[0120] 2. 2 < Hv ≤ 20, sub - healthy state. In this stage, the bearing and its lubrication state are poor. It is recommended to change the lubricating oil and monitor the operation. If the Hv value does not improve after changing the lubricating oil, replace the bearing at an appropriate time;

[0121] 3. Hv > 20, poor bearing and its lubrication state. It is recommended to replace the bearing.

[0122] Therefore, in a preferred embodiment, in the bearing state evaluation step, determine the health state result of the target rolling bearing according to the following evaluation and analysis rules:

[0123] If the bearing health state evaluation index Hv ≤ 2, it is determined to be in a good state. In this stage, the bearing and its lubrication state are good, and continue running;

[0124] If 2 < Hv ≤ 20, it is determined to be in a sub - healthy state. In this stage, the bearing and its lubrication state are poor. Change the lubricating oil and monitor the operation. If the Hv value does not improve after changing the lubricating oil, replace the bearing;

[0125] If Hv > 20, it is determined to be in a risk state, and replace the bearing.

[0126] The rolling bearing health status assessment method based on high-frequency bandpass filtering provided by the embodiment of the present invention is adopted, and the fault frequency characteristics of the envelope spectrum after high-frequency bandpass filtering are quantitatively described by dimensionless indicators. There is no need to collect historical bearing vibration monitoring data to build a bearing diagnosis database. The calculated bearing health status value is independent of physical parameters such as rotational speed and bearing model. The influence of poor bearing lubrication factors on the calculated bearing health status value is taken into account. The calculated bearing health status value can well standardize the bearing and its lubrication health status. In addition, the present invention is concise, easy to understand, and has strong engineering online application performance.

[0127] And the scheme of the present invention does not need to build a full life cycle data training model, and is not limited by the small amount of fault data in industrial production. In addition, the present invention does not need to optimize hyperparameters, does not need to extract multi-dimensional time domain, frequency domain, and time-frequency domain eigenvalues, has low computational complexity, and has good timeliness of health status assessment. In addition, the present invention verifies the effectiveness and reliability of the proposed method through engineering case data, and has broad prospects for practical engineering applications.

[0128] Embodiment 2:

[0129] Figure 2 The flowchart of the implementation principle of the bearing health status assessment method based on high-pass filtering provided in the second embodiment of the present invention is shown. Figure 2 It can be seen that the method includes the following steps.

[0130] On-site signal acquisition steps: Use data acquisition equipment to collect real-time vibration acceleration signals of rolling bearings on-site;

[0131] The first transformation operation step: performing a fast Fourier transform on the collected vibration acceleration signal data to obtain a complex sequence corresponding to the vibration acceleration signal, wherein the complex sequence includes time amplitude spectrum and phase spectrum information;

[0132] Filtering processing step: based on the complex sequence, high-frequency bandpass filters of different levels are used to perform high-frequency bandpass filtering processing in different frequency intervals to obtain corresponding multiple filtered complex sequences;

[0133] The second transformation operation step: performing inverse Fourier transformation on the multiple filtered complex sequences respectively to obtain multiple corresponding time domain waveform data;

[0134] Envelope spectrum peak index calculation steps: calculate the corresponding envelope spectrum for each time domain waveform data, and calculate the peak factor of each envelope spectrum based on the set observation frequency range, and select the maximum value from the peak factors of each envelope spectrum as the bearing health status assessment index representing the bearing health status information;

[0135] Bearing status assessment step: determining the health status result of the target rolling bearing based on the value of the bearing health status assessment index combined with preset assessment and analysis rules.

[0136] This embodiment is a variation of the above-mentioned embodiment 1. The specific technical features of the above-mentioned operations are consistent with those of the embodiment and will not be described in detail. Only the distinguishing technical features will be specifically described below.

