CN114371005A - Impact characteristic extraction method and device for rolling bearing - Google Patents
Impact characteristic extraction method and device for rolling bearing Download PDFInfo
- Publication number
- CN114371005A CN114371005A CN202111552015.5A CN202111552015A CN114371005A CN 114371005 A CN114371005 A CN 114371005A CN 202111552015 A CN202111552015 A CN 202111552015A CN 114371005 A CN114371005 A CN 114371005A
- Authority
- CN
- China
- Prior art keywords
- mode component
- rolling bearing
- vibration signal
- original vibration
- component
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M13/00—Testing of machine parts
- G01M13/04—Bearings
- G01M13/045—Acoustic or vibration analysis
Landscapes
- Physics & Mathematics (AREA)
- Acoustics & Sound (AREA)
- General Physics & Mathematics (AREA)
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
Abstract
The invention belongs to the field of nuclear power station rolling bearing fault diagnosis, and particularly relates to an impact characteristic extraction method and device of a rolling bearing, wherein VMD and a high-order frequency weighted energy operator are combined for extracting impact characteristics, firstly, signals are subjected to primary noise reduction through VMD decomposition, and the optimal mode component of the signals is selected; and secondly, suppressing the interference frequency in the signal energy spectrum by a high-order frequency weighted energy operator to achieve the effect of secondary noise reduction, thereby enhancing the periodic impact component in the signal.
Description
Technical Field
The application belongs to the technical field of nuclear power station rolling bearing fault diagnosis, and particularly relates to an impact characteristic extraction method and device for a rolling bearing.
Background
The rolling bearing is an important component in large-scale rotating machinery equipment, and once the rolling bearing fails, unnecessary economic loss and even casualties are caused, so that the evaluation of the running health state of the rolling bearing is realized by an effective modern signal processing means from the viewpoint of equipment maintenance mode and maintenance cost, and the rolling bearing has very important engineering value and practical significance.
The fault diagnosis of the rolling bearing is mainly realized by a classical signal processing method, and the core is as follows: and judging whether a fault exists or not by extracting the periodic impact characteristics representing fault information in the vibration signals of the rolling bearing. In practical industrial application, when the signal-to-noise ratio of a signal is low and the interference of an unrelated component (such as impulsive noise, harmonic waves and the like) is severe, the existing method is often difficult to accurately extract the periodic impact features in the signal.
Currently, algorithms such as Empirical Mode Decomposition (EMD), Local Mean Decomposition (LMD), Complementary Ensemble Empirical Mode Decomposition (CEEMD) and the like are widely used for signal decomposition. Although the algorithm has a good effect to a certain extent, some unavoidable problems still exist, such as incomplete elimination of the influence of the end effect, incomplete signal decomposition, low calculation efficiency, large influence of the signal-to-noise ratio, and the like.
Disclosure of Invention
The application aims to provide an impact feature extraction method and device for a rolling bearing, and the method and device are used for solving the problem that the existing method is difficult to accurately extract periodic impact features in signals when the signal-to-noise ratio of the signals is low and the interference of irrelevant components is serious.
The technical scheme for realizing the purpose of the application is as follows:
a first aspect of embodiments of the present application provides an impact feature extraction method for a rolling bearing, where the method includes:
collecting an original vibration signal of a rolling bearing;
decomposing the original vibration signal by utilizing a VMD algorithm to obtain a plurality of eigenmode components;
determining an eigenmode component corresponding to the maximum value of the time-frequency weighted kurtosis from the plurality of eigenmode components as a best mode component;
and demodulating the optimal mode component by using a high-order frequency weighted energy operator to extract periodic impact characteristics in the original vibration signal.
Optionally, the decomposing the original vibration signal by using the VMD algorithm to obtain a plurality of eigen-mode components specifically includes:
obtaining the most appropriate penalty factor of each mode component according to the mapping relation between the penalty factors and the center frequency of the mode component;
and respectively iterating and updating the mode component, the central frequency and the Lagrange multiplier based on the most appropriate penalty factor, and decomposing the original vibration signal into the plurality of eigen-mode components by adopting an energy loss coefficient and a Pearson correlation coefficient as iteration termination conditions.
Optionally, the energy loss coefficient is calculated according to the following formula:
where xi is the energy loss coefficient, f is the original vibration signal, ukIs the k-th mode component.
Optionally, the pearson correlation coefficient is calculated according to the following formula:
wherein r is the Pearson correlation coefficient, ukFor the k-th mode component,
is the mean of the k-th mode component, f is the original vibration signal,
is the mean value of the original vibration signal.
Optionally, the using the energy loss coefficient and the pearson correlation coefficient as the iteration termination condition specifically includes:
when the energy loss coefficient or pearson correlation coefficient of the kth mode component reaches the corresponding threshold, the iteration is stopped.
