CN109632291A - A kind of Fault Diagnosis of Gear Case method based on polynary mode decomposition-transfer entropy - Google Patents
A kind of Fault Diagnosis of Gear Case method based on polynary mode decomposition-transfer entropy Download PDFInfo
- Publication number
- CN109632291A CN109632291A CN201811472958.5A CN201811472958A CN109632291A CN 109632291 A CN109632291 A CN 109632291A CN 201811472958 A CN201811472958 A CN 201811472958A CN 109632291 A CN109632291 A CN 109632291A
- Authority
- CN
- China
- Prior art keywords
- shaft end
- signal
- polynary
- transfer entropy
- mode decomposition
- 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
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
- G01M13/02—Gearings; Transmission mechanisms
- G01M13/021—Gearings
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/02—Preprocessing
- G06F2218/04—Denoising
-
- 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
Abstract
The present invention relates to a kind of Fault Diagnosis of Gear Case methods based on polynary mode decomposition-transfer entropy, belong to gear distress analysis technical field.Technical solution is: using polynary mode decomposition noise reduction algorithm extraction system main feature first, then transfer entropy is used for the description of system complexity, and be used in Gear Fault Diagnosis.The present invention assists polynary empirical mode decomposition-transfer entropy to analyze gearbox fault vibration information transmission characteristic using noise, it is nonlinear between output shaft end and input shaft end signal frequency range in the case of quantitative description gearbox fault to couple and information transfer characteristic, facilitate the transmission path of exploration gearbox fault vibration signal, the gear-box state evaluation index based on polynary empirical mode decomposition-transfer entropy is established, provides new effective means for the fault diagnosis of rotating machinery, performance degradation state recognition and trend prediction.
Description
Technical field
The present invention relates to a kind of Fault Diagnosis of Gear Case methods based on polynary mode decomposition-transfer entropy, belong to gear event
Hinder analysis technical field.
Technical background
Practical, reliable mechanical fault diagnosis technology is studied, is prevention mechanical equipment fault and guarantee mechanical equipment
Operational safety, stable technical foundation.Mechanical power transmission system is as the widely applied one kind of national defence and national economy field
Important technology equipment, safety reliability are most important.Gear exists as the critical component long-term work of power drive system
Under the high revolving speed of top load, vulnerable to damage and break down, the fortune of entire mechanical system will be directly influenced by breaking down
Turn, therefore improves that rotating machinery reliability, promptly and accurately detection is out of order and has become more and more important.Event occurs in gear-box
When barrier, the periodic shock power that gear distress generates can cause the Non-Linear Vibration of mechanical system, this makes the vibration letter of acquisition
Number have non-stationary, non-gaussian and nonlinear characteristic.Simultaneously as the complexity and running environment of machine driven system
Diversity, the vibration signal that scene measures often contain much noise, and gear-box fault features are fainter, are easy quilt
Noise floods, and therefore, how to effectively inhibit the noise in mechanical fault signals, and the accurate fault signature that extracts is a problem.
With the raising of gear box structure complexity, state that is how objective, accurate and effective and meeting gearbox fault characteristic
Identification technology has become question of common concern.
In order to extract the fault characteristic information of gear, domestic and international researcher proposes many effective methods, such as short
When Fourier transformation (STFT), wavelet transformation, Hilbert-Huang transformation and local mean value decompose etc., wavelet transformation is in machinery
Preferable application effect is had been achieved in fault diagnosis, still, how to select the wavelet basis with preferable time frequency resolution
The standard of function is still not clear at present, needs artificial threshold value；EMD has very strong nonlinear properties capacity of decomposition, can be with
Multi -components mixed signal is adaptively decomposed into several IMF components from high frequency to low frequency, is similar to multiple adaptive band logicals
The filter effect of filter.But due to more time-frequency dimensional properties of single IMF component and different IMF component time-frequency scale phases
Like property so that its when decomposing the abnormal signal that there is intermittent ingredient and impulse disturbances ingredient, be easy to appear modal overlap,
The problems such as end effect.
