CN110595811A - Method for constructing health state characteristic diagram of mechanical equipment - Google Patents
Method for constructing health state characteristic diagram of mechanical equipment Download PDFInfo
- Publication number
- CN110595811A CN110595811A CN201910860506.2A CN201910860506A CN110595811A CN 110595811 A CN110595811 A CN 110595811A CN 201910860506 A CN201910860506 A CN 201910860506A CN 110595811 A CN110595811 A CN 110595811A
- Authority
- CN
- China
- Prior art keywords
- frequency
- time
- spectrum
- fmap
- characteristic diagram
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01H—MEASUREMENT OF MECHANICAL VIBRATIONS OR ULTRASONIC, SONIC OR INFRASONIC WAVES
- G01H17/00—Measuring mechanical vibrations or ultrasonic, sonic or infrasonic waves, not provided for in the preceding groups
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M99/00—Subject matter not provided for in other groups of this subclass
-
- 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
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- 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
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
Abstract
The method for constructing the health state characteristic diagram of the mechanical equipment comprises the steps of firstly carrying out time-frequency transformation on vibration signals collected on the mechanical equipment, carrying out compression, normalization and graying on the obtained time-frequency spectrum, obtaining a time-frequency characteristic gray value, and filling the time-frequency characteristic gray value into the health state characteristic diagram; then, Fourier transform is carried out on the vibration signal, frequency band equal division and graying processing are carried out on the frequency spectrum, a frequency domain characteristic gray value is obtained, and the frequency domain characteristic gray value is filled into a health state characteristic diagram; and finally, calculating a time domain dimensionless index of the vibration signal, acquiring a time domain characteristic gray value, and filling the time domain characteristic gray value into the health state characteristic diagram. The health state characteristic diagram obtained by the invention is small in size, can fully reflect the health state information of mechanical equipment, and improves the fault diagnosis accuracy.
Description
Technical Field
The invention relates to a method for constructing a health state characteristic diagram of mechanical equipment.
Background
With the continuous development of technologies such as computers, sensors, communication, artificial intelligence and the like, intelligent health state monitoring and fault diagnosis technologies of mechanical equipment are also rapidly developed. The intelligent fault diagnosis technology overcomes the contradiction between the massive health state monitoring data of the mechanical equipment operation and the relative scarcity of fault diagnosis professionals, can automatically process and analyze the health state data of the mechanical equipment, and can give out an accurate fault diagnosis result. In recent years, deep learning has been rapidly developed in academic and industrial fields by virtue of its strong pattern recognition capability. The fault diagnosis of mechanical equipment is also essentially pattern recognition, so that a plurality of scholars and experts at home and abroad research and utilize a deep learning technology to realize intelligent fault diagnosis of the equipment.
The basic network models commonly used for deep learning include a Deep Belief Network (DBN), a Convolutional Neural Network (CNN), a stacked automatic coding machine (SAE), a Recurrent Neural Network (RNN), a generative countermeasure network (GAN), and the like. Among them, CNN is very suitable for processing and learning of mass data as a supervised deep learning model, and has been successfully implemented in commercial applications at present. Since the input of the CNN is a two-dimensional graph, and the two-dimensional graph is usually an N × N square graph for the sake of processing simplicity, one of the key points in applying the CNN to intelligent fault diagnosis of mechanical equipment is how to obtain a two-dimensional graph that sufficiently reflects health status information from equipment health status data (usually vibration data).
Wuchuanzhi, Qujianling, Populi Hongbai and the like adopt one-dimensional time domain vibration signals as input of CNN, and are respectively used for fault diagnosis of a gear box, a rolling bearing and a reciprocating compressor. Luyang king et al utilize vibration signal frequency domain data as CNN input for gearbox health monitoring. The Liheng and Wanglihua utilize short-time Fourier transform time frequency spectrum as CNN input, and are respectively used for fault diagnosis of bearings and motors. Chenrenxiang and the like utilize discrete wavelet transformation to construct a two-dimensional time-frequency matrix which is used as CNN input for fault diagnosis of the rolling bearing. And the wavelet time-frequency graph is input as CNN by Yuan-Jian and the like and is used for fault diagnosis of the rolling bearing. And constructing a time-frequency diagram by using a Fourier transform method for the Liu (registered public network) set and the like as the input of the CNN, wherein the time-frequency diagram is used for diagnosing the fault of the rolling bearing.