Optionally, the most suitable penalty factor for each mode component is obtained according to the following formula:
in the formula, alphakPenalty factor for the k-th mode component, fkcIs the center frequency, f, of the k-th mode componentsIs the signal sampling frequency.
Optionally, the determining, as the best mode component, an eigen-mode component corresponding to the maximum time-frequency weighted kurtosis value from the multiple eigen-mode components specifically includes:
obtaining a time-domain spectral kurtosis and an envelope spectral kurtosis of the plurality of eigenmode components;
obtaining time-frequency weighted kurtosis of the plurality of eigen-mode components according to the time-domain spectral kurtosis and the envelope spectral kurtosis;
and determining the eigenmode component corresponding to the maximum value of the time-frequency weighted kurtosis as the best mode component.
Optionally, the demodulating the optimal mode component by using a high-order frequency weighted energy operator to extract periodic impact features in the original vibration signal specifically includes:
obtaining a derivation order m of the high-order frequency weighted energy operator;
performing 1-m-order demodulation on the optimal mode component according to the expression of the high-order frequency weighted energy operator, and drawing a corresponding energy spectrum;
and extracting periodic impact characteristics in the original vibration signal based on the energy spectrum.
Optionally, the expression of the higher-order frequency weighted energy operator specifically includes:
ξ[f(t),i]＝S[fi(t)]＝|fi(t)+jH[fi(t)]|2＝fi(t)2+H[fi(t)]2
where f (t) is the best mode component, ξ [ f (t), i]Weighting the energy operator for the i-order frequency of the best mode component, i ∈ [1, m]And is an integer, fi(t) is the i-th derivative of the best mode component, H [ f ]i(t)]A Hilbert transform of the i-th derivative of the best mode component.
A second aspect of the embodiments of the present application provides an impact characteristic extraction device for a rolling bearing, the device including:
the acquisition unit is used for acquiring an original vibration signal of the rolling bearing;
the decomposition unit is used for decomposing the original vibration signal by utilizing a VMD algorithm to obtain a plurality of eigen-mode components;
a determining unit, configured to determine, as a best mode component, an eigenmode component corresponding to a maximum value of the time-frequency weighted kurtosis from the plurality of eigenmode components;
and the extraction unit is used for demodulating the optimal mode component by using a high-order frequency weighted energy operator so as to extract the periodic impact characteristics in the original vibration signal.
The beneficial technical effect of this application lies in:
(1) according to the method and the device for extracting the impact characteristics of the rolling bearing, the punishment factors are adaptively determined by establishing the mapping relation between the punishment factors and the center frequency of the modal component, and the problem that the center frequency of each mode component is inaccurate due to the fact that the punishment factors are manually selected is solved.
(2) According to the impact feature extraction method and device for the rolling bearing, the mode component number of signal decomposition is adaptively determined according to the energy loss coefficient and the Pearson correlation coefficient after the signal decomposition, and the problem of over-decomposition or under-decomposition of a signal in the traditional VMD algorithm is solved.
(3) The method and the device for extracting the impact characteristics of the rolling bearing provided by the embodiment of the application construct a time-frequency index-time-frequency weighting kurtosis for measuring the size of the impact component in a signal. Compared with the traditional kurtosis index, the index has the good characteristics of sensitivity to periodic impact and insensitivity to random impact, and the problem of misselection of the optimal mode component after traditional VMD decomposition is effectively avoided.
(4) According to the method and the device for extracting the impact characteristics of the rolling bearing, the optimal mode component is demodulated by adopting the high-order frequency weighted energy operator, and the anti-interference characteristic is better than that of a Hilbert demodulation (HT) and a Teager Energy Operator (TEO). By increasing the weight of the instantaneous frequency part, harmonic waves and noise components in the signal can be further suppressed, and periodic impact is enhanced.
(5) The embodiment of the application provides an impact feature extraction method and device of a rolling bearing, VMD and a high-order frequency weighted energy operator are combined for extracting impact features, firstly, noise reduction is carried out on signals through VMD decomposition, and the best mode component of the signals is selected; and secondly, suppressing the interference frequency in the signal energy spectrum by a high-order frequency weighted energy operator to achieve the effect of secondary noise reduction, thereby enhancing the periodic impact component in the signal.
