CN113449630A - Bearing fault diagnosis method, system and medium for improving modulation double spectrum - Google Patents
Bearing fault diagnosis method, system and medium for improving modulation double spectrum Download PDFInfo
- Publication number
- CN113449630A CN113449630A CN202110709423.0A CN202110709423A CN113449630A CN 113449630 A CN113449630 A CN 113449630A CN 202110709423 A CN202110709423 A CN 202110709423A CN 113449630 A CN113449630 A CN 113449630A
- Authority
- CN
- China
- Prior art keywords
- kurtosis
- imfs
- signal
- correlation
- bispectrum
- 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
- 238000000034 method Methods 0.000 title claims abstract description 55
- 238000003745 diagnosis Methods 0.000 title claims abstract description 27
- 238000001228 spectrum Methods 0.000 title claims abstract description 11
- 238000004364 calculation method Methods 0.000 claims abstract description 22
- 238000010183 spectrum analysis Methods 0.000 claims abstract description 8
- 238000000354 decomposition reaction Methods 0.000 claims description 43
- 238000000605 extraction Methods 0.000 claims description 11
- 238000004458 analytical method Methods 0.000 claims description 10
- 238000012935 Averaging Methods 0.000 claims description 8
- 238000003860 storage Methods 0.000 claims description 6
- 238000010606 normalization Methods 0.000 claims description 3
- 238000005096 rolling process Methods 0.000 abstract description 16
- 230000007547 defect Effects 0.000 abstract description 3
- 238000010586 diagram Methods 0.000 description 15
- 230000003044 adaptive effect Effects 0.000 description 11
- 230000006870 function Effects 0.000 description 8
- 230000008569 process Effects 0.000 description 8
- 230000003595 spectral effect Effects 0.000 description 8
- 238000004590 computer program Methods 0.000 description 7
- 230000004927 fusion Effects 0.000 description 7
- 238000012360 testing method Methods 0.000 description 7
- 230000000694 effects Effects 0.000 description 5
- 238000005070 sampling Methods 0.000 description 5
- 238000012545 processing Methods 0.000 description 4
- 230000001133 acceleration Effects 0.000 description 2
- 238000002474 experimental method Methods 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 238000003672 processing method Methods 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 230000009977 dual effect Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000000737 periodic effect Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 230000001052 transient effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/08—Feature extraction
- G06F2218/10—Feature extraction by analysing the shape of a waveform, e.g. extracting parameters relating to peaks
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/22—Source localisation; Inverse modelling
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Artificial Intelligence (AREA)
- Signal Processing (AREA)
- General Engineering & Computer Science (AREA)
- Acoustics & Sound (AREA)
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
Abstract
The invention relates to a bearing fault diagnosis method, a system and a medium for improving modulation double spectrums, which comprise the following steps: decomposing an original vibration signal of the bearing by adopting a weighted self-adaptive white noise average ensemble empirical mode to obtain a series of IMFs; calculating the kurtosis and the correlation coefficient of each IMFs; obtaining a weighted average correlation-kurtosis value of each IMFs according to the calculation result, and reconstructing a signal; and carrying out modulation double-spectrum analysis on the reconstructed signal, and extracting fault characteristic frequency. The invention overcomes the defects of the traditional bispectrum, can effectively detect the nonlinear components in the signals, can inhibit the influence of noise and accurately extract the fault characteristics of the rolling bearing; the invention can be widely applied to the technical field of fault diagnosis of mechanical equipment.
Description
Technical Field
The invention relates to the technical field of mechanical equipment fault diagnosis, in particular to a bearing fault diagnosis method based on weighted self-adaptive white noise average overall empirical mode decomposition and modulation bispectrum fusion.
Background
At present, a practical method for detecting bearing faults is a vibration analysis method, various signal processing methods are applied to bearing fault diagnosis, Empirical Mode Decomposition (EMD) can effectively decompose nonstationary signals into a series of intrinsic mode components (IMFs) according to the self characteristics of the nonstationary signals, and problems of mode aliasing, false components and the like easily occur in the decomposition process; lumped mean-time empirical mode decomposition (EEMD) is a signal processing method, which is generally used to process non-stationary, non-linear signals and has been widely noticed in fault diagnosis, but EEMD decomposition has the problem of excessive residual components; the subsequent adaptive white noise average ensemble empirical mode decomposition (CEEMDAN) is an improvement on the method, and effectively overcomes the problems of end point effect and modal aliasing, and meanwhile, the residual noise signal is less.
