CN114742095A - Rolling bearing multichannel data fault identification method - Google Patents

Rolling bearing multichannel data fault identification method Download PDF

Info

Publication number
CN114742095A
CN114742095A CN202210281811.8A CN202210281811A CN114742095A CN 114742095 A CN114742095 A CN 114742095A CN 202210281811 A CN202210281811 A CN 202210281811A CN 114742095 A CN114742095 A CN 114742095A
Authority
CN
China
Prior art keywords
rolling bearing
equation
emmwpe
random forest
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.)
Pending
Application number
CN202210281811.8A
Other languages
Chinese (zh)
Inventor
王晓东
张爽
马军
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.)
Kunming University of Science and Technology
Original Assignee
Kunming University of Science and Technology
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 Kunming University of Science and Technology filed Critical Kunming University of Science and Technology
Priority to CN202210281811.8A priority Critical patent/CN114742095A/en
Publication of CN114742095A publication Critical patent/CN114742095A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
    • G06F2218/12—Classification; Matching
    • 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
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/243—Classification techniques relating to the number of classes
    • G06F18/24323—Tree-organised classifiers

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • General Engineering & Computer Science (AREA)
  • Artificial Intelligence (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • Signal Processing (AREA)
  • Acoustics & Sound (AREA)
  • Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)

Abstract

The invention discloses a rolling bearing multichannel data fault identification method, which comprises the steps of firstly, calculating fault characteristic vectors of multichannel signals by using an EMMWPE method; secondly, completing the dimensionality reduction and redundant feature removal of the feature vector through t-SNE; finally, inputting the feature set subjected to dimension reduction into a random forest model to complete rolling fault identification; the rolling bearing fault detection method can effectively identify the fault state of the rolling bearing, and has important significance for ensuring the normal operation of equipment.

Description

Rolling bearing multichannel data fault identification method
Technical Field
The invention belongs to the technical field of mechanical fault diagnosis, and particularly relates to an EMMWPE and random forest based rolling bearing multichannel data fault identification method.
Background
The rolling bearing is an indispensable key part in mechanical equipment and has the advantages of high mechanical efficiency, convenience in disassembly, assembly and maintenance and the like. However, bearings often operate in high temperature, high pressure and complex mechanical environments, and as operating time increases, the bearings experience various failures. When the bearing is out of order and not maintained in time, the whole mechanical equipment can not work normally or even be damaged, and huge loss is caused.
When monitoring the operating state of the rolling bearing, a plurality of sensors are often required to acquire data from a plurality of directions. Fault information contained in different data channels is different, and in order to extract more comprehensive fault information from a plurality of data channels, feature fusion and extraction are often performed on multi-channel signals through MMPE. However, the MMPE method ignores the amplitude information of the data sequence, and there are problems such as information loss on multiple scales, which may result in a drastic decrease in the stability of the MMPE method as the scale factor increases. Therefore, the invention provides a multi-channel data feature extraction method for EMMWPE, and the problems of MMPE are solved. Finally, the EMMWPE high-dimensional vector feature set is subjected to t-SNE dimension reduction and input into a random forest model for training and recognition, and the method provided by the invention can be found to have higher recognition accuracy.
Disclosure of Invention
In order to solve the technical problem, the invention provides a rolling bearing multichannel data fault identification method based on EMMWPE and a random forest, which is characterized in that a multichannel rolling bearing signal is taken as a basis, EMMWPE is used for extracting multichannel signal characteristics, t-SNE is used for reducing the dimension, and the obtained low-dimensional characteristics are input into a random forest model for fault identification;
the invention is realized by the following technical scheme: a rolling bearing multi-channel data fault identification method based on EMMWPE and random forest comprises the following steps:
s1: firstly, the multichannel signal is subjected to an enhanced multiscale coarse graining process, and for a time sequence with a given length n, X is [ X1, X2, X3 … xn ]]It can be processed into tau different rough physicochemical sequences respectively according to the given tau and the formula (1)
Figure BDA0003557221980000021
Wherein i is 1,2, …, τ;
Figure BDA0003557221980000022
s2: respectively calculating multiple weighted arrangement entropies of tau rough physicochemical sequences, and then calculating the average value of the tau entropy values, namely the average value is equal to the enhanced multiple multi-scale weighted arrangement entropies, as shown in a formula (2):
Figure BDA0003557221980000023
s3: performing dimension reduction on the EMMWPE value through a t-SNE algorithm, removing redundant features, and visualizing the calculated features;
s4: dividing each multi-channel data set with different fault types and damage degrees into N groups, wherein the training samples comprise M groups, and the testing samples comprise N-M groups;
s5: training and testing the sample based on a random forest model to realize fault identification of different types and damage degrees of the rolling bearing;
preferably, the MWPE function in S2 is calculated as follows:
s2.1: defining a time sequence with m dimension and length of N as
Figure BDA0003557221980000031
Figure BDA0003557221980000032
Introducing an embedding dimension d and a time delay coefficient tau to carry out phase space reconstruction; performing phase space reconstruction according to the formula (3) to obtain a reconstruction sequence
Figure BDA0003557221980000033
Figure BDA0003557221980000034
S2.2: the reconstructed time sequence
Figure BDA0003557221980000035
Arranged in ascending order, the result after the sorting is shown as the formula (4):
[xi(l+(k1-1)τ)≤xi(l+(k2-1)τ)≤...≤xi(l+(kd-1)τ)] (4)
s2.3: consider that
Figure BDA0003557221980000036
Wherein elements with the same value are present, are arranged by comparing the values of k, which results in
Figure BDA0003557221980000037
Arrangement of all elements in (1):
π＝[k1,k2,...,kd] (5)
s2.4: for a phase space with an embedding dimension d, the total number of permutations that can occur is d! (ii) a N is a radical of hydrogenjRepresents the number of occurrences of the jth ordering, where 1 ≦ j ≦ d! (ii) a
Figure BDA0003557221980000038
And
Figure BDA0003557221980000039
respectively represent
Figure BDA00035572219800000310
Mean and weight values of:
Figure BDA00035572219800000311
Figure BDA00035572219800000312
s2.5: weighted probability p of jth permutationi,jThe calculation can be made from equation (8):
Figure BDA00035572219800000313
s2.6: for a time series of m dimensions, pi,jSatisfy the requirement of
Figure BDA0003557221980000041
Weighted probability of jth ordering in m-dimensional time series
Figure BDA0003557221980000042
As shown in formula (9):
Figure BDA0003557221980000043
s2.7: the value of MWPE can be calculated from equation (10) according to the definition of shannon entropy.
Figure BDA0003557221980000044
Preferably, the t-SNE algorithm in S3 comprises the following steps:
s3.1: let data X ═ X1,x2,…,xnCalculating the degree of complexity of Perp affinity-sparseness P by equation (11)jI, Perp are cost function parameters.
Figure BDA0003557221980000045
In the formula, σiIs xiThe gaussian variance of (c).
S3.2: is provided with
Figure BDA0003557221980000046
From N (0, 10)-4I) Obtaining an initial solution y of the sample(0)＝{y1,y2,…,yn}。
S3.3: calculation of Low-dimensional affinity-phobicity q by equation (12)ij：
Figure BDA0003557221980000047
S3.4: calculating the gradient by equation (13)
Figure BDA0003557221980000048
Figure BDA0003557221980000049
S3.4: the low latitude data is obtained by equation (14):
Figure BDA0003557221980000051
in the formula: the learning rate eta and the momentum alpha (t) are optimization parameters.
S3.5: repeating the iteration S.33-3.5 until T is added from 1 to T, and outputting low-dimensional data y(T)＝{y1,y2,…,yn}。
Preferably, the technical features of the random forest model in S5 are as follows:
s5.1: performing permutation sampling on the training set to obtain k training subsets D ═ D { with the same size as the training set1,D2,…,Dk}。
S5.2: each sample of a training subset contains n features. Firstly, m (m is less than or equal to n) features are selected from n features to construct a subspace S, secondly, the optimal splitting point of the decision tree node is calculated, and the node is generated according to S. And repeating the process until the stopping criterion is met, and finishing the training of the decision tree. After the k training subsets are trained in this way, k decision trees DT ═ DT1,DT2,…,DTk}。
S5.3: each decision tree is tested by each sample of the test set to obtain k classification results R ═ R1,R2,…,Rk}。
The invention has the beneficial effects that:
the EMMWPE method is provided by combining the multi-scale rough physicochemical process and the MWPE, so that the defects of the MMPE method are overcome, and the stability of the MMPE in extracting the fault characteristics of the multi-channel signals is improved; and high-dimensional redundant features are removed through t-SNE dimension reduction, and the obtained fault feature set is input into a random forest model to finish automatic fault identification, so that the method has important significance for ensuring the normal operation of equipment.
Drawings
FIG. 1 is a flow chart of the steps of the present invention;
FIG. 2 is a bearing data center laboratory bench of the university of Kaiser storage, USA, in an embodiment of the invention;
FIG. 3 is a time domain signal diagram in an embodiment of the invention;
FIG. 4 is a graph of EMMWPE calculation characteristics in an embodiment of the present invention;
FIG. 5 is a low-dimensional feature diagram after the dimension reduction of t-SNE in the embodiment of the invention.
Detailed Description
In order to clearly and completely describe the technical scheme and the effects of the invention, the following embodiments are used for detailed description;
example 1
The embodiment of the invention provides a rolling bearing multi-channel data fault identification method based on EMMWPE and random forest, and a basic flow diagram is shown in figure 1; in the embodiment, data collected by a CWRU rolling bearing fault simulation experiment platform is used as an experiment object, as shown in FIG. 2; the specific parameters are as follows: the bearing model is SKF6205, the rotating speed is 1750r/min, and the sampling frequency is 12 kHz;
s1: when the load is selected to be 0, the vibration signals of the outer ring, the rolling body and the driving end and the fan end of the inner ring with different fault diameters of the rolling bearing are subjected to experimental analysis, as shown in table 1, wherein data of each fault type are divided into 50 groups, the data length of each group is 2048, and the time domain graph is shown in fig. 3;
TABLE 1
Figure BDA0003557221980000061
Figure BDA0003557221980000071
S2: calculating EMMWPE values of the multi-channel signals, wherein the calculation results are shown in FIG. 4;
s3: performing dimension reduction through a T-sne algorithm to obtain a low-dimensional feature set and performing visualization, as shown in FIG. 5; the characteristic clustering characteristic of each fault type extracted by EMMWPE is good, the extracted characteristic quality is high, and fault states can be well distinguished;
s4: for each type, 400 groups of 40 groups of characteristics are selected for training, 100 groups are selected for testing, and the test results are shown in table 2; it can be seen that the proposed method only misclassifies one cup out of 100 test samples, which is apparent to have a very high accuracy;
TABLE 2
Recognition rate Correct number of Number of errors Label (R)
100％ 20 0 NM
100％ 20 0 ORF1
100％ 20 0 ORF2
100％ 20 0 ORF3
100％ 20 0 BF1
100％ 20 0 BF2
100％ 20 0 BF3
100％ 20 0 IRF1
95％ 19 1 IRF2
100％ 20 0 IRF3
In the description herein, references to the description of "one embodiment," "an example," "a specific example" or the like are intended to mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, the schematic representations of the terms used above do not necessarily refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.
The preferred embodiments of the invention disclosed above are intended to be illustrative only. The preferred embodiments are not intended to be exhaustive or to limit the invention to the precise embodiments disclosed. Obviously, many modifications and variations are possible in light of the above teaching. The embodiments were chosen and described in order to best explain the principles of the invention and the practical application, to thereby enable others skilled in the art to best utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims (4)

1. A rolling bearing multichannel data fault identification method based on EMMWPE and random forest is characterized by comprising the following steps:
s1: firstly, the multichannel signal is subjected to an enhanced multiscale coarse graining process, and for a time sequence with a given length n, X is [ X1, X2, X3 … xn ]]It can be processed into tau different coarse physicochemical sequences, respectively, given tau and equation (1)
Figure FDA0003557221970000011
Wherein i is 1,2, …, τ;
Figure FDA0003557221970000012
s2: respectively calculating multiple weighted arrangement entropies of tau coarse physicochemical sequences, and then solving the average value of the tau entropy values to be equal to the enhanced multiple multi-scale weighted arrangement entropy, wherein the formula (2) is as follows:
Figure FDA0003557221970000013
s3: performing dimension reduction on the EMMWPE value through a t-SNE algorithm, removing redundant features, and visualizing the calculated features;
s4: dividing each multi-channel data set with different fault types and damage degrees into N groups, wherein the training samples comprise M groups, and the testing samples comprise N-M groups;
s5: based on the random forest model, the samples are trained and tested, and different types of rolling bearings and fault recognition of damage degrees are achieved.
2. The rolling bearing multichannel data fault identification method based on EMMWPE and random forest as claimed in claim 1, wherein the calculation method of MWPE function in S2 is as follows:
s2.1: defining a time sequence with m dimension and length of N as
Figure FDA0003557221970000014
t is 1,2, …, N, and an embedding dimension d and a time delay coefficient tau are introduced for phase space reconstruction; performing phase space reconstruction according to the formula (3) to obtain a reconstruction sequence
Figure FDA0003557221970000021
Figure FDA0003557221970000022
S2.2: the reconstructed time sequence
Figure FDA0003557221970000023
Arranged in ascending order, the result after the sorting is shown as the formula (4):
[xi(l+(k1-1)τ)≤xi(l+(k2-1)τ)≤...≤xi(l+(kd-1)τ)] (4)
s2.3: consider that
Figure FDA0003557221970000024
Wherein elements with the same value are present, are arranged by comparing the values of k, which results in
Figure FDA0003557221970000025
Arrangement of all elements in (1):
π＝[k1,k2,...,kd] (5)
s2.4: for a phase space with an embedding dimension d, the total number of permutations that can occur is d! (ii) a N is a radical of hydrogenjRepresents the number of occurrences of the jth ordering, where 1 ≦ j ≦ d! (ii) a
Figure FDA0003557221970000026
And
Figure FDA0003557221970000027
respectively represent
Figure FDA0003557221970000028
Mean and weight values of (c):
Figure FDA0003557221970000029
Figure FDA00035572219700000210
s2.5: weighted probability p of jth permutationi,jThe calculation can be made from equation (8):
Figure FDA00035572219700000211
s2.6: for a time series of m dimensions, pi,jSatisfy the requirements of
Figure FDA00035572219700000212
Weighted probability of jth ordering in m-dimensional time series
Figure FDA00035572219700000213
As shown in formula (9):
Figure FDA0003557221970000031
s2.7: the value of MWPE can be calculated from equation (10) according to the definition of shannon entropy.
Figure FDA0003557221970000032
3. The EMMWPE and random forest based rolling bearing multichannel data fault identification method as claimed in claim 1, wherein the t-SNE algorithm in S3 comprises the following steps:
s3.1: let data X ═ X1,x2,…,xnP, complexity Perp affinity-sparseness calculated by equation (11)jI, Perp are cost function parameters.
Figure FDA0003557221970000033
In the formula, σiIs xiThe gaussian variance of (c).
S3.2: is provided with
Figure FDA0003557221970000034
From N (0, 10)-4I) Obtaining an initial solution y of the sample(0)＝{y1,y2,…,yn}。
S3.3: calculation of Low-dimensional affinity-phobicity q by equation (12)ij：
Figure FDA0003557221970000035
S3.4: calculating the gradient by equation (13)
Figure FDA0003557221970000036
Figure FDA0003557221970000037
S3.4: the low latitude data is obtained by equation (14):
Figure FDA0003557221970000041
in the formula: the learning rate eta and the momentum alpha (t) are optimization parameters.
S3.5: repeating the iteration S.33-3.5 until T is added from 1 to T, and outputting low-dimensional data y(T)＝{y1,y2,…,yn}。
4. The rolling bearing multichannel data fault identification method based on EMMWPE and random forest as claimed in claim 1, wherein the technical features of the random forest model in S5 are as follows:
s5.1: performing permutation sampling on the training set to obtain k training subsets D ═ D { D } with the same size as the training set1,D2,…,Dk}。
S5.2: each sample of a training subset contains n features. Firstly, m (m is less than or equal to n) features are selected from n features to construct a subspace S, secondly, the optimal splitting point of the decision tree node is calculated, and the node is generated according to S. And repeating the process until the stop criterion is met, and finishing the training of the decision tree. After the k training subsets are trained in this way, k decision trees DT ═ DT1,DT2,…,DTk}。
S5.3: each decision tree is tested by each sample of the test set to obtain k classification results R ═ R1,R2,…,Rk}。
CN202210281811.8A 2022-03-21 2022-03-21 Rolling bearing multichannel data fault identification method Pending CN114742095A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210281811.8A CN114742095A (en) 2022-03-21 2022-03-21 Rolling bearing multichannel data fault identification method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210281811.8A CN114742095A (en) 2022-03-21 2022-03-21 Rolling bearing multichannel data fault identification method

Publications (1)

Publication Number Publication Date
CN114742095A true CN114742095A (en) 2022-07-12

Family

ID=82277346

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210281811.8A Pending CN114742095A (en) 2022-03-21 2022-03-21 Rolling bearing multichannel data fault identification method

Country Status (1)

Country Link
CN (1) CN114742095A (en)

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108388860A (en) * 2018-02-12 2018-08-10 大连理工大学 A kind of Aeroengine Ball Bearings method for diagnosing faults based on power entropy-spectrum-random forest
CN109916628A (en) * 2019-04-04 2019-06-21 哈尔滨理工大学 Based on the Fault Diagnosis of Roller Bearings for improving multiple dimensioned amplitude perception arrangement entropy
CN110991422A (en) * 2019-12-25 2020-04-10 安徽工业大学 Rolling bearing fault diagnosis method based on multi-element time-shifting multi-scale permutation entropy
CN111191740A (en) * 2020-01-10 2020-05-22 福州大学 Fault diagnosis method for rolling bearing
WO2021135630A1 (en) * 2019-12-31 2021-07-08 福州大学 Rolling bearing fault diagnosis method based on grcmse and manifold learning

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108388860A (en) * 2018-02-12 2018-08-10 大连理工大学 A kind of Aeroengine Ball Bearings method for diagnosing faults based on power entropy-spectrum-random forest
CN109916628A (en) * 2019-04-04 2019-06-21 哈尔滨理工大学 Based on the Fault Diagnosis of Roller Bearings for improving multiple dimensioned amplitude perception arrangement entropy
CN110991422A (en) * 2019-12-25 2020-04-10 安徽工业大学 Rolling bearing fault diagnosis method based on multi-element time-shifting multi-scale permutation entropy
WO2021135630A1 (en) * 2019-12-31 2021-07-08 福州大学 Rolling bearing fault diagnosis method based on grcmse and manifold learning
CN111191740A (en) * 2020-01-10 2020-05-22 福州大学 Fault diagnosis method for rolling bearing

Non-Patent Citations (8)

* Cited by examiner, † Cited by third party
Title
YUAN, XUYI等: "Research on Twin Extreme Learning Fault Diagnosis Method Based on Multi-Scale Weighted Permutation Entropy", ENTROPY, vol. 24, no. 9, 7 September 2022 (2022-09-07) *
ZHENG, JINDE: "Composite multi-scale weighted permutation entropy and extreme learning machine based intelligent fault diagnosis for rolling bearing", MEASUREMENT, vol. 143, 14 June 2019 (2019-06-14), pages 69 - 80, XP085701646, DOI: 10.1016/j.measurement.2019.05.002 *
丁嘉鑫等: "广义复合多尺度加权排列熵与参数优化支持向量机的滚动轴承故障诊断", 中国机械工程, vol. 32, no. 2, 31 January 2021 (2021-01-31), pages 147 - 155 *
克里斯托弗·拉尔森: "5G网络规划设计与优化", 31 July 2020, 机械工业出版社, pages: 117 - 118 *
吕麦丝等: "基于改进复合多元尺度加权排列熵的轴承故障识别研究", 中国工程机械学报, vol. 21, no. 1, 28 February 2023 (2023-02-28), pages 73 - 78 *
吴海滨等: "改进多尺度幅值感知排列熵与随机森林结合的滚动轴承故障诊断", 光学精密工程, vol. 28, no. 3, 31 March 2020 (2020-03-31), pages 621 - 631 *
肖洁;黎敬涛;邓超;罗志刚;林华峰;: "基于ITD与LLTSA的轴承故障诊断方法", 电子测量技术, no. 08, 23 April 2020 (2020-04-23), pages 189 - 194 *
赵家浩等: "基于改进多元多尺度加权排列熵的齿轮箱故障诊断", 组合机床与自动化加工技术, no. 12, 31 December 2022 (2022-12-31), pages 48 - 52 *

Similar Documents

Publication Publication Date Title
CN107368809B (en) A kind of bearing fault sorting technique based on rarefaction representation and dictionary learning
CN109582003B (en) Bearing fault diagnosis method based on pseudo label semi-supervised kernel local Fisher discriminant analysis
CN108803520B (en) Dynamic process monitoring method based on variable nonlinear autocorrelation rejection
CN110674892A (en) Fault feature screening method based on weighted multi-feature fusion and SVM classification
CN111191740B (en) Fault diagnosis method for rolling bearing
CN109993105A (en) A kind of improved adaptive sparse sampling Fault Classification
CN112257530A (en) Rolling bearing fault diagnosis method based on blind signal separation and support vector machine
CN111523081B (en) Aeroengine fault diagnosis method based on enhanced gate control circulating neural network
CN112577743B (en) Rolling bearing fault diagnosis method based on maximum local boundary criterion
CN110082136B (en) Rotary machine fault diagnosis method based on cloud genetic algorithm optimization support vector machine
CN110991422B (en) Rolling bearing fault diagnosis method based on multi-element time-shifting multi-scale permutation entropy
CN115510965A (en) Bearing imbalance fault diagnosis method based on generated data fusion
CN112633098A (en) Fault diagnosis method and system for rotary machine and storage medium
CN114925809B (en) LSTM-based printer bearing fault diagnosis method and device
CN115221930A (en) Fault diagnosis method for rolling bearing
CN114091504A (en) Rotary machine small sample fault diagnosis method based on generation countermeasure network
CN112380932B (en) Vibration signal characteristic value selection method and elevator health state evaluation or fault diagnosis method
CN112541524B (en) BP-Adaboost multisource information motor fault diagnosis method based on attention mechanism improvement
CN112149804B (en) Novel convolutional neural network model and application
CN117171544B (en) Motor vibration fault diagnosis method based on multichannel fusion convolutional neural network
CN112362292B (en) Method for anomaly detection of wind tunnel test data
CN114742095A (en) Rolling bearing multichannel data fault identification method
CN111382792B (en) Rolling bearing fault diagnosis method based on double-sparse dictionary sparse representation
CN111275003A (en) Microseismic signal identification method based on class-optimal Gaussian kernel multi-classification support vector machine
CN108898157B (en) Classification method for radar chart representation of numerical data based on convolutional neural network

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