Drawings
Fig. 1 is a schematic flow chart of an impact characteristic extraction method for a rolling bearing according to an embodiment of the present application;
fig. 2a and fig. 2b are a time domain waveform diagram and a hilbert transform envelope spectrum of an original vibration signal provided in an embodiment of the present application, respectively;
FIGS. 3a and 3b are a time domain waveform diagram and a Hilbert transform envelope spectrum, respectively, of another original vibration signal provided in an embodiment of the present application;
fig. 4a and 4b are graphs respectively illustrating the energy loss coefficient and the pearson correlation coefficient variation of two original vibration signals provided by the embodiment of the present application;
FIGS. 5 and 6 are time domain waveforms of eigenmode components of two original vibration signals provided by embodiments of the present application;
FIGS. 7a and 7b are time domain waveforms of best mode components of two original vibration signals provided by embodiments of the present application;
FIGS. 8a and 8b are first and second order HFWEO energy spectra of the best mode component of an original vibration signal, respectively, according to embodiments of the present application;
FIGS. 9a and 9b are first and second order HFWEO energy spectra, respectively, of the best mode component of another raw vibration signal provided in embodiments of the present application;
fig. 10 is a schematic structural diagram of an impact characteristic extraction device of a rolling bearing according to an embodiment of the present application.
Detailed Description
In order to make the technical solutions in the embodiments of the present application more comprehensible to those skilled in the art, the following description will be made in detail and completely with reference to the accompanying drawings in the embodiments of the present application. It should be apparent that the embodiments described below are only some of the embodiments of the present application, and not all of them. All other embodiments that can be derived by a person skilled in the art from the embodiments described herein without inventive step are within the scope of the present application.
The inventor of the present application found in research that the Variational Mode Decomposition (VMD) is a new signal decomposition method. Compared with the traditional method EMD, the method has the excellent characteristic of decomposing a non-stationary and non-linear signal into a plurality of single-component eigen mode functions (IMF) signals, and overcomes the defects of mode mixing, overcasting and the like of the EMD. However, VMD contains two key parameters: the number of modes K and the penalty factor α need to be determined manually. Secondly, when the VMD is decomposed into a plurality of IMFs, only a small part of sensitive components containing fault characteristic frequencies exist, and the rest are interference signals containing noise, so that how to select the optimal IMF becomes an important problem for the method to be widely applied.
The VMD algorithm can well and accurately realize the positioning of the resonance frequency band of the vibration signal, and then demodulates the impact envelope in the resonance frequency band through Hilbert transform to realize the identification of the fault impact information. However, when the signal-to-noise ratio of the signal in the resonance frequency band is low and other interference frequencies exist, the hilbert demodulation effect is significantly reduced, and particularly when the amplitude of the interference frequency is large, the main spectral lines in the hilbert envelope spectrum are the interference frequency and its modulated frequency components.
Therefore, the embodiment of the application provides an impact feature extraction method and device of a rolling bearing, VMD and a high-order frequency weighted energy operator are combined for extracting impact features, firstly, noise reduction is carried out on signals through VMD decomposition, and the optimal mode component of the signals is selected; and secondly, suppressing the interference frequency in the signal energy spectrum by a high-order frequency weighted energy operator to achieve the effect of secondary noise reduction, thereby enhancing the periodic impact component in the signal.
Based on the above, in order to clearly and specifically explain the above advantages of the present application, the following description of the embodiments of the present application will be made with reference to the accompanying drawings.
Referring to fig. 1, the figure is a schematic flow chart of an impact feature extraction method for a rolling bearing according to an embodiment of the present application.
The method for extracting the impact characteristics of the rolling bearing provided by the embodiment of the application comprises the following steps:
s101: and collecting an original vibration signal of the rolling bearing.
In one example, a vibration signal of bearing outer ring fault and inner ring fault is used as an original vibration signal, the bearing sampling frequency is 20kHz, the sampling points are 20480, and the rotating speed fr2000rpm, the corresponding theoretical outer and inner ring fault signature frequencies are 236Hz and 295.7Hz, respectively. Time domain waveforms and Hilbert Transform (HT) envelope spectra of the original vibration signal of the bearing outer ring failure and the original vibration signal of the bearing inner ring failure are shown in fig. 2a, fig. 2b, fig. 3a, and fig. 3b, respectively. As can be seen from fig. 2b, the outer ring fault characteristic frequency and the frequency doubling component thereof exist in the bearing outer ring fault vibration signal HT envelope spectrum, but interference components are more in the spectrogram, and more spectral lines with prominent amplitudes are contained in the vicinity of 2-4 times of the bearing outer ring fault characteristic frequency, so that the fault characteristic is not prominent. Similarly, as can be seen from fig. 3b, the envelope spectrum of the fault vibration signal HT of the bearing inner ring includes 1 time and 2 times of the characteristic frequency of the fault of the inner ring, and there are side frequency bands with the center frequency and the rotation frequency as intervals. However, interference components in a spectrogram are more, and the fault characteristic is not prominent due to the fact that the frequency band on the left side of 2 times of the fault characteristic frequency of the inner circle is not obvious.
S102: and decomposing the original vibration signal by utilizing a VMD algorithm to obtain a plurality of eigenmode components.