In recent years, a Modulation Signal Bispectrum (MSB) method is applied to the field of fault diagnosis, which is an improvement of the conventional Bispectrum, and not only overcomes the disadvantages of the conventional Bispectrum, but also considers the information of high and low side bands, and can effectively analyze the Modulation Signal.
When CEEMDAN is used according to a traditional method, more selected IMFs are components with information as the leading component, so that the effective information in the IMFs cannot be fully utilized and the like, and the rolling bearing fault feature extraction method based on weighted adaptive white noise average ensemble empirical mode decomposition (WACEEMDAN) and Modulation Signal Bispectrum (MSB) is provided.
Disclosure of Invention
In view of the above problems, an object of the present invention is to provide a bearing fault diagnosis method, system and medium for weighted adaptive white noise average ensemble empirical mode decomposition and modulation bispectrum fusion, which overcome the disadvantages of the conventional bispectrum, can effectively detect nonlinear components in signals, can suppress the influence of noise, and accurately extract the fault characteristics of a rolling bearing.
In order to achieve the purpose, the invention adopts the following technical scheme: a method of improving modulated bispectrum bearing fault diagnosis, comprising: step 1, decomposing an original vibration signal of a bearing by adopting a weighted self-adaptive white noise average ensemble empirical mode to obtain a series of IMFs; step 2, calculating the kurtosis and the correlation coefficient of each IMFs; step 3, obtaining a weighted average correlation-kurtosis value of each IMFs according to the calculation result of the step 2, and reconstructing a signal; and 4, carrying out modulation double-spectrum analysis on the reconstructed signal, and extracting fault characteristic frequency.
Further, in the step 2, a correlation coefficient ρ of each IMFsxyComprises the following steps:
wherein N is the number of decompositions, yiRepresenting the ith original vibration signal and,
representing the average, x, of all the original vibration signalsiRepresenting the IMF component;
the mean value of the IMF components is represented.
Further, in step 2, the kurtosis q of each IMFs is calculated as:
wherein x isiRepresenting the IMF component;
the mean value of the IMF components is represented.
Further, in the step 3, the calculating and reconstructing include the following steps:
step 3.1, calculating the product of kurtosis and correlation coefficient of each IMFs, namely a correlation-kurtosis index s (i);
step 3.2, calculating average correlation-kurtosis S (i) according to the correlation-kurtosis index s (i);
and 3.3, carrying out weighted reconstruction on the decomposed signals according to the average correlation-kurtosis S (i):
further, in step 3.2, the mean correlation-kurtosis s (i):
where s (i) represents the correlation-kurtosis value of the ith IMF component.
Further, in step 4, the modulation bispectrum is modified by eliminating the substantial influence of the carrier frequency component through amplitude normalization, and the modified modulation bispectrum sideband estimator is written as MSB-SE:
wherein, BMS(fc0) represents fxThe square estimate of the power spectrum when 0.
Further, in the step 4, a modulation bispectrum detector is arranged; by sideband estimator
And obtaining a modulation dual-spectrum detector to improve the reliability of fault feature extraction.
Further, the method for acquiring the modulation bispectrum detector comprises the following steps:
the amplitude of the bispectrum of the modulated signal is distributed in the frequency direction, and f is sliced to obtain the carrier frequencyxEffective modulation of direction dual-spectral amplitude averaging:
wherein Δ f represents fxFrequency resolution of the direction; b (f)c) Representing carrier frequency slice, M is the total number of effective amplitude values;
averaging the modulated bispectral slices therein to obtain a modulated bispectral detector B (f)x)：
L represents the total number of slices selected.
