CN110672324B - Bearing fault diagnosis method and device based on supervised LLE algorithm - Google Patents
Bearing fault diagnosis method and device based on supervised LLE algorithm Download PDFInfo
- Publication number
- CN110672324B CN110672324B CN201910824743.3A CN201910824743A CN110672324B CN 110672324 B CN110672324 B CN 110672324B CN 201910824743 A CN201910824743 A CN 201910824743A CN 110672324 B CN110672324 B CN 110672324B
- Authority
- CN
- China
- Prior art keywords
- fault
- training data
- dimension
- dimension reduction
- data
- 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
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
- G01M13/045—Acoustic or vibration analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/16—Matrix or vector computation, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Computational Mathematics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Computing Systems (AREA)
- Acoustics & Sound (AREA)
- Algebra (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
Abstract
The invention relates to the technical field of fault diagnosis, in particular to a bearing fault diagnosis method and device based on a supervised LLE algorithm.
Description
Technical Field
The invention relates to the technical field of fault diagnosis, in particular to a bearing fault diagnosis method and device based on a supervised LLE algorithm.
Background
As an emerging comprehensive marginal discipline, the bearing fault diagnosis technology initially forms a relatively complete discipline system. As for the technical means, the vibration diagnosis technology has become the mainstream technology of bearing fault diagnosis. The rapid progress of the computer technology and the signal information processing technology greatly promotes the development of the bearing fault diagnosis and monitoring technology towards the direction of scientification and practicability.
However, in the field of current bearing fault diagnosis, large-scale data concurrency often exists, great challenges are brought to the real-time requirement of fault diagnosis, and the online prediction rate of bearing fault diagnosis needs to be improved urgently.
Disclosure of Invention
The invention aims to provide a bearing fault diagnosis method and device based on a supervised LLE algorithm, aiming at improving the online prediction rate of bearing fault diagnosis.
In order to achieve the purpose, the invention provides the following technical scheme:
a bearing fault diagnosis method based on a supervised LLE algorithm comprises the following steps:
acquiring training data, wherein the training data are historical data representing bearing vibration signals, and extracting characteristic values of the training data and fault types corresponding to the characteristic values;
determining optimal dimension reduction training data of the training data, wherein the optimal dimension reduction training data has the largest ratio of inter-class dispersion and intra-class dispersion of all fault types;
calculating a mean value and a covariance matrix corresponding to each fault type in the optimized dimension reduction training data;
performing dimension reduction on the test data received in real time to obtain dimension reduction test data;
and calculating probability values of the dimension reduction data under all fault types according to the mean value and the covariance matrix, and taking the fault type with the maximum probability value as a fault type of bearing fault diagnosis.
Further, the characteristic values comprise vibration displacement, vibration speed, vibration acceleration and high-frequency acceleration, and the fault types comprise wear failure, fatigue failure and corrosion failure.
Further, the determining the preferred dimension-reduced training data of the training data includes:
performing dimensionality reduction on the training data by using an LLE algorithm to obtain dimensionality reduction training data, and determining the preferred neighbor number and the preferred fault dimensionality of the dimensionality reduction training data;
and taking the dimension reduction training data corresponding to the preferred neighbor number and the preferred fault dimension as preferred dimension reduction training data.