[0137] In order to comprehensively consider the impact of poor bearing lubrication and excessive vibration signal noise in the late life of the bearing on the calculated bearing health status value, the researchers of the present invention ensure that the calculated bearing health status value can well standardize the bearing and its lubrication health status. The bearing health status assessment method based on high-pass filtering provided in this embodiment also includes:

[0138] Noise reduction optimization step: Before the bearing status evaluation step, the collected vibration acceleration signal is subjected to noise reduction processing, and the corresponding signal-to-noise ratio is calculated based on the vibration acceleration signal and the noise signal before noise reduction, and then the signal-to-noise ratio is applied to optimize the calculated bearing health status evaluation index, and the optimized bearing health status evaluation index is used as the final bearing health status evaluation index.

[0139] In a preferred embodiment, in the noise reduction optimization step, a fixed threshold wavelet method is used to perform denoising processing on the vibration acceleration signal to obtain a noise-reduced vibration acceleration signal and a noise signal.

[0140] In practical application, the real-time vibration acceleration signal X0 is processed by using fixed threshold wavelet denoising to obtain the vibration acceleration signal X0' after noise reduction and the noise signal X0" = X0-X0'.

[0141] Further, in an optional embodiment, in the noise reduction optimization step, the process of calculating the signal-to-noise ratio includes:

[0142] Calculate the power of the original vibration acceleration signal and the power of the noise signal before noise reduction processing respectively;

[0143] Then the signal-to-noise ratio is calculated according to the following formula:

[0144] SNR=10lg(Power signal / Power niose )

[0145] In the formula, SNR represents the signal-to-noise ratio, Power signal Represents the power of the vibration acceleration signal before noise reduction, Power niose Represents the power of the noise signal.

[0146] In practical applications, the power of the original vibration acceleration signal X0 and the noise signal X0” are calculated respectively. signal and Power noise Then, the signal-to-noise ratio SNR is calculated based on the above formula.

[0147] Specifically, the maximum value among {F1, F2, F3, F4} is divided by the signal-to-noise ratio SNR, and is defined as the optimized bearing health status value Hv. Therefore, in one embodiment, in the noise reduction optimization step, the signal-to-noise ratio is applied to optimize the bearing health status evaluation index according to the following formula to obtain the final bearing health status evaluation index:

[0148] Hv'=Hv / SNR

[0149] Hv=max{F1,F2,F3,F4}

[0150] In the formula, Hv represents the bearing health status evaluation index before optimization, SNR represents the calculated signal-to-noise ratio, Hv' represents the bearing health status evaluation index after optimization, and F i is the peak factor of the 0-1kHz spectrum line range of the i-th envelope spectrum.

[0151] The bearing health status assessment method based on high-pass filtering of the embodiment of the present invention does not require a historical data training model, does not require expert experience and knowledge to set hyperparameters and identify bearing health status, does not require hyperparameter optimization, will not be affected by expert subjective factors and historical data quality, and does not require extraction of multi-dimensional time domain and frequency domain eigenvalues. The operation is simple, there is no unreliable intermediate parameter calculation process, and it has good industrial online application performance.

[0152] In addition, this method is used to realize bearing health status inspection, monitoring and evaluation. By exploring the frequency domain characteristics of the vibration signal, dimensionless indicators are used to well characterize the health status of the bearing, so that the calculated bearing health status value is not affected by physical parameters such as rotational speed and bearing signal. It has the characteristics of simple principle, easy understanding and strong interpretability.

[0153] In addition, the solution of this embodiment further comprehensively considers the impact of poor bearing lubrication and excessive vibration signal noise in the later stage of bearing life on the bearing health status. The calculated bearing health status value can not only characterize the bearing lubrication status, but also overcome the impact of the low signal-to-noise ratio of the vibration signal on the frequency domain characteristics.

[0154] The solution of the present invention can calculate the bearing health status value that is not affected by the bearing speed and model while meeting many restrictions on actual industrial applications; it can be widely used in the field of rolling bearing status monitoring of rotating equipment in the petroleum, chemical, and nuclear power fields, and can better guide enterprises to carry out predictive maintenance of rolling bearings, reduce production accidents in enterprises, reduce production costs, optimize spare parts storage, and provide technical support for mechanical equipment integrity management, with good application prospects.