At present, the input of a fault diagnosis CNN model of mechanical equipment usually only considers one of three analysis domains, namely a vibration signal time domain, a vibration signal frequency domain or a vibration signal time-frequency domain, and characteristic information in each analysis domain cannot be comprehensively considered, so that the health state of the mechanical equipment is difficult to be fully reflected, and the accuracy of the CNN fault diagnosis model for judging the health state of the mechanical equipment is influenced.
Disclosure of Invention
The invention provides a method for constructing a health state characteristic diagram of mechanical equipment, aiming at overcoming the defects of a method for acquiring the health state characteristic diagram of the mechanical equipment in the prior art.
The invention discloses a method for constructing a health state characteristic diagram of mechanical equipment, which comprises the following steps of:
step one, presetting the dimension of a health state characteristic diagram FMap as NXN;
step two, collecting a vibration signal x (L) with the length of L on mechanical equipment, wherein L is 1,2, …, L;
step three, acquiring a time-frequency characteristic gray value of the vibration signal x (l) and filling the time-frequency characteristic gray value into a health state characteristic diagram FMap, wherein the method specifically comprises the following steps:
(31) performing time-frequency transform processing on x (L) to obtain a time spectrum TFP (t, f), wherein t is 1,2, …, L, f is 1,2, …, L/2;
(32) the time axis N of the time spectrum TFP (t, f) is equally divided, the time point number of each equal division is delta t equal to L/N, the frequency point number of each equal division is delta f for the frequency axis N/21For (L/2)/(N/2) L/N, so that for time spectrum TFP (t, f) the block is divided, defining the position (t)d,fd) The time-frequency block of (d) is TFP ((t)d-1)·△t+i,(fd-1)·△f1+j1) Wherein t isd=1,2,…,N,fd=1,2,…,N/2,i=1,2,…,△t,j1=1,2,…,△f1;
(33) The element values in each time-frequency block are summed to obtain a compressed time frequency spectrum TFPC(td,fd) Wherein the coordinate (t)d,fd) Has an element value of
(34) Calculating a compressed time spectrum TFPC(td,fd) Maximum value of (P)cpThe compressed time spectrum TFPC(td,fd) Each element in (1) divided by PcpTo obtain a normalized spectrum NTFPC(td,fd);
(35) Calculating normalized spectra NTFPC(td,fd) Maximum value of (P)ncpAnd a minimum value VncpNormalized spectrum NTFPC(td,fd) Conversion to grey spectrum GTFPC(td,fd) Coordinate (t)d,fd) At a gray value of
Wherein [. cndot. ] represents a rounding symbol;
(36) will grey scale spectrum GTFPC(td,fd) Filling the state of health characteristic diagram FMap from line N to line N/2+ 1;
step four, acquiring frequency domain characteristic gray values of the vibration signals x (l) and filling the frequency domain characteristic gray values into the health state characteristic diagram FMap, wherein the method specifically comprises the following steps:
(41) performing fourier transform processing on x (L) to obtain a frequency spectrum x (f), wherein f is 1,2, …, L/2;
(42) equally dividing the frequency axis N of the frequency spectrum X (f) to obtain N frequency bands, wherein the frequency point number of each frequency band is delta f2(L/2)/N, and the kth frequency band is defined as X ((k-1) · Δ f)2+j2) Where k is 1,2, …, N, j2=1,2,…,△f2;
(43) Calculating the energy ratio of each frequency band in the whole frequency spectrum X (f), wherein the energy ratio of the k-th frequency band is
(44) Calculating an energy ratio sequence Er(k) Maximum value of (P)ErAnd a minimum value VErComparing the energy to the sequence Er(k) Conversion to Gray spectra GEr(k),GEr(k) The gray value of is calculated by
Wherein [. cndot. ] represents a rounding symbol;
(45) gray scale GEr(k) Copying and filling into the lines 6 to N/2 of the health state characteristic diagram FMap, wherein the specific method is that the h line FMap (h, k) ═ GE of FMapr(k) Wherein h is 6,7, …, N/2;
step five, acquiring a time domain characteristic gray value of the vibration signal x (l) and filling the time domain characteristic gray value into the health state characteristic diagram FMap, wherein the method specifically comprises the following steps:
(51) 5 time domain dimensionless indexes of x (l) are calculated, and are respectively peak value indexes CFPulse index IFKurtosis index KFMargin index MFSum waveform index WFThereby obtaining a dimensionless index set Fnd={CF,IF,KF,MF,WF};
(52) F is to bendConverted into a time-domain index gray value set GFnd={GCF,GIF,GKF,GMF,GWFThe concrete method is that
Wherein iF=1,2,…,5;
(53) Gray value GF of time domain indexnd(iF) Copy and fill to (6-i) th of the health status feature map FMapF) In the middle of the line, i.e., FMap (6-i)F) All values of a line are GFnd(iF) Thus, a complete health status characteristic diagram FMap is obtained.