Further, the performing dimension reduction on the training data by using the LLE algorithm to obtain dimension reduction training data, and determining a preferred neighbor number and a preferred fault dimension of the dimension reduction training data includes:
step 310, setting a value range of the neighbor number p and a value range of the fault dimension q;
step 320, selecting a p value and a q value as a parameter group, and forming a parameter set by all the parameter groups, wherein the parameter set comprises all combination forms of the p value and the q value;
step 330, selecting a parameter set in turn as the neighbor number p and the fault dimension q of the training data;
step 340, performing dimension reduction on the training sample data obtained in the step 330 by using an LLE algorithm to obtain a data set Y and a fault set phi after dimension reduction, wherein the data set Y is { Y ═ Y1，y2，...，yNY is an N multiplied by m matrix, N is the number of samples, and m is the dimension of the fault; set of faults
s is the total number of fault categories;
step 350, calculating an evaluation index F by using the dimensionality reduced data set and the fault set, specifically:
the mean vector c for each fault category is calculated by the following formulai：
Calculating the in-class dispersion matrix S of all classes by the following formulai：
Summing all the intra-class dispersion matrices to obtain a mixed intra-class dispersion matrix Sw:
Sw＝S1+S2+...+Ss；
calculating an inter-class dispersion matrix by the following formula:
the evaluation index F was calculated by the following formula:
F＝Sb/Sw；
step 360, judging whether all the parameter groups in the parameter set calculate evaluation indexes, if not, skipping to step 330, and if so, executing the following steps;
step 370 compares the magnitudes of the evaluation indexes in the parameter groups, selects the parameter group with the largest evaluation index as the preferred parameter group, uses the p value of the parameter group as the preferred neighbor number, and uses the q value of the parameter group as the preferred failure dimension.
Further, the performing dimension reduction on the test data received in real time to obtain dimension reduction test data includes:
and taking the preferred neighbor number as the neighbor number of the test data, taking the preferred fault dimension as the fault dimension of the test data, and performing dimension reduction on the test data by using an LLE algorithm to obtain dimension reduction test data.
A supervised LLE algorithm based bearing fault diagnosis apparatus, the apparatus comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, the processor executing the computer program to operate in modules of:
the extraction module is used for acquiring training data, wherein the training data are historical data representing bearing vibration signals, and extracting characteristic values of the training data and fault types corresponding to the characteristic values;
the determining module is used for determining the optimal dimension reduction training data of the training data, wherein in the optimal dimension reduction training data, the ratio of the inter-class dispersion and the intra-class dispersion of all fault types is the largest;
the calculation module is used for calculating a mean value and a covariance matrix corresponding to each fault type in the optimal dimension reduction training data;
the dimension reduction module is used for reducing the dimension of the test data received in real time to obtain dimension reduction test data;
and the diagnosis module is used for calculating the probability values of the dimension reduction data under each fault type according to the mean value and the covariance matrix, and taking the fault type with the maximum probability value as the fault type of the bearing fault diagnosis.
Further, the characteristic values comprise vibration displacement, vibration speed, vibration acceleration and high-frequency acceleration, and the fault types comprise wear failure, fatigue failure and corrosion failure.
Further, the determining module is specifically configured to:
performing dimensionality reduction on the training data by using an LLE algorithm to obtain dimensionality reduction training data, and determining the preferred neighbor number and the preferred fault dimensionality of the dimensionality reduction training data;
and taking the dimension reduction training data corresponding to the preferred neighbor number and the preferred fault dimension as preferred dimension reduction training data.
The invention has the beneficial effects that: the invention discloses a bearing fault diagnosis method and device based on a supervised LLE algorithm, which comprises the steps of firstly obtaining training data, wherein the training data are historical data representing bearing vibration signals, extracting characteristic values of the training data and fault types corresponding to the characteristic values, then determining optimal dimension reduction training data of the training data, further calculating a mean value and a covariance matrix corresponding to each fault type in the optimal dimension reduction training data, obtaining dimension reduction test data by reducing dimensions of the test data received in real time, calculating probability values of the dimension reduction data under each fault type according to the mean value and the covariance matrix, and taking the fault type with the maximum probability value as the fault type of bearing fault diagnosis. The invention improves the online prediction rate of bearing fault diagnosis.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings needed to be used in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings without inventive exercise.
FIG. 1 is a schematic flow chart of a bearing fault diagnosis method based on a supervised LLE algorithm according to an embodiment of the present invention;
FIG. 2 is a flowchart illustrating step S200 according to an embodiment of the present invention;
FIG. 3 is a flowchart illustrating step S210 according to an embodiment of the present invention;
fig. 4 is a schematic structural diagram of a bearing fault diagnosis device based on a supervised LLE algorithm according to an embodiment of the present invention.