[0155] The present invention is further described below in conjunction with examples of implementation. The scope of the present invention is not limited by the examples, but is set forth in the claims.

[0156] The specific implementation method of this patent is explained in combination with the performance degradation case data of the drive end bearing of the P3409A centrifugal pump of a Chinese petrochemical company's hydrogenation degradation unit. Figure 3 As shown, this centrifugal pump is driven by a motor, the speed of the centrifugal pump is 2980r / min, and two horizontal and vertical acceleration sensors are installed on the bearing of the centrifugal pump driving end model 6217, the sampling frequency is 25600Hz, and the sampling interval is 2 hours. In actual application, if multiple sensors are installed, the data of multiple sensors are calculated using the method of the embodiment of the present invention, and each sensor can obtain a bearing health status value, which is analyzed as the bearing health status value.

[0157] The original time domain acceleration waveform data of this centrifugal pump from December 15, 2013 to January 12, 2014 is obtained, with a total of 332 data files, each of which consists of 16,384 points. The acceleration effective value curve is calculated, such as Figure 4 As shown: when running to the 226th data file, the vibration amplitude was too large and triggered the alarm value control line; when running to the 329th data file, the dangerous value alarm control line was triggered, and then the machine was shut down for maintenance, and it was found that the outer ring of the driving end rolling bearing was faulty.

[0158] The specific process of using the method provided by the present invention based on the centrifugal pump drive end bearing is as follows:

[0159] S1: Simulate the process of collecting vibration acceleration data on site to obtain the first data of the drive end bearing of the P3409A centrifugal pump X p1 {x1,x2,x3,…x 16384}.

[0160] S2: Obtained bearing vibration acceleration signal X p1 Perform fast Fourier transform to obtain the complex sequence Y containing the time-amplitude spectrum and phase spectrum information p1 .

[0161] S3: Use high frequency bandpass filter {C1, C2, C3, C4} to p1 Bandpass filtering is performed at 0.9k~1.5kHz, 1.5k~5.5kHz, 3.5k~7.5kHz, and 5.5k~10kHz respectively. The bandpass filter sets the frequency outside the passband to zero, and the frequency within the passband remains unchanged. The four complex sequences {Y p1 ,Y p2 ,Y p3 ,Y p4}.

[0162] S4: Perform inverse Fourier transform on the filtered complex sequence to obtain 4 time domain waveform data after filtering with different frequency bands.

[0163] S5: Use Hilbert transform to calculate the envelope spectrum of the four filtered time domain waveform data.

[0164] S6: Calculate the peak factor {F p1 ,F p2 ,F p3 ,F p4 ,}

[0165] S7: Calculate X p1 The signal-to-noise ratio SNR.

[0166] S7.1 Using fixed threshold wavelet denoising to process X p1 , obtain the vibration acceleration signal after noise reduction and the noise signal;

[0167] S7.2 calculates the original vibration acceleration signal X p1 and the power of the noise signal signal and Power noise ;

[0168] S7.3 Calculate the vibration acceleration signal X p1 The signal-to-noise ratio SNR.

[0169] S8: Use the crest factor {F p1 ,F p2 ,F p3 ,F p4 ,} is divided by the signal-to-noise ratio SNR and defined as the bearing health status value Hv.

[0170] Hv=max{F1,F2,F3,F4} / SNR

[0171] S9: Determine the bearing health status based on the bearing health status value Hv calculated in S8, and determine the bearing health status online based on the value of Hv:

[0172] If Hv ≤ 2, the bearing and its lubrication state are good, continue to run, collect the next vibration data, and repeat S1 to S8.

[0173] If 2 < Hv ≤ 20, the operating state of the equipment should be closely monitored, the lubricating oil should be replaced or refilled. If the Hv value does not improve after replacing the lubricating oil, replace the bearing at an appropriate time.

[0174] S10: Actually collect the bearing health state values of all 332 groups of data, and plot their development trend as an axis diagram, as Figure 5 shown.

