CN108827634A - Manifold merges empirical mode decomposition method - Google Patents
Manifold merges empirical mode decomposition method Download PDFInfo
- Publication number
- CN108827634A CN108827634A CN201810662526.4A CN201810662526A CN108827634A CN 108827634 A CN108827634 A CN 108827634A CN 201810662526 A CN201810662526 A CN 201810662526A CN 108827634 A CN108827634 A CN 108827634A
- Authority
- CN
- China
- Prior art keywords
- failure
- manifold
- failure modalities
- component
- signal
- 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.)
- Granted
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
- G01M13/04—Bearings
- G01M13/045—Acoustic or vibration analysis
Landscapes
- Physics & Mathematics (AREA)
- Acoustics & Sound (AREA)
- General Physics & Mathematics (AREA)
- Debugging And Monitoring (AREA)
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
Abstract
The present invention relates to a kind of manifolds to merge empirical mode decomposition method, including：Analysis signal in be added mean value be 0, the random white noise that standard deviation is σ, obtain noisy signal；EMD processing is carried out to the noisy signal, obtains the IMF comprising fault message, i.e. failure modalities component；Change the value of σ, repeat the above steps n times, obtains N number of failure modalities component with different noise intensities, wherein N is positive integer；N number of failure modalities component is merged according to given manifold learning, obtains the inherent manifold structure of higher-dimension failure modalities component, i.e. failure transient ingredient.Above-mentioned manifold merges empirical mode decomposition method, takes different values to the standard deviation of each random white noise being added in analysis signal, using the outstanding feature mining ability of manifold learning, extracts the transient components with rock-steady structure from higher-dimension failure modalities component.
Description
Technical field
The present invention relates to mechanical fault diagnosis, merge empirical mode decomposition method more particularly to manifold.
Background technique
Rotating machinery just develops towards the direction of enlargement, precise treatment and automation, this is just to whole equipment system
The manufactures of middle all parts, installation and daily maintenance maintenance propose more strict requirements, and one of any component is subtle
Damage or concussion dislocation, are likely to influence the normal work of whole system, or even cause major accident.Rotary part goes out
When existing failure, there are transient impact response component in vibration signal, the successful detection to transient components in signal be effectively into
The conventional means of row rotary machinery fault diagnosis.But due to the complexity of working environment, characteristic of rotating machines vibration signal often table
Now it is non-stationary, and contains multi-frequency ingredient, including a large amount of ambient noise, transient components related with failure is caused to exist
Seem very faint in signal, to make troubles to fault diagnosis.
Empirical mode decomposition (EMD) method is a kind of nonstationary random response method, it can be adaptively by sophisticated signal
It is decomposed into limited intrinsic mode functions (IMF).Each IMF meets two conditions：A) function is in entire time range, part
The number of extreme point and zero crossing is equal, or at most difference one；B) point, the envelope of local maximum are (upper to wrap at any time
Winding thread) and the average value of envelope (lower envelope line) of local minimum be zero.Original is contained by each IMF that EMD method obtains
The local feature information of the different time scales of signal.Since the vibration transient components that rotating machinery fault is excited meet IMF
Condition, so EMD mechanical oscillation signal transient components detection in be widely used.But there are moulds for EMD method
State Aliasing Problem, i.e. a time scale sequence are distributed in two IMF or there are multiple time scale sequences in an IMF
Column.The main reason for causing modal overlap phenomenon is that the interruption of signal is discontinuous, for example noise, impact arteries and veins are doped in signal
Punching and intermittency signal.Modal overlap problem causes the transient components information obtained by EMD method imperfect or is mingled with dry
Component is disturbed, the identification of failure is unfavorable for.