Detailed Description
The technical solutions of the present invention will be described clearly and completely with reference to the accompanying drawings, and it should be understood that the described embodiments are some, but not all embodiments of the present invention. Based on the embodiments of the present invention, other embodiments obtained by a person of ordinary skill in the art without any creative effort belong to the protection scope of the present invention.
Referring to fig. 1, a bearing fault diagnosis method based on a supervised LLE algorithm provided by an embodiment of the present invention includes the following steps:
s100, acquiring training data, wherein the training data are historical data representing bearing vibration signals, and extracting characteristic values of the training data and fault types corresponding to the characteristic values;
s200, determining optimal dimension reduction training data of the training data, wherein the optimal dimension reduction training data has the largest ratio of inter-class dispersion and intra-class dispersion of all fault types;
step S300, calculating a mean value and a covariance matrix corresponding to each fault type in the optimized dimension reduction training data;
s400, performing dimension reduction on the test data received in real time to obtain dimension reduction test data;
and S500, calculating probability values of the dimension reduction data under all fault types according to the mean value and the covariance matrix, and taking the fault type with the maximum probability value as a fault type of bearing fault diagnosis.
The embodiment realizes bearing fault diagnosis by using a supervised dimension reduction method. By training the training data, the characteristic values and the fault types in the high-dimensional data are extracted, so that the training data have very good discrimination in a low-dimensional space, the parameter types required to be stored in the embodiment are fewer, the prediction speed is higher, and the method is suitable for online prediction.
In one embodiment, the characteristic values include vibration displacement (peak-to-peak value), vibration velocity (true effective value), vibration acceleration (peak value), high frequency acceleration, and the fault types include wear failure, fatigue failure, corrosion failure.
Referring to fig. 2, as a further modification of the present embodiment, the step S200 includes the steps of:
step S210, performing dimensionality reduction on the training data by using an LLE algorithm to obtain dimensionality reduction training data, and determining the preferred neighbor number and the preferred fault dimensionality of the dimensionality reduction training data;
and S220, taking the dimension reduction training data corresponding to the preferred neighbor number and the preferred fault dimension as preferred dimension reduction training data.
Referring to fig. 3, as a further improvement of the present embodiment, the step S210 includes:
and S211, setting the value range of the neighbor number p and the value range of the fault dimensionality q.
In this embodiment, p neighboring points of each training data in the dimension-reduced training data need to be found. The p training data points with the nearest Euclidean distance in each training data point are found out, and p is the so-called neighbor number.
The value range of the neighbor number p and the value range of the fault dimension q can be set manually according to historical records or diagnosis requirements of bearing faults, the larger the value ranges of p and q are, the longer the training time is, the more comprehensive the diagnosis is, and the smaller the value ranges of p and q are, the shorter the training time is.
The neighbor number p is the first important parameter in the LLE algorithm. The LLE algorithm assumes that each training data point is locally linear, i.e., each training data point can be expressed by a linear combination of its neighboring points, and the neighboring relationship between the training data is maintained in the process of mapping from the high dimension to the low dimension. The too large value of p makes the local linear range too large, and the local characteristics of the LLE algorithm cannot be well embodied. When the value of p is too small, the LLE algorithm can hardly ensure the topological structure of the training data in the low-dimensional space.
The fault dimension q is the second important parameter in the LLE algorithm, and if the value of the fault dimension q is too large, too much redundancy will be included in the training data after dimensionality reduction, whereas if the value of the fault dimension q is too small, the training data separated from each other in the high-dimensional space are overlapped in the low-dimensional space.
Step S212, selecting a p value and a q value as a parameter group, and forming a parameter set by all the parameter groups, wherein the parameter set comprises all the combination forms of the p value and the q value.
And step S213, selecting a parameter group in sequence as the neighbor number p and the fault dimension q of the training data.
And S214, performing dimensionality reduction on the training sample data obtained in the step S213 by using an LLE algorithm to obtain a dimensionality-reduced data set and a dimensionality-reduced fault set.
Wherein the data set Y ═ { Y ═ Y1，y2，...，yNY is an N multiplied by m matrix, N is the number of samples, and m is the dimension of the fault; set of faults
s is the total number of fault categories;
training the classifier again by using the training sample data after dimensionality reduction, and directly determining the fault type of the test data through sample distribution of different fault type data.