The invention has the following positive effects: in the health state characteristic diagram, time-frequency characteristic gray values depict time-frequency domain detail information of vibration signals, the frequency domain characteristic gray values represent frequency domain information of the vibration signals, and the time-domain characteristic gray values represent time-domain overall information of the vibration signals.
Drawings
FIG. 1 is a flow chart of an embodiment of the present invention;
FIG. 2 is a time domain waveform of vibration data of a rolling bearing according to an embodiment of the present invention;
fig. 3 is a health state characteristic diagram constructed by vibration data of the rolling bearing in the embodiment of the invention.
Detailed Description
The invention is further illustrated by the following figures and examples.
Referring to fig. 1, the method for constructing the health state characteristic diagram of the mechanical equipment comprises the following steps:
step one, presetting the dimension of a health state characteristic diagram FMap as NXN;
step two, collecting a vibration signal x (L) with the length of L on mechanical equipment, wherein L is 1,2, …, L;
step three, acquiring a time-frequency characteristic gray value of the vibration signal x (l) and filling the time-frequency characteristic gray value into a health state characteristic diagram FMap, wherein the method specifically comprises the following steps:
(31) performing time-frequency transform processing on x (L) to obtain a time spectrum TFP (t, f), wherein t is 1,2, …, L, f is 1,2, …, L/2;
(32) the time axis N of the time spectrum TFP (t, f) is equally divided, the time point number of each equal division is delta t equal to L/N, the frequency point number of each equal division is delta f for the frequency axis N/21For (L/2)/(N/2) L/N, so that for time spectrum TFP (t, f) the block is divided, defining the position (t)d,fd) The time-frequency block of (d) is TFP ((t)d-1)·△t+i,(fd-1)·△f1+j1) Which isMiddle td=1,2,…,N,fd=1,2,…,N/2,i=1,2,…,△t,j1=1,2,…,△f1;
(33) The element values in each time-frequency block are summed to obtain a compressed time frequency spectrum TFPC(td,fd) Wherein the coordinate (t)d,fd) Has an element value of
(34) Calculating a compressed time spectrum TFPC(td,fd) Maximum value of (P)cpThe compressed time spectrum TFPC(td,fd) Each element in (1) divided by PcpTo obtain a normalized spectrum NTFPC(td,fd);
(35) Calculating normalized spectra NTFPC(td,fd) Maximum value of (P)ncpAnd a minimum value VncpNormalized spectrum NTFPC(td,fd) Conversion to grey spectrum GTFPC(td,fd) Coordinate (t)d,fd) At a gray value of
Wherein [. cndot. ] represents a rounding symbol;
(36) will grey scale spectrum GTFPC(td,fd) Filling the state of health characteristic diagram FMap from line N to line N/2+ 1;
step four, acquiring frequency domain characteristic gray values of the vibration signals x (l) and filling the frequency domain characteristic gray values into the health state characteristic diagram FMap, wherein the method specifically comprises the following steps:
(41) performing fourier transform processing on x (L) to obtain a frequency spectrum x (f), wherein f is 1,2, …, L/2;
(42) equally dividing the frequency axis N of the frequency spectrum X (f) to obtain N frequency bands, wherein the frequency point number of each frequency band is delta f2(L/2)/N, and the kth frequency band is defined as X ((k-1) · Δ f)2+j2) Where k is 1,2, …, N, j2=1,2,…,△f2;