In order to solve the problems, such as modal overlap existing for EMD method, existing technology mainly gathers empirical mode decomposition
(EEMD) method.The principle of this method is that white noise is added in original signal, using the uniform distribution properties of white noise spectrum, is made
Signal all has a continuity on different time scales, in this way to add the signal made an uproar to carry out EMD processing can be mixed to avoid mode
Folded problem.The noise being added in the signal can be distributed in each IMFs, in order to remove the noise of these introducings, what EEMD was used
Method is to carry out repeatedly plus respectively EMD processing after making an uproar, and then multiple IMFs with adjacent frequency bands range are sought with average, utilization
The noise that the zero mean characteristic removal of noise introduces.The realization step of EEMD is：First analysis signal in be added mean value be 0,
Standard deviation is the random white noise of σ；Then EMD processing is carried out to noisy signal, obtains the subsignal being made of n IMFs
Group；Then it repeats above step M times, obtains M subsignal group；Finally seek that there is adjacent frequency bands range in all subsignal groups
IMFs mean value, obtain n removal introduce noise IMFs.
There are following technical problems for traditional technology：
EEMD method solves the problems, such as modal overlap by the way that white noise is added, and the standard deviation sigma for gathering number M and white noise is
EEMD method needs two parameters being manually set, and different parameters can generate certain influence to the decomposition result of signal.M's
Value increase can reduce the content that noise is introduced in (but cannot completely eliminate) finally obtained IMFs, but can also increase simultaneously
Add calculation amount, and own noise of the IMFs respectively in frequency band can not be disappeared that is, with interior own noise because of the increase of M
It removes；σ is too small, cannot inhibit modal overlap problem, excessive, not only will increase the IMFs quantity of decomposition and increases calculation amount, and
And will cause the radio-frequency component in signal be difficult to decompose and IMFs in introduce the remaining excessive problem of noise.Therefore, existing
EEMD method be primarily present two problems：A) limited ensemble average cannot completely eliminate the introducing noise in each IMF, more
Cannot eliminate in each IMF with interior own noise；B) standard deviation sigma for introducing white noise is difficult to determine suitable value.
Summary of the invention
Based on this, it is necessary in view of the above technical problems, provide a kind of manifold fusion empirical mode decomposition method, this method
Aiming at the problem that ensemble average limited in EEMD method cannot effectively eliminate and introduce noise and own noise, and introduce noise
Standard deviation be difficult to determining problem, using manifold learning method comprising fault message it is multiple have different noise intensities
IMFs merged, without determining the standard deviation for introducing noise, and the introducing noise preferably in cancellation band and own
Noise effectively detects the transient components in signal.
A kind of manifold fusion empirical mode decomposition method, including：
Analysis signal in be added mean value be 0, the random white noise that standard deviation is σ, obtain noisy signal；
EMD processing is carried out to the noisy signal, obtains the IMF comprising fault message, i.e. failure modalities component；
Change the value of σ, repeat the above steps n times, obtains N number of failure modalities component with different noise intensities, wherein
N is positive integer；
N number of failure modalities component is merged according to given manifold learning, obtains higher-dimension failure modalities point
The inherent manifold structure of amount, i.e. failure transient ingredient.
Above-mentioned manifold merges empirical mode decomposition method, to the standard deviation of each random white noise being added in analysis signal
Different values is taken, using the outstanding feature mining ability of manifold learning, extracts to have from higher-dimension failure modalities component and stablize
The transient components of structure, removal not rock-steady structure with interior introducings noise and intrinsic noise, and because introducing noise intensity not
The modal overlap problem bring non-faulting ingredient caused by realizes effective detection to failure transient ingredient in signal.It should
Technical method has at least the following advantages：It does not need to determine and introduces the standard deviation of noise, there is no modal overlap problem, band can be removed
Interior own noise can obtain higher signal-to-noise ratio etc..
In other one embodiment, " analysis signal in be added mean value be 0, the random white noise that standard deviation is σ,
Obtain noisy signal；" in, the standard deviation sigma is 0.01 to 1 times of the analysis signal standards difference.
In other one embodiment, " EMD processing is carried out to the noisy signal, obtaining one includes fault message
IMF, i.e. failure modalities component；" the failure modalities component is from multiple IMFs that EMD is handled according to given failure
Mode determines that method is select.
In other one embodiment, the given failure modalities determine that method includes but is not limited to utilize kurtosis, light
The sliding factor, sparse value, related coefficient, energy and their combination etc. can be selected described from multiple IMFs that EMD is obtained
The method of failure modalities component.