And S215, calculating an evaluation index F by using the data set and the fault set after the dimensionality reduction.
The method specifically comprises the following steps:
the mean vector c for each fault category is calculated by the following formulai：
Calculating the in-class dispersion matrix S of all classes by the following formulai：
Summing all the intra-class dispersion matrices to obtain a mixed intra-class dispersion matrix Sw:
Sw＝S1+S2+...+Ss；
calculating an inter-class dispersion matrix by the following formula:
the evaluation index F was calculated by the following formula:
F＝Sb/Sw；
and S216, judging whether all the parameter groups in the parameter set calculate evaluation indexes, if not, jumping to S213, and if so, executing the following steps.
Step S217 compares the magnitudes of the evaluation indexes in the parameter groups, selects the parameter group with the largest evaluation index as the preferred parameter group, uses the p value of the parameter group as the preferred neighbor number, and uses the q value of the parameter group as the preferred failure dimension.
The preferred set of parameters is selected to maximize the inter-class spacing between data of different fault types and to minimize the intra-class spacing between data of different fault types.
The LLE algorithm is a typical unsupervised learning method, and in the embodiment, the inter-class distance of different fault category data is maximized and the intra-class distance is minimized by traversing the neighbor number p and the fault dimension q. Therefore, the feature selection after dimension reduction is guided through the known feature value and fault type, and the supervised LLE algorithm is realized. Compared with the traditional bearing fault diagnosis method, the method has the advantages that the parameter types needing to be stored are fewer, the prediction speed is higher, and the method is suitable for online prediction.
As a further improvement of this embodiment, the step S400 includes:
and taking the preferred neighbor number as the neighbor number of the test data, taking the preferred fault dimension as the fault dimension of the test data, and performing dimension reduction on the test data by using an LLE algorithm to obtain dimension reduction test data.
Referring to fig. 4, the present embodiment also provides a bearing fault diagnosis apparatus based on a supervised LLE algorithm, the apparatus including: a memory, a processor, and a computer program stored in the memory and executable on the processor, the processor executing the computer program to operate in modules of:
the extraction module 100 is configured to acquire training data, where the training data is historical data representing a bearing vibration signal, and extract a feature value of the training data and a fault type corresponding to the feature value;
a determining module 200, configured to determine preferred dimension-reduced training data of the training data, where in the preferred dimension-reduced training data, a ratio of inter-class dispersion to intra-class dispersion of all fault types is the largest;
a calculating module 300, configured to calculate a mean value and a covariance matrix corresponding to each fault type in the preferred dimension reduction training data;
the dimension reduction module 400 is used for performing dimension reduction on the test data received in real time to obtain dimension reduction test data;
and the diagnosis module 500 is configured to calculate probability values of the dimension reduction data under each fault type according to the mean value and the covariance matrix, and use the fault type with the maximum probability value as the fault type of the bearing fault diagnosis.
As a further improvement of the embodiment, the characteristic values include vibration displacement, vibration speed, vibration acceleration, and high-frequency acceleration, and the failure types include wear failure, fatigue failure, and corrosion failure.
As a further improvement of this embodiment, the determining module 200 is specifically configured to:
performing dimensionality reduction on the training data by using an LLE algorithm to obtain dimensionality reduction training data, and determining the preferred neighbor number and the preferred fault dimensionality of the dimensionality reduction training data;
and taking the dimension reduction training data corresponding to the preferred neighbor number and the preferred fault dimension as preferred dimension reduction training data.
The bearing fault diagnosis device based on the supervised LLE algorithm can be operated in computing equipment such as desktop computers, mobile phones, notebooks, tablet computers, cloud servers and the like. The bearing fault diagnosis device based on the supervised LLE algorithm can be operated by a system comprising but not limited to a processor and a memory. It will be understood by those skilled in the art that the example is merely an example of a supervised LLE algorithm based bearing fault diagnosis apparatus, and does not constitute a limitation of a supervised LLE algorithm based bearing fault diagnosis apparatus, and may include more or less components than a certain proportion, or combine certain components, or different components, for example, the supervised LLE algorithm based bearing fault diagnosis apparatus may further include an input-output device, a network access device, a bus, etc.