It is a kind of new non-linear unstable signal adaptive Time-frequency Decomposition side that noise, which assists polynary empirical mode decomposition,
The problems such as method, which overcome the modal overlap of EEMD and big operands, polynary empirical mode decomposition method overcomes EMD and LMD
The methods of in multi-channel data analysis the limitation that is theoretically unsound, standard EMD is extended to multi-channel signal processing neck
Domain realizes the multi-channel synchronous Conjoint Analysis of multi channel signals oscillation mode, obtains the common mode in different channels, it is ensured that
Intrinsic mode function component matches in quantity and scale, solves the problems, such as the model calibration of multi channel signals, remains
Important cross-correlation information between each channel signal, while this method has good adaptivity and Time-Frequency Localization ability.
Method for diagnosing faults based on polynary empirical mode decomposition has stronger adaptive ability, while it is mixed to solve EMD mode
The problems such as folded, model calibration, the failure-frequency and its frequency multiplication information extracted is also apparent, inhibits noise effects more preferable and excellent
In EMD and EEMD method.
Comentropy correlation theory achieves preferable application effect in the state recognition of gear-box and accident analysis field,
When rotating machinery breaks down, the dynamic behavior of system can show as nonlinear dissipation.Common description system is multiple
The characteristic parameter of polygamy includes: fractal dimension, Lyapunov index and K-S entropy etc..These Nonlinear Dynamics are all logical
Phase space reconfiguration is crossed to describe system performance, still, the selection in phase space reconfiguration about delay time is a problem, right
The selection of the closest point of initial phase point is then more difficult.Pincus proposed a kind of new measure time sequence in 1991
Method-approximate entropy of complexity, approximate entropy are a kind of based on marginal probability distribution statistic quantification time series amplitude degree
Method, approximate entropy is bigger, illustrates that the probability for generating new model is bigger, time series is more complicated.Sample Entropy has required data
The advantages that length is short, anti-noise ability is strong is suitable for random signal and deterministic signal, raw in brain electricity, electrocardio etc. since proposition
Object field of signal processing and mechanical signal process field are used widely.Although approximate entropy only needs less data can
With the complexity of measuring period sequence, but lead to its meter since approximate entropy has the intrinsic comparison to data section
Calculation can generate deviation, and the sensibility that the complexity that this deviation causes it small to system changes is poor.Meanwhile these algorithms
Enough data lengths are required, and the data length of characteristic of rotating machines vibration signal is generally limited.Therefore, these are non-thread
Property kinetic index it is in practical engineering applications and inconvenient, need to find it is more effective, stablize and to data length requirement compared with
Low new method, above- mentioned information entropy method and can not output shaft in the case of comprehensive and accurate description quantitative description gearbox fault
End couples and information transfer characteristic with nonlinear between input shaft end signal frequency range.
Summary of the invention
It is an object of that present invention to provide a kind of Fault Diagnosis of Gear Case methods based on polynary mode decomposition-transfer entropy, fixed
Non-linear synchronous coupling between gearbox fault signal output axis signal between quantity research difference time-frequency scale and input axis signal
Feature and information transmitting are closed, polynary mode decomposition noise reduction algorithm extraction system main feature is used first, then uses transfer entropy
It in the description of system complexity, and is used in Gear Fault Diagnosis, solves the above problem existing for prior art.
The technical scheme is that
A kind of Fault Diagnosis of Gear Case method based on polynary mode decomposition-transfer entropy uses polynary mode decomposition first
Transfer entropy, is then used for the description of system complexity by noise reduction algorithm extraction system main feature, and is used for gear distress
In diagnosis.