In other one embodiment, " N number of failure modalities component is melted according to given manifold learning
It closes, obtains the inherent manifold structure of higher-dimension failure modalities component, i.e. failure transient ingredient." in, the given manifold learning
Including but not limited to local tangent space alignment algorithm, Isometric Maps algorithm, Local Liner Prediction, laplacian eigenmaps
Algorithm or local retaining projection algorithm.
A kind of computer equipment can be run on a memory and on a processor including memory, processor and storage
The step of computer program, the processor realizes the method when executing described program.
A kind of computer readable storage medium, is stored thereon with computer program, realization when which is executed by processor
The step of the method.
A kind of processor, the processor is for running program, wherein the method is executed when described program is run.
Detailed description of the invention
Fig. 1 is the flow chart that manifold disclosed by the embodiments of the present invention merges empirical mode decomposition method.
Fig. 2 is the time domain waveform of bearing sound vibration signal provided in an embodiment of the present invention.
Fig. 3 is the failure transient ingredient obtained after being handled using EEMD method signal described in Fig. 2.
Fig. 4 is the failure transient ingredient obtained after being handled using technology signal described in Fig. 2 disclosed by the invention.
Specific embodiment
In order to make the objectives, technical solutions, and advantages of the present invention clearer, with reference to the accompanying drawings and embodiments, right
The present invention is further elaborated.It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, and
It is not used in the restriction present invention.
A kind of manifold fusion empirical mode decomposition method, including：
Analysis signal in be added mean value be 0, the random white noise that standard deviation is σ, obtain noisy signal；
EMD processing is carried out to the noisy signal, obtains the IMF comprising fault message, i.e. failure modalities component；
Change the value of σ, repeat the above steps n times, obtains N number of failure modalities component with different noise intensities, wherein
N is positive integer；
N number of failure modalities component is merged according to given manifold learning, obtains higher-dimension failure modalities point
The inherent manifold structure of amount, i.e. failure transient ingredient.
Above-mentioned manifold merges empirical mode decomposition method, to the standard deviation of each random white noise being added in analysis signal
Different values is taken, using the outstanding feature mining ability of manifold learning, extracts to have from higher-dimension failure modalities component and stablize
The transient components of structure, removal not rock-steady structure with interior introducings noise and intrinsic noise, and because introducing noise intensity not
The modal overlap problem bring non-faulting ingredient caused by realizes effective detection to failure transient ingredient in signal.It should
Technical method has at least the following advantages：It does not need to determine and introduces the standard deviation of noise, there is no modal overlap problem, band can be removed
Interior own noise can obtain higher signal-to-noise ratio etc..
In other one embodiment, " analysis signal in be added mean value be 0, the random white noise that standard deviation is σ,
Obtain noisy signal；" in, the standard deviation sigma is 0.01 to 1 times of the analysis signal standards difference.
In other one embodiment, " EMD processing is carried out to the noisy signal, obtaining one includes fault message
IMF, i.e. failure modalities component；" the failure modalities component is from multiple IMFs that EMD is handled according to given failure
Mode determines that method is select.
In other one embodiment, the given failure modalities determine that method includes but is not limited to utilize kurtosis, light
The sliding factor, sparse value, related coefficient, energy and their combination etc. can be selected described from multiple IMFs that EMD is obtained
The method of failure modalities component.
In other one embodiment, " N number of failure modalities component is melted according to given manifold learning
It closes, obtains the inherent manifold structure of higher-dimension failure modalities component, i.e. failure transient ingredient." in, the given manifold learning
Including but not limited to local tangent space alignment algorithm, Isometric Maps algorithm, Local Liner Prediction, laplacian eigenmaps
Algorithm or local retaining projection algorithm.
It is appreciated that the given manifold learning can also be other methods with Dimensionality Reduction function.
A kind of computer equipment can be run on a memory and on a processor including memory, processor and storage
The step of computer program, the processor realizes the method when executing described program.
A kind of computer readable storage medium, is stored thereon with computer program, realization when which is executed by processor
The step of the method.
A kind of processor, the processor is for running program, wherein the method is executed when described program is run.