[0175] As Figure 5 shown, except for a few points, the health state values of the driving end bearing of the centrifugal pump P3409A exceed 2 starting from point 125 during the performance degradation process, then slightly improve. After point 142, the bearing health state values have exceeded 2 and are in a sub-healthy state. Subsequently, after point 216, they exceed 20 and are in a nearly failed bearing health state.

[0176] Figure 6 The schematic diagram of the bearing fault characteristic frequency distribution curve at different points of the envelope spectrum is shown in Figure 6 shown. As shown, the bearing fault characteristic frequencies are found in the envelope spectra at points 125, 142, and 200. However, at point 216, the bearing fault characteristic frequency is not obvious, but the Hv value is still high, indicating that the proposed method can not only detect the early bearing faults in a timely manner but also track the occurrence process of bearing faults well when the noise signal interference is large and the fault characteristic frequency is not obvious in the late stage of rolling bearing faults.

[0177] As Figure 3 shown, the method proposed by the present invention alarms at point 226, 202 hours earlier than the traditional fixed threshold line, and 34 hours earlier than the mutation point of the effective value of vibration acceleration (point 142), which can reserve sufficient time for the enterprise to carry out predictive maintenance work.

[0178] For the foregoing method embodiments, for the sake of simple description, they are all expressed as a series of action combinations. However, those skilled in the art should know that the present invention is not limited by the described action sequence, because according to the present invention, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily essential to the present invention.

[0179] It should be pointed out that in other embodiments of the present invention, the method can also obtain a new bearing health status assessment method by combining one or several of the above embodiments to achieve safe operation, maintenance and high-quality monitoring of bearings in the engineering field.

[0180] It should be noted that, based on the method in any one or more of the above-mentioned embodiments of the present invention, the present invention also provides a storage medium, which stores program code that can implement the method as described in any one or more of the above-mentioned embodiments, and when the code is executed by the operating system, it can implement the bearing health status assessment method based on high-pass filtering as described above.

[0181] Embodiment three:

[0182] The method is described in detail in the embodiments disclosed in the present invention. The method of the present invention can be implemented by various devices or systems. Therefore, based on other aspects of the method described in any one or more embodiments, the present invention also provides a bearing health status assessment system based on high-pass filtering, which is used to execute the bearing health status assessment method based on high-pass filtering described in any one or more embodiments. Specific embodiments are given below for detailed description.

[0183] Specifically, Figure 7 FIG. 4 shows a schematic diagram of the structure of a bearing health status assessment system based on high-pass filtering provided in an embodiment of the present invention. Figure 7 As shown, the system includes:

[0184] A field signal acquisition module, which is configured to acquire real-time vibration acceleration signals of rolling bearings on-site based on data acquisition equipment;

[0185] A first transformation operation module is configured to perform a fast Fourier transform on the collected vibration acceleration signal data to obtain a complex sequence corresponding to the vibration acceleration signal, wherein the complex sequence includes time amplitude spectrum and phase spectrum information;

[0186] A filtering processing module, which is configured to perform high-frequency bandpass filtering processing in different frequency intervals using high-frequency bandpass filters of different levels based on the complex sequence to obtain corresponding multiple filtered complex sequences;

[0187] A second transformation operation module is configured to perform inverse Fourier transformation on the multiple filtered complex sequences respectively to obtain multiple corresponding time domain waveform data;

[0188] An envelope spectrum peak index calculation module is configured to calculate the corresponding envelope spectrum for each time domain waveform data, and calculate the peak factor of each envelope spectrum based on a set observation frequency range, and select the maximum value from the peak factors of each envelope spectrum as a bearing health status assessment indicator representing the bearing health status information;

[0189] The bearing status assessment module is configured to determine the health status result of the target rolling bearing based on the value of the bearing health status assessment index combined with preset assessment and analysis rules.

[0190] Preferably, in one embodiment, the field signal acquisition module is configured to collect real-time vibration acceleration signal data of the rolling bearing according to set sampling configuration parameters using a vibration sensor whose distance from the target bearing meets set requirements.