In the embodiment of the present application, the number K of initial decomposition modalities may be set0And a penalty factor alpha as initial input, typically K0＝3，α＝2000。
In some possible implementation manners of the embodiment of the present application, step S102 may specifically include:
obtaining the most appropriate penalty factor of each mode component according to the mapping relation between the penalty factors and the center frequency of the mode component; based on the most appropriate penalty factors, respectively iterating and updating the mode components, the center frequency and the Lagrange multiplier, and decomposing the original vibration signal into a plurality of eigenmode components by adopting an energy loss coefficient and a Pearson correlation coefficient as iteration termination conditions.
As an example, the most suitable penalty factor for each mode component can be specifically obtained according to the following formula (1):
in the formula, alphakPenalty factor for the k-th mode component, fkcIs the center frequency, f, of the k-th mode componentsIs the signal sampling frequency.
In the specific example described above, the signal sampling frequency fsAfter VMD decomposition, the original vibration signal of the bearing outer ring fault and the original vibration signal of the bearing inner ring fault are respectively as shown in table 1 and table 2 according to the above formula, and the center frequency of each mode component and the penalty factor corresponding to the center frequency can be obtained.
TABLE 1 penalty factor and center frequency for each mode component of the original vibration signal for bearing outer ring faults
TABLE 2 penalty factor and center frequency for each mode component of the original vibration signal for bearing inner race faults
In one example, the energy loss coefficient is calculated according to the following equation:
where xi is the energy loss coefficient and f isOriginal vibration signal ukIs the k-th mode component.
In another example, the pearson correlation coefficient may be calculated according to the following equation (3):
wherein r is the Pearson correlation coefficient, ukFor the k-th mode component,
is the mean of the k-th mode component, f is the original vibration signal,
is the mean of the original vibration signal.
In some possible implementation manners of the embodiment of the present application, the using the energy loss coefficient and the pearson correlation coefficient as iteration termination conditions specifically includes:
when the energy loss coefficient or pearson correlation coefficient of the kth mode component reaches the corresponding threshold, the iteration is stopped.
As an example, the threshold value corresponding to the energy loss coefficient ξ may be set to 0.01 and the threshold value corresponding to the pearson correlation coefficient r may be set to 0.995, while the initial values are set, the initial value of the energy loss coefficient ξ is 1 and the initial value of the pearson correlation coefficient r is 0.5. Stopping decomposition when xi is less than or equal to 0.01 or r is more than or equal to 0.995, thereby obtaining the optimal modal component number of signal decomposition.
Continuing with the specific example described above, the energy loss coefficient and pearson correlation coefficient change curves of the original vibration signal of the bearing outer ring failure and the original vibration signal of the bearing inner ring failure are shown in fig. 4a and 4b, respectively, and it can be found that when K is 7, the decomposition termination condition is reached, i.e., the optimum mode component number K is 7. Time domain waveforms of eigenmode components of the original vibration signal of the bearing outer ring fault and the original vibration signal of the bearing inner ring fault are respectively shown in fig. 5 and 6, and time domain waveforms of the best mode component are respectively shown in fig. 7a and 7 b.
S103: and determining the eigenmode component corresponding to the maximum value of the time-frequency weighted kurtosis from the plurality of eigenmode components as the best mode component.
In some possible implementation manners of the embodiment of the present application, step S103 may specifically include:
obtaining a time domain spectral kurtosis and an envelope spectral kurtosis of a plurality of eigenmode components; obtaining time-frequency weighted kurtosis of a plurality of eigenmode components according to the time-domain spectral kurtosis and the envelope spectral kurtosis; and determining the eigenmode component corresponding to the maximum value of the time-frequency weighted kurtosis as the best mode component.
As an example, the time-domain spectral kurtosis SK may be specifically obtained by the following equation (4):
wherein T is a signal observation interval, f (T) is a signal amplitude corresponding to the time T, and σfIs the standard deviation of the signal f (t),
is the mean of the original vibration signal.
The kurtosis HSK of the envelope spectrum can be specifically obtained by the following formula (5):
wherein y (t) is the envelope spectrum of the original signal f (t) after Hilbert transform,
is the mean value of y (t), σyIs the standard deviation of the signal y (t).
Then, the time-frequency weighted kurtosis TFSK can be obtained by the following formula (6):
TFSK＝log2(1+a×SK+b×HSK) (6)
in the formula, a is a time domain spectrum kurtosis weight coefficient, and b is an envelope spectrum kurtosis weight coefficient. As an example, the temporal kurtosis weight coefficient a and the envelope spectrum kurtosis weight coefficient b may take 0.3 and 0.7, respectively.