A concrete application scene of the invention is described below：
It can be seen from background technology that existing EEMD method removes the introducing noise in IMFs using ensemble average.But
The ensemble average of limited times cannot completely remove introducing noise, can not remove with interior own noise, less can solve because introducing
Modal overlap problem caused by noise intensity is improper.
Therefore, the signal transient ingredient the invention discloses manifold fusion empirical mode decomposition method and based on this method
Detection device.This method merges multiple failure modalities components with different noise intensities using manifold learning, therefrom
Extract the failure transient ingredient in signal.Since manifold learning has outstanding feature mining ability, it can extract failure wink
The intrinsic manifold structure of state ingredient, to remove with interior introducing noise, own noise and draw because introducing noise intensity is improper
The modal overlap problem bring non-faulting ingredient risen realizes effective detection to failure transient ingredient in signal.
According to foregoing invention content and the manifold of attached drawing 1 fusion empirical mode decomposition method and based on the signal of this method
The flow chart of transient components detection device, the technology specifically include：
Step 101：Analysis signal in be added mean value be 0, the random white noise that standard deviation is σ, obtain noisy signal；
The excessive and too small modal overlap problem that not can be well solved EMD of the value of the σ, and the optimal value of σ
It can change because of the difference of the analysis signal.In order to avoid discussing the optimal value of σ, the method for the present invention takes σ different
Value, range are 0.01 to 1 times of the analysis signal standards difference, σ uniform value in the range.
Step 102：EMD processing is carried out to noisy signal, obtains the IMF comprising fault message, i.e. failure modalities point
Amount；
After EMD is handled, failure transient ingredient is primarily present in the failure modalities component, therefore the method for the present invention
It is only further to the failure modalities component to be handled.The failure modalities component is the multiple IMFs handled from EMD
According to given failure modalities determine that method is select.
The given failure modalities determine that method includes but is not limited to utilize kurtosis, smoothing factor, sparse value, phase relation
The method that number, energy and their combination etc. can select the failure modalities component from multiple IMFs that EMD is obtained.
Step 103：Change the value of σ, repeat the above steps n times, obtains N number of failure modalities with different noise intensities point
Amount；
Since the method for the present invention is not to remove introducing noise using statistical property, so the number of repetition N is not required to
It is very much.In order to reduce calculation amount, the value of General N is 10.
Step 104：N number of failure modalities component is merged according to given manifold learning, obtains higher-dimension failure mould
The inherent manifold structure of state component, i.e. failure transient ingredient.
Manifold learning is a kind of Dimensionality Reduction and data digging method, can be used for extracting the inherence being embedded in high dimensional data
Low dimensional manifold structure.N number of failure modalities component can form N-dimensional data.Wherein, failure transient ingredient is present in per one-dimensional
In data, there is stable structure, the manifold structure of N-dimensional data can be regarded as, can be retained in manifold learning result；
And other ingredients, including introduce noise, own noise and the modal overlap problem because caused by introducing noise intensity is improper
Bring non-faulting ingredient, is different from every one-dimensional data, without stable structure, meeting quilt in manifold learning result
It rejects.Therefore, after the higher-dimension failure modalities component is merged by given manifold learning, available signal-to-noise ratio
High failure transient ingredient.
The given manifold learning includes but is not limited to local tangent space alignment algorithm, Isometric Maps algorithm, part
The methods with Dimensionality Reduction function such as linearly embedding algorithm, laplacian eigenmaps algorithm, local retaining projection algorithm.
In order to clearly understand technical solution of the present invention and its effect, below with reference to a specific embodiment into
Row is described in detail.
By taking the detection of bearing early-stage weak fault as an example, which is N306E, is turned using motor driven bearing inner ring
Dynamic, revolving speed 1464.6rpm fixes sound pressure sensor near bearing to acquire the sound vibration signal of bearing, sample frequency
For 20kHz.
Firstly, its major failure eigenperiod is calculated according to bearing inner race rotation speed and bearing geometric dimension, tie
Fruit is as shown in table 1.
