CN104729853B - A kind of rolling bearing performance degradation assessment device and method - Google Patents
A kind of rolling bearing performance degradation assessment device and method Download PDFInfo
- Publication number
- CN104729853B CN104729853B CN201510166910.1A CN201510166910A CN104729853B CN 104729853 B CN104729853 B CN 104729853B CN 201510166910 A CN201510166910 A CN 201510166910A CN 104729853 B CN104729853 B CN 104729853B
- Authority
- CN
- China
- Prior art keywords
- svdd
- signal
- wavelet packet
- performance degradation
- rolling bearing
- 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.)
- Expired - Fee Related
Links
Abstract
A kind of rolling bearing performance degradation assessment device and method, described device includes acceleration transducer, data acquisition module, characteristic extracting module, SVDD evaluation modules and authentication module.Be used for for acceleration transducer to gather the vibration signal of bearing to be measured and vibration signal is converted into analog signal by methods described;Data acquisition module is used to being amplified analog signal, filter etc. process after be converted to data signal, data signal is then sent to computer;Characteristic extracting module is used to extract the wavelet packet singular spectrum entropy of vibration signal as input feature vector vector, for the use of SVDD assessment models;SVDD evaluation modules are used to set up self adaptation SVDD models, and the performance degradation process of rolling bearing is estimated by self adaptation SVDD models, obtain performance degradation index DI;Authentication module uses the correctness of the Hilbert envelope demodulation method validation assessment results based on EMD.The present invention is applied to the performance degradation assessment of rolling bearing in life cycle management.
Description
Technical field
A kind of rolling bearing performance degradation assessment device and method, belongs to mechanical product quality reliability assessment and fault diagnosis
Technical field.
Background technology
With the fast development of industrial requirement, plant equipment is constantly improved at the aspect such as complicated, efficient, light-duty, together
When also face harsher working environment.Once the critical component of equipment breaks down, it is possible to can influence entirely to produce
Journey, causes huge economic loss, the problems such as result even in casualties.Therefore, plant-maintenance system is just determined by traditional
Phase repairs or correction maintenance changes to the condition maintenarnce based on state, and as setting up the premise of rational maintenance strategy, equipment
Energy degradation assessment also begins to receive much concern.
Rolling bearing directly affects whole and sets as one of key components and parts in rotating machinery, the quality of its performance state
Standby operational reliability.In general, rolling bearing can all experience from normally to degeneration up to the mistake for failing in use
Journey, and a series of different performance degradation states are generally experienced during this.If the mistake that can be degenerated in rolling bearing performance
The degree that bearing performance is degenerated is monitored in journey, then just can targetedly organize production and formulate rational maintenance meter
Draw, the generation for preventing unit exception from failing.
At present, time domain index is commonly used in engineering to monitor the running status of rolling bearing.Stability in time domain index refers to
Mark (such as root-mean-square value, root amplitude) can gradually increase with fault progression, but cannot judge the position of initial damage.And
Sensitiveness index (such as kurtosis index) is fallen after rising though the position of initial damage can be recognized as the development of failure can be presented
Trend, and do not meet the development trend of bearing fault degree.Accordingly, it would be desirable to the performance degradation index for building novelty is come comprehensively
Reflect the performance degradation process of rolling bearing.
The bearing vibration signal that actually measures is often non-linear, non-stationary signal, and WAVELET PACKET DECOMPOSITION can vibration letter
Number decomposed on different frequency bands, and not only the low frequency part to signal is decomposed, also the HFS to signal is carried out
Decompose, therefore WAVELET PACKET DECOMPOSITION can realize portray more fine to signal.
Wavelet packet singular spectrum entropy is theoretical based on singular value decomposition, by vibration signal through the coefficient matrix after wavelet package transforms
A series of singular values that can reflect former coefficient matrix essential characteristic are decomposed into, the statistical property of comentropy is recycled to singular value collection
Conjunction carries out analysis on Uncertainty, therefore wavelet packet singular spectrum entropy can provide a determination to the complexity of original vibration signal
Measure.