(43) Calculating the energy ratio of each frequency band in the whole frequency spectrum X (f), wherein the energy ratio of the k-th frequency band is
(44) Calculating an energy ratio sequence Er(k) Maximum value of (P)ErAnd a minimum value VErComparing the energy to the sequence Er(k) Conversion to Gray spectra GEr(k),GEr(k) The gray value of is calculated by
Wherein [. cndot. ] represents a rounding symbol;
(45) gray scale GEr(k) Copying and filling into the lines 6 to N/2 of the health state characteristic diagram FMap, wherein the specific method is that the h line FMap (h, k) ═ GE of FMapr(k) Wherein h is 6,7, …, N/2;
step five, acquiring a time domain characteristic gray value of the vibration signal x (l) and filling the time domain characteristic gray value into the health state characteristic diagram FMap, wherein the method specifically comprises the following steps:
(51) 5 time domain dimensionless indexes of x (l) are calculated, and are respectively peak value indexes CFPulse index IFKurtosis index KFMargin index MFSum waveform index WFThereby obtaining a dimensionless index set Fnd={CF,IF,KF,MF,WF};
(52) F is to bendConverted into a time-domain index gray value set GFnd={GCF,GIF,GKF,GMF,GWFThe concrete method is that
Wherein iF=1,2,…,5;
(53) When going toField index gray value GFnd(iF) Copy and fill to (6-i) th of the health status feature map FMapF) In the middle of the line, i.e., FMap (6-i)F) All values of a line are GFnd(iF) Thus, a complete health status characteristic diagram FMap is obtained.
The invention is applied to the construction of the health state characteristic diagram of the SKF 6205-2RS type rolling bearing. The bearing vibration data is sourced from a bearing data center website of the university of western storage, and drive end bearing fault data is selected. The inner ring of the test bearing was machined with a dimple having a fault diameter of 0.18 mm. The load is set to be 0.75kW, the rotating speed is 1772r/min, the vibration data sampling frequency is 12k, and the steps of constructing the health state characteristic diagram of the rolling bearing by using the method are as follows.
Step one, the dimension of the preset health state characteristic diagram FMap is 32 multiplied by 32.
And step two, acquiring a vibration signal x (L) with the length L of 2048 on a mechanical device, wherein L is 1,2, … and 2048.
Step three, acquiring a time-frequency characteristic gray value of the vibration signal x (l) and filling the time-frequency characteristic gray value into a health state characteristic diagram FMap, wherein the method specifically comprises the following steps:
(31) performing short-time fourier transform on x (l) to obtain a time-frequency spectrum TFP (t, f), wherein t is 1,2, …,2048, f is 1,2, …, 1024;
(32) the time axis 32 of the time-frequency spectrum TFP (t, f) is equally divided, the number of time points per equal division is 64, the frequency axis 16 is equally divided, the number of frequency points per equal division is 64, and therefore the position (t, f) is defined for the TFP (t, f) blocksd,fd) The time-frequency block of (d) is TFP ((t)d-1)×64+i,(fd-1)×64+j1) Wherein t isd=1,2,…,32,fd=1,2,…,16,i=1,2,…,64,j1=1,2,…,64;