[0191] Furthermore, in one embodiment, the system further comprises:

[0192] The noise reduction optimization module is configured to perform noise reduction processing on the collected vibration acceleration signal before bearing status evaluation, calculate the corresponding signal-to-noise ratio according to the vibration acceleration signal and the noise signal before noise reduction, and then use the signal-to-noise ratio to optimize the calculated bearing health status evaluation index, and use the optimized bearing health status evaluation index as the final bearing health status evaluation index.

[0193] In an optional embodiment, the vibration sensor sampling frequency f s Not less than 20480Hz, sampling length n not less than 16384.

[0194] In actual application, in a preferred embodiment, the filtering processing module is configured as follows: based on the complex sequence, high-frequency bandpass filters C1, C2, C3, and C4 are used to perform bandpass filtering at 0.9k~1.5kHz, 1.5k~5.5kHz, 3.5k~7.5kHz, and 5.5k~10kHz respectively.

[0195] Optionally, in one embodiment, the envelope spectrum peak index calculation module calculates the corresponding envelope spectrum using Hilbert transform for each item of time domain waveform data.

[0196] Furthermore, in one embodiment, the envelope spectrum peak index calculation module is configured to calculate the peak factor of the 0-1kHz spectrum line range for each envelope spectrum, wherein the peak factor of the envelope spectrum {f1, f2, f3, f4} is calculated according to the following formula:

[0197] F1=max(f1) / rms(f1)

[0198] F2 = max(f2) / rms(f2)

[0199] F3 = max(f3) / rms(f3)

[0200] F4 = max(f4) / rms(f4)

[0201] wherein, f i is the envelope spectrum corresponding to the i-th time-domain waveform data, max() represents calculating the maximum value among the spectral lines of the envelope spectrum from 0 to 1 kHz, and F i is the peak factor within the spectral line range of 0 to 1 kHz of the i-th envelope spectrum, and rms() represents calculating the root mean square value within the spectral lines of the envelope spectrum from 0 to 1 kHz.

[0202] Further, in a preferred embodiment, the bearing condition assessment module determines the health state result of the target rolling bearing according to the following assessment and analysis rules:

[0203] If the bearing health state assessment index Hv ≤ 2, it is determined to be in a good state. At this stage, the bearing and its lubrication state are good, and continue to operate;

[0204] If 2 < Hv ≤ 20, it is determined to be in a sub-healthy state. At this stage, the bearing and its lubrication state are poor. Replace the lubricating oil and monitor the operation. If the Hv value does not improve after replacing the lubricating oil, replace the bearing;

[0205] If Hv > 20, it is determined to be in a risk state, and replace the bearing.

[0206] Optionally, in an embodiment, the noise reduction and optimization module is configured to perform noise reduction processing on the vibration acceleration signal using the fixed threshold wavelet method to obtain the noise-reduced vibration acceleration signal and the noise signal.

[0207] In actual application, in a preferred embodiment, the noise reduction and optimization module calculates the signal-to-noise ratio according to the following operations:

[0208] Calculate the power of the original vibration acceleration signal before noise reduction processing and the power of the noise signal respectively;

[0209] Then calculate the signal-to-noise ratio according to the following formula:

[0210] SNR = 10lg(Power signal / Power niose )

[0211] wherein, SNR represents the signal-to-noise ratio value, and Power signal represents the power of the vibration acceleration signal before noise reduction, and Power niose represents the power of the noise signal.

[0212] Further, the noise reduction optimization module is configured to apply the signal-to-noise ratio to perform optimization calculation on the bearing health status evaluation index according to the following formula to obtain a final bearing health status evaluation index:

[0213] Hv'=Hv / SNR

[0214] Hv=max{F1,F2,F3,F4}

[0215] Where Hv represents the bearing health status evaluation index before optimization, SNR represents the calculated signal-to-noise ratio, Hv' represents the bearing health status evaluation index after optimization, and F i is the peak factor of the 0-1kHz spectrum line range of the i-th envelope spectrum.

[0216] In the bearing health status assessment system based on high-pass filtering provided by the embodiment of the present invention, each module or unit structure can operate independently or in combination according to actual signal acquisition requirements and data calculation requirements to achieve corresponding technical effects.