With continued reference to the specific example described above, the time-domain spectral kurtosis, the envelope spectral kurtosis, and the time-frequency weighted kurtosis of each mode component after VMD decomposition of the original vibration signal of the bearing outer ring fault and the original vibration signal of the bearing inner ring fault may be specifically as in tables 3 and 4. As can be seen from tables 3 and 4, the original vibration signal of the bearing outer ring fault is decomposed into 7 eigenmode components IMF, where the IMF4 has the largest time-frequency weighted kurtosis value, which is the best mode component. Similarly, the original vibration signal of the bearing inner ring fault is decomposed into 7 eigenmode components IMF, wherein the IMF6 time-frequency weighted kurtosis value is the largest and is the best mode component.
TABLE 3 calculation results of component indexes of each mode of original vibration signals of bearing outer ring faults
TABLE 4 calculation results of component indexes of each mode of original vibration signals of bearing inner ring faults
S104: and demodulating the optimal mode component by using a high-order frequency weighted energy operator to extract periodic impact characteristics in the original vibration signal.
It should be noted that, by demodulating the best mode component using a high-order frequency weighted energy operator (HFWEO), harmonics and noise components in the signal can be suppressed, and periodic impacts in the signal can be enhanced.
As an example, the expression of the higher-order frequency weighted energy operator is specifically the following formula (7):
ξ[f(t),i]＝S[fi(t)]＝|fi(t)+jH[fi(t)]|2＝fi(t)2+H[fi(t)]2 (7)
where f (t) is the best mode component, ξ [ f (t), i]Frequency weighted energy operator of order i for best mode component, i ∈ [1, m]And is an integer, fi(t) is the i-th derivative of the best mode component, H [ f ]i(t)]The hilbert transform of the i-th derivative of the best mode component.
To achieve discrete signal processing, a discrete version of HFWEO is obtained. The m-order derivative of the discrete signal can be obtained by a forward difference formula, and the discrete form of HFWEO is the following formula (8):
ξ[f(n),m]＝|fm(n)2+H[fm(n)]|2＝[fm-1(n+1)-fm-1(n)]2+H[fm-1(n+1)-fm-1(n)]2 (8)
where f (n) is a discrete version of signal f (t), and n is a corresponding amplitude sequence of signal f (t).
In some possible implementation manners of the embodiment of the present application, step S104 may specifically include:
obtaining a derivation order m of a high-order frequency weighted energy operator; performing 1-m order demodulation on the optimal mode component according to an expression of a high-order frequency weighted energy operator, and drawing a corresponding energy spectrum; periodic impact features in the original shock signal are extracted based on the energy spectrum.
In some possible implementations of the embodiment of the present application, a value range of m may be 1 to 9. In practical application, since the HFWEO demodulation effect is deteriorated when m is greater than 2, only the first two-order energy spectrum is calculated here, i.e. m takes values of 1 and 2.
Continuing with the specific example described above, bearing outer ring failure the first and second order HFWEO energy spectra of the best mode components of the original vibration signal of the outer ring failure are shown in fig. 8a and 8 b.
As can be seen from fig. 8a, the larger interference component in the first-order HFWEO energy spectrum is eliminated, however, there are spectral lines with outstanding amplitude values beside 3 times and 4 times of the bearing outer ring fault characteristic frequency. As can be seen from FIG. 8b, spectral lines with prominent amplitude values beside 3 times and 4 times of bearing outer ring fault characteristic frequency in the second-order HFWEO energy spectrum are suppressed, and the fault characteristics are obvious.
The best mode component first and second order HFWEO energy spectra of the original vibration signal for bearing inner race faults are shown in fig. 9a and 9 b. As can be seen from fig. 9a, the spectral lines of the 1-fold and 2-fold inner ring fault characteristic frequencies in the first-order HFWEO energy spectrum are not prominent, and a large number of interference frequencies are accompanied by the spectral lines. As can be seen from FIG. 9b, the amplitude of the fault characteristic frequency of the bearing inner ring is 1 and 2 times higher in the second-order HFWEO energy spectrum, the spectral line with the outstanding amplitude beside the second-order HFWEO energy spectrum is also greatly suppressed, and the fault characteristic is obvious.
The embodiment of the application provides an impact characteristic extraction method of a rolling bearing, VMD and a high-order frequency weighted energy operator are combined for extracting impact characteristics, firstly, primary noise reduction is carried out on signals through VMD decomposition, and the optimal mode component of the signals is selected; and secondly, suppressing the interference frequency in the signal energy spectrum by a high-order frequency weighted energy operator to achieve the effect of secondary noise reduction, thereby enhancing the periodic impact component in the signal.
Based on the method for extracting the impact characteristics of the rolling bearing provided by the embodiment, the embodiment of the application also provides an impact characteristic extraction device of the rolling bearing.
Referring to fig. 10, the figure is a schematic structural diagram of an impact feature extraction device of a rolling bearing according to an embodiment of the present application.