A modulated bispectral bearing fault diagnosis system comprising: the device comprises a decomposition module, a first calculation module, a second calculation module and an analysis and extraction module; the decomposition module is used for decomposing the original vibration signal of the bearing by adopting a weighted self-adaptive white noise average ensemble empirical mode to obtain a series of IMFs; the first calculating module is used for calculating the kurtosis and the correlation coefficient of each IMFs; the second calculation module obtains a weighted average correlation-kurtosis value of each IMFs according to the calculation result of the first calculation module, and reconstructs a signal; and the analysis and extraction module is used for carrying out modulation double-spectrum analysis on the reconstructed signal and extracting fault characteristic frequency.
A computer readable storage medium storing one or more programs, the one or more programs comprising instructions, which when executed by a computing device, cause the computing device to perform any of the methods as described above.
Due to the adoption of the technical scheme, the invention has the following advantages:
1. the weighting method based on the correlation-kurtosis index can highlight the sensitive IMFs.
2. The modulation bispectrum method adopted by the invention overcomes the defects of the traditional bispectrum, also considers the information of high and low side frequency bands, can effectively detect the nonlinear component in the signal, can inhibit the influence of noise, and can clearly reflect the demodulated modulation component.
3. Compared with the rapid spectral kurtosis in the prior art, the method for weighted self-adaptive white noise average overall empirical mode decomposition and modulation bispectrum fusion has better superiority and can more accurately extract the fault characteristics of the rolling bearing.
Drawings
FIG. 1 is a schematic overall flow diagram of the fault diagnosis method of the present invention;
FIG. 2 is a time domain diagram of a bearing in embodiment 1 of the present invention;
FIG. 3a is a graph of the kurtosis of each component in example 1 of the present invention;
FIG. 3b is a graph of correlation coefficients of components in embodiment 1 of the present invention;
FIG. 4a is a diagram showing the analysis results of the modulation bispectral sideband estimator in embodiment 1 of the invention;
FIG. 4b is a graph of the results of the modulated bispectrum detector in example 1 of the present invention;
FIG. 5a is a graph of the fast spectral kurtosis in example 1 of the present invention;
FIG. 5b is a graph of the results of the modulated bispectrum detector in example 1 of the present invention;
FIG. 6 is a time domain diagram of a bearing in embodiment 2 of the present invention;
FIG. 7a is a graph of the kurtosis of each component in example 2 of the present invention;
FIG. 7b is a graph of correlation coefficients of components in embodiment 2 of the present invention;
FIG. 8a is a diagram showing the analysis results of the modulation bispectral sideband estimator in embodiment 2 of the invention;
FIG. 8b is a graph of the results of the modulated bispectrum detector in example 2 of the present invention;
FIG. 9a is a fast spectral kurtosis graph of the spectral kurtosis method of embodiment 2 of the present invention;
FIG. 9b is an envelope spectrum of the spectral kurtosis method of example 2 of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the drawings of the embodiments of the present invention. It is to be understood that the embodiments described are only a few embodiments of the present invention, and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the described embodiments of the invention, are within the scope of the invention.
The invention provides a rolling bearing fault diagnosis method based on weighted adaptive white noise average general empirical mode decomposition and modulation signal bispectrum fusion, which is characterized in that a rolling bearing fault impact signal of an electromechanical system has the characteristics of non-stationarity and easiness in being submerged by noise, and in order to improve the extraction rate of the traditional adaptive white noise average general empirical mode decomposition (CEEMDAN) on a bearing fault signal and reduce the distortion of a reconstructed signal. Firstly, decomposing an acquired vibration signal into a series of intrinsic mode components (IMFs) by using weighted self-adaptive white noise average ensemble empirical mode decomposition; then constructing a novel index for emphasizing sensitive components, and weighting each IMFs by using the index and reconstructing signals; and finally, carrying out modulation double-spectrum analysis on the reconstructed signal and extracting the bearing fault characteristic frequency. The electromechanical system bearing fault characteristic frequency extracted by the fault diagnosis method is more accurate.
In a first embodiment of the present invention, a bearing fault diagnosis method based on weighted adaptive white noise average ensemble empirical mode decomposition and modulation signal bispectrum fusion is provided, as shown in fig. 1, which includes the following steps:
in the embodiment, vibration signals of the electromechanical system bearing are acquired through the acceleration sensor.