The Processor may be a Central-Processing Unit (CPU), other general-purpose Processor, a Digital Signal Processor (DSP), an Application-Specific-Integrated-Circuit (ASIC), an off-the-shelf Programmable Gate Array (FPGA) or other Programmable logic device, a discrete Gate or transistor logic device, a discrete hardware component, or the like. The general purpose processor may be a microprocessor or the processor may be any conventional processor or the like, the processor is a control center of the operation system of the bearing fault diagnosis device based on the supervised LLE algorithm, and various interfaces and lines are used for connecting various parts of the whole operation system of the bearing fault diagnosis device based on the supervised LLE algorithm.
The memory can be used for storing the computer program and/or the module, and the processor can realize various functions of the bearing fault diagnosis device based on the supervised LLE algorithm by operating or executing the computer program and/or the module stored in the memory and calling the data stored in the memory. The memory may mainly include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required by at least one function (such as a sound playing function, an image playing function, etc.), and the like; the storage data area may store data (such as audio data, a phonebook, etc.) created according to the use of the cellular phone, and the like. In addition, the memory may include high speed random access memory, and may also include non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a Smart-Media-Card (SMC), a Secure-Digital (SD) Card, a Flash-memory Card (Flash-Card), at least one magnetic disk storage device, a Flash memory device, or other volatile solid state storage device.
While the present disclosure has been described in considerable detail and with particular reference to a few illustrative embodiments thereof, it is not intended to be limited to any such details or embodiments or any particular embodiments, but it is to be construed with references to the appended claims so as to provide a broad, possibly open interpretation of such claims in view of the prior art, and to effectively encompass the intended scope of the disclosure. Furthermore, the foregoing describes the disclosure in terms of embodiments foreseen by the inventor for which an enabling description was available, notwithstanding that insubstantial modifications of the disclosure, not presently foreseen, may nonetheless represent equivalent modifications thereto.
Claims (5)
1. A bearing fault diagnosis method based on a supervised LLE algorithm is characterized by comprising the following steps:
acquiring training data, wherein the training data are historical data representing bearing vibration signals, and extracting characteristic values of the training data and fault types corresponding to the characteristic values;
determining optimal dimension reduction training data of the training data, wherein the optimal dimension reduction training data has the largest ratio of inter-class dispersion and intra-class dispersion of all fault types;
calculating a mean value and a covariance matrix corresponding to each fault type in the optimized dimension reduction training data;
performing dimension reduction on the test data received in real time to obtain dimension reduction test data;
calculating probability values of the dimension reduction test data under each fault type according to the mean value and the covariance matrix, and taking the fault type with the maximum probability value as a fault type of bearing fault diagnosis;
wherein the determining of the preferred dimension-reduced training data of the training data comprises:
performing dimensionality reduction on the training data by using an LLE algorithm to obtain dimensionality reduction training data, and determining the preferred neighbor number and the preferred fault dimensionality of the dimensionality reduction training data;
using the dimension reduction training data corresponding to the preferred neighbor number and the preferred fault dimension as preferred dimension reduction training data;
the using the LLE algorithm to perform dimension reduction on the training data to obtain dimension reduction training data, and determining a preferred neighbor number and a preferred fault dimension of the dimension reduction training data includes:
step 310, setting a value range of the neighbor number p and a value range of the fault dimension q;
step 320, selecting a p value and a q value as a parameter group, and forming a parameter set by all the parameter groups, wherein the parameter set comprises all combination forms of the p value and the q value;
step 330, selecting a parameter set in turn as the neighbor number p and the fault dimension q of the training data;
step 340, performing dimension reduction on the training sample data obtained in the step 330 by using an LLE algorithm to obtain a data set Y and a fault set phi after dimension reduction, wherein the data set Y is { Y ═ Y1，y2，...，yNY is an N multiplied by m matrix, N is the number of samples, and m is the dimension of the fault after dimension reduction; set of faults
s is the total number of fault categories;
step 350, calculating an evaluation index F by using the dimensionality reduced data set and the fault set, specifically:
the mean vector c for each fault category is calculated by the following formulai：
Through the following disclosureCalculating the in-class dispersion matrix S of all classesi：
Summing all the intra-class dispersion matrices to obtain a mixed intra-class dispersion matrix Sw:
Sw＝S1+S2+...+Ss；
calculating an inter-class dispersion matrix by the following formula:
the evaluation index F was calculated by the following formula:
F＝Sb/Sw；
step 360, judging whether all the parameter groups in the parameter set calculate evaluation indexes, if not, skipping to step 330, and if so, executing the following steps;
step 370 compares the magnitudes of the evaluation indexes in the parameter groups, selects the parameter group with the largest evaluation index as the preferred parameter group, uses the p value of the parameter group as the preferred neighbor number, and uses the q value of the parameter group as the preferred failure dimension.