(33) The element values in each time-frequency block are summed to obtain a compressed time-frequency spectrum TFPC(td,ffd) Wherein the coordinate (t)d,fd) Has an element value of
(34) Calculating a compressed time spectrum TFPC(td,fd) Maximum value Pcp0.0092, the compressed time spectrum TFPC(td,fd) Each element in (1) divided by PcpTo obtain normalized time frequency spectrum NTFPC(td,fd);
(35) Calculating normalized spectra NTFPC(td,fd) Maximum value Pncp1 and minimum value VncpNormalized spectrum NTFP is normalized to 0.0021C(td,fd) Converting into time-frequency gray spectrum GTFPC(td,fd) Coordinate (t)d,fd) At a gray value of
Wherein [. cndot. ] represents a rounding symbol;
(36) time-frequency gray scale spectrum GTFPC(td,fd) Wherein t isd=1,2,…,32,fd1,2, …,16, filled into lines 32 through 17 of the state of health feature map FMap;
step four, acquiring frequency domain characteristic gray values of the vibration signals x (l) and filling the frequency domain characteristic gray values into the health state characteristic diagram FMap, wherein the method specifically comprises the following steps:
(41) fourier transforming x (l) to obtain a spectrum x (f), wherein f is 1,2, …, 1024;
(42) dividing the frequency axis 32 of the frequency spectrum X (f) equally to obtain 32 frequency bands, wherein the frequency point number of each frequency band is delta f2The kth band is X ((k-1) × 32+ j) × 322) Where k is 1,2, …,32, j2=1,2,…,32;
(43) Calculating the energy ratio of each frequency band in the whole frequency spectrum X (f), wherein the energy ratio of the k-th frequency band is
(44) Calculating an energy ratio sequence Er(k) Maximum value of (P)Er0.0471 and minimum value VEr0.0005, energy is compared to sequence Er(k) Conversion to frequency domain gray spectrum GEr(k),GEr(k) Has a gray value of
(45) GE frequency domain gray scale spectrumr(k) Copying and filling the data into lines 6 to 16 of the health state characteristic diagram FMap by lines, wherein the specific method is that the h line FMap (h, k) ═ GE of the health state characteristic diagram FMapr(k) Wherein h is 6,7, …, 16;
step five, acquiring a time domain characteristic gray value of the vibration signal x (l) and filling the time domain characteristic gray value into the health state characteristic diagram FMap, wherein the method specifically comprises the following steps:
(51) calculating 5 time domain dimensionless indexes of x (l), wherein the peak index CF4.8716 pulse index IF250.8961 kurtosis index KF5.4728, margin index MF8.6028 and waveform index WF51.5022, obtaining a dimensionless index set Fnd={4.8716,250.8961,5.4728,8.6028,51.5022};
(52) F is to bendConverted into a time-domain index gray value set GFnd={GCF,GIF,GKF,GMF,GWFThe concrete method is that
Wherein iF=1,2,…,5;
(53) Gray value GF of time domain indexnd(iF) Copy and fill to (6-i) th of the health status feature map FMapF) In the row, i.e., (6-i) of FMapF) All values of a line are GFnd(iF) Thus, a complete health status characteristic diagram FMap is obtained.
The embodiments described in this specification are merely illustrative of implementations of the inventive concept and the scope of the present invention should not be considered limited to the specific forms set forth in the embodiments but rather by the equivalents thereof as may occur to those skilled in the art upon consideration of the present inventive concept.