Table 1：The bearing fault characteristics period
| The inner ring fault signature period | Ti=0.0068s |
| The outer ring fault signature period | To=0.0102s |
| The rolling element fault signature period | Tb=0.0085s |
With reference to attached drawing 2, Fig. 2 is the time domain waveform of bearing sound vibration signal provided in an embodiment of the present invention.From the wave
It is observed that some transient pulse ingredients in shape figure, but is submerged in there are also transient pulse ingredient and is difficult to know in noise
Not, so a bearing fault characteristics period cannot be found from figure.
It is handled using EEMD method signal described in Fig. 2, the standard deviation for introducing noise takes common empirical value, i.e., 0.2
Original signal standard deviation again, ensemble average number are 200.Fig. 3 shows the failure modalities component in processing result, i.e. failure wink
State ingredient.It is observed that periodic transient pulse ingredient from the figure, but still remains between each pulse and largely makes an uproar
Sound does not remove, so obtained failure transient ingredient signal-to-noise ratio is not high.
It is handled using technology signal described in Fig. 2 disclosed by the invention, given failure modalities determine that method is each mould
The method that state component time domain and frequency domain kurtosis value combine, given manifold learning are local tangent space alignment algorithms.Fig. 4 gives
Processing result is gone out.Noise in figure between transient pulse is almost all removed, so that the periodicity of pulse is more obvious, is led to
It crosses and is divided into 0.0068s between average pulse is calculated, it is identical as the 1 middle (center) bearing inner ring fault signature period of table, therefore can be assumed that
The inner ring existing defects of test bearing.In fact, before the experiments were performed, being artificially provided in inner ring in test bearing
One rift defect, defect width are 0.5mm.It therefore can be accurately from the noisy sound of bearing using technology disclosed by the invention
Bearing fault transient components are detected in vibration signal.
In conclusion then being carried out at EMD respectively by the noise for introducing varying strength to bearing sound vibration signal
Reason finally merges obtained higher-dimension failure modalities component using manifold learning, can remove in failure modalities component
Various noise contributions, to effectively detect bearing fault transient components.The method overcome existing EEMD technologies to be difficult to determine
It introduces and noise intensity and is difficult to the problem of removing with interior own noise, the transient components flooded by noise can be extracted, to making an uproar by force
Signal transient composition detection under sound background is of great significance.
Each technical characteristic of embodiment described above can be combined arbitrarily, for simplicity of description, not to above-mentioned reality
It applies all possible combination of each technical characteristic in example to be all described, as long as however, the combination of these technical characteristics is not deposited
In contradiction, all should be considered as described in this specification.
The embodiments described above only express several embodiments of the present invention, and the description thereof is more specific and detailed, but simultaneously
It cannot therefore be construed as limiting the scope of the patent.It should be pointed out that coming for those of ordinary skill in the art
It says, without departing from the inventive concept of the premise, various modifications and improvements can be made, these belong to protection of the invention
Range.Therefore, the scope of protection of the patent of the invention shall be subject to the appended claims.
Claims (8)
1. a kind of manifold merges empirical mode decomposition method, which is characterized in that including：
In the analysis signal be added mean value be 0, the random white noise that standard deviation is σ, obtain noisy signal；
EMD processing is carried out to the noisy signal, obtains the IMF comprising fault message, i.e. failure modalities component；
Change the value of σ, repeat the above steps n times, obtains the N number of failure modalities component with different noise intensities, wherein N is
Positive integer；
N number of failure modalities component is merged according to given manifold learning, obtains higher-dimension failure modalities component
Inherent manifold structure, i.e. failure transient ingredient.
2. manifold according to claim 1 merges empirical mode decomposition method, which is characterized in that " add in analysis signal
Enter the random white noise that mean value is 0, standard deviation is σ, obtains noisy signal；" in, the standard deviation sigma is the analysis signal post
0.01 to 1 times of quasi- difference.
3. manifold according to claim 1 merges empirical mode decomposition method, which is characterized in that " to the noisy signal
EMD processing is carried out, the IMF comprising fault message, i.e. failure modalities component are obtained；" the failure modalities component is from EMD
It handles in obtained multiple IMFs and determines that method is select according to given failure modalities.