And 2, calculating the kurtosis and the correlation coefficient of each IMFs.
the product of the kurtosis value of each IMFs and the correlation coefficient value is called a correlation-kurtosis index, and after the indexes are summed and averaged, each IMFs is weighted to obtain a weighted average correlation-kurtosis value.
And 4, carrying out modulation double-spectrum analysis on the reconstructed signal, and extracting fault characteristic frequency.
In the step 1, white noise is adaptively added in the traditional CEEMDAN overall decomposition process, and the IMF component is obtained through the unique residual signal, so that the problem of modal aliasing is solved to a great extent, but the IMF component with dominant information is selected in actual use, and then the IMF component with dominant noise is abandoned, so that effective information in the IMFs cannot be fully utilized, and the overall signal noise reduction effect is influenced. Therefore, a weighted adaptive white noise average ensemble empirical mode decomposition is employed in the present embodiment.
The weighted self-adaptive white noise average ensemble empirical mode decomposition is that white noise is added to a vibration signal in a self-adaptive mode according to the characteristics of the vibration signal at each stage of vibration signal decomposition, and an IMF component is obtained by calculating a unique residual signal, so that the mode aliasing problem can be effectively solved, and the completeness and the decomposition efficiency of the decomposition process are improved.
In the weighted self-adaptive white noise average ensemble empirical mode decomposition method, the input signal is xi(n) adaptively adding a noise coefficient εk，wi(N) white noise following N (0, 1) distribution, and Ek(. The) is defined as the k-th order IMF component obtained by EMD decomposition, i.e. the IMFkDefined as the k-th order IMF component obtained by weighted adaptive white noise average ensemble empirical mode decomposition, the signal decomposition steps are as follows:
step 1.1, for input signal xi(n)＝x(n)+ε0wi(N) performing EMD decomposition, where i is 1,2, and N, N is the decomposition number of the whole process, and obtaining the IMF component of order 1 through EMD decomposition, that is:
wherein, x (n) is an original signal;
step 1.2, computing residual signal component r1(n)：
r1(n)＝x(n)-IMF1(n) (2)
Step 1.3, in the process of performing decomposition for N times, performing EMD decomposition on the input signal each time, and then the 2 nd order IMF component is:
step 1.4, calculating the remaining decomposition process by using step 1.3, and obtaining a k +1 order modal component by using a kth residual component, where the kth residual component and the modal component are:
rk(n)＝rk-1(n)-IMFn(n) (4)
step 1.5, when the number of extreme points of the residual signal is less than or equal to 2, at this time, decomposition is completed, and when decomposition is completed, a K-order IMF component is obtained, at this time, the original signal x (n) can be represented as:
in the step 2, the method for calculating the correlation coefficient and kurtosis value of each IMFs includes:
(1) kurtosis: the kurtosis can reflect the transient degree of a signal, and the index can effectively detect the fault impact characteristic.
The kurtosis q is calculated as:
wherein x isiRepresenting the IMF component;
the mean value of the IMF components is represented.
(2) Correlation coefficient: calculating the correlation coefficient of each IMFs component and the original vibration signal, and correlating the cross-correlation coefficient rho of the original vibration signal y (n) and the signal x (n)xyIs defined as:
in the formula, yiRepresenting the ith original vibration signal and,
representing the average of all the original vibration signals.
In the step 3, the calculation and reconstruction includes the following steps:
step 3.1, calculating the product of kurtosis and correlation coefficient of each IMFs, which is called correlation-kurtosis index s (i):
s(i)＝q(i)×ρ(i) (9)
step 3.2, calculating the average correlation-kurtosis S (i) according to the correlation-kurtosis index s (i):
where s (i) represents the correlation-kurtosis value of the ith IMF component.