2. The supervised LLE algorithm based bearing fault diagnosis method as recited in claim 1, wherein the characteristic values comprise vibration displacement, vibration velocity, vibration acceleration, high frequency acceleration, and the fault types comprise wear failure, fatigue failure, and corrosion failure.
3. The method as claimed in claim 2, wherein the step of performing dimension reduction on the test data received in real time to obtain dimension-reduced test data comprises:
receiving test data in real time, taking the preferred neighbor number as the neighbor number of the test data, taking the preferred fault dimension as the fault dimension of the test data, and performing dimension reduction on the test data by using an LLE algorithm to obtain dimension reduction test data.
4. A supervised LLE algorithm based bearing fault diagnosis device, the device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, the processor executing the computer program to operate in modules of:
the extraction module is used for acquiring training data, wherein the training data are historical data representing bearing vibration signals, and extracting characteristic values of the training data and fault types corresponding to the characteristic values;
the determining module is used for determining the optimal dimension reduction training data of the training data, wherein in the optimal dimension reduction training data, the ratio of the inter-class dispersion and the intra-class dispersion of all fault types is the largest;
the calculation module is used for calculating a mean value and a covariance matrix corresponding to each fault type in the optimal dimension reduction training data;
the dimension reduction module is used for reducing the dimension of the test data received in real time to obtain dimension reduction test data;
the diagnosis module is used for calculating probability values of the dimension reduction test data under each fault type according to the mean value and the covariance matrix, and taking the fault type with the maximum probability value as the fault type of bearing fault diagnosis;
wherein the determining module is specifically configured to:
performing dimensionality reduction on the training data by using an LLE algorithm to obtain dimensionality reduction training data, and determining the preferred neighbor number and the preferred fault dimensionality of the dimensionality reduction training data;
using the dimension reduction training data corresponding to the preferred neighbor number and the preferred fault dimension as preferred dimension reduction training data;
wherein the determining module is further specifically configured to:
step 310, setting a value range of the neighbor number p and a value range of the fault dimension q;
step 320, selecting a p value and a q value as a parameter group, and forming a parameter set by all the parameter groups, wherein the parameter set comprises all combination forms of the p value and the q value;
step 330, selecting a parameter set in turn as the neighbor number p and the fault dimension q of the training data;
step 340, performing dimension reduction on the training sample data obtained in the step 330 by using an LLE algorithm to obtain a data set Y and a fault set phi after dimension reduction, wherein the data set Y is { Y ═ Y1，y2，...，yNY is an N multiplied by m matrix, N is the number of samples, and m is the dimension of the fault after dimension reduction; set of faults
s is the total number of fault categories;
step 350, calculating an evaluation index F by using the dimensionality reduced data set and the fault set, specifically:
the mean vector c for each fault category is calculated by the following formulai：
Calculating the in-class dispersion matrix S of all classes by the following formulai：
Summing all the intra-class dispersion matrices to obtain a mixed intra-class dispersion matrix Sw:
Sw＝S1+S2+...+Ss；
calculating an inter-class dispersion matrix by the following formula:
the evaluation index F was calculated by the following formula:
F＝Sb/Sw；
step 360, judging whether all the parameter groups in the parameter set calculate evaluation indexes, if not, skipping to step 330, and if so, executing the following steps;
step 370 compares the magnitudes of the evaluation indexes in the parameter groups, selects the parameter group with the largest evaluation index as the preferred parameter group, uses the p value of the parameter group as the preferred neighbor number, and uses the q value of the parameter group as the preferred failure dimension.
5. The supervised LLE algorithm based bearing fault diagnosis device according to claim 4, wherein said characteristic values include vibrational displacement, vibrational velocity, vibrational acceleration, high frequency acceleration, and said fault types include wear failure, fatigue failure, corrosion failure.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910824743.3A CN110672324B (en) | 2019-09-02 | 2019-09-02 | Bearing fault diagnosis method and device based on supervised LLE algorithm |
| PCT/CN2020/087799 WO2021042749A1 (en) | 2019-09-02 | 2020-04-29 | Bearing fault diagnosis method and apparatus based on supervised lle algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910824743.3A CN110672324B (en) | 2019-09-02 | 2019-09-02 | Bearing fault diagnosis method and device based on supervised LLE algorithm |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110672324A CN110672324A (en) | 2020-01-10 |
| CN110672324B true CN110672324B (en) | 2021-03-26 |
Family
ID=69075909
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910824743.3A Active CN110672324B (en) | 2019-09-02 | 2019-09-02 | Bearing fault diagnosis method and device based on supervised LLE algorithm |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN110672324B (en) |
| WO (1) | WO2021042749A1 (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110672324B (en) * | 2019-09-02 | 2021-03-26 | 佛山科学技术学院 | Bearing fault diagnosis method and device based on supervised LLE algorithm |
| CN111896254A (en) * | 2020-08-10 | 2020-11-06 | 山东大学 | Fault prediction system and method for variable-speed variable-load large rolling bearing |
| CN114239867A (en) * | 2021-11-08 | 2022-03-25 | 深圳供电局有限公司 | Fault type determination method and device, computer equipment and readable storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH04254013A (en) * | 1990-12-27 | 1992-09-09 | Yaskawa Electric Corp | Bearing breakage preventing method |
| CN105760839A (en) * | 2016-02-22 | 2016-07-13 | 重庆大学 | Bearing fault diagnosis method based on multi-feature manifold learning and support vector machine |
| CN109582003A (en) * | 2018-12-03 | 2019-04-05 | 东北林业大学 | Based on pseudo label semi-supervised kernel part Fei Sheer discriminant analysis bearing failure diagnosis |
| CN109827777A (en) * | 2019-04-01 | 2019-05-31 | 哈尔滨理工大学 | Rolling bearing fault prediction technique based on Partial Least Squares extreme learning machine |
| CN110031226A (en) * | 2019-04-12 | 2019-07-19 | 佛山科学技术学院 | A kind of diagnostic method and device of bearing fault |
Family Cites Families (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP4236022B2 (en) * | 2000-06-16 | 2009-03-11 | Ｎｔｎ株式会社 | Monitoring, diagnosis and sales system for machine parts |
| CN102930285A (en) * | 2012-09-18 | 2013-02-13 | 四川大学 | Early failure identification method based on SILLE (Supervised Increment Locally Linear Embedding) dimensionality reduction |
| CN105043766B (en) * | 2015-06-24 | 2017-07-07 | 北京航空航天大学 | A kind of bearing variable working condition method for diagnosing faults being locally linear embedding into based on Hessian |
| CN105973594B (en) * | 2016-04-25 | 2018-05-29 | 西北工业大学 | A kind of rolling bearing fault Forecasting Methodology based on continuous depth confidence network |
| CN107036817B (en) * | 2017-04-05 | 2019-03-08 | 哈尔滨理工大学 | SVR rolling bearing performance decline prediction technique based on krill group's algorithm |
| CN107024352A (en) * | 2017-05-03 | 2017-08-08 | 哈尔滨理工大学 | A kind of Rolling Bearing Fault Character extracting method based on slip entropy ICA algorithm |
| CN110163261B (en) * | 2019-04-28 | 2024-06-21 | 平安科技（深圳）有限公司 | Unbalanced data classification model training method, device, equipment and storage medium |
| CN110672324B (en) * | 2019-09-02 | 2021-03-26 | 佛山科学技术学院 | Bearing fault diagnosis method and device based on supervised LLE algorithm |
-
2019
- 2019-09-02 CN CN201910824743.3A patent/CN110672324B/en active Active
-
2020
- 2020-04-29 WO PCT/CN2020/087799 patent/WO2021042749A1/en active Application Filing
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH04254013A (en) * | 1990-12-27 | 1992-09-09 | Yaskawa Electric Corp | Bearing breakage preventing method |
| CN105760839A (en) * | 2016-02-22 | 2016-07-13 | 重庆大学 | Bearing fault diagnosis method based on multi-feature manifold learning and support vector machine |
| CN109582003A (en) * | 2018-12-03 | 2019-04-05 | 东北林业大学 | Based on pseudo label semi-supervised kernel part Fei Sheer discriminant analysis bearing failure diagnosis |
| CN109827777A (en) * | 2019-04-01 | 2019-05-31 | 哈尔滨理工大学 | Rolling bearing fault prediction technique based on Partial Least Squares extreme learning machine |
| CN110031226A (en) * | 2019-04-12 | 2019-07-19 | 佛山科学技术学院 | A kind of diagnostic method and device of bearing fault |
Non-Patent Citations (1)
| Title |
|---|
| 基于LLE及其改进距离算法的轴承故障诊断模型;魏永合 等;《组合机床与自动化加工技术》;20180731(第7期);基于LLE及其改进距离算法的轴承故障诊断模型 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110672324A (en) | 2020-01-10 |
| WO2021042749A1 (en) | 2021-03-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110672324B (en) | Bearing fault diagnosis method and device based on supervised LLE algorithm | |
| US20160104058A1 (en) | Generic object detection in images | |
| CN110245132B (en) | Data anomaly detection method, device, computer readable storage medium and computer equipment | |
| CN110672323B (en) | Bearing health state assessment method and device based on neural network | |
| CN113435509B (en) | Small sample scene classification and identification method and system based on meta-learning | |
| CN110602120B (en) | Network-oriented intrusion data detection method | |
| Shrivastava et al. | An effective scheme for image texture classification based on binary local structure pattern | |
| CN111489517B (en) | Screw locking abnormality warning method, device, computer device and storage medium | |
| CN110148117B (en) | Power equipment defect identification method and device based on power image and storage medium | |
| CN111798047A (en) | Wind control prediction method and device, electronic equipment and storage medium | |
| CN111046969A (en) | Data screening method and device, storage medium and electronic equipment | |
| CN116030237A (en) | Industrial defect detection method and device, electronic equipment and storage medium | |
| CN110287311A (en) | File classification method and device, storage medium, computer equipment | |
| CN114332500A (en) | Image processing model training method and device, computer equipment and storage medium | |
| CN111222005B (en) | Voiceprint data reordering method and device, electronic equipment and storage medium | |
| CN112529908A (en) | Digital pathological image segmentation method based on cascade convolution network and model thereof | |
| US9424484B2 (en) | Feature interpolation | |
| EP4064186A1 (en) | Methods and systems for generating synthetic microstructure images of material with desired features | |
| CN111127407B (en) | Fourier transform-based style migration forged image detection device and method | |
| Amid et al. | Unsupervised feature extraction for multimedia event detection and ranking using audio content | |
| Ferraz et al. | Object classification using a local texture descriptor and a support vector machine | |
| CN108021935B (en) | Dimension reduction method and device based on big data technology | |
| Sharma et al. | Regularization and feature selection for large dimensional data | |
| CN110647796A (en) | Iris identification method and device | |
| Park et al. | k-fold Subsampling based Sequential Backward Feature Elimination. |
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 |