Claims (1)
1. The method for constructing the health state characteristic diagram of the mechanical equipment comprises the following steps:
step one, presetting the dimension of a health state characteristic diagram FMap as NXN;
step two, collecting a vibration signal x (L) with the length of L on mechanical equipment, wherein L is 1,2, …, L;
step three, acquiring a time-frequency characteristic gray value of the vibration signal x (l) and filling the time-frequency characteristic gray value into a health state characteristic diagram FMap, wherein the method specifically comprises the following steps:
(31) performing time-frequency transform processing on x (L) to obtain a time spectrum TFP (t, f), wherein t is 1,2, …, L, f is 1,2, …, L/2;
(32) the time axis N of the time spectrum TFP (t, f) is equally divided, the time point number of each equal division is delta t equal to L/N, the frequency point number of each equal division is delta f for the frequency axis N/21For (L/2)/(N/2) L/N, so that for time spectrum TFP (t, f) the block is divided, defining the position (t)d,fd) The time-frequency block of (d) is TFP ((t)d-1)·△t+i,(fd-1)·△f1+j1) Wherein t isd=1,2,…,N,fd=1,2,…,N/2,i=1,2,…,△t,j1=1,2,…,△f1;
(33) The element values in each time-frequency block are summed to obtain a compressed time frequency spectrum TFPC(td,fd) Wherein the coordinate (t)d,fd) Has an element value of
(34) Calculating a compressed time spectrum TFPC(td,fd) Maximum value of (P)cpThe compressed time spectrum TFPC(td,fd) Each element in (1) divided by PcpTo obtain a normalized spectrum NTFPC(td,fd);
(35) Calculating normalized spectra NTFPC(td,fd) Maximum value of (P)ncpAnd a minimum value VncpNormalized spectrum NTFPC(td,fd) Conversion to grey spectrum GTFPC(td,fd) Coordinate (t)d,fd) At a gray value of
Wherein [. cndot. ] represents a rounding symbol;
(36) will grey scale spectrum GTFPC(td,fd) Filling the state of health characteristic diagram FMap from line N to line N/2+ 1;
step four, acquiring frequency domain characteristic gray values of the vibration signals x (l) and filling the frequency domain characteristic gray values into the health state characteristic diagram FMap, wherein the method specifically comprises the following steps:
(41) performing fourier transform processing on x (L) to obtain a frequency spectrum x (f), wherein f is 1,2, …, L/2;
(42) equally dividing the frequency axis N of the frequency spectrum X (f) to obtain N frequency bands, wherein the frequency point number of each frequency band is delta f2(L/2)/N, and the kth frequency band is defined as X ((k-1) · Δ f)2+j2) Where k is 1,2, …, N, j2=1,2,…,△f2;
(43) Calculating the energy ratio of each frequency band in the whole frequency spectrum X (f), wherein the energy ratio of the k-th frequency band is
(44) Calculating an energy ratio sequence Er(k) Maximum value of (P)ErAnd a minimum value VErComparing the energy to the sequence Er(k) Conversion to Gray spectra GEr(k),GEr(k) The gray value of is calculated by
Wherein [. cndot. ] represents a rounding symbol;
(45) gray scale GEr(k) Copying and filling into the lines 6 to N/2 of the health state characteristic diagram FMap, wherein the specific method is that the h line FMap (h, k) ═ GE of FMapr(k) Wherein h is 6,7, …, N/2;
step five, acquiring a time domain characteristic gray value of the vibration signal x (l) and filling the time domain characteristic gray value into the health state characteristic diagram FMap, wherein the method specifically comprises the following steps:
(51) 5 time domain dimensionless indexes of x (l) are calculated, and are respectively peak value indexes CFPulse index IFKurtosis index KFMargin index MFSum waveform index WFThereby obtaining a dimensionless index set Fnd={CF,IF,KF,MF,WF};
(52) F is to bendConverted into a time-domain index gray value set GFnd={GCF,GIF,GKF,GMF,GWFThe concrete method is that
Wherein iF=1,2,…,5;
(53) Gray value GF of time domain indexnd(iF) Copy and fill to (6-i) th of the health status feature map FMapF) In the middle of the line, i.e., FMap (6-i)F) All values of a line are GFnd(iF) Thus, a complete health status characteristic diagram FMap is obtained.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910860506.2A CN110595811B (en) | 2019-09-11 | 2019-09-11 | Method for constructing health state characteristic diagram of mechanical equipment |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910860506.2A CN110595811B (en) | 2019-09-11 | 2019-09-11 | Method for constructing health state characteristic diagram of mechanical equipment |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110595811A true CN110595811A (en) | 2019-12-20 |