And 3.3, carrying out weighted reconstruction on the decomposed signals according to the average correlation-kurtosis S (i):
in the step 4, the inherent modulation component in the signal can be demodulated by suppressing the influence of noise and interference components by using a modulation bispectrum method, and according to the discrete fourier transform form x (f) of the original signal x (n), the modulation bispectrum can be represented as:
BMS(fc，fx)＝E<X(fc+fx)X(fc-fx)X*(fc)X*(fc)> (12)
in the formula, BMS(fc,fx) Bispectrum of modulated signal, E, of original signal x (t)<·>Is a mathematical expectation; f. ofcIs the carrier frequency; f. ofxTo modulate frequency, (f)c+fx) And (f)c-fx) Respectively, the frequencies of the high and low sidebands.
Modulating the total phase phi of the dual spectrumMS(fc,fx) Comprises the following steps:
ΦMS(fc,fx)＝Φ(fc+fx)+Φ(fc-fx)-Φ(fc)-Φ(fc) (13)
when f iscAnd fxWhen coupled, its phase is:
bringing equation (13) into equation (14) yields a modulation bispectrum with a total phase of 0 and whose amplitude determination is calculated from the product of the amplitudes of the 4 components, so that at (f)c，fx) Where a double spectral peak occurs. In addition, if the components of the random noise are randomly distributed rather than coupled, the value of the modulation bispectrum will be close to 0. Therefore, the double spectrum of the modulation signal can well inhibit random noise and non-periodic components in the signal, thereby more clearly representing components related to the modulation effect.
In order to quantify the sideband amplitude more accurately, the modulated bispectrum may be modified by amplitude normalization to eliminate the substantial effect of the carrier frequency component, and in order to distinguish the modified modulated bispectrum from the original modulated bispectrum, the modified modulated bispectrum sideband estimator may be written as MSB-SE, as follows:
wherein, BMS(fc0) represents fxThe square estimate of the power spectrum when 0.
In the step 4, a method of setting a modulation bispectrum detector is further included. By sideband estimator
The modulation bispectrum detector can further improve the reliability of fault feature extraction.
In the above embodiment, the modulation bispectrum sideband estimator is used to select the optimal carrier frequency range, and the slice averaging of the modulation bispectrum can improve the modulation bispectrum detector, so that the bearing fault characteristic frequency and the frequency multiplication can be clearly observed.
Wherein the amplitude of the bispectrum of the modulated signal is distributed along the frequency direction, and f is sliced to obtain the carrier frequencyxEffective modulation of direction dual-spectral amplitude averaging:
wherein Δ f represents fxFrequency resolution of the direction; b (f)c) Representing the carrier frequency slice, and M is the total number of effective amplitudes.
To make the results more reliable, a modulated bispectral detector B (f) may be retrieved by averaging the modulated bispectral slices thereinx)：
L represents the total number of slices selected.
In a second embodiment of the present invention, a bearing fault diagnosis system of weighted adaptive white noise average ensemble empirical mode decomposition and modulation signal bispectrum fusion is provided, which comprises: the device comprises a decomposition module, a first calculation module, a second calculation module and an analysis and extraction module;
the decomposition module is used for decomposing the original vibration signal of the bearing by adopting a weighted self-adaptive white noise average ensemble empirical mode to obtain a series of IMFs;
the first calculation module is used for calculating the kurtosis and the correlation coefficient of each IMFs;
the second calculation module is used for obtaining a weighted average correlation-kurtosis value of each IMFs according to the calculation result of the first calculation module and reconstructing a signal;
and the analysis and extraction module is used for carrying out modulation double-spectrum analysis on the reconstructed signal and extracting fault characteristic frequency.
In a third embodiment of the invention, there is provided a computer readable storage medium storing one or more programs, the one or more programs comprising instructions, which when executed by a computing device, cause the computing device to perform any of the methods as in the first embodiment.
Example 1:
vibration data of a rolling bearing life test released by the Xian traffic university-Shengyang science and technology combined laboratory are adopted. The model of a test rolling bearing selected by the test bed is LDK UER204, relevant parameters of the test rolling bearing are shown in Table 1, the model of a selected acceleration sensor is PCB 352C33, the two sensors are respectively fixed at the horizontal position and the vertical position of a bearing to be tested, the model of a signal collector for collecting vibration signals is DT9837, the rotating speed of a main shaft during testing is 2100r/min, the radial force is 10KN, the testing time set sampling frequency is 25.6kHz, the sampling interval is 1min, and the sampling time is 1.28s each time.