Support Vector data description (Support Vector Data Description, SVDD) is one kind in statistics
Practise the monodrome sorting technique based on border thought grown up on the basis of theoretical and SVMs, it is only necessary to which normal sample is carried out
Model training, this provides solution route for the abnormal data scarcity problem in fault diagnosis.Additionally, the method has calculates speed
Degree is fast, strong robustness the features such as.
The content of the invention
The purpose of the present invention is, in order to obtain the performance degradation index of rolling bearing, the initial failure of bearing is found in time
Moment and failure moment, prevent the generation of major accident, the present invention provides a kind of rolling bearing performance degradation assessment device and side
Method.
Realize the technical scheme is that, the present invention provides a kind of rolling bearing performance degradation assessment device, including,
Acceleration transducer, for gathering the vibration signal of bearing to be measured and vibration signal being converted into analog signal;
Data acquisition module, for the analog signal treatment such as to be amplified, filtered after be converted to data signal, so
The data signal is sent to computer afterwards, and is stored as data file;
Characteristic extracting module, for extracting the wavelet packet singular spectrum entropy of vibration signal as input feature vector vector, for institute
State the use of assessment models in SVDD evaluation modules;
SVDD evaluation modules and authentication module, for setting up self adaptation SVDD models, and by the self adaptation SVDD moulds
Type is estimated to the performance degradation process of rolling bearing, obtains performance degradation index DI;Authentication module is used to verify that performance is moved back
Change the correctness of assessment result.
The acceleration transducer connects the input of data acquisition module;Data acquisition module passes through characteristic extracting module
Connection SVDD evaluation modules and authentication module;The data acquisition module includes NI SCXI accelerometers input module, NI
SCXI signal conditions cabinet and NI multifunctional data acquisition cards, wherein accelerometer input module are encapsulated in signal condition machine
In case, data collecting card can be using usb bus or the NI data collecting cards of pci bus, to adapt to different computer requirements.
The present invention provides a kind of rolling bearing performance degradation assessment method, including data acquisition, feature extraction, Performance Evaluation
With the checking to assessment result, rolling bearing performance is degenerated by rolling bearing performance degradation assessment device is estimated.Institute
The method of stating includes:
After bearing vibration signal is carried out into WAVELET PACKET DECOMPOSITION, the WAVELET PACKET DECOMPOSITION coefficient to last layer of each node enters respectively
Line reconstruction, then wavelet package reconstruction coefficient to obtaining carries out singular value decomposition, and then asks for last layer of wavelet packet of each node
Singular spectrum entropy;
Using wavelet packet singular spectrum entropy as the input feature vector vector of SVDD evaluation modules, set up certainly by input feature vector vector
Adapt to SVDD models and obtain performance degradation index DI;
Performance degradation assessment result is verified using the Hilbert envelope demodulations method based on EMD.
Feature extraction in the inventive method is comprised the following steps:
(1) WAVELET PACKET DECOMPOSITION, according to the vibration signal waveforms of rolling bearing, from db5 wavelet basis as wavelet basis function
Vibration signal to collecting carries out 4 layers of WAVELET PACKET DECOMPOSITION, obtains last layer of WAVELET PACKET DECOMPOSITION coefficient of each node;
(2) decomposition coefficient is reconstructed, the WAVELET PACKET DECOMPOSITION coefficient of last layer of each node is reconstructed, obtain weight
Structure coefficient;
(3) singular value decomposition is carried out to reconstruction coefficients, the wavelet package reconstruction coefficient of last layer of each node is carried out unusual
Value is decomposed, then each sample standard deviation can obtain 16 singular value r1,r2,…,r16, and these singular values are normalized,
(4) wavelet packet singular spectrum entropy is calculated, by the definition of comentropy, the wavelet packet singular spectrum entropy of vibration signal can be represented
For:Si=-gilog2gi;Then the vibration signal at each moment is comprising 16 wavelet packet singular spectrum entropy vectors.
Performance Evaluation in the inventive method is comprised the following steps:
(1) using the feature samples under normal condition as the input vector of SVDD model trainings, envelope normal sample is obtained
The suprasphere of feature space, and obtain the radius R of the suprasphere;
(2) new feature samples a is input into, this feature sample to the generalized distance R of SVDD supraspheres center d is calculateda;
(3) R is comparedaWith the size of R, if Ra- R≤0, then perform step (4), if Ra-R>0, then show SVDD model trainings
Terminate, perform step (5);
(4) input feature vector by the feature samples under feature samples a and normal condition together as SVDD model trainings is sweared
Amount, continues to update SVDD models by step (1)-(3);
(5) it is estimated in new feature sample being substituted into self adaptation SVDD models, calculates new feature sample to self adaptation
The generalized distance R of SVDD supraspheres center db, and try to achieve performance degradation index DI.
Assessment result is verified in the inventive method, is comprised the following steps:
(1) vibration signal to rolling bearing initial failure moment and failure moment carries out EMD decomposition, obtains limited originally
Levy modular function IMF;
(2) correlation analysis is carried out to each IMF component and primary signal respectively;
(3) choose the first two IMF component high with primary signal correlation and be overlapped reconstruct, obtain reconstruction signal;
(4) reconstruction signal is obtained into its signal envelope using Hilbert conversion;
(5) FFT is carried out to signal envelope, obtains the envelope spectrum of reconstruction signal, according to envelope spectrum Zhong Ke areas
Relation between the spectral line frequency and fault characteristic frequency that divide verifies the correctness of assessment result.
The beneficial effects of the invention are as follows the present invention uses wavelet packet singular spectrum entropy as the input feature vector of SVDD assessment models
Vector, can reflect non-linear, the non-stationary characteristic of bearing vibration signal, and complexity to vibration signal provides one
What is determined measures.It is estimated invention introduces self adaptation SVDD models so that SVDD supraspheres border is with sample to be tested
Increase and constantly update, the accuracy of assessment models can be greatly improved.The present invention is using the Hilbert envelopes based on EMD
Demodulation method carries out double verification to assessment result, it is ensured that the correctness and validity of assessment result.
The present invention is applicable the performance degradation assessment in rolling bearing life cycle management.
Brief description of the drawings
Fig. 1 is rolling bearing performance degradation assessment schematic device;
Fig. 2 is characterized extraction module flow chart;
Fig. 3 is self adaptation SVDD model algorithm flow charts;
Fig. 4 is authentication module flow chart.
Specific embodiment
Embodiments of the present invention are related to a kind of rolling bearing performance degradation assessment device, as shown in Figure 1.
The present embodiment be a kind of rolling bearing performance degradation assessment device, including acceleration transducer, data acquisition module,
Characteristic extracting module, SVDD evaluation modules and authentication module.
Acceleration transducer is used to gather the vibration signal of bearing to be measured and vibration signal is converted into analog signal;Data
Acquisition module be used to being amplified the analog signal, filter etc. process after be converted to data signal, then by the numeral
Signal is sent to computer, and is stored as data file;Characteristic extracting module is used to extract the wavelet packet singular spectrum of vibration signal
Entropy as input feature vector vector, for the use of assessment models in the SVDD evaluation modules;SVDD evaluation modules are used to set up
Self adaptation SVDD models, and the performance degradation process of rolling bearing is estimated by the self adaptation SVDD models, obtain
Performance degradation index DI, is that can determine that the initial failure moment and failure moment of bearing by the index;The authentication module is used
In the correctness of checking performance degradation assessment result.
The installation site of the acceleration transducer of the present embodiment device must be fixed, to ensure the comparativity of signal.NI
SCXI accelerometers input module can select SCXI-1531 accelerometer input modules, and accelerometer input module is encapsulated in
In SCXI signal condition cabinets.Accelerometer input module is mainly analog signal the conditioning such as is amplified, filtered.To accelerate
Core bus and cable adaptor input data capture card that the signal of degree meter input module output passes through signal condition cabinet, number
Can be inserted directly into using pci bus or the NI data collecting cards of usb bus, the data collecting card of wherein pci bus according to capture card
In the PCI slot of computer motherboard, the data collecting card of usb bus is connected with the USB interface of computer, just can so fit
Answer different computer requirements.Characteristic extracting module, SVDD evaluation modules and authentication module are completed in a computer.Computer
MATLAB softwares and LABVIEW softwares should be carried.By the collection of LABVIEW programming realization bearing vibration signals, collection
Vibration signal be stored as data file.
The specific embodiment of rolling bearing performance degradation assessment method is realized by device in the present invention.And in device
Characteristic extracting module, SVDD evaluation modules and authentication module realize that specific embodiment is as follows by MATLAB softwares:
1st, feature extraction
Characteristic extracting module flow chart as shown in Fig. 2 implement according to the following steps:
(1) WAVELET PACKET DECOMPOSITION
According to the vibration signal waveforms of rolling bearing, the vibration from db5 wavelet basis as wavelet basis function to collecting
Signal carries out 4 layers of WAVELET PACKET DECOMPOSITION, obtains last layer of WAVELET PACKET DECOMPOSITION coefficient of each node.Wavelet transformation may be considered profit
Vibration signal is approached with a series of wavelet basis function, it is general in actually calculating to use Mallat fast algorithms, should
Algorithm is decomposed by constructing a series of low pass filter groups and high-pass filter group to approximation signal:
Wherein a0,k=x (i), i=0,1,2 ..., N-1, N count for signal sampling, and x (i) is discrete time signal, and j is
The WAVELET PACKET DECOMPOSITION number of plies, k=0,1,2 ..., 15, p (n), q (n) they are the impulse response of conjugate mirror filter P, Q, aj,k(i)、
bj,kI () is respectively low frequency, high-frequency decomposition coefficient.
(2) decomposition coefficient is reconstructed
The WAVELET PACKET DECOMPOSITION coefficient of last layer of each node is reconstructed, restructing algorithm is as follows:
(3) singular value decomposition is carried out to reconstruction coefficients
Wavelet package reconstruction coefficient to last layer of each node carries out singular value decomposition, then each sample standard deviation is available 16
Singular value r1,r2,…,r16, and these singular values are normalized
(4) wavelet packet singular spectrum entropy is calculated
By the definition of comentropy, the wavelet packet singular spectrum entropy of vibration signal is represented by
Si=-gilog2gi (5)
Then the vibration signal at each moment is comprising 16 wavelet packet singular spectrum entropy vectors.
2nd, SVDD assessments
SVDD evaluation modules are mainly sets up self adaptation SVDD models by input feature vector vector, and the basic thought of SVDD is just
It is one minimal hyper-sphere of generation, makes it as far as possible comprising all of normal characteristics sample, various can be constructed by following
SVDD models:
Wherein d is the suprasphere centre of sphere of SVDD models, and R is suprasphere radius, and C is penalty factor, and ξ is relaxation factor, and L is
Object function, is introduced into Lagrange multiplier and constraints is imported into object function, and inner product is replaced using kernel function, can obtain
Following quadratic programming formula:
Wherein αiIt is Lagrange multiplier, meets 0≤αiThe sample of≤C conditions is supporting vector, so as to obtain SVDD surpass
The centre of sphere d and radius R of spheroid.R can be determined by following formula:
Wherein, xsIt is supporting vector, αiIt is Lagrange multiplier, xiIt is target sample.
Adaptive process, self adaptation SVDD model algorithms flow chart such as Fig. 3 are increased during SVDD models are set up
Shown, the specific implementation step of algorithm is as follows:
(1) the wavelet packet singular spectrum entropy vector under the normal condition that will be measured offline is obtained as input feature vector vector
SVDD supraspheres, and obtain the radius R of the suprasphere;
(2) the feature samples a that Input Online is measured, calculates this feature sample to the generalized distance of SVDD supraspheres center d
Ra, RaCan be determined by following formula:
(3) R is comparedaWith the size of R, if Ra- R≤0, then show that feature samples a belongs to the feature samples under normal condition,
Step (4) is performed, if Ra-R>0, then show that self adaptation SVDD model trainings terminate, perform step (5);
(4) input feature vector by the feature samples under feature samples a and normal condition together as SVDD model trainings is sweared
Amount, continues to press step (1) to (3), updates SVDD models, and the radius that renewal terminates rear self adaptation SVDD models is R1;
(5) the new feature sample that will be measured online is estimated in substituting into self adaptation SVDD models, calculates new feature sample
To the generalized distance R of self adaptation SVDD supraspheres center db, RbIt is same to be determined by formula (11), can finally try to achieve a series of performances and move back
Change index DI values, DI values are determined by following formula:
Then the decision criteria at rolling bearing initial failure moment and failure moment is:If DI values are all higher than after sometime
0, and always ascendant trend, then the moment can determine that to be the bearing initial failure moment;If DI value curves rise to sometime
When slope between later moment in time and the moment reach maximum, then the moment can determine that to be the bearing failure moment.Additionally, DI values
Conspicuousness turning point of the curve in uphill process is regarded as the turning point of different phase in bearing performance degenerative process.
3rd, assessment result is verified
The rolling bearing initial failure moment and failure the moment determine after, in order to ensure assessment result correctness and effectively
Property, double verification can be carried out to assessment result by authentication module.Authentication module flow chart is as shown in figure 4, real according to the following steps
Apply:
(1) the vibration signal data file for having been determined as the initial failure moment is transferred, EMD decomposition is carried out to vibration signal,
Determine all of maximum point of signal and minimum point, be fitted maximum point and minimum point respectively by cubic spline, obtain
The coenvelope line and lower envelope line of signal, and the averaged curve of coenvelope line and lower envelope line is sought, adopted further according to the averaged curve
With screening, principle is by signal decomposition and obtains limited intrinsic mode functions IMF;
(2) correlation analysis is carried out to each IMF component and primary signal respectively, and obtains coefficient correlation;
(3) the maximum the first two IMF components of coefficient correlation are chosen and is overlapped reconstruct, obtain reconstruction signal;
(4) reconstruction signal is obtained into its signal envelope using Hilbert conversion;
(5) FFT is carried out to signal envelope, obtains the envelope spectrum of reconstruction signal, and calculate rolling to be measured
The fault characteristic frequency at each position of bearing (including inner ring, outer ring, rolling element and retainer), according to differentiable spectrum in envelope spectrum
Relation between line frequency and fault characteristic frequency verifies the correctness of assessment result.
Checking the bearing failure moment then transfer have been determined as fail the moment vibration signal data file and by step (1) extremely
(5) verified.
During the checking initial failure moment, if presence can distinguish the spectral line and its frequency multiplication of " carpet " noise in signal envelope spectrum
Spectral line, and the spectral line frequency and a certain position of rolling bearing fault characteristic frequency very close to then can determine that in the moment bearing
Initial failure is occurred in that, i.e. the moment is the initial failure moment.
During the checking failure moment, if there is very prominent spectral line and its again spectrum line in signal envelope spectrum, now envelope spectrum
In " carpet " noise very little, and the spectral line frequency and a certain position of rolling bearing fault characteristic frequency very close to then can be true
The moment bearing failure, the i.e. moment are scheduled on for the bearing failure moment.
Claims (1)
1. a kind of rolling bearing performance degradation assessment method, it is characterised in that methods described includes:Data acquisition, feature extraction,
Performance Evaluation and the checking to assessment result;
After bearing vibration signal is carried out into WAVELET PACKET DECOMPOSITION, the WAVELET PACKET DECOMPOSITION coefficient to last layer of each node carries out weight respectively
Structure, then wavelet package reconstruction coefficient to obtaining carries out singular value decomposition, and then it is unusual to ask for the wavelet packet of last layer of each node
Spectrum entropy;
Using wavelet packet singular spectrum entropy as the input feature vector vector of SVDD evaluation modules, self adaptation is set up by input feature vector vector
SVDD models simultaneously obtain performance degradation index DI;
Performance degradation assessment result is verified using the Hilbert envelope demodulations method based on EMD;
The feature extraction is comprised the following steps:
(1) WAVELET PACKET DECOMPOSITION, according to the vibration signal waveforms of rolling bearing, from db5 wavelet basis as wavelet basis function to adopting
The vibration signal for collecting carries out 4 layers of WAVELET PACKET DECOMPOSITION, obtains last layer of WAVELET PACKET DECOMPOSITION coefficient of each node;
(2) decomposition coefficient is reconstructed, the WAVELET PACKET DECOMPOSITION coefficient of last layer of each node is reconstructed;
(3) singular value decomposition is carried out to reconstruction coefficients, the wavelet package reconstruction coefficient to last layer of each node carries out singular value point
Solve, then each sample standard deviation can obtain 16 singular value r1,r2,…,r16, and these singular values are normalized,
(4) wavelet packet singular spectrum entropy is calculated, by the definition of comentropy, the wavelet packet singular spectrum entropy of vibration signal is represented by:Si
=-gilog2gi;Then the vibration signal at each moment is comprising 16 wavelet packet singular spectrum entropy vectors;
The Performance Evaluation includes step in detail below:
(1) using the feature samples under normal condition as the input vector of SVDD model trainings, envelope normal sample feature is obtained
The suprasphere in space, and obtain the radius R of the suprasphere;
(2) new feature samples a is input into, this feature sample to the generalized distance R of SVDD supraspheres center d is calculateda;
(3) R is comparedaWith the size of R, if Ra- R≤0, then perform step (4), if Ra-R>0, then show SVDD model training knots
Beam, performs step (5);
(4) by the feature samples under feature samples a and normal condition together as SVDD model trainings input feature vector vector, after
It is continuous to update SVDD models by step (1)-(3);
(5) it is estimated in new feature sample being substituted into self adaptation SVDD models, calculates new feature sample super to self adaptation SVDD
The generalized distance R of ball centre db, and try to achieve performance degradation index DI;
It is described that assessment result is verified, comprise the following steps:
(1) vibration signal to rolling bearing initial failure moment and failure moment carries out EMD decomposition, obtains limited eigen mode
Function IMF;
(2) correlation analysis is carried out to each IMF component and primary signal respectively;
(3) choose the first two IMF component high with primary signal correlation and be overlapped reconstruct, obtain reconstruction signal;
(4) reconstruction signal is obtained into its signal envelope using Hilbert conversion;
(5) FFT is carried out to signal envelope, obtains the envelope spectrum of reconstruction signal, according to differentiable in envelope spectrum
Relation between spectral line frequency and fault characteristic frequency verifies the correctness of assessment result.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510166910.1A CN104729853B (en) | 2015-04-10 | 2015-04-10 | A kind of rolling bearing performance degradation assessment device and method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510166910.1A CN104729853B (en) | 2015-04-10 | 2015-04-10 | A kind of rolling bearing performance degradation assessment device and method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN104729853A CN104729853A (en) | 2015-06-24 |
CN104729853B true CN104729853B (en) | 2017-06-06 |
Family
ID=53453951