It specifically includes the following steps:
Step 1: being placed on bearing block using multiple ICP vibration acceleration sensors, synchronous detection pinion unit
Transverse direction/longitudinal direction/axial vibration response and tank surface vibratory response of epicyclic gearbox when system has typical initial failure,
Experimental section is carried out using virtual instrument system and LabView software；
Step 2: polynary empirical mode decomposition is assisted to the noise that the multi-channel data of acquisition improves, it is logical to be set as n
Road, steps are as follows:
(1) m vector white noise sequence composition is added into n channel vibration data sequence of acquisitionSignal length is T,It indicates to tie up at (n+m-1) single
Deflection in the ball of positionDirection vector collection.K=1,2,3 ..., K；
(2) Hammersley sequential sampling method is used, suitable uniform sampling point set is obtained on (n+m-1) n-dimensional sphere n,
Obtain the direction vector of n+m dimension space；
(3) input signal v (t) is calculated in each direction vectorOn mapping
(4) mapping signal of all direction vectors is determinedThe corresponding instantaneous moment of extreme valueL is indicated
Extreme point position, l ∈ [1, T]；
(5) multivariate spline interpolating function value extreme point is usedObtain K polynary envelopesTo ball sky
Between K direction vector, n signal mean value m (t) are as follows:
(6) intrinsic mode function h (t) is extracted by h (t)=v (t)-m (t), detects h1(t) whether contain high frequency intermittent
Signal and noise signal；If so, (1)-(2) step is continued to execute, until h1It (t) is not abnormal signal；If h1(t) row
Column entropy is greater than given threshold value β0, then it is assumed that the signal is abnormal signal；If arranging entropy is less than β0, then the signal is not abnormal
Signal；If h (t) meets polynary IMF judgment criteria, just believe the result of v (t)-h (t) as the input of (2) step
Number, continue the iterative calculation of (2)-(5) step, extracts new polynary IMF component h (t)；Otherwise, by h (t) as the defeated of (2) step
Enter signal, continues (2)-(5) step iteration.By NAMEMD decomposable process, former n+m signalIt is broken down into a series ofIt is as follows with the adduction form of r (t):
In formula, d indicates to decomposite the polynary IMF number of plies h (t) come, and isForCorrespond respectively to the n+m group signal IMFs component and n+m surplus of n+m member signal；Finally
The corresponding IMFs of m channel noise is deleted from (m+n) member IMFs, retains n channel IMFs of useful signal；
Step 3: preceding 7 IMFs of preceding 7 IMFs and input axis channel to output shaft end channel are carried out between any two
Transfer entropy calculate, calculating be repeatedly averaged, standard deviation and calculating the time；Choose output shaft end -> input shaft end average value
Between 0.4-2, input shaft end -> output shaft end transfer entropy average value between 0.1-2, standard deviation less than 0.001,
Scale parameter of the single calculation time less than 5 seconds carries out subsequent analysis.
The subsequent analysis of the step 3 is made a concrete analysis of as follows:
(1) it based on the algorithm of transfer entropy, constructs output shaft end IMFs and (is set as x (t) to input shaft IMFs (being set as y (t))
Noise assist polynary empirical mode decomposition-transfer entropy NMTEx->y, formula is as follows:
In formula, u is predicted time；Joint probability of the p () between variable；WithRespectively indicate output shaft end
With the delay vector of IMF1, IMF2, IMF3, IMF4, IMF5 component of input shaft end；ForFor forecasting sequence；
(2)NMTEy->xThen indicate the IMFs (x of input shaft endi(t)) output shaft end IMFs (y is arrivedi(t)) noise auxiliary
Polynary empirical mode decomposition-transfer entropy, expression formula are as follows:
In formula, u is predicted time；Joint probability of the p () between variable；WithRespectively indicate output shaft end
With the delay vector of IMF1, IMF2, IMF3, IMF4, IMF5 component of input shaft end；ForFor forecasting sequence；Transmitting
Entropy is bigger, illustrates that Vibration Fault Signal coupling is stronger between this frequency range.
Transfer entropy (transfer entropy) is the recent amount of transmitted information portrayed between two time serieses, transmitting
Entropy can quantify the information exchange intensity between two systems out, can more calculate the flow direction of information, in addition to physical communication field
Be widely applied it is outer, transfer entropy and associated delivery entropy be can between analysis system information interaction a kind of effective tool,
And non-linear and asymmetric system can be solved the problems, such as simultaneously.
The effect of mechanical equipment is exactly the transmitting and conversion of energy, and the quality of equipment state directly influences transmitting and turns
The loss of additional-energy during changing, the good then transmitting of equipment state and the additional-energy loss in conversion process are small, on the contrary,
Transmitting and the additional-energy loss in conversion process are big, and transmitting passes through vibration with the additional-energy loss in conversion process mostly
Parameter shows.Transmitting increases with additional-energy loss in conversion process with the increase of equipment fault degree.Therefore,
The situation of change of monitoring transmitting and additional-energy loss in conversion process is to understand an important channel of equipment fault degree.