4. manifold according to claim 3 merges empirical mode decomposition method, which is characterized in that the given failure modalities
The method of determination includes but is not limited to utilize the energy such as kurtosis, smoothing factor, sparse value, related coefficient, energy and their combination
Enough methods that the failure modalities component is selected from multiple IMFs that EMD is obtained.
5. manifold according to claim 1 merges empirical mode decomposition method, which is characterized in that " according to given manifold
Learning method merges N number of failure modalities component, obtains the inherent manifold structure of higher-dimension failure modalities component, i.e. failure
Transient components." in, the given manifold learning include but is not limited to local tangent space alignment algorithm, Isometric Maps algorithm,
Local Liner Prediction, laplacian eigenmaps algorithm or local retaining projection algorithm.
6. a kind of computer equipment including memory, processor and stores the meter that can be run on a memory and on a processor
Calculation machine program, which is characterized in that the processor realizes any one of claims 1 to 5 the method when executing described program
Step.
7. a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that the program is held by processor
The step of any one of claims 1 to 5 the method is realized when row.
8. a kind of processor, which is characterized in that the processor is for running program, wherein right of execution when described program is run
Benefit requires 1 to 5 described in any item methods.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810662526.4A CN108827634B (en) | 2018-06-25 | 2018-06-25 | Manifold merges empirical mode decomposition method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810662526.4A CN108827634B (en) | 2018-06-25 | 2018-06-25 | Manifold merges empirical mode decomposition method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108827634A true CN108827634A (en) | 2018-11-16 |
| CN108827634B CN108827634B (en) | 2019-12-03 |
Family
ID=64138392
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810662526.4A Active CN108827634B (en) | 2018-06-25 | 2018-06-25 | Manifold merges empirical mode decomposition method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108827634B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110261305A (en) * | 2019-06-18 | 2019-09-20 | 南京东南建筑机电抗震研究院有限公司 | Based on across footpaths continuous bridge damnification recognition methods such as the multispan for influencing line |
| CN110346143A (en) * | 2019-08-06 | 2019-10-18 | 江苏今创车辆有限公司 | Hyperfrequency noise assists empirical mode decomposition method |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101706370A (en) * | 2009-04-29 | 2010-05-12 | 深圳市普禄科智能检测设备有限公司 | Device and method for diagnosing cylinder of diesel engine |
| CN101750210A (en) * | 2009-12-24 | 2010-06-23 | 重庆大学 | Fault diagnosis method based on OLPP feature reduction |
| CN102095588A (en) * | 2010-12-28 | 2011-06-15 | 中国人民解放军军事交通学院 | Intelligent fault diagnosis method for diesel engine |
| CN201993234U (en) * | 2011-01-14 | 2011-09-28 | 华南理工大学 | Diesel engine state detecting device based on energy consumption |
| CN102494899A (en) * | 2011-11-25 | 2012-06-13 | 华南理工大学 | Composite fault diagnosis method for diesel engine and diagnosis system |
| CN104897403A (en) * | 2015-06-24 | 2015-09-09 | 北京航空航天大学 | Self-adaption fault diagnosis method based on permutation entropy (PE) and manifold-based dynamic time warping (MDTW) |
-
2018
- 2018-06-25 CN CN201810662526.4A patent/CN108827634B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101706370A (en) * | 2009-04-29 | 2010-05-12 | 深圳市普禄科智能检测设备有限公司 | Device and method for diagnosing cylinder of diesel engine |
| CN101750210A (en) * | 2009-12-24 | 2010-06-23 | 重庆大学 | Fault diagnosis method based on OLPP feature reduction |
| CN102095588A (en) * | 2010-12-28 | 2011-06-15 | 中国人民解放军军事交通学院 | Intelligent fault diagnosis method for diesel engine |
| CN201993234U (en) * | 2011-01-14 | 2011-09-28 | 华南理工大学 | Diesel engine state detecting device based on energy consumption |
| CN102494899A (en) * | 2011-11-25 | 2012-06-13 | 华南理工大学 | Composite fault diagnosis method for diesel engine and diagnosis system |