The embodiment of the application provides an impact characteristic extraction element of antifriction bearing includes:
the acquisition unit 100 is used for acquiring an original vibration signal of the rolling bearing;
the decomposition unit 200 is configured to decompose the original vibration signal by using a VMD algorithm to obtain a plurality of eigen-mode components;
a determining unit 300, configured to determine, as a best mode component, an eigenmode component corresponding to a maximum value of the time-frequency weighted kurtosis from the plurality of eigenmode components;
an extracting unit 400, configured to demodulate the best mode component by using a high-order frequency weighted energy operator to extract a periodic impact feature in the original vibration signal.
The embodiment of the application provides an impact feature extraction device of a rolling bearing, wherein a VMD (minimum mean square) and a high-order frequency weighted energy operator are combined for extracting impact features, firstly, the signals are subjected to primary noise reduction through VMD decomposition, and the optimal mode components of the signals are selected; and secondly, suppressing the interference frequency in the signal energy spectrum by a high-order frequency weighted energy operator to achieve the effect of secondary noise reduction, thereby enhancing the periodic impact component in the signal.
The present application has been described in detail with reference to the drawings and examples, but the present application is not limited to the above examples, and various changes can be made within the knowledge of those skilled in the art without departing from the spirit of the present application. The prior art can be used for all the matters not described in detail in this application.
Claims (10)
1. An impact characteristic extraction method of a rolling bearing, characterized by comprising:
collecting an original vibration signal of a rolling bearing;
decomposing the original vibration signal by utilizing a VMD algorithm to obtain a plurality of eigenmode components;
determining an eigenmode component corresponding to the maximum value of the time-frequency weighted kurtosis from the plurality of eigenmode components as a best mode component;
and demodulating the optimal mode component by using a high-order frequency weighted energy operator to extract periodic impact characteristics in the original vibration signal.
2. The method for extracting impact characteristics of a rolling bearing according to claim 1, wherein the decomposing the original vibration signal by using the VMD algorithm to obtain a plurality of eigen-mode components specifically comprises:
obtaining the most appropriate penalty factor of each mode component according to the mapping relation between the penalty factors and the center frequency of the mode component;
and respectively iterating and updating the mode component, the central frequency and the Lagrange multiplier based on the most appropriate penalty factor, and decomposing the original vibration signal into the plurality of eigen-mode components by adopting an energy loss coefficient and a Pearson correlation coefficient as iteration termination conditions.
3. The impact characteristic extraction method of a rolling bearing according to claim 2, wherein the energy loss coefficient is calculated according to the following formula:
where xi is the energy loss coefficient, f is the original vibration signal, ukIs the k-th mode component.
4. The impact characteristic extraction method of a rolling bearing according to claim 2, wherein the pearson correlation coefficient is calculated according to the following formula:
wherein r is the Pearson correlation coefficient, ukFor the k-th mode component,
is the mean of the k-th mode component, f is the original vibration signal,
is the mean value of the original vibration signal.
5. The impact characteristic extraction method of a rolling bearing according to claim 2, wherein the using the energy loss coefficient and the pearson correlation coefficient as iteration termination conditions specifically includes:
when the energy loss coefficient or pearson correlation coefficient of the kth mode component reaches the corresponding threshold, the iteration is stopped.
6. The method for extracting impact characteristics of a rolling bearing according to claim 2, wherein the penalty factor most suitable for each mode component is obtained according to the following formula:
in the formula, alphakPenalty factor for the k-th mode component, fkcIs the center frequency, f, of the k-th mode componentsIs the signal sampling frequency.
7. The method for extracting impact features of a rolling bearing according to any one of claims 1 to 6, wherein the determining, as a best mode component, an eigenmode component corresponding to a time-frequency weighted kurtosis maximum value from the plurality of eigenmode components specifically comprises:
obtaining a time-domain spectral kurtosis and an envelope spectral kurtosis of the plurality of eigenmode components;
obtaining time-frequency weighted kurtosis of the plurality of eigen-mode components according to the time-domain spectral kurtosis and the envelope spectral kurtosis;
and determining the eigenmode component corresponding to the maximum value of the time-frequency weighted kurtosis as the best mode component.
8. The method for extracting impact characteristics of a rolling bearing according to any one of claims 1 to 6, wherein the demodulating the best mode component by using a higher-order frequency weighted energy operator to extract periodic impact characteristics in the original vibration signal comprises:
obtaining a derivation order m of the high-order frequency weighted energy operator;
performing 1-m-order demodulation on the optimal mode component according to the expression of the high-order frequency weighted energy operator, and drawing a corresponding energy spectrum;
and extracting periodic impact characteristics in the original vibration signal based on the energy spectrum.
9. The impact characteristic extraction method of a rolling bearing according to claim 8, wherein the expression of the higher-order frequency weighted energy operator is specifically:
ξ[f(t),i]＝S[fi(t)]＝|fi(t)+jH[fi(t)]|2＝fi(t)2+H[fi(t)]2
where f (t) is the best mode component, ξ [ f (t), i]Weighting the energy operator for the i-order frequency of the best mode component, i ∈ [1, m]And is an integer, fi(t) is the i-th derivative of the best mode component, H [ f ]i(t)]A Hilbert transform of the i-th derivative of the best mode component.