Gear-box is under operating condition, and along with the generation and development of its internal fault, Gearbox vibration signal is from linearly becoming non-thread
Property, it eventually becomes strong non-linear, therefore gear distress need to be identified using nonlinear research approach.The present invention is first
Using polynary mode decomposition noise reduction algorithm extraction system main feature, then transfer entropy to be used for the description of system complexity, and
It is used in Gear Fault Diagnosis.The present invention is existing for the feature extraction of rotating machinery Weak fault and EMD etc. in background of making an uproar by force
There are Time-Frequency Analysis Method, existing modal overlap problem, for the gearbox fault signal between quantitative study difference time-frequency scale
It exports axis signal and inputs the non-linear synchronous coupling feature between axis signal and information transmitting, propose based on polynary mode point
Solution-transfer entropy gearbox fault analysis method.
The invention has the advantages that: assist polynary empirical mode decomposition-transfer entropy analysis gearbox fault to vibrate using noise
Information transmission characteristic, nonlinear coupling between output shaft end and input shaft end signal frequency range in the case of quantitative description gearbox fault
Conjunction and information transfer characteristic facilitate the transmission path for exploring gearbox fault vibration signal, establish and be based on polynary empirical modal
Decomposition-transfer entropy gear-box state evaluation index is fault diagnosis, performance degradation state recognition and the trend of rotating machinery
Prediction provides new effective means；The present invention is to in-depth gear train assembly fault diagnosis technology system and method, guidance key
The design of component has important theoretical significance and practical application value.
Detailed description of the invention
Fig. 1 is flow diagram of the embodiment of the present invention；
Fig. 2 is Gear Fault Diagnosis of embodiment of the present invention testing stand and sensor mounting location schematic diagram；
Fig. 3 is gear distress of embodiment of the present invention vibration signal time domain waveform schematic diagram (output shaft end Y-direction vibration letter
Number time domain waveform)；
Fig. 4 is gear distress of embodiment of the present invention vibration signals spectrograph figure (output shaft end Y-direction vibration signals spectrograph)；
Fig. 5 is output shaft end of embodiment of the present invention Y-direction vibration signal NAMEMD decomposition result schematic diagram (first 5)；
Fig. 6 is output shaft end of embodiment of the present invention Y-direction vibration signal NAMEMD time-frequency spectrum schematic diagram；
Fig. 7 is that polynary empirical mode decomposition-transfer entropy of output shaft end gear-box under normal condition of the embodiment of the present invention shows
It is intended to (the transfer entropy Y-direction of each IMF component of output shaft end IMF component each to input shaft end)；
Fig. 8 is that polynary empirical mode decomposition-transfer entropy of input shaft end gear-box under normal condition of the embodiment of the present invention shows
It is intended to (the transfer entropy Y-direction of each IMF component of input shaft end IMF component each to output shaft end)；
Fig. 9 is that polynary empirical mode decomposition-transfer entropy of output shaft end gear-box under malfunction of the embodiment of the present invention shows
It is intended to (the transfer entropy Y-direction of each IMF component of output shaft end IMF component each to input shaft end)；
Figure 10 is polynary empirical mode decomposition-transfer entropy of input shaft end gear-box under malfunction of the embodiment of the present invention
Schematic diagram (the transfer entropy Y-direction of each IMF component of input shaft end IMF component each to output shaft end)；
Figure 11 is 7 IMF before output shaft end of the embodiment of the present invention to the transfer entropy schematic diagram of input shaft end IMF1；
Figure 12 is 7 IMF before output shaft end of the embodiment of the present invention to the transfer entropy schematic diagram of input shaft end IMF2.
In the figure, it is marked as driving motor 1, shaft coupling 2, input shaft end 3 (2 directions acceleration transducer X, Y), output
Shaft end 4 (2 acceleration transducer X, Y-directions), magnetic powder retarder 5, gear-box 6.
Specific embodiment:
The present invention will be further explained below with reference to the attached drawings.
