CN111476339B - Rolling bearing fault feature extraction method, intelligent diagnosis method and system - Google Patents
Rolling bearing fault feature extraction method, intelligent diagnosis method and system Download PDFInfo
- Publication number
- CN111476339B CN111476339B CN202010313588.1A CN202010313588A CN111476339B CN 111476339 B CN111476339 B CN 111476339B CN 202010313588 A CN202010313588 A CN 202010313588A CN 111476339 B CN111476339 B CN 111476339B
- Authority
- CN
- China
- Prior art keywords
- training set
- signal data
- feature
- signal
- training
- 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
Links
- 238000000034 method Methods 0.000 title claims abstract description 62
- 238000000605 extraction Methods 0.000 title claims abstract description 34
- 238000005096 rolling process Methods 0.000 title claims abstract description 28
- 238000003745 diagnosis Methods 0.000 title claims abstract description 14
- 238000012549 training Methods 0.000 claims abstract description 93
- 238000012360 testing method Methods 0.000 claims abstract description 92
- 230000009467 reduction Effects 0.000 claims abstract description 29
- 238000001228 spectrum Methods 0.000 claims abstract description 26
- 238000007781 pre-processing Methods 0.000 claims abstract description 15
- 238000013144 data compression Methods 0.000 claims abstract description 9
- 238000000354 decomposition reaction Methods 0.000 claims description 33
- 239000002245 particle Substances 0.000 claims description 28
- 239000011159 matrix material Substances 0.000 claims description 27
- 238000012545 processing Methods 0.000 claims description 14
- 239000013598 vector Substances 0.000 claims description 10
- 108010076504 Protein Sorting Signals Proteins 0.000 claims description 7
- 238000007906 compression Methods 0.000 claims description 4
- 230000006835 compression Effects 0.000 claims description 4
- 230000008569 process Effects 0.000 claims description 4
- 238000010606 normalization Methods 0.000 claims description 3
- 238000003860 storage Methods 0.000 claims description 3
- 238000007405 data analysis Methods 0.000 abstract 1
- 238000012706 support-vector machine Methods 0.000 description 20
- 238000005457 optimization Methods 0.000 description 11
- 230000006870 function Effects 0.000 description 10
- 230000001133 acceleration Effects 0.000 description 2
- 230000005484 gravity Effects 0.000 description 2
- 208000032544 Cicatrix Diseases 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 230000009977 dual effect Effects 0.000 description 1
- 238000013100 final test Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 231100000241 scar Toxicity 0.000 description 1
- 230000037387 scars Effects 0.000 description 1
- 230000003595 spectral effect Effects 0.000 description 1
- 238000004611 spectroscopical analysis Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 238000013519 translation Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M13/00—Testing of machine parts
- G01M13/04—Bearings
- G01M13/045—Acoustic or vibration analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
- G06N20/10—Machine learning using kernel methods, e.g. support vector machines [SVM]
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Mathematical Physics (AREA)
- Computing Systems (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- Acoustics & Sound (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
Abstract
The invention provides a rolling bearing fault feature extraction method, an intelligent diagnosis method and a system, comprising the following steps: collecting bearing signal data of an electric driving end, preprocessing the signal data, and dividing the signal data into a test set and a training set; extracting characteristics of signal data by adopting a wavelet band energy method to respectively obtain first characteristic matrixes of a training set and a testing set; extracting characteristics of signal data by adopting a wavelet packet-AR spectrum estimation method to respectively obtain second characteristic matrixes of a training set and a testing set; extracting features of the signal data by adopting an EMD-SVD method to respectively obtain third feature matrixes of a training set and a testing set; the first, second and third feature matrixes are spliced, data compression is carried out through PCA to obtain a training set and a testing set after dimension reduction, the training set and the testing set are input into a classifier to carry out data analysis, faults are diagnosed, completeness and sparsity of signal features can be effectively guaranteed, generalization accuracy of feature extraction is improved, and generalization capability of the classifier is improved.
Description
Technical Field
The invention belongs to the field of mechanical fault diagnosis, and particularly relates to a rolling bearing fault feature extraction method, an intelligent diagnosis method and a system which are combined with various feature processing methods and fault recognition means.
Background
Mechanical fault diagnosis is a method for automatically monitoring and maintaining equipment based on measurable signal characteristics in the shortest expected period of fault occurrence. Rolling bearings have been attracting attention in industry as the most important and most easily damaged components in rotary machines.
Existing bearing fault diagnosis techniques typically include three steps: digital signals are collected, processed and classified by a classifier. Where the acquisition of digital signals is mainly achieved by acceleration of the sensor acquisition, while the classifier usually employs a machine supervised learning algorithm for fault identification, which, although such algorithms are quite sophisticated today, is highly dependent on the extracted signal features. The more complete and representative the extracted features are, the stronger the fault recognition capability is, and the traditional method for extracting the single features cannot obtain the comprehensive and effective features, so that the classification of the classifier is difficult to express good generalization capability.
Disclosure of Invention
Aiming at the defects existing in the prior art, the invention adopts various methods to extract the characteristics from different angles and performs dimension reduction treatment by a certain means, thereby ensuring the completeness and sparsity of the extracted characteristics, and in the classification and identification stage of the classifier, aiming at the parameter optimization problem of SVM (support vector machine), the invention proposes to adopt PSO (particle swarm optimization) to accelerate the parameter optimization speed.
In a first aspect, the present invention provides a method for extracting fault characteristics of a rolling bearing, including the steps of: collecting bearing signal data of an electric driving end, preprocessing the signal data, and dividing the signal data into a test set and a training set;
extracting characteristics of signal data by adopting a wavelet band energy method to respectively obtain first characteristic matrixes of a training set and a testing set;
extracting characteristics of signal data by adopting a wavelet packet-AR spectrum estimation method to respectively obtain second characteristic matrixes of a training set and a testing set;
extracting features of the signal data by adopting an EMD-SVD method to respectively obtain third feature matrixes of a training set and a testing set;
and splicing the first, second and third feature matrixes, and performing data compression through PCA to obtain a training set and a testing set after dimension reduction, namely bearing fault characteristics.
In a second aspect, the present invention also provides an intelligent diagnosis method for a rolling bearing fault, including obtaining a training set and a test set after dimension reduction by using the rolling bearing fault feature extraction method according to the first aspect, inputting the training set and the test set after dimension reduction into a classifier for training, and diagnosing the bearing fault;
the specific steps of inputting the training set and the testing set after the dimension reduction into the classifier for training comprise the following steps: initializing a particle swarm, and initializing the particle speed and the position;
calculating the fitness of particles, and calculating the SVM recognition accuracy under the current penalty term coefficient and the width of the kernel function;
searching an extremum; updating particle velocity and position; judging whether the SVM classification error meets a termination condition; if the termination condition is met, putting the test set into a classifier for classification, and analyzing classification results of the training set and the test set; if not, returning to the position where the extremum is found to continue processing.
In a third aspect, the present invention also provides a rolling bearing fault feature extraction system, including:
and the acquisition module is used for: is configured to collect bearing signal data for the electric drive end;
and a pretreatment module: is configured to pre-process the signal data, and divide the signal data into a test set and a training set;
and the feature extraction module is used for: the method comprises the steps of extracting characteristics of signal data by adopting a wavelet band energy method to respectively obtain first characteristic matrixes of a training set and a testing set; extracting characteristics of signal data by adopting a wavelet packet-AR spectrum estimation method to respectively obtain second characteristic matrixes of a training set and a testing set; extracting features of the signal data by adopting an EMD-SVD method to respectively obtain third feature matrixes of a training set and a testing set; and splicing the first, second and third feature matrixes, and performing data compression through PCA to obtain a training set and a testing set after dimension reduction.
In a fourth aspect, the present invention also provides a computer readable storage medium storing computer instructions which, when executed by a processor, perform the rolling bearing fault feature extraction method according to the first aspect.
In a fifth aspect, the invention also provides an intelligent diagnosis system for faults of the rolling bearing, which comprises a classifier, an acquisition module, a preprocessing module and a feature extraction module, wherein the acquisition module, the preprocessing module and the feature extraction module are described in the third aspect; the classifier is configured to: inputting a training set and a testing set subjected to dimension reduction by the feature extraction module; initializing a particle swarm, and initializing the punishment item coefficient of the particle, the speed and the position of the width of the kernel function; under the current punishment item coefficient and the width of the kernel function, calculating the SVM identification accuracy; searching an extremum; updating particle velocity and position; judging whether the SVM classification error meets a termination condition; if the termination condition is met, putting the test set into a classifier for classification, and analyzing classification results of the training set and the test set; if not, returning to the position where the extremum is found to continue processing.
Compared with the prior art, the invention has the following beneficial effects:
1. the invention adopts various methods to extract the characteristics of the signals, adopts PCA (singular value decomposition) to carry out dimension reduction treatment, and can effectively ensure the completeness and sparsity of the characteristics of the signals and improve the generalization accuracy of the characteristic extraction compared with the single method for carrying out the characteristic extraction.
2. Aiming at the parameter optimization problem of the SVM, the optimization speed of two important parameters C and g of the SVM is realized by using a PSO algorithm, and compared with a conventional SVM classifier, the method can effectively avoid the problems of over-fitting and under-fitting and improve the generalization capability of the classifier.
3. The invention can extract accurate and representative characteristics from a large amount of data through preprocessing, adopts various methods to extract the characteristics from different angles, and carries out dimension reduction processing through a certain means, thereby ensuring the completeness and sparsity of the extracted characteristics, and in the classification and identification stage of the classifier, aiming at the parameter optimization problem of the SVM (support vector machine), proposes to adopt PSO (particle swarm optimization) to accelerate the parameter optimization speed.
Drawings
The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this application, illustrate embodiments of the application and together with the description serve to explain the application and do not constitute an undue limitation to the application.
FIG. 1 is a schematic flow chart of bearing vibration signal processing according to the present invention;
FIG. 2 is a wavelet decomposition tree of the present invention;
FIG. 3 is a wavelet packet decomposition tree of the present invention;
fig. 4 is a flowchart of the SVM-PSO algorithm of the present invention.
The specific embodiment is as follows:
the invention will be further described with reference to the drawings and examples.
It should be noted that the following detailed description is exemplary and is intended to provide further explanation of the present application. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.
It is noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of example embodiments in accordance with the present application. As used herein, the singular is also intended to include the plural unless the context clearly indicates otherwise, and furthermore, it is to be understood that the terms "comprises" and/or "comprising" when used in this specification are taken to specify the presence of stated features, steps, operations, devices, components, and/or combinations thereof.
In the present invention, terms such as "upper", "lower", "left", "right", "front", "rear", "vertical", "horizontal", "side", "bottom", etc. refer to an orientation or a positional relationship based on that shown in the drawings, and are merely relational terms, which are used for convenience in describing structural relationships of various components or elements of the present invention, and do not denote any one of the components or elements of the present invention, and are not to be construed as limiting the present invention.
In the present invention, terms such as "fixedly attached," "connected," "coupled," and the like are to be construed broadly and refer to either a fixed connection or an integral or removable connection; can be directly connected or indirectly connected through an intermediate medium. The specific meaning of the terms in the present invention can be determined according to circumstances by a person skilled in the relevant art or the art, and is not to be construed as limiting the present invention.
Example 1
In order to extract accurate and representative features from a large amount of data, the invention adopts various methods to extract the features from different angles and performs dimension reduction processing by a certain means, thereby ensuring the completeness and sparsity of the extracted features, and in the classification and identification stage of the classifier, aiming at the parameter optimization problem of an SVM (support vector machine), the invention proposes to adopt a PSO (particle swarm optimization) to accelerate the parameter optimization speed.
In order to achieve the above object, the solution of the present invention is as follows:
the method for extracting the fault characteristics of the rolling bearing comprises the following steps: collecting bearing signal data of an electric driving end, preprocessing the signal data, and dividing the signal data into a test set and a training set;
extracting characteristics of signal data by adopting a wavelet band energy method to respectively obtain first characteristic matrixes of a training set and a testing set;
extracting characteristics of signal data by adopting a wavelet packet-AR spectrum estimation method to respectively obtain second characteristic matrixes of a training set and a testing set;
extracting features of the signal data by adopting an EMD-SVD method to respectively obtain third feature matrixes of a training set and a testing set;
and splicing the first, second and third feature matrixes, and performing data compression through PCA to obtain a training set and a testing set after dimension reduction, namely bearing fault characteristics.
Further, the step of preprocessing the signal includes: the signal sequence to be used in the bearing data signal is F 1 ~F i F is to F i Suitably split to form l signal sequences, F i ={f i1 ,f i2 ,f i3 …f il Of f, where f i1 A first subset sequence segmented for an ith failure sequence; random slave F i Select a part of the messagesThe number subsequence is used as an original training set; the other part is used as the original test set.
Further, the specific step of extracting the characteristics of the signal data by using the wavelet band energy method comprises the following steps:
carrying out wavelet decomposition on the bearing signal data;
decomposing different components of the bearing signal data into corresponding frequency bands, and calculating the energy characteristics of each frequency band;
and carrying out normalization processing on the energy feature vectors to respectively obtain first feature matrixes of the training set and the testing set.
Further, the specific step of extracting the characteristics of the signal data by adopting the wavelet packet-AR spectrum estimation method comprises the following steps: carrying out wavelet packet decomposition on the bearing signals;
decomposing different components of the signal into corresponding frequency bands, and carrying out wavelet packet reconstruction on each component to obtain a reconstructed signal;
AR spectrum estimation is carried out on the reconstructed signals, and AR power spectrum densities of all frequency bands are calculated;
and calculating the AR spectrum energy characteristics of each frequency band to respectively obtain second characteristic matrixes of the training set and the testing set.
Further, the specific steps of extracting the characteristics of the signal data by adopting the EMD-SVD method comprise the following steps:
EMD is carried out on the signals to obtain a plurality of IMF components;
calculating the energy duty ratio of the IMF component;
setting a threshold value, and extracting a plurality of IMF components with energy duty ratio exceeding the threshold value;
and (3) carrying out compression processing on the extracted IMF components by adopting SVD, and respectively obtaining a training set and a third feature matrix of a test set by taking the compressed IMF components as signal features.
Further, the specific steps of the first, second and third feature matrix splicing process include: obtaining a training set alpha and a test set beta after the splicing treatment, wherein alpha= [ alpha ] 1 ,α 2 ,α 3 ] A1×M ,β=[β 1 ,β 2 ,β 3 ] A2×M Tables A1 and A2 respectivelyShowing the number of signals of the training set and the test set, M representing the feature dimension alpha 1 、α 2 And alpha 3 First, second and third feature matrices, beta, respectively, of the training set 1 、β 2 And beta 3 The first, second and third feature matrices of the test set, respectively.
Further, the specific steps of data compression by PCA include: for a training data set alpha, calculating a covariance matrix of the training data set alpha; singular value decomposition is carried out on the covariance matrix; outputting the training set after dimension reduction; and the same treatment is carried out on the test set beta to obtain the test set after dimension reduction.
Example 2
The invention is based on the bearing signal data set of the driving end of the motor of the West university, and needs to point out that the bearing fault data signals are divided into three types of inner rings, outer rings and rolling bodies, wherein each type of bearing fault data signals also comprises artificially manufactured spark scars with diameters of 0.1778,0.3556 and 0.5334mm, and the normal bearing data signals are divided into ten cases in total, so that the signal sequence to be used is determined to be F 1 ~F i Where i=1, 2 … 10.
Fig. 1 is a flow chart of feature extraction and intelligent diagnosis of a bearing vibration signal according to the present invention, as shown in fig. 1, the feature extraction and intelligent diagnosis method of a bearing vibration signal according to the present invention specifically comprises the following steps:
(1) Data preprocessing
Will F i Suitably split to form 1 signal sequence, F i ={f i1 ,f i2 ,f i3 …f il Of f, where f i1 A first subset sequence partitioned for an ith failure sequence. Next randomly follow F i Selecting a part of signal subsequences as an original training set; the other part is used as the original test set. To represent a matrix in which a training set is conveniently a1×b1, and a test set is a2×b2, where A1, A2 are the number of signals and B1, B2 represent the feature dimension of the signals
(2) Multi-resolution analysis by wavelet transformation to extract energy characteristics of signals in different frequency bands
And carrying out wavelet decomposition on the bearing signals, so as to decompose different components of the signals into corresponding frequency bands, and finally calculating the energy characteristics of each frequency band. The method comprises the following steps:
(2-1) wavelet decomposition: selecting discrete wavelet functionsj epsilon Z (1) is a wavelet basis function, and the signal is decomposed, and the specific mathematical expression is as follows: />
Wherein j is a decomposition scale, x is a decomposition position parameter, W 2j f (x) is the wavelet decomposition coefficient of the signal f at the decomposition scale 2j at the position x.
And (3) carrying out M1-layer decomposition on the signals based on the mathematical expression of the wavelet decomposition of the formula (2), and extracting the signal characteristics contained in each leaf node. Assuming that the original signal is f, (M, n) represents an nth node of an mth layer of the wavelet decomposition tree (as shown in fig. 2), where m=0, 1,2,3 … M1, and n=0, 1; each node represents a certain frequency band characteristic, (0, 0) represents an original signal, (1, 0) and (1, 1) represent low-frequency coefficients S of wavelet decomposition respectively 10 And high frequency coefficient S 11 And so on (m, n) represents the nth node coefficient S of the mth layer mn 。
(2-2) calculating band energy: calculating the energy E of each leaf node of the signal f mn The calculation formula is as follows:
wherein, the value of n is constrained by M, and M is more than 0 and less than or equal to M1; when M < M1, n=1; when m=m1, n=0 or n=1 (in the practical sense, M1-th layer signal decomposition results in a low-frequency leaf node and a high-frequency leaf node). At this time k represents S mn Feature dimension of (a)
(2-3) feature vector normalization: feature vectorNormalizing, energy specific gravity of each frequency band is taken as +.>Is a characteristic component of->The following is shown:
wherein the range of values of m and n is the same as that of formula (3).
(2-4) obtaining a feature matrix: extracting features of each signal according to the steps to obtain the alpha of the first A1 x M1 1 Beta of training set and A2 x M1 1 Test set.
(3) Extracting AR spectrum energy characteristics of signals in each frequency band by wavelet packet-AR spectrum estimation method
And carrying out wavelet packet decomposition on the bearing signals so as to decompose different components of the signals into corresponding frequency bands, reconstructing each component, and finally calculating AR spectrum energy characteristics of each frequency band through AR spectrum estimation. The method comprises the following steps:
(3-1) wavelet packet decomposition: selecting a proper wavelet basis function, and determining the decomposition layer number of signals, wherein the specific mathematical expression is as follows:
wherein k is a translation time factor; h represents a low-pass filter coefficient; g represents a high pass filter coefficient;an mth decomposition sequence obtained by decomposing the signal f (x) through the j-layer wavelet packet; />Is the original digital signal f (x).
And (3) performing M2-layer decomposition on the signals based on the mathematical expression of the (5) wavelet packet decomposition, and extracting the signal characteristics contained in the last layer of leaf nodes. Assuming that the original signal is f, (M, n) represents an nth node of an mth layer of a wavelet packet decomposition tree (shown in fig. 3), where m=0, 1,2,3 … M2, n=0, 1,..2 M2 -1; each node represents a certain frequency band characteristic, (0, 0) represents an original signal, (1, 0) and (1, 1) represent a low-frequency coefficient S10 and a high-frequency coefficient S11 of wavelet decomposition respectively, and the like (m, n) represents an nth node coefficient S representing an mth layer mn 。
(3-2) wavelet packet reconstruction: and carrying out signal reconstruction according to the coefficient sequence obtained in the M2 layer. RS (Reed-Solomon) mn Representing the coefficient S mn Is used to reconstruct the signal. Wavelet reconstruction mathematics are expressed as follows:
wherein:and->Dual operators h and g, respectively, +.>The m-th reconstructed sequence formed for the j-1 layer reconstruction.
(3-3) AR Spectrometry: for RS mn AR spectrum estimation is carried out on the reconstructed signals to obtain AR spectrum power density P of each frequency band mn The method is characterized by comprising the following steps:
assuming that the reconstructed signal sequence is x (n), the autoregressive model of the sequence can be expressed as:
where w (x) is zero mean and varianceIs a normal distribution white noise of (2); q is the order of the model. Through relevant signal processing knowledge, AR model parameters aj (j=1, 2 … N) and +.>Then, the power spectral density of the signal x (n) is calculated from the self-transfer function as:
(3-4) solving AR spectrum energy of each frequency band signal: let P be mn Corresponding energy Ep mn The following steps are:
wherein k is a feature dimension; m=m2; n=0, 1,2 … 2 M2 -1。
(3-5) obtaining feature vectorsEnergy of each band as +.>Is a characteristic component of->The following is shown:
(3-6) obtaining a feature matrix: extracting features of each signal according to the above steps to obtain a signal with a size of A1 x (2 M2 Alpha of-1) 2 Training set and A2 (2 M2 Beta of 1) 2 Test set.
(4) Signal characteristic extraction method based on EMD-SVD
EMD is carried out on the signals to obtain a plurality of IMF components; calculating the energy duty ratio of the IMF component; extracting a plurality of IMF components with larger energy specific gravity; and (3) compressing the extracted IMF component by SVD, and taking the compressed IMF component as a signal characteristic.
The method comprises the following steps:
(4-1) extracting the IMF components which are arranged in front and have energy sum of more than 98% by using EMD decomposition, specifically as follows:
find all local extremum points of the original signal f (t)
Fitting all the found local extreme points by using an interpolation algorithm to obtain an upper envelope line f max (t) and lower envelope f min (t)
The mean value a is obtained from the upper envelope curve and the lower envelope curve 1
Further pass through h 1 =f(t)-a 1 Judging h 1 Whether two conditions of IMF are satisfied, if so, h 1 The first order IMF component; if not, in h 1 Repeating the steps a) to d) as new f (t) to obtain h 11 =h 1 -a 11 Wherein a is 11 The mean value of the upper envelope curve and the lower envelope curve of h 1; if h 11 If not, repeating the above steps until h 1k Meets the condition and is marked as C 1 =h 1k For the first order IMF component
After separating out the first order IMF component, r is obtained 1 =f(t)-C 1 Will r 1 As new f (t), repeating steps a) to d), and calculating for multiple times to obtain nIMF component and residual r of order n And finally, EMD decomposition is realized, wherein the formula is as follows:
calculating the energy Eci of the different IMF components Ci, where i=1, 2,3..n, then finding the IMF component S that occupies a larger specific energy weight and has an energy sum over 98% m 。
(4-2) obtaining the IMF feature matrix U m×τ : u as each IMF component m×τ A characteristic component, U m×τ The following is shown:
U m×τ m in the matrix represents the number of vectors and τ represents the feature dimension.
(4-3) further decomposing the characteristic value by SVD to realize U m×τ The compression of the matrix, the decomposition formula is as follows:
q in the above m×m Is a left singular matrix which is a matrix of the singular,right singular matrix; b (B) m×τ Is a singular value matrix, and it is a diagonal matrix, as follows:
the first singular values with larger values are taken as the characteristics of the signals according to the conditions, and the extracted characteristic vectors are taken asAnd the dimension is M3, thus the following can be obtained: />
(4-4) obtaining a feature matrix: extracting features of each signal according to the steps to obtain alpha with the size of A1-M3 3 Beta of training set and A2 x M3 3 Test set.
(5) Dimensionality reduction of an obtained dataset Using PCA
(5-1) obtaining a training set α and a test set β, wherein α= [ α ] 1 ,α 2 ,α 3 ] A1×M ,β=[β 1 ,β 2 ,β 3 ] A2×M A1, A2 represent the number of signals of the training set and the test set, respectively, M represents the feature dimension, and m=m1+2 M2 -1+m3 further expansion of α and β is represented specifically as follows:
wherein Tri and Tei represent vectors of A1 x 1 and A2 x 1, respectively.
(5-2) compression of alpha and beta Using PCA
The method comprises the following steps:
for the training dataset α, its covariance matrix Q is calculated as follows:
singular value decomposition is performed on the covariance matrix: [ U, S, V ] =svd (Q), where U represents a left singular matrix, V represents a right singular matrix, and S represents a singular value matrix.
Searching a k value corresponding to the characteristic duty ratio and meeting the threshold value, and judging the conditions as follows:wherein ev i Representing the ith singular value of Q.
The first k columns of U are selected as projection matrices ptr= [ U1, U1, … uk ], where the kth column is the eigenvector corresponding to the kth singular value.
Outputting the training data set Trian after dimension reduction A1×k =Ptr×α,k<M
The same treatment is carried out on the Test set beta to obtain a final Test set Test A2×r =Pte×β,r<M
(6) Classifying fault signals by SVM method
(6-1) classifier training
As shown in the flowchart of the SVM-PSO algorithm of fig. 4, the following is explained in detail:
(a) Initializing a particle swarm, wherein the particle swarm scale is N; iteration number It; the position of the penalty term coefficient ci is Pci, and the speed is Vci; the position of the kernel width gi is Pgi, and the speed is Vgi; definition of SVM Classification error
(b) Calculating the fitness Fit i of the particles, namely calculating the accuracy of SVM classification under the current ci and gi;
(c) Comparing the Fit [ i ] with an individual optimum Pbest [ i ], if Fit [ i ] > Pbest [ i ], pbest [ i ] =fit [ i ]; comparing Fit [ i ] with individual optimal Gbest [ i ], if Fit [ i ] > Gbest [ i ], gbest [ i ] = Fit [ i ]
(d) Updates (Pci, vci) and (Pgi, vgi), i.e., correspond to updates ci and gi; then, the fitness Fit [ i ] of the particles is calculated, and the update formula is shown as (13) (14):
v im =w·v im (k)+c 1 r 1 ·(P im (k)-x id (k))+c 2 r 2 ·(P gm (k)-x im (k)) (13)
x im (k+1)=x im (k)+v im (k+1) (14)
in the above formula, w is called inertial weight, c1 and c2 represent acceleration constants, r1 and r2 are random numbers between (0, 1), and k is the iteration number.
(e) Judging whether the current error is small enough or the current iteration item is larger than It, exiting, and outputting c and g parameters and target parameter values of w and b in the target function (15); otherwise, returning to the second step.
Subject to:y i (W T x i +b)≥1-ξ i ,ξ i ≥0
(6-2) placing the test set into a classifier for classification
(6-3) analyzing the classification results of the training set Train and the Test set Test
The above is a specific flow of the present invention.
In other embodiments, there is also provided:
an intelligent diagnosis method for rolling bearing faults comprises the steps of obtaining a training set and a test set after dimension reduction by adopting the rolling bearing fault feature extraction method in the embodiment 1, inputting the training set and the test set after dimension reduction into a classifier for training, and diagnosing the bearing faults.
Further, the specific step of inputting the training set and the test set after the dimension reduction into the classifier for training comprises the following steps: initializing a particle swarm, and initializing the speed and the position of a penalty term coefficient ci and a kernel function width gi of the particle; calculating particle fitness (SVM recognition accuracy) under the current penalty term coefficient ci and the kernel function width gi; searching an extremum; updating particle velocity and position; judging whether the SVM classification error meets a termination condition; if the termination condition is met, putting the test set into a classifier for classification, and analyzing classification results of the training set and the test set; if not, returning to the position where the extremum is found to continue processing.
A rolling bearing fault signature extraction system comprising:
and the acquisition module is used for: is configured to collect bearing signal data for the electric drive end;
and a pretreatment module: is configured to pre-process the signal data, and divide the signal data into a test set and a training set;
and the feature extraction module is used for: the method comprises the steps of extracting characteristics of signal data by adopting a wavelet band energy method to respectively obtain first characteristic matrixes of a training set and a testing set; extracting characteristics of the signal data by adopting a wavelet packet-AR spectrum estimation method to respectively obtain second characteristic matrixes of a training set and a testing set; extracting features of the signal data by adopting an EMD-SVD method to respectively obtain third feature matrixes of a training set and a testing set; and splicing the first, second and third feature matrixes, and performing data compression through PCA to obtain a training set and a testing set after dimension reduction.
A computer-readable storage medium storing computer instructions that, when executed by a processor, perform the rolling bearing fault signature extraction method of embodiment 1.
An intelligent diagnosis system for rolling bearing faults comprises a classifier, an acquisition module, a preprocessing module and a feature extraction module, wherein the acquisition module, the preprocessing module and the feature extraction module are described in the embodiment; the classifier is configured to: inputting a training set and a testing set subjected to dimension reduction by the feature extraction module; initializing a particle swarm, and initializing the speeds and positions of particles c and g; calculating particle fitness (SVM recognition accuracy) under the current penalty term coefficient c and the kernel function width g; searching an extremum; updating particle velocity and position; judging whether the SVM classification error meets a termination condition; if the termination condition is met, putting the test set into a classifier for classification, and analyzing classification results of the training set and the test set; if not, returning to the position where the extremum is found to continue processing.
While the foregoing description of the embodiments of the present invention has been presented in conjunction with the drawings, it should be understood that it is not intended to limit the scope of the invention, but rather, it is intended to cover all modifications or variations within the scope of the invention as defined by the claims of the present invention.
Claims (9)
1. The method for extracting the fault characteristics of the rolling bearing is characterized by comprising the following steps of: collecting bearing signal data of an electric driving end, preprocessing the signal data, and dividing the signal data into a test set and a training set;
extracting characteristics of signal data by adopting a wavelet band energy method to respectively obtain first characteristic matrixes of a training set and a testing set;
extracting characteristics of signal data by adopting a wavelet packet-AR spectrum estimation method to respectively obtain second characteristic matrixes of a training set and a testing set;
extracting features of the signal data by adopting an EMD-SVD method to respectively obtain third feature matrixes of a training set and a testing set;
the first, second and third feature matrixes are spliced, data compression is carried out through PCA, and a training set and a testing set after dimension reduction are obtained, namely, the training set and the testing set are bearing fault features;
the specific steps of extracting the characteristics of the signal data by adopting the wavelet packet-AR spectrum estimation method comprise the following steps: carrying out wavelet packet decomposition on the bearing signals;
decomposing different components of the signal into corresponding frequency bands, and carrying out wavelet packet reconstruction on each component to obtain a reconstructed signal;
AR spectrum estimation is carried out on the reconstructed signals, and AR spectrum energy characteristics of each frequency band are calculated;
and obtaining feature vectors of the AR spectrum energy features, and respectively obtaining second feature matrixes of the training set and the testing set.
2. The rolling bearing failure feature extraction method according to claim 1, wherein the step of preprocessing the signal data includes: the signal sequence to be used in the bearing data signal is F 1 ~F i F is to F i Suitably split to form l signal sequences, F i =f i1 ,f i2 ,f i3 …f il Of f, where f i1 A first subset sequence segmented for an ith failure sequence; random slave F i Selecting a part of signal subsequences as an original training set; the other part is used as the original test set.
3. The rolling bearing fault signature extraction method as recited in claim 1, wherein the specific step of extracting the signature of the signal data using wavelet band energy method comprises: carrying out wavelet decomposition on the bearing signal data;
decomposing different components of the bearing signal data into corresponding frequency bands, and calculating the energy characteristics of each frequency band;
and carrying out normalization processing on the energy feature vectors to respectively obtain first feature matrixes of the training set and the testing set.
4. The method for extracting the fault characteristics of the rolling bearing according to claim 1, wherein the specific step of extracting the characteristics of the signal data by using the EMD-SVD method comprises the steps of:
EMD is carried out on the signals to obtain a plurality of IMF components;
calculating the energy duty ratio of the IMF component;
setting a threshold value, and extracting a plurality of IMF components with energy duty ratio exceeding the threshold value;
and (3) carrying out compression processing on the extracted IMF components by adopting SVD, and respectively obtaining a training set and a third feature matrix of a test set by taking the compressed IMF components as signal features.
5. The rolling bearing fault signature extraction method as recited in claim 1, wherein the specific step of performing a stitching process on the first, second and third feature matrices comprises: obtaining a training set alpha and a test set beta after the splicing treatment, wherein alpha= [ alpha ] 1 ,α 2 ,α 3 ] A1×M ,β=[β 1 ,β 2 ,β 3 ] A2×M A1 and A2 respectively represent the signal numbers of the training set and the test set, and M represents the feature dimension alpha 1 、α 2 And alpha 3 First, second and third feature matrices, beta, respectively, of the training set 1 、β 2 And beta 3 The first feature matrix, the second feature matrix and the third feature matrix are respectively of the test set;
the specific steps of data compression by PCA include: for a training data set alpha, calculating a covariance matrix of the training data set alpha; singular value decomposition is carried out on the covariance matrix; outputting the training set after dimension reduction; and the same treatment is carried out on the test set beta to obtain the test set after dimension reduction.
6. An intelligent diagnosis method for a rolling bearing fault is characterized by comprising the steps of obtaining a training set and a test set after dimension reduction by adopting the rolling bearing fault feature extraction method according to claims 1-5, inputting the training set and the test set after dimension reduction into a classifier for training, and diagnosing the bearing fault;
the specific steps of inputting the training set and the testing set after the dimension reduction into the classifier for training comprise the following steps: initializing a particle swarm, and initializing the punishment item coefficient of the particle, the speed and the position of the width of the kernel function; under the current punishment item coefficient and the width of the kernel function, calculating the SVM identification accuracy;
searching an extremum; updating particle velocity and position; judging whether the SVM classification error meets a termination condition;
if the termination condition is met, putting the test set into a classifier for classification, and analyzing classification results of the training set and the test set; if not, returning to the position where the extremum is found to continue processing.
7. A rolling bearing fault signature extraction system comprising:
an acquisition module configured to: collecting bearing signal data of an electric driving end;
a preprocessing module configured to: preprocessing signal data, and dividing the signal data into a test set and a training set;
a feature extraction module configured to: extracting characteristics of signal data by adopting a wavelet band energy method to respectively obtain first characteristic matrixes of a training set and a testing set; extracting characteristics of signal data by adopting a wavelet packet-AR spectrum estimation method to respectively obtain second characteristic matrixes of a training set and a testing set; extracting features of the signal data by adopting an EMD-SVD method to respectively obtain third feature matrixes of a training set and a testing set; splicing the first, second and third feature matrixes, and performing data compression through PCA to obtain a training set and a testing set after dimension reduction;
the specific steps of extracting the characteristics of the signal data by adopting the wavelet packet-AR spectrum estimation method comprise the following steps: carrying out wavelet packet decomposition on the bearing signals;
decomposing different components of the signal into corresponding frequency bands, and carrying out wavelet packet reconstruction on each component to obtain a reconstructed signal;
AR spectrum estimation is carried out on the reconstructed signals, and AR spectrum energy characteristics of each frequency band are calculated;
and obtaining feature vectors of the AR spectrum energy features, and respectively obtaining second feature matrixes of the training set and the testing set.
8. A computer readable storage medium storing computer instructions which, when executed by a processor, perform the rolling bearing fault signature extraction method of claims 1-5.
9. An intelligent diagnostic system for a rolling bearing failure, comprising: a classifier and an acquisition module, a preprocessing module and a feature extraction module in the feature extraction system as claimed in claim 7; the classifier is configured to: inputting a training set and a testing set subjected to dimension reduction by the feature extraction module; initializing a particle swarm, and initializing the punishment item coefficient of the particle, the speed and the position of the width of the kernel function; under the current punishment item coefficient and the width of the kernel function, calculating the SVM identification accuracy; searching an extremum; updating particle velocity and position; judging whether the SVM classification error meets a termination condition; if the termination condition is met, putting the test set into a classifier for classification, and analyzing classification results of the training set and the test set; if not, returning to the position where the extremum is found to continue processing.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010313588.1A CN111476339B (en) | 2020-04-20 | 2020-04-20 | Rolling bearing fault feature extraction method, intelligent diagnosis method and system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010313588.1A CN111476339B (en) | 2020-04-20 | 2020-04-20 | Rolling bearing fault feature extraction method, intelligent diagnosis method and system |
Publications (2)
Publication Number | Publication Date |
---|---|
CN111476339A CN111476339A (en) | 2020-07-31 |
CN111476339B true CN111476339B (en) | 2023-05-05 |
Family
ID=71754323
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202010313588.1A Active CN111476339B (en) | 2020-04-20 | 2020-04-20 | Rolling bearing fault feature extraction method, intelligent diagnosis method and system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN111476339B (en) |
Families Citing this family (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112633333A (en) * | 2020-12-11 | 2021-04-09 | 广州致新电力科技有限公司 | Method for identifying partial discharge defects |
CN113076844B (en) * | 2021-03-26 | 2022-05-20 | 华中科技大学 | Method for constructing fault diagnosis model of rotating part and application |
CN113705100A (en) * | 2021-08-31 | 2021-11-26 | 哈尔滨理工大学 | Gas detection method based on combination of temperature modulation detection and PSO-SVM algorithm |
CN116893314A (en) * | 2023-09-04 | 2023-10-17 | 国网浙江省电力有限公司余姚市供电公司 | Non-invasive power load monitoring method, device, equipment and storage medium |
CN117686226B (en) * | 2024-02-04 | 2024-04-16 | 南京凯奥思数据技术有限公司 | Automatic bearing fault diagnosis method and system based on energy ratio and energy sum |
Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110220711A (en) * | 2019-05-22 | 2019-09-10 | 北京化工大学 | A kind of piston-mode motor shock characteristic extracting method based on EMD |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CA3200604A1 (en) * | 2017-03-24 | 2018-09-24 | Canadian Pacific Railway Company | Condition based maintenance of railcar roller bearings using predictive wayside alerts based on acoustic bearing detector measurements |
-
2020
- 2020-04-20 CN CN202010313588.1A patent/CN111476339B/en active Active
Patent Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110220711A (en) * | 2019-05-22 | 2019-09-10 | 北京化工大学 | A kind of piston-mode motor shock characteristic extracting method based on EMD |
Non-Patent Citations (2)
Title |
---|
JIAYU CHEN.An Active Learning Method Based on Uncertainty and Complexity for Gearbox Fault Diagnosis.《IEEE》.2018,9022-9031. * |
张玲玲.基于小波包 -AR 谱的变速器轴承故障特征提取.《振动、测试与诊断》.2011,492-537. * |
Also Published As
Publication number | Publication date |
---|---|
CN111476339A (en) | 2020-07-31 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111476339B (en) | Rolling bearing fault feature extraction method, intelligent diagnosis method and system | |
CN108319962B (en) | Tool wear monitoring method based on convolutional neural network | |
CN110840402B (en) | Atrial fibrillation signal identification method and system based on machine learning | |
CN111523509B (en) | Equipment fault diagnosis and health monitoring method integrating physical and depth expression characteristics | |
CN112036301B (en) | Driving motor fault diagnosis model construction method based on intra-class feature transfer learning and multi-source information fusion | |
CN111460956B (en) | Unbalanced electrocardiogram sample classification method based on data enhancement and loss weighting | |
CN105841961A (en) | Bearing fault diagnosis method based on Morlet wavelet transformation and convolutional neural network | |
CN112257530B (en) | Rolling bearing fault diagnosis method based on blind signal separation and support vector machine | |
CN111753891B (en) | Rolling bearing fault diagnosis method based on unsupervised feature learning | |
CN113176092B (en) | Motor bearing fault diagnosis method based on data fusion and improved experience wavelet transform | |
CN108509701B (en) | Rotary machine fault direct intelligent diagnosis method based on vibration signal | |
CN112577743B (en) | Rolling bearing fault diagnosis method based on maximum local boundary criterion | |
CN111238843A (en) | Fan health evaluation method based on rapid spectrum kurtosis analysis | |
CN111914703A (en) | Mechanical rotating part fault diagnosis method based on wavelet transformation and transfer learning GoogLeNet | |
CN114169377A (en) | G-MSCNN-based fault diagnosis method for rolling bearing in noisy environment | |
CN115290326A (en) | Rolling bearing fault intelligent diagnosis method | |
CN115221926A (en) | Heart beat signal classification method based on CNN-GRU network model | |
CN113155462A (en) | Bearing fault diagnosis method based on octyl geometric mode decomposition and graph structure enhanced dynamic time warping | |
CN116842460A (en) | Cough-related disease identification method and system based on attention mechanism and residual neural network | |
CN112945556B (en) | Bearing fault diagnosis method based on wavelet packet decomposition and optimal rejection classification strategy | |
CN113758709A (en) | Rolling bearing fault diagnosis method and system combining edge calculation and deep learning | |
CN113295413A (en) | Traction motor bearing fault diagnosis method based on indirect signals | |
CN116340812A (en) | Transformer partial discharge fault mode identification method and system | |
CN116482526A (en) | Analysis system for non-fault phase impedance relay | |
CN116429426A (en) | Bearing fault diagnosis method, device and medium for multi-domain feature fusion |
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 | ||
TR01 | Transfer of patent right |
Effective date of registration: 20231122 Address after: No. 2151, 21st Floor, Building 2, No. 35 Jianguo Road, Chaoyang District, Beijing, 100123 Patentee after: Beijing Xiniu Technology Co.,Ltd. Address before: 250014 No. 88, Wenhua East Road, Lixia District, Shandong, Ji'nan Patentee before: SHANDONG NORMAL University |
|
TR01 | Transfer of patent right |