[0217] It should be understood that the embodiments disclosed in the present invention are not limited to the specific structures, processing steps or materials disclosed herein, but should be extended to equivalent substitutions of these features understood by ordinary technicians in the relevant field. It should also be understood that the terms used herein are only used for the purpose of describing specific embodiments and are not meant to be limiting.

[0218] The "one embodiment" mentioned in the specification means that a particular feature, structure or characteristic described in conjunction with the embodiment is included in at least one embodiment of the present invention. Therefore, the phrase "one embodiment" appearing in various places throughout the specification does not necessarily refer to the same embodiment.

[0219] Although the embodiments disclosed in the present invention are as above, the above contents are only embodiments adopted for facilitating the understanding of the present invention and are not intended to limit the present invention. Any technician in the technical field to which the present invention belongs can make any modifications and changes in the form and details of the implementation without departing from the spirit and scope disclosed in the present invention, but the patent protection scope of the present invention shall still be subject to the scope defined in the attached claims.

Claims

1. A bearing health status assessment method based on high-pass filtering, characterized in that: The method comprises: On-site signal acquisition steps: Use data acquisition equipment to collect real-time vibration acceleration signals of rolling bearings on-site; The first transformation operation step: performing a fast Fourier transform on the collected vibration acceleration signal data to obtain a complex sequence corresponding to the vibration acceleration signal, wherein the complex sequence includes time amplitude spectrum and phase spectrum information; Filtering processing step: based on the complex sequence, high-frequency bandpass filters of different levels are used to perform high-frequency bandpass filtering processing in different frequency intervals to obtain corresponding multiple filtered complex sequences; The second transformation operation step: performing inverse Fourier transformation on the multiple filtered complex sequences respectively to obtain multiple corresponding time domain waveform data; Envelope spectrum peak index calculation steps: calculate the corresponding envelope spectrum for each time domain waveform data, and calculate the peak factor of each envelope spectrum based on the set observation frequency range, and select the maximum value from the peak factors of each envelope spectrum as the bearing health status assessment index representing the bearing health status information; Bearing status assessment step: determining the health status result of the target rolling bearing based on the value of the bearing health status assessment index combined with preset assessment and analysis rules.

2. The method according to claim 1, characterized in that In the on-site signal acquisition step, a vibration sensor whose distance from the target bearing meets the set requirements is used to acquire real-time vibration acceleration signal data of the rolling bearing according to the set sampling configuration parameters.

3. The method according to claim 1, characterized in that The method further comprises: Noise reduction optimization step: Before the bearing status evaluation step, the collected vibration acceleration signal is subjected to noise reduction processing, and the corresponding signal-to-noise ratio is calculated based on the vibration acceleration signal and the noise signal before noise reduction, and then the signal-to-noise ratio is applied to optimize the calculated bearing health status evaluation index, and the optimized bearing health status evaluation index is used as the final bearing health status evaluation index.

4. The method according to claim 2, characterized in that In the field signal acquisition step, the vibration sensor sampling frequency f s Not less than 20480Hz, sampling length n not less than 16384.

5. The method according to claim 1, characterized in that In the filtering process, high frequency bandpass filters C1, C2, C3, and C4 are used to perform bandpass filtering at 0.9k to 1.5kHz, 1.5k to 5.5kHz, 3.5k to 7.5kHz, and 5.5k to 10kHz, respectively, based on the complex sequence.

6. The method according to claim 1, characterized in that In the envelope spectrum peak index calculation step, the corresponding envelope spectrum is calculated using Hilbert transform for each item of time domain waveform data.

7. The method according to claim 1, characterized in that In the envelope spectrum peak index calculation step, the peak factor of the 0-1kHz spectrum line range is calculated for each envelope spectrum, wherein the peak factor of the envelope spectrum {f1, f2, f3, f4} is calculated according to the following formula: F1=max(f1) / rms(f1) F2=max(f2) / rms(f2) F3=max(f3) / rms(f3) F4=max(f4) / rms(f4) In the formula, f i is the envelope spectrum corresponding to the i-th time domain waveform data, max() means calculating the maximum value of the envelope spectrum in the 0-1kHz spectrum line, F i is the peak factor of the 0-1kHz spectrum line range of the i-th envelope spectrum, and rms() represents the root mean square value within the 0-1kHz spectrum line of the envelope spectrum.