TABLE 1 Main parameters of the Rolling bearings
Since the horizontal position contains more vibration state information than the vertical position due to the force applied in the horizontal position, the fault characteristic frequency of the bearing can be obtained according to the data of table 1 by using the data of the horizontal vibration as the original vibration signal analyzed by the bearing.
Bearing rotation frequency frComprises the following steps:
failure frequency f of bearing inner raceiComprises the following steps:
failure frequency f of bearing outer ringoComprises the following steps:
frequency f of rolling element failurebComprises the following steps:
wherein d is the diameter of the rolling body; d is the pitch diameter of the bearing; alpha is the contact angle of the rolling body; z is the number of rolling elements.
Taking an outer ring fault of the rolling bearing as an example, partial data is extracted as fault data of the bearing, and an extracted signal time domain diagram is shown in fig. 2.
Then, the signal is decomposed by weighted adaptive white noise average ensemble empirical mode decomposition to obtain 16 intrinsic mode functions, and correlation coefficients and kurtosis values of the intrinsic mode functions are calculated, and the results are shown in fig. 3a and 3 b.
Calculating correlation-kurtosis indexes, reconstructing signals, analyzing the reconstructed signals by a modulation bispectrum method, wherein the analysis result is shown in fig. 4a and 4b, a modulation bispectrum sideband estimator is used for selecting an optimal carrier frequency range, and a modulation bispectrum detector can be improved by averaging slices of a modulation bispectrum, so that bearing fault characteristic frequency and frequency multiplication can be clearly observed, and the effectiveness of the method is illustrated.
In order to prove the superiority of the method, a rapid spectral kurtosis method is used for analyzing signals, and the results are shown in fig. 5a and 5 b.
Example 2:
the bearing fault simulation test bed adopted in the embodiment consists of three parts, namely a driving part, a loading part and a fault bearing. The rolling bearing with the outer ring having a fault is adopted for an experiment, the selected bearing is a cylindrical roller bearing with the model number of N1004, basic parameters are shown in the table 2, in the experiment, the set sampling frequency is 12000Hz, the sampling time duration is 10s, the extracted signal time domain diagram is shown in fig. 6, the actual rotating speed of the motor is 1210r/min, and the outer ring fault characteristic frequency is 99.63Hz according to a corresponding calculation formula.
TABLE 2N 1004 Cylinder roller bearing Main parameters
Then, the signal is decomposed by a weighted adaptive white noise average ensemble empirical mode decomposition algorithm to obtain 19 intrinsic mode functions, and correlation coefficients and kurtosis values of the intrinsic mode functions are calculated, and the results are shown in fig. 7a and 7 b.
And calculating a correlation-kurtosis index, reconstructing a signal, analyzing the reconstructed signal by a modulation bispectrum method, and clearly observing bearing fault characteristic frequency and frequency multiplication as shown in fig. 8a and 8b, thereby illustrating the effectiveness of the method.
The results of the fast spectral kurtosis method are shown in fig. 9a and fig. 9b, and comparison shows that the method disclosed by the invention has better effect and can more clearly extract the fault characteristics of the bearing.
In conclusion, the bearing fault diagnosis method based on weighted self-adaptive white noise average ensemble empirical mode decomposition and modulation signal bispectrum fusion can highlight sensitive IMFs; the adopted modulation bispectrum method overcomes the defects of the traditional bispectrum, considers the information of high and low side frequency bands, can effectively detect the nonlinear components in the signals, can inhibit the influence of noise, and can clearly reflect the demodulated modulation components.
As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
Claims (10)
1. A bearing fault diagnosis method for improving modulation bispectrum is characterized by comprising the following steps:
step 1, decomposing an original vibration signal of a bearing by adopting a weighted self-adaptive white noise average ensemble empirical mode to obtain a series of IMFs;
step 2, calculating the kurtosis and the correlation coefficient of each IMFs;
step 3, obtaining a weighted average correlation-kurtosis value of each IMFs according to the calculation result of the step 2, and reconstructing a signal;
and 4, carrying out modulation double-spectrum analysis on the reconstructed signal, and extracting fault characteristic frequency.