In embodiment, Fig. 1 is flow diagram of the embodiment of the present invention；Fig. 2 is Gear Fault Diagnosis of the embodiment of the present invention
Testing stand and sensor mounting location schematic diagram；Fig. 3 is gear distress of embodiment of the present invention vibration signal time domain waveform schematic diagram
(output shaft end Y-direction vibration signal time domain waveform)；Fig. 4 is the (output of gear distress of embodiment of the present invention vibration signals spectrograph figure
Shaft end Y-direction vibration signals spectrograph).
NAMEMD analysis and EEMD analysis are carried out to the gear distress signal that 4 sensors acquire respectively, and carried out pair
Than, by NAMEMD and EEMD to the comparison of the decomposition result of gear tooth breakage Test to Failure signal it is found that in EEMD decomposition result,
(identical frequency band signal distributions are mainly reflected in there is significantly overlapping phenomenon in the time-frequency distributions boundary Relative Fuzzy of different IMF
In multiple IMF), and the time-frequency distributions boundary of NAMEMD decomposition result is relatively clear.The decomposition result of above two method is all
There are a degree of frequency aliasing, compare EEMD, the frequency distribution for the IMF that NAMEMD is obtained is ideal, modal overlap
Degree is lower.
Fig. 5 is output shaft end of embodiment of the present invention Y-direction vibration signal NAMEMD decomposition result schematic diagram (first 9)；Figure
6 be output shaft end of embodiment of the present invention Y-direction vibration signal NAMEMD decomposition result spectrum distribution schematic diagram；Fig. 7 is the present invention
Embodiment output shaft end Y-direction vibration signal NAMEMD time-frequency spectrum schematic diagram；Fig. 8 is gear tooth breakage of embodiment of the present invention failure
Signal EEMD decomposition result schematic diagram (first 12)；Fig. 9 is gear tooth breakage of embodiment of the present invention fault-signal EEMD decomposition result
Spectrum distribution schematic diagram.
Preceding 7 IMFs of preceding 7 IMFs and input axis channel to output shaft end channel carry out transfer entropy meter between any two
Calculate, calculating be repeatedly averaged, standard deviation and calculating the time.Choose output shaft end -> input shaft end average value between 0.4~
Between 2, input shaft end -> output shaft end transfer entropy average value is between 0.1~2, and standard deviation is less than 0.001, single meter
Scale parameter of the evaluation time less than 5 seconds carries out subsequent analysis.Polynary empirical mode decomposition-transmitting of normal condition lower tooth roller box
Entropy is on different coupling directions, the coupling information between different frequency range is as shown in the picture.By attached drawing it is found that gear-box normal condition
Under same frequency range output shaft end -> input shaft end transmitting entropy difference it is smaller, illustrate that signal is under each frequency range in this case
Stiffness of coupling it is more average.
Polynary empirical mode decomposition-transfer entropy of gear distress state lower tooth roller box is on different coupling directions, different frequencies
Coupling information between section is as shown in the picture.By attached drawing it is found that under gearbox fault state same frequency range output shaft end -> input
Shaft end transmit entropy difference it is larger, illustrate in this case stiffness of coupling of the signal under each IMF3, IMF4, IMF5 frequency range compared with
To concentrate.
Coupled relation between further description vibration signal extracts under gear-box different conditions 7 before output shaft end
IMF component is to the transfer entropy of input shaft end IMF1, IMF2, as a result as shown in table 1, table 2.By attached drawing it is found that discovery gear-box event
Output shaft end IMF3, IMF4, IMF5 under barrier state to the stiffness of coupling (transfer entropy) of input shaft end IMF1, IMF2 generally
Higher than the transfer entropy of same direction identical frequency scale under gear-box normal condition.The research of patent embodies gear-box difference
Coupling feature and information transmitting between position failure signal different frequency range, provide to explore the coupling mechanism of gear distress signal
Theoretical foundation.