| CN104897403A (en) * | 2015-06-24 | 2015-09-09 | 北京航空航天大学 | Self-adaption fault diagnosis method based on permutation entropy (PE) and manifold-based dynamic time warping (MDTW) |
Non-Patent Citations (2)
| Title |
|---|
| 吕世鹏: "基于集合经验模态分解的转子故障特征提取和分类识别", 《中国优秀硕士学位论文全文数据库》 * |
| 蔡江林 等: "一种基于经验模态分解和流形学习的滚动轴承故障诊断方法", 《井冈山大学学报（自然科学版）》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110261305A (en) * | 2019-06-18 | 2019-09-20 | 南京东南建筑机电抗震研究院有限公司 | Based on across footpaths continuous bridge damnification recognition methods such as the multispan for influencing line |
| CN110346143A (en) * | 2019-08-06 | 2019-10-18 | 江苏今创车辆有限公司 | Hyperfrequency noise assists empirical mode decomposition method |
| CN110346143B (en) * | 2019-08-06 | 2021-02-09 | 江苏今创车辆有限公司 | Ultrahigh frequency noise-assisted empirical mode decomposition method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108827634B (en) | 2019-12-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108760316B (en) | Information fusion method is joined in the change of variation mode decomposition | |
| Abdelkader et al. | Enhancement of rolling bearing fault diagnosis based on improvement of empirical mode decomposition denoising method | |
| He et al. | Automatic fault feature extraction of mechanical anomaly on induction motor bearing using ensemble super-wavelet transform | |
| Yin et al. | Weak fault feature extraction of rolling bearings based on improved ensemble noise-reconstructed EMD and adaptive threshold denoising | |
| Meng et al. | A hybrid fault diagnosis method using morphological filter–translation invariant wavelet and improved ensemble empirical mode decomposition | |
| Zhang et al. | Time-varying singular value decomposition for periodic transient identification in bearing fault diagnosis | |
| Ding et al. | Transient feature extraction based on time–frequency manifold image synthesis for machinery fault diagnosis | |
| CN108827634B (en) | Manifold merges empirical mode decomposition method | |
| Klein et al. | Bearing diagnostics using image processing methods | |
| Wang et al. | Fault Detection Enhancement in Rolling Element Bearings via Peak‐Based Multiscale Decomposition and Envelope Demodulation | |
| Li et al. | Incipient fault information determination for rolling element bearing based on synchronous averaging reassigned wavelet scalogram | |
| Sahu et al. | Fault diagnosis of rolling bearing based on an improved denoising technique using complete ensemble empirical mode decomposition and adaptive thresholding method | |
| GS et al. | Evaluation of effectiveness of wavelet based denoising schemes using ANN and SVM for bearing condition classification | |
| WO2023093315A1 (en) | Vibration signal feature extraction method, and device analysis method and apparatus | |
| Sun et al. | Application of wavelet soft threshold denoising algorithm based on EMD decomposition in vibration signals | |
| Imoru et al. | Diagnosis of stator shorted-turn faults in induction machines using discrete wavelet transform | |
| Fan et al. | A joint wavelet lifting and independent component analysis approach to fault detection of rolling element bearings | |
| Khan et al. | Intelligent fault detection via dilated convolutional neural networks | |
| JP2022087010A (en) | Time varying structure instantaneous frequency determination method, system, device, and storage medium | |
| Cao et al. | Remaining useful life prediction of wind turbine generator bearing based on EMD with an indicator | |
| Xu et al. | A novel method for extracting maximum kurtosis component and its applications in rolling bearing fault diagnosis | |
| Wang et al. | Fast Spectral Correlation Based on Sparse Representation Self‐Learning Dictionary and Its Application in Fault Diagnosis of Rotating Machinery | |
| Xu et al. | Peak‐Based Mode Decomposition for Weak Fault Feature Enhancement and Detection of Rolling Element Bearing | |
| CN114639391A (en) | Mechanical failure prompting method and device, electronic equipment and storage medium | |
| Guo et al. | Gear fault diagnosis based on narrowband demodulation with frequency shift and spectrum edit |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |