KR101823746B1 - Method for bearing fault diagnosis - Google Patents
Method for bearing fault diagnosis Download PDFInfo
- Publication number
- KR101823746B1 KR101823746B1 KR1020160015240A KR20160015240A KR101823746B1 KR 101823746 B1 KR101823746 B1 KR 101823746B1 KR 1020160015240 A KR1020160015240 A KR 1020160015240A KR 20160015240 A KR20160015240 A KR 20160015240A KR 101823746 B1 KR101823746 B1 KR 101823746B1
- Authority
- KR
- South Korea
- Prior art keywords
- calculating
- bearing
- value
- cluster
- cdf
- Prior art date
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
- G01M13/04—Bearings
-
- 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
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G06N99/005—
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Acoustics & Sound (AREA)
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
Abstract
The present invention accurately diagnoses a fault condition by using a clustering technique that adaptively re-learns even when a new state that has not been learned before is added to improve diagnosis performance and adaptively sets an accurate k value for adaptive state update The present invention relates to a method of diagnosing a bearing failure. The present invention includes: A step of extracting features from a signal according to a bearing fault and learning it in a machine learning algorithm; A step B for extracting features from an unknown signal of a bearing and classifying a current bearing state by comparing the determined characteristics with previously learned characteristics by the machine learning algorithm; And a step C for setting the number of clusters k through the clustering algorithm and the cluster distribution analysis from the features extracted from the unknown signal of the bearing and learning the new state information to the machine learning algorithm when new state information is detected.
Description
The present invention relates to a method for diagnosing a bearing failure, and more particularly, to a method for diagnosing a bearing failure to minimize technical, economic and safety damage by detecting and determining a failure or defect of the bearing in an early stage.
In the detection of defects in bearings, it is used for non-destructive inspection by using characteristics of noise and vibration in real industrial field, which is greatly helping maintenance maintenance such as extension of bearing life.
Causes of bearing failure include insufficient lubrication, improper use of lubricant, misalignment of bearings, and excessive deformation of shaft. In the past, experienced technicians diagnosed these problems and judged whether they were faulty. However, most of them have a long diagnosis time, subjective, and sometimes have to stop the operation of the equipment system. In recent years, a system capable of diagnosing a failure of a bearing has been required while maintaining the operation of a device system, and thus the bearing operating state is continuously diagnosed and developed as a type of technology capable of detecting an abnormality before a failure.
1 is a view for explaining a basic principle of a bearing failure diagnosis method according to the related art.
Fig. 1 (a) shows the structure of the bearing, (b) shows the bearing outer ring defect, and (c) shows the inner ring of the bearing. The sensor is attached to the position of the machine, , it is possible to extract the features from the acquired signal and diagnose the failure by analyzing the pattern of the features in the status diagnosis area as shown in (d).
On the other hand, it is important to select a frequency band that best reveals defect symptoms in addition to envelope analysis from input signals (vibration, current, voltage, acoustic emission, etc.) for reliable machine fault diagnosis.
As an example of such a conventional method for diagnosing a bearing failure, a k-means algorithm (hereinafter referred to as a clustering algorithm) was used.
The clustering algorithm first selects k samples from the samples and sets them as a center point (step 1), calculates the distance between the center point of each cluster and each sample, classifies the samples into a cluster having a shorter distance (
However, when the conventional machine learning algorithm is used for the diagnosis of the bearing failure, it is not possible to classify the state when the machine has not been learned beforehand. If the existing clustering algorithm is used to update the machine state, There is a problem in that performance degradation and judgment error may occur when the k value is selected incorrectly.
SUMMARY OF THE INVENTION The present invention has been made to overcome the above-described problems of the prior art, and it is an object of the present invention to adaptively re-learn even if a new state not previously learned is added to increase diagnostic performance, adaptively set an accurate k value The present invention relates to a method for diagnosing a failure of a bearing which can accurately diagnose a failure state by using a clustering technique.
According to an aspect of the present invention, there is provided a method for detecting a bearing fault, comprising the steps of: extracting features from a signal according to a bearing defect; A step B for extracting features from an unknown signal of a bearing and classifying a current bearing state by comparing the determined characteristics with previously learned characteristics by the machine learning algorithm; A step of setting the number of clusters (k) through a clustering algorithm and a cluster distribution analysis from the features extracted from the unknown signal of the bearing, and a step C of learning the new state information to the machine learning algorithm ≪ / RTI >
It is preferable that the features extracted from the signals of the bearing defects and the features extracted from the unknown signals of the bearings are the same.
The features may include a root-mean-square, a shape factor, a kurtosis value, a square-mean-root, a peak-to-peak value A skewness value, an impulse factor, and a crest factor, or a combination of features extracted from two or more signals.
The machine learning algorithm may be any one of algorithms that can be used in machine learning, such as a support vector machine, an artificial neural network, or a k-nearest neighbors classifier, It can learn previously acquired features and classify current feature values into specific states.
The step C includes the steps of classifying samples into k clusters through a clustering algorithm; Calculating an average value and a covariance for each cluster; Probability density function of each cluster
; For each cluster, a local distribution factor (LDF) ; Calculating a minimum value among the LDFs of the clusters and calculating a global density factor (GDF); Global separability factor (GSF) ; Calculating a CDF (Cluster Distribution Factor) by calculating a difference between the GDF and the GSF, calculating k CDF from 1 to a specific value, and setting the k value when the CDF is the smallest to be an optimal k value have.Where X is the samples, Σ is the covariance, d is the number of features (dimensionality), μ is the mean value, and ICD i, j is the distance between the ith cluster and the jth cluster.
Another aspect of the present invention is a method comprising: A step of extracting features from a signal according to a bearing fault and learning the clustering algorithm (k-mean algorithm); Extracting features from an unknown signal of the bearing and classifying the current bearing state by the clustering algorithm; And a step C of setting the number of clusters (k) again by analyzing the clustering algorithm and the cluster distribution from the features extracted from the unknown signal of the bearing and learning the new number of clusters when the new state information is detected. A diagnostic method can be provided.
According to the method for diagnosing a failure of a bearing according to the present invention configured as described above, it is possible to adaptively re-learn even if a new state that has not been previously learned is added, thereby improving diagnostic performance. , It is possible to accurately diagnose the fault state and to prevent the performance degradation and the judgment error from occurring.
BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a view for explaining a bearing failure diagnosis method according to a related art;
FIG. 2 is a control flowchart illustrating a method for diagnosing a bearing failure according to a preferred embodiment of the present invention. FIG.
FIG. 3 is a graph showing a cluster distribution analysis process of the present invention,
FIGS. 4 and 5 are graphs showing comparison of cluster distribution states according to k value selection,
6 is a control flowchart illustrating a method for diagnosing a bearing failure according to another embodiment of the present invention.
The present invention may have various modifications and various embodiments, and specific embodiments are illustrated in the drawings and described in detail in the detailed description. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, the present invention will be described in detail with reference to the accompanying drawings.
With reference to the accompanying drawings, preferred embodiments of the present invention will be described in detail.
FIG. 2 is a control flowchart illustrating a method for diagnosing a bearing failure according to an embodiment of the present invention. FIG. 3 is a graph illustrating a cluster distribution analysis process of the present invention.
As shown in the drawings, the method for diagnosing a bearing failure according to a preferred embodiment of the present invention includes steps A100 to S120 for extracting features from a signal according to a bearing fault and learning a machine learning algorithm; A step B (S130 to S150) of extracting features from an unknown signal of the bearing and comparing the current bearing state with the features learned in advance by the machine learning algorithm; (C) step (S160 ~ S170) for setting the number of clusters (k) through the clustering algorithm and cluster distribution analysis from the features extracted from the unknown signal of the bearing and learning the new state information to the machine learning algorithm .
The operation of the bearing fault diagnosis method according to the preferred embodiment of the present invention will be described in more detail as follows.
First, the features are extracted from the signal for each bearing defect, and are learned by the machine learning algorithm (S100 to S120).
The signal for each bearing defect may include signals such as vibration, current, voltage, acoustic emission, etc., obtained from various sensors attached to the bearing.
Next, the features are extracted from the defect-by-defect analysis signal and are learned by the machine learning algorithm. At this time, the features that can distinguish each defect from the previously obtained analysis signal for each defect are extracted, and the extracted features are learned in the machine learning algorithm.
That is, after calculating and extracting the characteristics of the defect information from the input signal having the defect information, a feature vector composed of the class of the defect information and the features of the extracted defect information is generated and the machine learning algorithm is learned.
At this time, the features may include a root-mean-square, a shape factor, a kurtosis value, a square-mean-root, a peak-to- a peak value, a skewness value, an impulse factor, and a crest factor, or a combination of features extracted from two or more signals.
Also, the machine learning algorithm may be any one of the algorithms that can be used for machine learning, including a support vector machine, an artificial neural network, or a k-nearest neighbors classifier Can be used to learn pre-acquired features and classify the current feature vector into a particular state.
Thereafter, the features are extracted from the unknown signal of the bearing, and the current bearing state is classified by comparing with the features learned in advance by the machine learning algorithm (S130 to S150).
At this time, it is preferable that the features extracted from the signals of the bearing defects and the features extracted from the unknown signals of the bearings are the same.
Then, the number of clusters k is set from the features extracted from the unknown signals of the bearings through the clustering algorithm and the cluster distribution analysis. If new state information is detected, the controller learns the new state information to the machine learning algorithm (S160 to S170).
The clustering algorithm sets k samples among the samples as a center point (step 1), calculates the distance between the center point of each cluster and each sample, (Step 2). After the average value of the samples of each classified cluster is calculated and reset to the center point (step 3), the above steps 1-3 are repeated until the change of the center point becomes lower than a specific threshold value ).
In this state, the present invention sets the number of clusters (k) through clustering algorithm and cluster distribution analysis.
The step of setting the number of clusters k will be described in more detail as follows.
First, the clustering algorithm classifies the samples into k clusters, and then calculates the mean and covariance for each cluster.
Then, the probability density function ρ (X) of each cluster is obtained by the following equation (1).
[Equation 1]
Where X is the samples, Σ is the covariance, d is the number of features (number of dimensions), and μ is the mean value.
Next, the local distribution factor (LDF) is calculated for each cluster by the following equation (2).
&Quot; (2) "
Next, the minimum value among the LDFs of the clusters is obtained, and the global density factor (GDF) is calculated.
&Quot; (3) "
Next, the global separability factor (GSF) is calculated by the following equation (4).
&Quot; (4) "
Here, ICD i, j ( Inter cluster distance) is the distance between the ith cluster and the jth cluster.
Next, the difference between the GDF and GSF is calculated to calculate a CDF (Cluster Distribution Factor).
&Quot; (5) "
Therefore, the value of k varies from 1 to a specific value, and the CDF is calculated to set the k value when the CDF is the smallest to the optimum k value.
As shown in FIG. 3, the effect of the cluster distribution analysis in the bearing failure diagnosis according to the present invention is as follows: (a) when an outer ring defect is issued in a steady state class k = 1, (b) That is, a class that has not been learned occurs. At this time, the k value is increased by the cluster distribution analysis, and (c) the newly learned outer-ring coupling class is generated. Thereafter, when the inner ring defect is issued in the state of (d) k = 2, (e) samples due to the inner ring defect, that is, an unlearned class occurs and the k value increases to 3 by the cluster distribution analysis, (f) A newly learned inner-wheel coupling class is generated.
As shown in the drawing, the present invention calculates an Euclidean distance between a feature vector of an acquired signal and a feature vector of an analysis signal, selects k neighboring vectors having a distance closest to the feature vector of the acquired signal, The feature vector of the acquired signal is classified in a state including the largest number of neighboring vectors among the two groups (
FIGS. 4 and 5 are graphs showing comparison of cluster distribution according to k value selection. FIG.
As shown in FIG. 4, the prior art has a problem in that it can not be classified if a non-learned state occurs beforehand from a bearing machine. Also, even if an existing clustering technique is used to update a machine state, As shown in FIG. 4 (b), there are three clusters. If k is set to 2 incorrectly, performance degradation will occur.
The present invention employs a clustering technique to adaptively re-learn even if a new state that has not been learned before is added to increase diagnostic performance and to adaptively set the k value for adaptive state update.
As shown in FIGS. 5A and 5B, the conventional clustering result (a) and the distribution diagram (b) of each cluster are shown. The number of clusters is actually 4, .
On the other hand, the clustering result (c) of the present invention and the distribution diagram (d) of each cluster are shown in FIGS. 5C and 5D, It shows that it is done properly.
Therefore, the present invention can improve the diagnostic performance by adaptively re-learning even if a new state not previously learned is added, and by using the clustering technique for adaptively setting the correct k value for the adaptive state update, And it is possible to prevent the occurrence of performance deterioration and judgment error.
6 is a control flowchart illustrating a method for diagnosing a bearing failure according to another embodiment of the present invention.
As shown, another aspect of the present invention is a method comprising: A step of extracting features from a signal according to a bearing fault and learning the clustering algorithm (k-mean algorithm); Extracting features from an unknown signal of the bearing and classifying the current bearing state by the clustering algorithm; And a step C of setting the number of clusters (k) again by analyzing the clustering algorithm and the cluster distribution from the features extracted from the unknown signal of the bearing and learning the new number of clusters when the new state information is detected. A diagnostic method can be provided.
Another aspect of the present invention is to simplify a configuration by directly applying a clustering algorithm to a machine learning algorithm, thereby enabling a quick and accurate diagnosis of a fault state and preventing the occurrence of performance degradation and judgment errors.
The embodiments of the present invention described in the present specification and the configurations shown in the drawings relate to the most preferred embodiments of the present invention and are not intended to encompass all of the technical ideas of the present invention so that various equivalents It should be understood that water and variations may be present. Therefore, it is to be understood that the present invention is not limited to the above-described embodiments, and that various modifications may be made without departing from the spirit and scope of the invention as defined in the appended claims. , Such changes shall be within the scope of the claims set forth in the claims.
Claims (6)
A step B for extracting features from an unknown signal of a bearing and classifying a current bearing state by comparing the determined characteristics with previously learned characteristics by the machine learning algorithm;
Setting a cluster number (k) through a clustering algorithm and a cluster distribution analysis from features extracted from an unknown signal of the bearing, and learning new state information to the machine learning algorithm when new state information is detected,
In the step C,
Classifying the samples into k clusters through a clustering algorithm;
Calculating an average value and a covariance for each cluster;
Obtaining a probability density function of each cluster;
Calculating a local distribution factor (LDF) for each cluster using the probability density function and the covariance;
Calculating a global density factor (GDF) based on a minimum value of the LDFs of the clusters;
Calculating a global separability factor (GSF) using the distance between the clusters; And
Calculating a CDF (Cluster Distribution Factor)
Calculating CDF by calculating a CDF by calculating a difference between the GDF and the GSF, calculating a CDF by changing the value of k from 1 to a specific value and calculating the CDF, Value of the bearing failure.
Wherein the features extracted from the signals of the bearing defects and the features extracted from the unknown signals of the bearings are identical.
The features may include a root-mean-square, a shape factor, a kurtosis value, a square-mean-root, a peak-to-peak value , A skewness value, an impulse factor, and a crest factor, or a combination of features extracted from two or more signals.
The machine learning algorithm
Any of the algorithms that can be used for machine learning, including a support vector machine, an artificial neural network or a k-nearest neighbors classifier approach, Characterized in that the feature is learned and the current feature value is classified into a specific state.
Extracting features from an unknown signal of the bearing and classifying the current bearing state by the clustering algorithm;
(C) resetting the number of clusters (k) through analysis of the clustering algorithm and cluster distribution from features extracted from the unknown signal of the bearing, and learning new state information to the clustering algorithm when new state information is detected,
In the step C,
Classifying the samples into k clusters through a clustering algorithm;
Calculating an average value and a covariance for each cluster;
Obtaining a probability density function of each cluster;
Calculating a local distribution factor (LDF) for each cluster using the probability density function and the covariance;
Calculating a global density factor (GDF) based on a minimum value of the LDFs of the clusters;
Calculating a global separability factor (GSF) using the distance between the clusters; And
Calculating a CDF (Cluster Distribution Factor)
Calculating CDF by calculating a CDF by calculating a difference between the GDF and the GSF, calculating a CDF by changing the value of k from 1 to a specific value and calculating the CDF, Value of the bearing failure.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
KR1020160015240A KR101823746B1 (en) | 2016-02-05 | 2016-02-05 | Method for bearing fault diagnosis |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
KR1020160015240A KR101823746B1 (en) | 2016-02-05 | 2016-02-05 | Method for bearing fault diagnosis |
Publications (2)
Publication Number | Publication Date |
---|---|
KR20170093613A KR20170093613A (en) | 2017-08-16 |
KR101823746B1 true KR101823746B1 (en) | 2018-01-30 |
Family
ID=59752506
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
KR1020160015240A KR101823746B1 (en) | 2016-02-05 | 2016-02-05 | Method for bearing fault diagnosis |
Country Status (1)
Country | Link |
---|---|
KR (1) | KR101823746B1 (en) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR102123522B1 (en) * | 2019-12-16 | 2020-06-16 | 주식회사 한국가스기술공사 | Failure diagnostic method based on cluster of fault data |
KR102141391B1 (en) * | 2019-12-16 | 2020-08-05 | 주식회사 한국가스기술공사 | Failure data management method based on cluster estimation |
KR20230083810A (en) | 2021-12-03 | 2023-06-12 | 건국대학교 산학협력단 | Method for building predictive model of roll bearing life in the roll-to-roll process |
KR20230134850A (en) | 2022-03-15 | 2023-09-22 | 한국과학기술원 | Current Data Imaging Method and Apparatus for Rotating Machinery Fault Diagnosis |
KR20230149109A (en) | 2022-04-19 | 2023-10-26 | 단국대학교 산학협력단 | Bearing Fault Diagnosis Device and Diagnosis Method Using First Order Deadbeat Observer |
Families Citing this family (29)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108267312B (en) * | 2017-12-25 | 2019-10-29 | 北京建筑大学 | A kind of subway train bearing intelligent diagnostic method based on fast search algorithm |
CN108398266B (en) * | 2018-01-22 | 2020-06-23 | 武汉科技大学 | Bearing fault diagnosis method based on integrated transfer learning |
KR102501883B1 (en) * | 2018-07-06 | 2023-02-21 | 에임시스템 주식회사 | System and method for fault classification of equipment based on machine learning |
CN109190464B (en) * | 2018-07-24 | 2021-09-17 | 南京航空航天大学 | Mechanical fault intelligent diagnosis method based on transfer learning under variable working conditions |
CN109506936B (en) * | 2018-11-05 | 2020-12-22 | 哈尔滨理工大学 | Bearing fault degree identification method based on flow chart and non-naive Bayes inference |
CN109253883A (en) * | 2018-11-12 | 2019-01-22 | 广西交通科学研究院有限公司 | A kind of rotating machinery rolling bearing intelligent diagnosing method based on incremental search cluster |
WO2020138549A1 (en) * | 2018-12-27 | 2020-07-02 | 울산대학교 산학협력단 | Acoustic emission-based defect signal detection device and method |
KR102198190B1 (en) * | 2018-12-31 | 2021-01-04 | 주식회사 원프레딕트 | Data standardization method considering operating contion for diagnosis of rotating machinery failure and diagnosis method rotating machinery failure using the same |
KR102166649B1 (en) * | 2019-01-30 | 2020-10-16 | 한국해양대학교 산학협력단 | Machine Diagnosis and Prediction System using Machine Learning |
CN109782603A (en) * | 2019-02-03 | 2019-05-21 | 中国石油大学(华东) | The detection method and monitoring system of rotating machinery coupling fault |
CN109827776B (en) * | 2019-03-15 | 2024-02-13 | 合肥工业大学 | Bearing fault detection method and system |
CN109827777B (en) * | 2019-04-01 | 2020-12-18 | 哈尔滨理工大学 | Rolling bearing fault prediction method based on partial least square method extreme learning machine |
KR102295879B1 (en) * | 2019-04-25 | 2021-08-31 | 울산대학교 산학협력단 | Method and System for Real-time Leakage State Recognition of Tubes |
KR102278702B1 (en) * | 2019-09-27 | 2021-07-16 | 광주과학기술원 | Method for selecting sensor signal features based on statistical indicator sensitive to outlier |
KR102360004B1 (en) * | 2019-11-15 | 2022-02-15 | 한국타이어앤테크놀로지 주식회사 | Management system of machine based on a vibration |
KR102230463B1 (en) | 2019-12-12 | 2021-03-19 | 울산대학교 산학협력단 | Diagnosis system and method of defect of equipment component |
KR102348387B1 (en) * | 2020-05-18 | 2022-01-06 | 한국항공대학교산학협력단 | Rotating body health evaluation system and method thereof |
JP6837612B1 (en) * | 2020-05-28 | 2021-03-03 | 三菱電機株式会社 | Equipment status monitoring device and equipment status monitoring method |
CN111811818B (en) * | 2020-06-02 | 2022-02-01 | 桂林电子科技大学 | Rolling bearing fault diagnosis method based on AP clustering algorithm of specified clustering number |
CN111898556B (en) * | 2020-08-01 | 2024-04-16 | 华东交通大学 | CK index consistency-based bearing transient impact feature extraction method |
CN112945556B (en) * | 2021-01-26 | 2022-07-05 | 大连海事大学 | Bearing fault diagnosis method based on wavelet packet decomposition and optimal rejection classification strategy |
CN113295413B (en) * | 2021-06-24 | 2022-10-25 | 北京交通大学 | Traction motor bearing fault diagnosis method based on indirect signals |
CN113790892B (en) * | 2021-09-13 | 2024-01-23 | 哈电发电设备国家工程研究中心有限公司 | Decision-stage fusion-based tilting-pad bearing pad fault diagnosis method for heavy-duty gas turbine, computer and storage medium |
CN114647992A (en) * | 2022-03-10 | 2022-06-21 | 北京航空航天大学 | Solid-liquid rocket engine fault diagnosis method based on improved Bayesian algorithm |
CN114323644B (en) * | 2022-03-14 | 2022-06-03 | 中国长江三峡集团有限公司 | Gear box fault diagnosis and signal acquisition method and device and electronic equipment |
CN114743039B (en) * | 2022-05-17 | 2024-02-13 | 合肥工业大学 | Fuzzy clustering bearing fault detection method based on feature reduction |
CN115898925B (en) * | 2022-10-27 | 2024-06-04 | 华能国际电力股份有限公司上海石洞口第二电厂 | Fan fault early warning method based on vibration signal multi-order moment |
CN116933170B (en) * | 2023-09-18 | 2024-01-02 | 福建福清核电有限公司 | Mechanical seal fault classification method |
CN117664576B (en) * | 2023-11-29 | 2024-06-21 | 东北大学 | Bearing health assessment method based on improved optimizer deep learning model |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2006039658A (en) * | 2004-07-22 | 2006-02-09 | Hitachi Software Eng Co Ltd | Image classification learning processing system and image identification processing system |
JP2014032455A (en) * | 2012-08-01 | 2014-02-20 | Hitachi Power Solutions Co Ltd | Equipment condition monitoring method and device thereof |
-
2016
- 2016-02-05 KR KR1020160015240A patent/KR101823746B1/en active IP Right Grant
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2006039658A (en) * | 2004-07-22 | 2006-02-09 | Hitachi Software Eng Co Ltd | Image classification learning processing system and image identification processing system |
JP2014032455A (en) * | 2012-08-01 | 2014-02-20 | Hitachi Power Solutions Co Ltd | Equipment condition monitoring method and device thereof |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR102123522B1 (en) * | 2019-12-16 | 2020-06-16 | 주식회사 한국가스기술공사 | Failure diagnostic method based on cluster of fault data |
KR102141391B1 (en) * | 2019-12-16 | 2020-08-05 | 주식회사 한국가스기술공사 | Failure data management method based on cluster estimation |
KR20230083810A (en) | 2021-12-03 | 2023-06-12 | 건국대학교 산학협력단 | Method for building predictive model of roll bearing life in the roll-to-roll process |
KR20230134850A (en) | 2022-03-15 | 2023-09-22 | 한국과학기술원 | Current Data Imaging Method and Apparatus for Rotating Machinery Fault Diagnosis |
KR20230149109A (en) | 2022-04-19 | 2023-10-26 | 단국대학교 산학협력단 | Bearing Fault Diagnosis Device and Diagnosis Method Using First Order Deadbeat Observer |
Also Published As
Publication number | Publication date |
---|---|
KR20170093613A (en) | 2017-08-16 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
KR101823746B1 (en) | Method for bearing fault diagnosis | |
KR101797400B1 (en) | Method and apparatus for diagnosing fault based on probabilistic density | |
US7930122B2 (en) | Evaluating anomaly for one-class classifiers in machine condition monitoring | |
CN110598851A (en) | Time series data abnormity detection method fusing LSTM and GAN | |
CN106528975B (en) | A kind of prognostic and health management method applied to Circuits and Systems | |
US20150219530A1 (en) | Systems and methods for event detection and diagnosis | |
CN109726730B (en) | Automatic optical detection image classification method, system and computer readable medium | |
CN107729985B (en) | Method for detecting process anomalies in a technical installation and corresponding diagnostic system | |
KR102362159B1 (en) | Method for Fault Detection and Fault Diagnosis in Semiconductor Manufacturing Process | |
KR102398046B1 (en) | Fault diagnosis device using unsupervised domain adaptation technique and fault diagnosis method using the same | |
CN102589884A (en) | Method for diagnosing failure of airplane generator bearing based on GentleBoost | |
CN108304348B (en) | Bearing residual life prediction method based on binary wiener process | |
JP2008090529A (en) | Abnormality detection device, abnormality detection method | |
KR102265298B1 (en) | Apparatus and method for fault diagnosis using fake data generated by machine learning | |
KR101808390B1 (en) | Method for machine fault diagnosis | |
Zhang et al. | Few-shot bearing anomaly detection via model-agnostic meta-learning | |
CN111506049A (en) | Multiple fault diagnosis method for aero-engine control system based on AANN network system | |
RU2445598C1 (en) | Diagnostic method of technical state of gas-turbine engine | |
TWI755468B (en) | Diagnostic methods for the classifiers and the defects captured by optical tools | |
CN114356897A (en) | Electromechanical actuator fault diagnosis method based on data fusion | |
US8271233B2 (en) | Method of multi-level fault isolation design | |
CN112434755B (en) | Data anomaly sensing method based on heterogeneous system | |
US7814034B2 (en) | Method and system for automatically developing a fault classification system by segregation of kernels in time series data | |
KR101811962B1 (en) | Apparatus and method for evaluating class discrimination of nonlinear data | |
KR20200053254A (en) | Method and device of detecting sensor fault |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
A201 | Request for examination | ||
E902 | Notification of reason for refusal | ||
E90F | Notification of reason for final refusal | ||
E701 | Decision to grant or registration of patent right | ||
GRNT | Written decision to grant |