10. An impact characteristic extraction device of a rolling bearing, characterized by comprising:
the acquisition unit is used for acquiring an original vibration signal of the rolling bearing;
the decomposition unit is used for decomposing the original vibration signal by utilizing a VMD algorithm to obtain a plurality of eigen-mode components;
a determining unit, configured to determine, as a best mode component, an eigenmode component corresponding to a maximum value of the time-frequency weighted kurtosis from the plurality of eigenmode components;
and the extraction unit is used for demodulating the optimal mode component by using a high-order frequency weighted energy operator so as to extract the periodic impact characteristics in the original vibration signal.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111552015.5A CN114371005A (en) | 2021-12-17 | 2021-12-17 | Impact characteristic extraction method and device for rolling bearing |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111552015.5A CN114371005A (en) | 2021-12-17 | 2021-12-17 | Impact characteristic extraction method and device for rolling bearing |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114371005A true CN114371005A (en) | 2022-04-19 |
Family
ID=81139462
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111552015.5A Pending CN114371005A (en) | 2021-12-17 | 2021-12-17 | Impact characteristic extraction method and device for rolling bearing |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114371005A (en) |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110295556A1 (en) * | 2008-12-22 | 2011-12-01 | S.P.M. Instrument Ab | Method and apparatus for analysing the condition of a machine having a rotating part |
| US20110301872A1 (en) * | 2008-12-22 | 2011-12-08 | S.P.M. Instrument Ab | Analysis system |
| WO2013115716A1 (en) * | 2012-01-30 | 2013-08-08 | S.P.M. Instrument Ab | Apparatus and method for analysing the condition of a machine having a rotating part |
| CN109633368A (en) * | 2018-12-03 | 2019-04-16 | 三峡大学 | The method of duration power quality disturbances containing distributed power distribution network based on VMD and DFA |
| US20200200648A1 (en) * | 2018-02-12 | 2020-06-25 | Dalian University Of Technology | Method for Fault Diagnosis of an Aero-engine Rolling Bearing Based on Random Forest of Power Spectrum Entropy |
| CN111898556A (en) * | 2020-08-01 | 2020-11-06 | 华东交通大学 | Method for extracting transient impact characteristics of bearings based on CK index consistency |
| CN112200015A (en) * | 2020-09-16 | 2021-01-08 | 昆明理工大学 | Rolling bearing fault diagnosis method based on improved VMD |
| US20210010893A1 (en) * | 2019-01-29 | 2021-01-14 | Nanjing Tech University | Device and method for pipeline leak detection using particle swarm optimization-variational mode decomposition algorithm |
| CN112461547A (en) * | 2020-12-22 | 2021-03-09 | 河北建投能源投资股份有限公司 | Bearing fault diagnosis method based on multi-island inheritance and variation modal decomposition |
| WO2021139331A1 (en) * | 2020-01-08 | 2021-07-15 | 重庆交通大学 | Bearing fault diagnosis method based on instantaneous frequency optimization vmd |
-
2021
- 2021-12-17 CN CN202111552015.5A patent/CN114371005A/en active Pending
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110295556A1 (en) * | 2008-12-22 | 2011-12-01 | S.P.M. Instrument Ab | Method and apparatus for analysing the condition of a machine having a rotating part |
| US20110301872A1 (en) * | 2008-12-22 | 2011-12-08 | S.P.M. Instrument Ab | Analysis system |
| WO2013115716A1 (en) * | 2012-01-30 | 2013-08-08 | S.P.M. Instrument Ab | Apparatus and method for analysing the condition of a machine having a rotating part |
| US20200200648A1 (en) * | 2018-02-12 | 2020-06-25 | Dalian University Of Technology | Method for Fault Diagnosis of an Aero-engine Rolling Bearing Based on Random Forest of Power Spectrum Entropy |
| CN109633368A (en) * | 2018-12-03 | 2019-04-16 | 三峡大学 | The method of duration power quality disturbances containing distributed power distribution network based on VMD and DFA |
| US20210010893A1 (en) * | 2019-01-29 | 2021-01-14 | Nanjing Tech University | Device and method for pipeline leak detection using particle swarm optimization-variational mode decomposition algorithm |
| WO2021139331A1 (en) * | 2020-01-08 | 2021-07-15 | 重庆交通大学 | Bearing fault diagnosis method based on instantaneous frequency optimization vmd |