8. The method according to claim 1, characterized in that: In the bearing condition assessment step, the health condition result of the target rolling bearing is determined according to the following assessment and analysis rules: If the bearing health status assessment index Hv≤2, it is determined to be in good condition. At this stage, the bearing and its lubrication are in good condition and continue to operate; If 2 < Hv ≤ 20, it is determined as the sub - healthy state. In this stage, the bearing and its lubrication state are poor. Replace the lubricating oil and monitor the operation. If the Hv value does not improve after replacing the lubricating oil, replace the bearing. If Hv > 20, it is determined as the risk state and the bearing is replaced.

9. The method according to claim 1, characterized in that: In the noise reduction and optimization step, the stationary - threshold wavelet method is used to denoise the vibration acceleration signal, and the denoised vibration acceleration signal and the noise signal are obtained.

10. The method according to claim 1, characterized in that In the noise reduction and optimization step, the process of calculating the signal - to - noise ratio includes: Calculate the power of the original vibration acceleration signal and the power of the noise signal before the noise reduction process respectively. Furthermore, calculate the signal - to - noise ratio according to the following formula: SNR=10lg(Power signal / Power niose ) In the formula, SNR represents the signal-to-noise ratio, Power signal Represents the power of the vibration acceleration signal before noise reduction, Power niose Represents the power of the noise signal.

11. The method according to claim 1, characterized in that: In the noise reduction and optimization step, the signal - to - noise ratio is applied to optimize the bearing health state evaluation index according to the following formula, and the final bearing health state evaluation index is obtained: Hv' = Hv / SNR Hv = max{F1, F2, F3, F4} In the formula, Hv represents the bearing health status evaluation index before optimization, SNR represents the calculated signal-to-noise ratio, Hv' represents the bearing health status evaluation index after optimization, and F i is the peak factor of the 0-1kHz spectrum line range of the i-th envelope spectrum.

12. A storage medium, characterized in that: The program code for implementing the method described in any one of claims 1 to 11 is stored on the storage medium.

13. A bearing health status assessment system based on high-pass filtering, characterized in that: The system executes the method described in any one of claims 1 to 11. The system includes: A field signal acquisition module configured to acquire the real - time vibration acceleration signal of the rolling bearing on - site based on the data acquisition device. A first transformation operation module configured to perform a fast Fourier transform on the acquired vibration acceleration signal data to obtain a complex sequence corresponding to the vibration acceleration signal, and the complex sequence contains time - amplitude spectrum and phase - spectrum information. A filtering processing module configured to perform high - frequency band - pass filtering processing on the complex sequence in different frequency intervals using high - frequency band - pass filters of different levels to obtain corresponding multiple filtered complex sequences. A second transformation operation module configured to perform an inverse Fourier transform on each of the multiple filtered complex sequences to obtain multiple corresponding time - domain waveform data. An envelope spectrum peak index operation module configured to calculate the corresponding envelope spectrum for each item of time - domain waveform data respectively, calculate the peak factor of each envelope spectrum based on the set observation frequency range, and select the maximum value from the peak factors of each envelope spectrum as the bearing health state evaluation index representing the bearing health state information. A bearing state evaluation module configured to determine the health state result of the target rolling bearing based on the value of the bearing health state evaluation index in combination with the preset evaluation and analysis rules.

14. The system according to claim 1, characterized in that The system further includes: A noise reduction and optimization module configured to perform noise reduction processing on the acquired vibration acceleration signal before bearing state evaluation, calculate the corresponding signal - to - noise ratio according to the vibration acceleration signal and the noise signal before noise reduction, and then apply the signal - to - noise ratio to optimize the calculated bearing health state evaluation index, and use the optimized bearing health state evaluation index as the final bearing health state evaluation index.

Images