Transfer entropy of 7 IMF components to input shaft end IMF1 before 1 output shaft end of table
Transfer entropy of 7 IMF components to input shaft end IMF2 before 2 output shaft end of table
Fig. 7 is that polynary empirical mode decomposition-transfer entropy of output shaft end gear-box under normal condition of the embodiment of the present invention shows
It is intended to (the transfer entropy Y-direction of each IMF component of output shaft end IMF component each to input shaft end)；Fig. 8 be the embodiment of the present invention just
(each IMF component of input shaft end is to defeated for polynary empirical mode decomposition-transfer entropy schematic diagram of input shaft end gear-box under normal state
The transfer entropy Y-direction of each IMF component in shaft end out)；Fig. 9 is the more of output shaft end gear-box under malfunction of the embodiment of the present invention
First empirical mode decomposition-transfer entropy schematic diagram (transfer entropy side Y of each IMF component of output shaft end IMF component each to input shaft end
To)；Figure 10 is polynary empirical mode decomposition-transfer entropy signal of input shaft end gear-box under malfunction of the embodiment of the present invention
Scheme (the transfer entropy Y-direction of each IMF component of input shaft end IMF component each to output shaft end)；Figure 11 is output of the embodiment of the present invention
Transfer entropy schematic diagram of 7 IMF to input shaft end IMF1 before shaft end；Figure 12 is 7 IMF before output shaft end of the embodiment of the present invention
To the transfer entropy schematic diagram of input shaft end IMF2.
Claims (3)
1. a kind of Fault Diagnosis of Gear Case method based on polynary mode decomposition-transfer entropy, which is characterized in that first using polynary
Transfer entropy, is then used for the description of system complexity, and be used for by mode decomposition noise reduction algorithm extraction system main feature
In Gear Fault Diagnosis.
2. a kind of Fault Diagnosis of Gear Case method based on polynary mode decomposition-transfer entropy according to claim 1, special
Sign is that steps are as follows:
Step 1: being placed on bearing block using multiple ICP vibration acceleration sensors, and synchronous detection gear train assembly exists
Transverse direction/longitudinal direction/axial vibration response and tank surface vibratory response of epicyclic gearbox when typical initial failure；
Step 2: polynary empirical mode decomposition is assisted to the noise that the multi-channel data of acquisition improves, n-channel is set as, walks
It is rapid as follows:
(1) m vector white noise sequence composition is added into n channel vibration data sequence of acquisitionSignal length is T,It indicates to tie up unit at (n+m-1)
Deflection in ballDirection vector collection.K=1,2,3 ..., K；
(2) Hammersley sequential sampling method is used, suitable uniform sampling point set is obtained on (n+m-1) n-dimensional sphere n, obtains n
The direction vector of+m-dimensional space；
(3) input signal v (t) is calculated in each direction vectorOn mapping
(4) mapping signal of all direction vectors is determinedThe corresponding instantaneous moment of extreme valueL indicates extreme point
Position, l ∈ [1, T]；
(5) multivariate spline interpolating function value extreme point is usedObtain K polynary envelopesTo spherical space K
A direction vector, n signal mean value m (t) are as follows:
(6) intrinsic mode function h (t) is extracted by h (t)=v (t)-m (t), detects h1(t) whether containing high frequency intermittent signal and
Noise signal；If so, (1)-(2) step is continued to execute, until h1It (t) is not abnormal signal；If h1(t) arrangement entropy
Greater than given threshold value β0, then it is assumed that the signal is abnormal signal；If arranging entropy is less than β0, then the signal is not abnormal signal；Such as
Fruit h (t) meets polynary IMF judgment criteria, then just continuing using the result of v (t)-h (t) as the input signal of (2) step
(2)-(5) step iterative calculation, extracts new polynary IMF component h (t)；Otherwise, the input signal by h (t) as (2) step, after
Continuous (2)-(5) step iteration.By NAMEMD decomposable process, former n+m signalQuilt
It is decomposed into a series of IMFIt is as follows with the adduction form of r (t):
In formula, d indicates to decomposite the polynary IMF number of plies h (t) come, and isR (t) isCorrespond respectively to the n+m group signal IMFs component and n+m surplus of n+m member signal；Finally from
(m+n) the corresponding IMFs of m channel noise is deleted in member IMFs, retains n channel IMFs of useful signal；
Step 3: preceding 7 IMFs of preceding 7 IMFs and input axis channel to output shaft end channel carry out transfer entropy between any two
Calculate, calculating be repeatedly averaged, standard deviation and calculating the time；Output shaft end -> input shaft end average value is chosen between 0.4-
Between 2, input shaft end -> output shaft end transfer entropy average value is between 0.1-2, and standard deviation is less than 0.001, single calculation
Scale parameter of the time less than 5 seconds carries out subsequent analysis.