| CN111898556A (en) * | 2020-08-01 | 2020-11-06 | 华东交通大学 | Method for extracting transient impact characteristics of bearings based on CK index consistency |
| CN112200015A (en) * | 2020-09-16 | 2021-01-08 | 昆明理工大学 | Rolling bearing fault diagnosis method based on improved VMD |
| CN112461547A (en) * | 2020-12-22 | 2021-03-09 | 河北建投能源投资股份有限公司 | Bearing fault diagnosis method based on multi-island inheritance and variation modal decomposition |
Non-Patent Citations (7)
| Title |
|---|
| 刘泽潮，张兵，易彩，吴文逸，黄晨光: "高阶频率加权能量算子在列车轴箱轴承故障诊断中的应用", 西安交通大学学报, vol. 53, no. 12, pages 46 - 56 * |
| 岳建辉: "基于变分模态分解的滚动轴承故障诊断方法研究", 中国优秀博硕士学位论文全文数据库, pages 2 * |
| 张龙;蔡秉桓;吴佳敏;熊国良;: "滚动轴承故障特征自适应增强的相关峭度联合降噪方法", 机械设计与研究, no. 02 * |
| 徐元博; 蔡宗琰; 胡永彪; 丁凯: "强噪声背景下频率加权能量算子和变分模态分解在轴承故障提取中的应用", 振动工程学报, vol. 31, no. 3, pages 513 - 522 * |
| 江志农;胡明辉;冯坤;贺雅;: "阶次跟踪能量算子与奇异值分解结合的滚动轴承故障诊断", 轴承, no. 11 * |
| 王琇峰，文俊: "基于噪声信号和改进VMD的滚动轴承故障诊断", 噪声与振动控制, vol. 41, no. 2, pages 118 - 124 * |
| 马增强; 李亚超; 刘政; 谷朝健: "基于变分模态分解和Teager能量算子的滚动轴承故障特征提取", 振动与冲击, vol. 35, no. 13, pages 134 - 139 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Li et al. | Research on test bench bearing fault diagnosis of improved EEMD based on improved adaptive resonance technology | |
| Yin et al. | Weak fault feature extraction of rolling bearings based on improved ensemble noise-reconstructed EMD and adaptive threshold denoising | |
| Zheng et al. | Incipient fault detection of rolling bearing using maximum autocorrelation impulse harmonic to noise deconvolution and parameter optimized fast EEMD | |
| Cheng et al. | Adaptive multipoint optimal minimum entropy deconvolution adjusted and application to fault diagnosis of rolling element bearings | |
| Li et al. | An enhanced rolling bearing fault detection method combining sparse code shrinkage denoising with fast spectral correlation | |
| CN109883706B (en) | Method for extracting local damage weak fault features of rolling bearing | |
| Liu et al. | A novel fault diagnosis approach for rolling bearing based on high-order synchrosqueezing transform and detrended fluctuation analysis | |
| Guo et al. | Generalized variational mode decomposition: A multiscale and fixed-frequency decomposition algorithm | |
| CN106404399B (en) | Method for Bearing Fault Diagnosis based on self-adaptive redundant Lifting Wavelet packet decomposition tree | |
| Cui et al. | An investigation of rolling bearing early diagnosis based on high-frequency characteristics and self-adaptive wavelet de-noising | |
| CN111769810B (en) | Fluid mechanical modulation frequency extraction method based on energy kurtosis spectrum | |
| Wang et al. | Weak fault diagnosis of rolling bearing under variable speed condition using IEWT-based enhanced envelope order spectrum | |
| Zhang et al. | Feature extraction by enhanced analytical mode decomposition based on order statistics filter | |
| CN103699513A (en) | Stochastic resonance method based on multi-scale noise adjustment | |
| CN104374575A (en) | Wind turbine main bearing fault diagnosis method based on blind source separation | |
| CN106908232A (en) | A kind of vibration signal fault recognition method of rotating machinery | |
| CN113033304A (en) | Multi-resonance-band amplitude demodulation analysis method for overcoming frequency domain overlapping interference | |
| Liang et al. | Periodicity measure of cyclo-stationary impulses based on low sparsity of Gini index and its application to bearing diagnosis | |
| Wang et al. | A novel optimal demodulation frequency band extraction method of fault bearing based on power spectrum screening combination-gram | |
| Li et al. | Incipient detection of bearing fault using impulse feature enhanced weighted sparse representation | |
| Hua et al. | The methodology of modified frequency band envelope kurtosis for bearing fault diagnosis | |
| Zhao et al. | Peak envelope spectrum Fourier decomposition method and its application in fault diagnosis of rolling bearings | |
| CN117349615B (en) | Self-adaptive enhancement envelope spectrum method for fault diagnosis of rolling bearing under strong noise condition | |
| CN117906958A (en) | Bearing fault diagnosis method based on bivariate weighted mapping diagram | |
| CN114371005A (en) | Impact characteristic extraction method and device for rolling bearing |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |