CN109000921B - Method for diagnosing main shaft fault of wind turbine generator - Google Patents

Method for diagnosing main shaft fault of wind turbine generator Download PDF

Info

Publication number
CN109000921B
CN109000921B CN201710423126.3A CN201710423126A CN109000921B CN 109000921 B CN109000921 B CN 109000921B CN 201710423126 A CN201710423126 A CN 201710423126A CN 109000921 B CN109000921 B CN 109000921B
Authority
CN
China
Prior art keywords
fault
model
modal
signal
main shaft
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.)
Active
Application number
CN201710423126.3A
Other languages
Chinese (zh)
Other versions
CN109000921A (en
Inventor
孙兆儒
梁秀广
宋红兵
霍锦
王子佳
杨松
征少卿
郭懿萱
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
China Datang Technologies and Engineering Co Ltd
Original Assignee
China Datang Technologies and Engineering Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by China Datang Technologies and Engineering Co Ltd filed Critical China Datang Technologies and Engineering Co Ltd
Priority to CN201710423126.3A priority Critical patent/CN109000921B/en
Publication of CN109000921A publication Critical patent/CN109000921A/en
Application granted granted Critical
Publication of CN109000921B publication Critical patent/CN109000921B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
    • G01M13/00—Testing of machine parts
    • G01M13/02—Gearings; Transmission mechanisms
    • G01M13/028—Acoustic or vibration analysis

Abstract

The invention provides a method for diagnosing a main shaft fault of a wind turbine generator, which comprises vibration signal acquisition, VMD variation modal decomposition, sensitive factor evaluation, signal reconstruction and VPMCD fault diagnosis, wherein a fan vibration signal is subjected to VMD analysis, the signal is decomposed into a group of modal components with different frequencies, and as the fault characteristics only appear in a specific frequency band, the modal components are evaluated and analyzed by adopting the sensitive factors, components containing the fault characteristics are screened, noise interference components are eliminated, the signal is reconstructed, and fault information is highlighted; and then, extracting the characteristic value of the reconstructed signal to form a characteristic identification vector, and performing fault diagnosis by establishing an intrinsic relation of the characteristic value by combining a VPMCD method, so that the processes of subjective parameter selection and optimization of the traditional classifier are avoided, and the fault diagnosis time is shortened.

Description

Method for diagnosing main shaft fault of wind turbine generator
Technical Field
The invention relates to the technical field of fault diagnosis, in particular to wind turbine generator main shaft fault diagnosis, and specifically relates to a wind turbine generator main shaft fault diagnosis method.
Background
With the emphasis on the development of non-fossil energy in China, the installed capacity of wind power generation is increased year by year. The direct-drive wind driven generator gradually becomes a mainstream trend due to the advantages of high power generation efficiency, low maintenance cost, excellent power grid access performance and the like. However, because the wind power plant is mostly located in a severe environment, the operation condition is complex, so that the operation parts of the unit are easy to damage, especially the main shaft breaks down, and if the main shaft cannot be found and processed in time, a shutdown accident is directly caused, and huge loss is caused. Therefore, in recent years, a plurality of scholars deeply research on fault diagnosis of the main shaft of the fan and obtain abundant results. Some wind turbine generator rolling bearing fault characteristics are extracted by Empirical Mode Decomposition (EMD); some of the rolling bearing fault types are classified by combining Local Mean Decomposition (LMD) and approximate entropy; some methods add adaptive high-frequency harmonic waves aiming at the modal aliasing problem of the LMD method and successfully extract the unbalanced fault characteristics of the shafting. Both the EMD method and the LMD method belong to recursive mode decomposition, and the decomposition mode has an endpoint effect, which causes envelope deformation and causes decomposition inaccuracy, thereby being a limitation of analyzing and processing signals by the methods.
Disclosure of Invention
In order to solve the technical problem, the invention provides a method for diagnosing a main shaft fault of a wind turbine generator, and aims to provide a method for diagnosing a main shaft fault by automatically identifying the type of the main shaft fault and solve the problem of inaccurate classification in the fault diagnosis process.
The invention provides a method for diagnosing faults of a main shaft of a wind turbine generator, which comprises the following steps:
step 1, vibration signal acquisition:
acquiring a vibration signal through a sensor arranged on a main shaft seat of the wind turbine generator;
step 2, VMD variational modal decomposition:
performing VMD (variable mode decomposition) on the vibration signal to obtain a plurality of modal components under different frequencies;
and 3, sensitive factor evaluation:
calculating a sensitivity factor of each modal component, and evaluating the fault characteristic degree of each modal component;
and 4, signal reconstruction:
noise interference is eliminated, signals are reconstructed, the characteristic vector of each reconstructed signal is calculated, and a modal analysis vector is constructed;
step 5, VPMCD fault diagnosis:
carrying out fault identification on the modal analysis vector by using a prediction model, calculating the error square sum of the main shaft state VPM, and carrying out classification identification by taking the minimum value as judgment;
the method comprises the steps of VMD (variable mode decomposition), VPMCD (virtual machine model) and main shaft state prediction, wherein the VMD is variable mode decomposition, the VPMCD is mode recognition of a variable prediction model, and the main shaft state comprises a normal state, an outer ring fault, an inner ring fault and a rolling body fault.
As a further improvement of the present invention, the sensor in step 1 is an acceleration sensor.
As a further improvement of the present invention, step 2 specifically includes:
step 201, decomposing the vibration signal into K single-component mode functions, where the single-component mode functions are limited band intrinsic mode functions:
Figure BDA0001315521120000021
wherein f (t) is a vibration signal, uk(t) is the kth band-limited intrinsic mode function, Ak(t) is the envelope curve and,
Figure BDA0001315521120000022
is a phase function, t is time;
step 202, estimating the frequency bandwidth of each single-component mode function, and establishing a variation constraint model:
Figure BDA0001315521120000023
wherein, wkFor the kth single-component mode function uk(t) center angular frequency, { u }k}＝{u1……uK}，{wk}＝{w1……wK}，
Figure DEST_PATH_GDA0001340976200000025
The sum of all single-component mode functions is obtained;
step 203, introducing the secondary penalty factor a and the Lagrangian operator r into a variation constraint model to obtain an augmented Lagrangian equation;
Figure BDA0001315521120000025
wherein a is a secondary penalty factor, and r is a Lagrange operator;
step 204, initialize uk、wkAnd r value, continuously iteratively updating by adopting an alternative direction multiplier method, searching saddle points of the augmented Lagrange in an iterative optimization sequence, and determining the judgment precision e1And according to the judgment formula, stopping iterative output of the K modal components for the condition of meeting the judgment formula.
As a further improvement of the present invention, the determination formula is specifically:
Figure BDA0001315521120000031
wherein, get e1＝1×10-6。
As a further improvement of the present invention, step 3 specifically comprises:
calculating the sensitivity factor lambda of each modal component obtained in step 2k，
Figure BDA0001315521120000032
Wherein, deltak＝βk－αkWherein αkCorrelation coefficient of modal component and fault signal, βkIs the correlation coefficient of the modal component and the normal signal.
As a further improvement of the present invention, step 4 specifically includes:
step 401, utilizing a sorting method to sort the sensitivity factor λkSequencing from big to small to obtain a new sensitive factor sequence lambda'k；
Step 402, according to a sensitive factor sequence lambda'kCalculating a sequence of difference spectrum evaluation components dk，
dk＝λ′k+1-λ′k；
Step 403, finding out the difference spectrum evaluation component sequence dkMaximum value of dj；
Step 404, passing through front jA dkCorresponding sensitivity factor lambda'kFinding out corresponding j modal components, and superposing the j modal components to obtain a reconstructed signal;
and 405, obtaining a feature vector by using a singular value decomposition method according to the reconstructed signal.
As a further improvement of the present invention, the sorting method in step 401 is bubble sorting method or cocktail sorting method or radix sorting method or hill sorting method.
As a further improvement of the present invention, step 5 specifically includes:
step 501, substituting the characteristic vector into a variable prediction model, wherein the variable prediction model is VPMi：
Xi＝f(Xj,b0,bj,bjj,bjk)+e2，
Wherein, b0,bj,bjj,bjkIs a model parameter; e.g. of the type2Is a prediction error;
step 502, calculating the sum of the squares of the errors of the states VPM, and performing classification and identification with the minimum value as a judgment.
As a further improvement of the invention, the model parameters are obtained by training through vibration data of various fault classes, and the prediction error is preset.
As a further improvement of the present invention, the prediction model is a linear model or a linear interaction model or a quadratic interaction model;
the linear model is:
Figure BDA0001315521120000041
the linear interaction model is as follows:
Figure BDA0001315521120000042
the secondary model is as follows:
Figure BDA0001315521120000043
the secondary interaction model is as follows:
Figure BDA0001315521120000044
the invention has the following beneficial effects: VMD analysis is carried out on the fan vibration signal, the signal is decomposed into a group of modal components with different frequencies, and because the fault characteristics only appear in a specific frequency band, the modal components are evaluated and analyzed by adopting a sensitive factor, components containing the fault characteristics are screened, noise interference components are eliminated, the signal is reconstructed, and the fault information is highlighted; and then, extracting the characteristic value of the reconstructed signal to form a characteristic identification vector, and performing fault diagnosis by establishing an intrinsic relation of the characteristic vector in combination with a VPMCD method, so that the processes of subjective parameter selection and optimization of the traditional classifier are avoided, and the fault diagnosis time is shortened.
Drawings
Fig. 1 is a flowchart of a method for diagnosing a failure of a main shaft of a wind turbine generator according to a first embodiment of the present invention;
FIG. 2 is a diagram of a descending order of sensitivity factor sequences and a spectrum of sensitivity factor difference according to a first embodiment of the present invention;
FIG. 3 is a vector trend chart of modal analysis of different status signal samples according to a first embodiment of the present invention.
Detailed Description
The invention is described in further detail below with reference to the following figures and specific embodiments.
Embodiment 1, as shown in fig. 1, a method for diagnosing a failure of a main shaft of a wind turbine generator includes:
step 1, vibration signal acquisition:
acquiring a vibration signal through a sensor arranged on a main shaft seat of the wind turbine generator;
the sensor is an acceleration sensor, the sampling frequency is 10kHz, and the number of sampling points is 1000.
Step 2, VMD variational modal decomposition:
performing VMD (variable mode decomposition) on the vibration signal to obtain a plurality of modal components under different frequencies;
wherein, the VMD is variation modal decomposition;
the method specifically comprises the following steps:
step 201, decomposing the vibration signal into K single-component mode functions, where the single-component mode functions are limited band intrinsic mode functions:
Figure BDA0001315521120000051
wherein f (t) is a vibration signal, uk(t) is the kth band-limited intrinsic mode function, Ak(t) is the envelope curve and,
Figure BDA0001315521120000052
is a phase function, t is time;
step 202, estimating the frequency bandwidth of each single-component mode function, and establishing a variation constraint model:
Figure BDA0001315521120000053
wherein, wkFor the kth single-component mode function uk(t) center angular frequency, { u }k}＝{u1……uK}，{wk}＝{w1……wK}，
Figure BDA0001315521120000054
δ (t) is the unit pulse signal for the summation of all single-component mode functions;
step 203, introducing the secondary penalty factor a and the Lagrangian operator r into a variation constraint model to obtain an augmented Lagrangian equation;
Figure BDA0001315521120000055
wherein a is a secondary penalty factor, the secondary penalty factor a is set as a default value of 2000, and r is a Lagrange operator;
step 204, initialize uk、wkAnd r value, continuously iteratively updating by adopting an alternative direction multiplier method, searching saddle points of the augmented Lagrange in an iterative optimization sequence, and determining the judgment precision e1According to the decision formula, stopping iteration to output K modal components for the condition of meeting the decision formula,
the judgment formula is specifically as follows:
Figure BDA0001315521120000056
wherein, get e1＝1×10-6；
Before step 201, step 2 further comprises:
step a, performing VMD decomposition according to sensor data with known faults, taking acceleration sensor data in an outer ring fault state as an example, and observing the change situation of the central frequency of a component modal function under different K values, as shown in the following table:
Figure BDA0001315521120000061
as can be seen from the above table, starting from the 4 th modal component, the center frequency tends to be a fixed value, indicating that an over-decomposition phenomenon occurs after the sample, therefore, the outer ring fault sample K value is 4, and similarly, the remaining state sample K values can be obtained, as shown in the following table,
Figure BDA0001315521120000062
and 3, sensitive factor evaluation:
calculating a sensitivity factor of each modal component, and evaluating the fault characteristic degree of each modal component;
the modal components obtained by VMD analysis include both effective components including fault characteristics and interference components such as noise. Evaluating the mode component by using the sensitive factor, and extracting an effective component;
calculating each of the modes obtained in step 2Sensitivity factor lambda of the state componentk，
Figure BDA0001315521120000063
Wherein, deltak＝βk－αkWherein αkCorrelation coefficient of modal component and fault signal, βkThe vibration signal is a vibration signal of the main shaft in the states of outer ring fault, inner ring fault and rolling body fault, and the normal signal is a vibration signal of the main shaft in the normal state;
the calculation of the modal component susceptibility factor amplitudes is shown in the table below,
Figure BDA0001315521120000064
and 4, signal reconstruction:
noise interference is eliminated, signals are reconstructed, the characteristic vector of each reconstructed signal is calculated, and a modal analysis vector is constructed;
the method specifically comprises the following steps:
step 401, utilizing a sorting method to process sensitivity factor lambdakSequencing from big to small to obtain a new sensitive factor sequence lambda'k；
The sorting method is a bubble method, a cocktail sorting method, a radix sorting method or a hill sorting method, in this embodiment, a bubble method is adopted to perform descending sorting on the sensitive factor sequences, and a sensitive factor difference spectrum is obtained, as shown in fig. 2, the difference spectrum in the diagram has a peak value at the second sensitive factor, so modal components corresponding to two sensitive factors before descending sorting are selected to be superposed, and a signal is reconstructed;
step 402, according to a sensitive factor sequence lambda'kCalculating a sequence of difference spectrum evaluation components dk，
dk＝λ′k+1-λ′k；
Step 403, find out the sequence d of the estimated component of the difference spectrumkMaximum value of dj；
Step (ii) of404, by the first j dkCorresponding sensitivity factor lambda'kFinding out corresponding j modal components, and superposing the j modal components to obtain a reconstructed signal;
step 405, obtaining a feature vector by using a singular value decomposition method according to the reconstructed signal;
the reconstructed signal is divided into 4 rows, and the feature vectors are calculated to form modal analysis vectors, as shown in fig. 3, the modal analysis vector trends of the signal samples in different principal axis states show that the feature vectors in different principal axis states reflect different feature distributions.
Step 5, VPMCD fault diagnosis:
carrying out fault identification on the modal analysis vector by using a prediction model, calculating the error square sum of the main shaft state VPM, and carrying out classification identification by taking the minimum value as judgment;
the VPMCD is mode identification of a variable prediction model, and the main shaft state comprises a normal state, an outer ring fault, an inner ring fault and a rolling body fault;
the method specifically comprises the following steps:
step 501, substituting the characteristic vector into a variable prediction model, wherein the variable prediction model is VPMi：
Xi＝f(Xj,b0,bj,bjj,bjk)+e2，
Wherein, b0,bj,bjj,bjkIs a model parameter; e.g. of the type2Is a prediction error;
the model parameters are obtained by utilizing the vibration data of various faults through training, and the prediction error is preset. The prediction model is a linear model or a linear interaction model or a secondary interaction model;
the linear model is:
Figure BDA0001315521120000071
the linear interaction model is:
Figure BDA0001315521120000081
the secondary model is:
Figure BDA0001315521120000082
the secondary interaction model is as follows:
Figure BDA0001315521120000083
the linear interaction model adopted in the embodiment is used for constructing a VPMCD recognition model, modal analysis vectors of training samples of all states are substituted into the linear interaction model, and b is estimated0，bj，bjj，bjkObtaining the sub-model corresponding to each feature vector by 7 parameter values in total
Figure BDA0001315521120000085
Constructing a VPMCD recognition model of each state, wherein i represents a state type, i is 1 and represents a VPM model of a normal state, 2 represents an outer ring fault, 3 represents an inner ring fault, and 4 represents a rolling body fault; k represents a feature value, k-1 represents the VPM model of the first feature value of a state, and so on.
Step 502, calculating the sum of squares of errors of VPMs in all states, and performing classification and identification by taking the minimum value as judgment;
modal analysis vectors for 20 sets of test samples for each state were calculated and identified using the VPMCD identification model, with the results shown in the following table:
Figure BDA0001315521120000084
Figure BDA0001315521120000091
the result shows that in the test samples, except 4 groups of outer ring fault samples, the rest samples are correctly identified, and the accumulated identification rate reaches 98.75%.
The invention has the following beneficial effects: VMD analysis is carried out on the fan vibration signal, the signal is decomposed into a group of modal components with different frequencies, and because the fault characteristics only appear in a specific frequency band, the modal components are evaluated and analyzed by adopting a sensitive factor, components containing the fault characteristics are screened, noise interference components are eliminated, the signal is reconstructed, and the fault information is highlighted; and then, extracting the characteristic value of the reconstructed signal to form a characteristic identification vector, and performing fault diagnosis by establishing an intrinsic relation of the characteristic value by combining a VPMCD method, so that the processes of subjective parameter selection and optimization of the traditional classifier are avoided, and the fault diagnosis time is shortened.
Various modifications may be made to the above without departing from the spirit and scope of the invention as defined by the appended claims. The scope of the invention is therefore intended to be limited not by the above description, but rather by the scope of the appended claims.

Claims (9)

1. A method for diagnosing faults of a main shaft of a wind turbine generator is characterized by comprising the following steps:
step 1, vibration signal acquisition:
acquiring a vibration signal through a sensor arranged on a main shaft seat of the wind turbine generator;
step 2, VMD variational modal decomposition:
performing VMD (variable mode decomposition) on the vibration signal to obtain a plurality of modal components under different frequencies;
and 3, sensitive factor evaluation:
calculating the sensitivity factor lambda of each modal component obtained in step 2k，
Figure FDA0002406926830000011
Wherein, deltak＝βk－αkWherein αkCorrelation coefficient of modal component and fault signal, βkEvaluating the degree of fault characteristics contained in each modal component for the correlation coefficient of the modal component and the normal signal;
and 4, signal reconstruction:
noise interference is eliminated, signals are reconstructed, the characteristic vector of each reconstructed signal is calculated, and a modal analysis vector is constructed;
step 5, VPMCD fault diagnosis:
carrying out fault identification on the modal analysis vector by using a prediction model, calculating the error square sum of the main shaft state VPM, and carrying out classification identification by taking the minimum value as judgment;
the method comprises the steps of VMD (variable mode decomposition), VPMCD (virtual machine model) and main shaft state prediction, wherein the VMD is variable mode decomposition, the VPMCD is mode recognition of a variable prediction model, and the main shaft state comprises a normal state, an outer ring fault, an inner ring fault and a rolling body fault.
2. The diagnostic method of claim 1, wherein the sensor in step 1 is an acceleration sensor.
3. The diagnostic method according to claim 1, wherein step 2 specifically comprises:
step 201, decomposing the vibration signal into K single-component mode functions, where the single-component mode functions are limited band intrinsic mode functions:
Figure FDA0002406926830000012
wherein f (t) is a vibration signal, uk(t) is the kth band-limited intrinsic mode function, Ak(t) is the envelope curve and,
Figure FDA0002406926830000013
is a phase function, t is time;
step 202, estimating the frequency bandwidth of each single-component mode function, and establishing a variation constraint model:
Figure FDA0002406926830000014
wherein, wkFor the kth single-component mode function uk(t) center angular frequency, { u }k}＝{u1……uK}，{wk}＝{w1……wK}，
Figure FDA0002406926830000021
δ (t) is the unit pulse signal for the summation of all single-component mode functions;
step 203, introducing the secondary penalty factor a and the Lagrangian operator r into a variation constraint model to obtain an augmented Lagrangian equation;
Figure FDA0002406926830000022
wherein a is a secondary penalty factor, and r is a Lagrange operator;
step 204, initialize uk、wkAnd r value, continuously iteratively updating by adopting an alternative direction multiplier method, searching saddle points of the augmented Lagrange in an iterative optimization sequence, and determining the judgment precision e1And according to the judgment formula, stopping iterative output of the K modal components for the condition of meeting the judgment formula.
4. The diagnostic method according to claim 3, characterized in that said decision formula is in particular:
Figure FDA0002406926830000023
wherein, get e1＝1×10-6。
5. The diagnostic method according to claim 1, wherein step 4 specifically comprises:
step 401, utilizing a sorting method to sort the sensitivity factor λkSequencing from big to small to obtain a new sensitive factor sequence lambda'k；
Step 402, according to a sensitive factor sequence lambda'kCalculating a sequence of difference spectrum evaluation components dk，
dk＝λ′k+1-λ′k；
Step 403, finding out the difference spectrum evaluation component sequence dkMaximum value of dj；
Step 404, pass the first j dkCorresponding sensitivity factor lambda'kFinding out corresponding j modal components, and superposing the j modal components to obtain a reconstructed signal;
and 405, obtaining a feature vector by using a singular value decomposition method according to the reconstructed signal.
6. The diagnostic method of claim 5, wherein the ranking method in step 401 is a bubble or cocktail ranking or a cardinal or hill ranking.
7. The diagnostic method according to claim 1, wherein step 5 specifically comprises:
step 501, substituting the characteristic vector into a variable prediction model, wherein the variable prediction model is VPMi：
Xi＝f(Xj，b0，bj，bjj，bjk)+e2，
Wherein, b0,bj,bjj,bjkIs a model parameter; e.g. of the type2Is a prediction error;
step 502, calculating the sum of the squares of the errors of the states VPM, and performing classification and identification with the minimum value as a judgment.
8. The diagnostic method of claim 7, wherein the model parameters are trained using fault-like vibration data, and the prediction error is predetermined.
9. The diagnostic method of claim 7, wherein the predictive model is a linear model or a linear interaction model or a quadratic interaction model;
the linear model is:
Figure FDA0002406926830000031
the linear interaction model is as follows:
Figure FDA0002406926830000032
the secondary model is as follows:
Figure FDA0002406926830000033
the secondary interaction model is as follows:
Figure FDA0002406926830000034
CN201710423126.3A 2017-06-07 2017-06-07 Method for diagnosing main shaft fault of wind turbine generator Active CN109000921B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201710423126.3A CN109000921B (en) 2017-06-07 2017-06-07 Method for diagnosing main shaft fault of wind turbine generator

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201710423126.3A CN109000921B (en) 2017-06-07 2017-06-07 Method for diagnosing main shaft fault of wind turbine generator

Publications (2)

Publication Number Publication Date
CN109000921A CN109000921A (en) 2018-12-14
CN109000921B true CN109000921B (en) 2020-05-08

Family

ID=64573004

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201710423126.3A Active CN109000921B (en) 2017-06-07 2017-06-07 Method for diagnosing main shaft fault of wind turbine generator

Country Status (1)

Country Link
CN (1) CN109000921B (en)

Families Citing this family (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109489977B (en) * 2018-12-28 2021-03-05 西安工程大学 KNN-AdaBoost-based bearing fault diagnosis method
CN110186682B (en) * 2019-07-08 2021-03-23 石家庄铁道大学 Rolling bearing fault diagnosis method based on fractional order variation modal decomposition
CN111413089A (en) * 2020-04-08 2020-07-14 北华大学 Gear fault diagnosis method based on combination of VMD entropy method and VPMCD
CN111985315A (en) * 2020-07-10 2020-11-24 合肥工业大学 Bearing fault signal intrinsic mode function decomposition and extraction method and device
CN112326245B (en) * 2020-10-21 2023-03-10 中国航空工业集团公司上海航空测控技术研究所 Rolling bearing fault diagnosis method based on variational Hilbert-Huang transform
CN116625688A (en) * 2023-05-24 2023-08-22 石家庄铁道大学 Rolling bearing health monitoring method based on multilayer noise reduction and self-encoder
CN116738372B (en) * 2023-08-15 2023-10-27 昆仑数智科技有限责任公司 Rolling bearing fault diagnosis method, device and equipment for refining centrifugal pump
CN117030268B (en) * 2023-10-07 2024-01-23 太原科技大学 Rolling bearing fault diagnosis method

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DK2810027T3 (en) * 2012-01-30 2017-08-28 Spm Instr Ab DEVICE AND PROCEDURE FOR ANALYZING THE STATE OF A MACHINE WITH A ROTATING PART
CN106017926A (en) * 2016-05-13 2016-10-12 山东理工大学 Rolling bearing fault diagnosis method based on variational mode decomposition
CN105758644A (en) * 2016-05-16 2016-07-13 上海电力学院 Rolling bearing fault diagnosis method based on variation mode decomposition and permutation entropy
CN106198015B (en) * 2016-06-29 2018-05-25 潍坊学院 A kind of VMD of rolling bearing, spectrum kurtosis and smooth iteration envelope Analysis Method

Also Published As

Publication number Publication date
CN109000921A (en) 2018-12-14

Similar Documents

Publication Publication Date Title
CN109000921B (en) Method for diagnosing main shaft fault of wind turbine generator
Jiang et al. An improved VMD with empirical mode decomposition and its application in incipient fault detection of rolling bearing
Huo et al. Incipient fault diagnosis of roller bearing using optimized wavelet transform based multi-speed vibration signatures
CN108760327B (en) Diagnosis method for rotor fault of aircraft engine
CN112200015B (en) Rolling bearing fault diagnosis method based on improved VMD
CN109827777B (en) Rolling bearing fault prediction method based on partial least square method extreme learning machine
CN111238843B (en) Fan health evaluation method based on rapid spectrum kurtosis analysis
CN110443117B (en) Wind turbine generator fault diagnosis method
CN109655266B (en) Wind turbine generator bearing fault diagnosis method based on AVMD and spectrum correlation analysis
Wang et al. Fault diagnosis of wind turbine bearing based on optimized adaptive chirp mode decomposition
Meng et al. Health indicator of bearing constructed by rms-CUMSUM and GRRMD-CUMSUM with multifeatures of envelope spectrum
CN113569990B (en) Strong noise interference environment-oriented performance equipment fault diagnosis model construction method
CN111665050B (en) Rolling bearing fault diagnosis method based on clustering K-SVD algorithm
CN111076934A (en) Method for diagnosing potential fault of bearing based on S transformation
CN112067298A (en) Rolling bearing fault diagnosis method based on hierarchical global fuzzy entropy
Jianhui et al. New fault diagnosis method for rolling bearing based on PCA
CN112183260A (en) One-way valve fault diagnosis method based on total variation noise reduction and RQA
CN114894479A (en) Fan main shaft bearing fault diagnosis method based on VMD parameter optimization
CN114897016A (en) Fan bearing fault intelligent diagnosis method based on multi-source frequency spectrum characteristics
Geng et al. Fault identification of rolling bearing with variable speed based on generalized broadband mode decomposition and DET
CN114611551A (en) Electromechanical fault classification method based on wavelet packet energy spectrum entropy
CN114818815A (en) Method and system for acquiring timed arrival time of blade tip for blade vibration measurement
CN114136604A (en) Rotary equipment fault diagnosis method and system based on improved sparse dictionary
CN115326396A (en) Bearing fault diagnosis method and device
CN113281047A (en) Bearing inner and outer ring fault quantitative trend diagnosis method based on variable-scale Lempel-Ziv

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