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201510166910.1A Expired - Fee Related CN104729853B (en) | 2015-04-10 | 2015-04-10 | A kind of rolling bearing performance degradation assessment device and method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN104729853B (en) |
Families Citing this family (24)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105956514B (en) * | 2016-04-15 | 2019-07-23 | 南京航空航天大学 | A kind of lifting airscrew method for detecting abnormality of vibration data driving |
CN106053066A (en) * | 2016-05-23 | 2016-10-26 | 华东交通大学 | Antifriction bearing performance degradation assessment method based on empirical mode decomposition and logistic regression |
CN109635307A (en) * | 2016-06-29 | 2019-04-16 | 潍坊学院 | WT, spectrum kurtosis and the smooth iterative demodulation method for diagnosing faults of rolling bearing |
CN109635306A (en) * | 2016-06-29 | 2019-04-16 | 潍坊学院 | Rotary machinery fault diagnosis method based on wavelet decomposition and spectrum kurtosis |
CN106096198B (en) * | 2016-06-29 | 2019-02-22 | 潍坊学院 | A kind of envelope Analysis Method based on variation Mode Decomposition and spectrum kurtosis |
CN106198746A (en) * | 2016-07-19 | 2016-12-07 | 中南林业科技大学 | A kind of ultrasonic intelligent detection wood members glue-line defect method |
CN106338385B (en) * | 2016-08-25 | 2019-03-19 | 东南大学 | A kind of rotary machinery fault diagnosis method based on unusual spectral factorization |
CN106355148B (en) * | 2016-08-29 | 2019-06-07 | 广东工业大学 | A kind of denoising method and device based on SSA and EMD combination |
CN106644481B (en) * | 2016-12-27 | 2018-10-30 | 哈尔滨理工大学 | Rolling bearing Predicting Reliability method based on mathematical morphology and IFOA-SVR |
CN106844957A (en) * | 2017-01-19 | 2017-06-13 | 西安工业大学 | A kind of engine operation reliability evaluation method under condition of small sample |
CN107144430B (en) * | 2017-06-27 | 2019-02-01 | 电子科技大学 | A kind of Method for Bearing Fault Diagnosis based on incremental learning |
CN108181098A (en) * | 2017-12-15 | 2018-06-19 | 天津金岸重工有限公司 | A kind of high pedestal jib crane low-speed heave-load unit failure feature extracting method |
CN108614940A (en) * | 2018-05-02 | 2018-10-02 | 湖南大学 | A kind of permanent magnet synchronous motor performance degradation assessment method and system |
CN108629348A (en) * | 2018-05-09 | 2018-10-09 | 江南大学 | A kind of Trouble Diagnostic Method of Machinery Equipment based on significance test |
CN109374298B (en) * | 2018-11-06 | 2020-08-28 | 石家庄铁道大学 | Bearing fault diagnosis method based on cross-correlation singular value |
CN109974985A (en) * | 2019-02-22 | 2019-07-05 | 昆明理工大学 | A kind of check valve performance degradation assessment device and its diagnostic method |
CN109974782B (en) * | 2019-04-10 | 2021-03-02 | 郑州轻工业学院 | Equipment fault early warning method and system based on big data sensitive characteristic optimization selection |
CN110132596A (en) * | 2019-04-24 | 2019-08-16 | 昆明理工大学 | A method of the rolling bearing fault diagnosis based on wavelet packet and GWO-SVM |
CN110595751B (en) * | 2019-09-19 | 2020-12-18 | 华东理工大学 | Early fault characteristic wavelet reconstruction method guided by Gini index and application thereof |
CN110836776B (en) * | 2019-12-25 | 2021-08-10 | 济南大学 | Bearing fault identification method and system based on multilayer information entropy characteristic matrix |
CN111504647A (en) * | 2020-06-09 | 2020-08-07 | 华东交通大学 | AR-MSET-based performance degradation evaluation method for rolling bearing |
CN111896260B (en) * | 2020-08-01 | 2022-05-13 | 华东交通大学 | NGAs synchronous optimization wavelet filter and MCKD bearing fault diagnosis method |
CN114252266A (en) * | 2021-12-27 | 2022-03-29 | 上海应用技术大学 | Rolling bearing performance degradation evaluation method based on DBN-SVDD model |
CN116242612B (en) * | 2023-01-09 | 2023-11-21 | 广东海洋大学 | Fault diagnosis method, device, medium and equipment |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
DE102008021360A1 (en) * | 2008-04-29 | 2009-11-05 | Siemens Aktiengesellschaft | Method and device for detecting bearing damage |
CN102721545A (en) * | 2012-05-25 | 2012-10-10 | 北京交通大学 | Rolling bearing failure diagnostic method based on multi-characteristic parameter |
CN102778358A (en) * | 2012-06-04 | 2012-11-14 | 上海东锐风电技术有限公司 | Failure prediction model establishing method and system as well as fan monitoring pre-warning system and method |
CN202582898U (en) * | 2011-12-20 | 2012-12-05 | 东南大学 | Flush type bearing failure intelligent diagnosing device based on ARM and DSP |