3. a kind of Fault Diagnosis of Gear Case method based on polynary mode decomposition-transfer entropy according to claim 2, special
Sign is the subsequent analysis of the step 3, makes a concrete analysis of as follows:
(1) based on the algorithm of transfer entropy, construct output shaft end IMFs (be set as x (t) to input shaft IMFs (being set as y (t)) noise
Assist polynary empirical mode decomposition-transfer entropy NMTEx->y, formula is as follows:
In formula, u is predicted time；Joint probability of the p () between variable；WithRespectively indicate output shaft end and input
The delay vector of IMF1, IMF2, IMF3, IMF4, IMF5 component of shaft end；ForFor forecasting sequence；
(2)NMTEy->xThen indicate the IMFs (x of input shaft endi(t)) output shaft end IMFs (y is arrivedi(t)) noise assists polynary warp
Test mode decomposition-transfer entropy, expression formula are as follows:
In formula, u is predicted time；Joint probability of the p () between variable；WithRespectively indicate output shaft end and input
The delay vector of IMF1, IMF2, IMF3, IMF4, IMF5 component of shaft end；ForFor forecasting sequence；Transmitting entropy is got over
Greatly, illustrate that Vibration Fault Signal coupling is stronger between this frequency range.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811472958.5A CN109632291A (en) | 2018-12-04 | 2018-12-04 | A kind of Fault Diagnosis of Gear Case method based on polynary mode decomposition-transfer entropy |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811472958.5A CN109632291A (en) | 2018-12-04 | 2018-12-04 | A kind of Fault Diagnosis of Gear Case method based on polynary mode decomposition-transfer entropy |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109632291A true CN109632291A (en) | 2019-04-16 |
Family
ID=66070951
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811472958.5A Pending CN109632291A (en) | 2018-12-04 | 2018-12-04 | A kind of Fault Diagnosis of Gear Case method based on polynary mode decomposition-transfer entropy |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109632291A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111141518A (en) * | 2019-12-16 | 2020-05-12 | 西安交通大学 | Model-based asymmetric rotor bearing system unbalance identification method |
| CN111397868A (en) * | 2020-02-27 | 2020-07-10 | 广西电网有限责任公司电力科学研究院 | Breaker fault analysis method based on aggregation empirical mode decomposition algorithm |
| CN112747921A (en) * | 2020-12-24 | 2021-05-04 | 武汉科技大学 | Multi-sensor mechanical fault diagnosis method based on NA-MEMD |
| CN112881006A (en) * | 2021-01-12 | 2021-06-01 | 北华大学 | Gear fault diagnosis method |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105022912A (en) * | 2015-05-28 | 2015-11-04 | 北京交通大学 | Rolling bearing fault prediction method based on wavelet principal component analysis |
| CN105354587A (en) * | 2015-09-25 | 2016-02-24 | 国网甘肃省电力公司电力科学研究院 | Fault diagnosis method for gearbox of wind generation unit |
| CN106197999A (en) * | 2016-07-08 | 2016-12-07 | 安徽德衍智控科技有限公司 | A kind of planetary gear method for diagnosing faults |
| CN106383028A (en) * | 2016-09-05 | 2017-02-08 | 北京航空航天大学 | Gear case fault diagnosis method |
-
2018
- 2018-12-04 CN CN201811472958.5A patent/CN109632291A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105022912A (en) * | 2015-05-28 | 2015-11-04 | 北京交通大学 | Rolling bearing fault prediction method based on wavelet principal component analysis |
| CN105354587A (en) * | 2015-09-25 | 2016-02-24 | 国网甘肃省电力公司电力科学研究院 | Fault diagnosis method for gearbox of wind generation unit |
| CN106197999A (en) * | 2016-07-08 | 2016-12-07 | 安徽德衍智控科技有限公司 | A kind of planetary gear method for diagnosing faults |
| CN106383028A (en) * | 2016-09-05 | 2017-02-08 | 北京航空航天大学 | Gear case fault diagnosis method |
Non-Patent Citations (2)
| Title |
|---|
| 武哲等: "基于多元经验模态分解的旋转机械早期故障诊断方法", 《仪器仪表学报》 * |
| 陈子瑜等: "传递熵算法在风电机组故障穿越特征分析中的应用", 《中国电机工程学报》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111141518A (en) * | 2019-12-16 | 2020-05-12 | 西安交通大学 | Model-based asymmetric rotor bearing system unbalance identification method |
| CN111397868A (en) * | 2020-02-27 | 2020-07-10 | 广西电网有限责任公司电力科学研究院 | Breaker fault analysis method based on aggregation empirical mode decomposition algorithm |
| CN111397868B (en) * | 2020-02-27 | 2022-02-08 | 广西电网有限责任公司电力科学研究院 | Breaker fault analysis method based on aggregation empirical mode decomposition algorithm |
| CN112747921A (en) * | 2020-12-24 | 2021-05-04 | 武汉科技大学 | Multi-sensor mechanical fault diagnosis method based on NA-MEMD |
| CN112881006A (en) * | 2021-01-12 | 2021-06-01 | 北华大学 | Gear fault diagnosis method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109632291A (en) | A kind of Fault Diagnosis of Gear Case method based on polynary mode decomposition-transfer entropy | |
| Wang et al. | Early fault diagnosis for planetary gearbox based on adaptive parameter optimized VMD and singular kurtosis difference spectrum | |
| CN103034170B (en) | Numerical control machine tool machining performance prediction method based on intervals | |
| Hou et al. | A tacholess order tracking method for wind turbine planetary gearbox fault detection | |
| CN105136454A (en) | Wind turbine gear box fault recognition method | |
| CN104698837A (en) | Method and device for identifying operating modal parameters of linear time-varying structure and application of the device | |
| CN104122086B (en) | Fault Diagnosis of Gearbox for Wind Turbine method based on Kriging model | |
| CN106441896A (en) | Characteristic vector extraction method for rolling bearing fault mode identification and state monitoring | |
| CN103499437A (en) | Rotary machine fault detection method of dual-tree complex wavelet transformation with adjustable quality factors | |
| Mohanty et al. | Adaptive fault identification of bearing using empirical mode decomposition–principal component analysis‐based average kurtosis technique | |
| CN105956290A (en) | High-voltage circuit breaker mechanical fault diagnosis method based on multi-data fusion technology | |
| CN103822786A (en) | Wind turbine generator set mechanical equipment state diagnosis method based on multivariate statistical analysis | |
| Miao et al. | An enhanced multifeature fusion method for rotating component fault diagnosis in different working conditions | |
| CN108362488A (en) | OLTC mechanical failure diagnostic methods based on MPE and SVM | |
| Li et al. | A review on the signal processing methods of rotating machinery fault diagnosis | |
| CN106777611A (en) | Complicated cyclic train Weak fault identification and performance degradation monitoring system and method | |
| CN107229269A (en) | A kind of wind-driven generator wheel-box method for diagnosing faults of depth belief network | |
| Li et al. | A local weighted multi-instance multilabel network for fault diagnosis of rolling bearings using encoder signal | |
| CN102778354A (en) | Method for diagnosing failure of wind-powered rotary support based on wavelet analysis | |
| Huang et al. | Feature extraction for rolling element bearing faults using resonance sparse signal decomposition | |
| Sheng et al. | Research on gear crack fault diagnosis model based on permanent magnet motor current signal | |
| CN113255458A (en) | Bearing fault diagnosis method based on multi-view associated feature learning | |
| Song et al. | A new hybrid method for bearing fault diagnosis based on CEEMDAN and ACPSO-BP neural network | |
| Yao et al. | A low-frequency fault detection method for low-speed planetary gearbox based on acoustic signals | |
| Min et al. | Probabilistic neural network motor bearing fault diagnosis based on improved feature extraction |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20190416 |
|
| WD01 | Invention patent application deemed withdrawn after publication |