2. The bearing fault diagnosis method according to claim 1, wherein in the step 2, the correlation coefficient p of each IMFsxyComprises the following steps:
wherein N is the number of decompositions, yiRepresenting the ith original vibration signal and,
representing the average, x, of all the original vibration signalsiRepresenting the IMF component;
the mean value of the IMF components is represented.
3. The bearing fault diagnosis method as claimed in claim 1, wherein in the step 2, the kurtosis q of each IMFs is calculated by:
wherein x isiRepresenting the IMF component;
the mean value of the IMF components is represented.
4. The bearing fault diagnosis method according to claim 1, wherein in the step 3, the calculating and reconstructing includes the steps of:
step 3.1, calculating the product of kurtosis and correlation coefficient of each IMFs, namely a correlation-kurtosis index s (i);
step 3.2, calculating average correlation-kurtosis S (i) according to the correlation-kurtosis index s (i);
and 3.3, carrying out weighted reconstruction on the decomposed signals according to the average correlation-kurtosis S (i):
5. a bearing fault diagnosis method as claimed in claim 4, wherein in step 3.2, the mean correlation-kurtosis S (i):
where s (i) represents the correlation-kurtosis value of the ith IMF component.
6. The bearing fault diagnosis method according to claim 1, characterized in that in step 4, the modulation bispectrum is modified by amplitude normalization to eliminate substantial influence of the carrier frequency component, and the modified modulation bispectrum sideband estimator is written as MSB-SE:
wherein, BMS(fc0) represents fxThe square estimate of the power spectrum when 0.
7. The bearing fault diagnosis method according to claim 6, wherein in the step 4, a modulation bispectrum detector is provided; by sideband estimator
And obtaining a modulation dual-spectrum detector to improve the reliability of fault feature extraction.
8. The bearing fault diagnosis method according to claim 7, wherein the modulated bispectrum detector is obtained by:
the amplitude of the bispectrum of the modulated signal is distributed in the frequency direction, and f is sliced to obtain the carrier frequencyxEffective modulation of direction dual-spectral amplitude averaging:
wherein Δ f represents fxFrequency resolution of the direction; b (f)c) Representing carrier frequency slice, M is the total number of effective amplitude values;
averaging the modulated bispectral slices therein to obtain a modulated bispectral detector B (f)x)：
L represents the total number of slices selected.
9. A modulated bispectral bearing fault diagnosis system, comprising: the device comprises a decomposition module, a first calculation module, a second calculation module and an analysis and extraction module;
the decomposition module is used for decomposing the original vibration signal of the bearing by adopting a weighted self-adaptive white noise average ensemble empirical mode to obtain a series of IMFs;
the first calculating module is used for calculating the kurtosis and the correlation coefficient of each IMFs;
the second calculation module obtains a weighted average correlation-kurtosis value of each IMFs according to the calculation result of the first calculation module, and reconstructs a signal;
and the analysis and extraction module is used for carrying out modulation double-spectrum analysis on the reconstructed signal and extracting fault characteristic frequency.
10. A computer readable storage medium storing one or more programs, the one or more programs comprising instructions, which when executed by a computing device, cause the computing device to perform any of the methods of claims 1-8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110709423.0A CN113449630A (en) | 2021-06-25 | 2021-06-25 | Bearing fault diagnosis method, system and medium for improving modulation double spectrum |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110709423.0A CN113449630A (en) | 2021-06-25 | 2021-06-25 | Bearing fault diagnosis method, system and medium for improving modulation double spectrum |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113449630A true CN113449630A (en) | 2021-09-28 |
Family
ID=77812787
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110709423.0A Pending CN113449630A (en) | 2021-06-25 | 2021-06-25 | Bearing fault diagnosis method, system and medium for improving modulation double spectrum |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113449630A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113776836A (en) * | 2021-10-25 | 2021-12-10 | 长沙理工大学 | Self-adaptive synchronous average bearing fault quantitative diagnosis method |
| CN117112984A (en) * | 2023-08-15 | 2023-11-24 | 北京理工大学珠海学院 | Fault diagnosis method for belt transmission system of non-invasive lathe |
| CN118094477A (en) * | 2024-04-25 | 2024-05-28 | 山东大学 | Signal processing method and system based on multi-sensor information fusion |
-
2021
- 2021-06-25 CN CN202110709423.0A patent/CN113449630A/en active Pending
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113776836A (en) * | 2021-10-25 | 2021-12-10 | 长沙理工大学 | Self-adaptive synchronous average bearing fault quantitative diagnosis method |
| CN113776836B (en) * | 2021-10-25 | 2024-01-02 | 长沙理工大学 | Self-adaptive synchronous average bearing fault quantitative diagnosis method |
| CN117112984A (en) * | 2023-08-15 | 2023-11-24 | 北京理工大学珠海学院 | Fault diagnosis method for belt transmission system of non-invasive lathe |
| CN117112984B (en) * | 2023-08-15 | 2024-05-03 | 北京理工大学珠海学院 | Fault diagnosis method for belt transmission system of non-invasive lathe |
| CN118094477A (en) * | 2024-04-25 | 2024-05-28 | 山东大学 | Signal processing method and system based on multi-sensor information fusion |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Hemmati et al. | Roller bearing acoustic signature extraction by wavelet packet transform, applications in fault detection and size estimation | |
| CN113449630A (en) | Bearing fault diagnosis method, system and medium for improving modulation double spectrum | |
| WO2019179340A1 (en) | Eemd- and msb-based failure feature extraction method for rolling-element bearing | |
| Cong et al. | Short-time matrix series based singular value decomposition for rolling bearing fault diagnosis | |
| Yan et al. | Improved Hilbert–Huang transform based weak signal detection methodology and its application on incipient fault diagnosis and ECG signal analysis | |
| Zhang et al. | Feature extraction method based on adaptive and concise empirical wavelet transform and its applications in bearing fault diagnosis | |
| Feng et al. | Time-varying demodulation analysis for rolling bearing fault diagnosis under variable speed conditions | |
| Yu et al. | An adaptive sensitive frequency band selection method for empirical wavelet transform and its application in bearing fault diagnosis | |
| Hao et al. | Morphological undecimated wavelet decomposition for fault diagnostics of rolling element bearings | |
| Nikolaou et al. | Rolling element bearing fault diagnosis using wavelet packets | |
| Qu et al. | A new acoustic emission sensor based gear fault detection approach | |
| JP6133112B2 (en) | Rolling bearing diagnostic device and rolling bearing diagnostic method | |
| Du et al. | Feature identification with compressive measurements for machine fault diagnosis | |
| CN109029999B (en) | Rolling bearing fault diagnosis method based on enhanced modulation bispectrum analysis | |
| Klausen et al. | Multi-band identification for enhancing bearing fault detection in variable speed conditions | |
| Klein et al. | Bearing diagnostics using image processing methods | |
| Guo et al. | An enhanced modulation signal bispectrum analysis for bearing fault detection based on non-Gaussian noise suppression | |
| Zhang et al. | A two-step denoising strategy for early-stage fault diagnosis of rolling bearings | |
| Zheng et al. | Faults diagnosis of rolling bearings based on shift invariant K-singular value decomposition with sensitive atom nonlocal means enhancement | |
| Zhang et al. | Improved local cepstrum and its applications for gearbox and rolling bearing fault detection | |
| CN107831013A (en) | A kind of Method for Bearing Fault Diagnosis for strengthening cyclic bispectrum using probability principal component analysis | |
| Lee | A weighting function for improvement of spectral coherence based envelope spectrum | |
| Kumar et al. | A combined approach for weak fault signature extraction of rolling element bearing using Hilbert envelop and zero frequency resonator | |
| CN114486263B (en) | Noise reduction demodulation method for vibration signal of rolling bearing of rotary machine | |
| CN112098093A (en) | Bearing fault feature identification method and system |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210928 |
|
| RJ01 | Rejection of invention patent application after publication |