CN104112072A (en) * | 2014-07-15 | 2014-10-22 | 华侨大学 | Operating modal parameter identification method for principal component analysis on basis of wavelet threshold denoising |
-
2015
- 2015-04-10 CN CN201510166910.1A patent/CN104729853B/en not_active Expired - Fee Related
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
DE102008021360A1 (en) * | 2008-04-29 | 2009-11-05 | Siemens Aktiengesellschaft | Method and device for detecting bearing damage |
CN202582898U (en) * | 2011-12-20 | 2012-12-05 | 东南大学 | Flush type bearing failure intelligent diagnosing device based on ARM and DSP |
CN102721545A (en) * | 2012-05-25 | 2012-10-10 | 北京交通大学 | Rolling bearing failure diagnostic method based on multi-characteristic parameter |
CN102778358A (en) * | 2012-06-04 | 2012-11-14 | 上海东锐风电技术有限公司 | Failure prediction model establishing method and system as well as fan monitoring pre-warning system and method |
CN104112072A (en) * | 2014-07-15 | 2014-10-22 | 华侨大学 | Operating modal parameter identification method for principal component analysis on basis of wavelet threshold denoising |
Non-Patent Citations (3)
Title |
---|
Performance degradation assessment of rolling bearing based on bispectrum and support vector data description;Hongchao Wang et al.;《Journal of Vibration and Control》;20141231;第20卷;第2032-2041页 * |
基于声波信号小波包奇异谱熵和支持向量机的断路器故障诊断研究;谭振沛;《价值工程》;20121231(第25期);第28-30页 * |
小波包-支持向量数据描述在轴承性能退化评估中的应用研究;潘玉娜等;《振动与冲击》;20090430;第28卷(第4期);第164-167页 * |
Also Published As
Publication number | Publication date |
---|---|
CN104729853A (en) | 2015-06-24 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN104729853B (en) | A kind of rolling bearing performance degradation assessment device and method | |
Li et al. | The entropy algorithm and its variants in the fault diagnosis of rotating machinery: A review | |
CN105716857B (en) | A kind of epicyclic gearbox health state evaluation method | |
Yang et al. | Vibration condition monitoring system for wind turbine bearings based on noise suppression with multi-point data fusion | |
CN101900789B (en) | Tolerance analog circuit fault diagnosing method based on wavelet transform and fractal dimension | |
CN104655423B (en) | A kind of Fault Diagnosis of Roller Bearings based on time-frequency domain multi-dimensional vibration Fusion Features | |
Wenyi et al. | Wind turbine fault diagnosis method based on diagonal spectrum and clustering binary tree SVM | |
CN102866027B (en) | Based on the rotary machine fault characteristic extraction method of LMD and local time-frequency entropy | |
WO2022156330A1 (en) | Fault diagnosis method for rotating device | |
CN106934125A (en) | A kind of exponential model plant equipment method for predicting residual useful life of trapezoidal noise profile | |
CN108426713A (en) | Rolling bearing Weak fault diagnostic method based on wavelet transformation and deep learning | |
CN106197999B (en) | A kind of planetary gear method for diagnosing faults | |
Gao et al. | Total variation on horizontal visibility graph and its application to rolling bearing fault diagnosis | |
CN106885697A (en) | The performance degradation assessment method of the rolling bearing based on FCM HMM | |
CN110298085A (en) | Analog-circuit fault diagnosis method based on XGBoost and random forests algorithm | |
CN103115789A (en) | Second generation small-wave support vector machine assessment method for damage and remaining life of metal structure | |
Shi et al. | A two-stage sound-vibration signal fusion method for weak fault detection in rolling bearing systems | |
CN102179722A (en) | Estimation method for operational reliability of numerical control machine based on proportional hazards model | |
CN103018043A (en) | Fault diagnosis method of variable-speed bearing | |
CN106769051B (en) | A kind of rolling bearing remaining life prediction technique based on MCEA-KPCA and combination S VR | |
CN106934126A (en) | Component of machine health indicator building method based on Recognition with Recurrent Neural Network fusion | |
CN108444696A (en) | A kind of gearbox fault analysis method | |
CN109813547A (en) | Rotating machinery local type method for diagnosing faults based on sparse decomposition optimization algorithm | |
CN103822786A (en) | Wind turbine generator set mechanical equipment state diagnosis method based on multivariate statistical analysis | |
CN102682180A (en) | Evaluation method for performance degradation of rotary mechanical equipment |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20170606 Termination date: 20180410 |
|
CF01 | Termination of patent right due to non-payment of annual fee |