CN110595811B CN110595811B (en) | 2021-04-09 |
Family
ID=68858904
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910860506.2A Active CN110595811B (en) | 2019-09-11 | 2019-09-11 | Method for constructing health state characteristic diagram of mechanical equipment |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110595811B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113822257A (en) * | 2021-11-24 | 2021-12-21 | 航天智控(北京)监测技术有限公司 | Abnormal point detection method based on combination of dimensionless features and virtual samples |
Citations (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102222324A (en) * | 2011-06-17 | 2011-10-19 | 电子科技大学 | Symmetry property-based method for detecting salient regions of images |
CN105654037A (en) * | 2015-12-21 | 2016-06-08 | 浙江大学 | Myoelectric signal gesture recognition method based on depth learning and feature images |
US20160283841A1 (en) * | 2015-03-27 | 2016-09-29 | Google Inc. | Convolutional neural networks |
CN108564005A (en) * | 2018-03-26 | 2018-09-21 | 电子科技大学 | A kind of human body tumble discrimination method based on convolutional neural networks |
CN108764601A (en) * | 2018-04-03 | 2018-11-06 | 哈尔滨工业大学 | A kind of monitoring structural health conditions abnormal data diagnostic method based on computer vision and depth learning technology |
US20190104357A1 (en) * | 2017-09-29 | 2019-04-04 | Apple Inc. | Machine learning based sound field analysis |
CN109934206A (en) * | 2019-04-08 | 2019-06-25 | 中国矿业大学(北京) | A kind of rotary machinery fault diagnosis method under non-stationary operating condition |
CN109994203A (en) * | 2019-04-15 | 2019-07-09 | 江南大学 | A kind of epilepsy detection method based on EEG signal depth multi-angle of view feature learning |
CN110108431A (en) * | 2019-05-22 | 2019-08-09 | 西安因联信息科技有限公司 | A kind of Trouble Diagnostic Method of Machinery Equipment based on machine learning classification algorithm |
CN110197127A (en) * | 2019-05-06 | 2019-09-03 | 安徽继远软件有限公司 | Wireless signal detection and electromagnetic interference categorizing system and method based on deep learning |
-
2019
- 2019-09-11 CN CN201910860506.2A patent/CN110595811B/en active Active
Patent Citations (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102222324A (en) * | 2011-06-17 | 2011-10-19 | 电子科技大学 | Symmetry property-based method for detecting salient regions of images |
US20160283841A1 (en) * | 2015-03-27 | 2016-09-29 | Google Inc. | Convolutional neural networks |
CN105654037A (en) * | 2015-12-21 | 2016-06-08 | 浙江大学 | Myoelectric signal gesture recognition method based on depth learning and feature images |
US20190104357A1 (en) * | 2017-09-29 | 2019-04-04 | Apple Inc. | Machine learning based sound field analysis |
CN108564005A (en) * | 2018-03-26 | 2018-09-21 | 电子科技大学 | A kind of human body tumble discrimination method based on convolutional neural networks |
CN108764601A (en) * | 2018-04-03 | 2018-11-06 | 哈尔滨工业大学 | A kind of monitoring structural health conditions abnormal data diagnostic method based on computer vision and depth learning technology |
CN109934206A (en) * | 2019-04-08 | 2019-06-25 | 中国矿业大学(北京) | A kind of rotary machinery fault diagnosis method under non-stationary operating condition |
CN109994203A (en) * | 2019-04-15 | 2019-07-09 | 江南大学 | A kind of epilepsy detection method based on EEG signal depth multi-angle of view feature learning |
CN110197127A (en) * | 2019-05-06 | 2019-09-03 | 安徽继远软件有限公司 | Wireless signal detection and electromagnetic interference categorizing system and method based on deep learning |
CN110108431A (en) * | 2019-05-22 | 2019-08-09 | 西安因联信息科技有限公司 | A kind of Trouble Diagnostic Method of Machinery Equipment based on machine learning classification algorithm |
Non-Patent Citations (3)
Title |
---|
刘炳集等: "基于时频图和CNN的滚动轴承故障诊断", 《南昌航空大学学报:自然科学版》 * |
袁建虎等: "基于小波时频图和CNN的滚动轴承智能故障诊断方法", 《机械设计与研究》 * |
郭远晶等: "S变换用于滚动轴承故障信号冲击特征提取", 《振动、测试与诊断》 * |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113822257A (en) * | 2021-11-24 | 2021-12-21 | 航天智控(北京)监测技术有限公司 | Abnormal point detection method based on combination of dimensionless features and virtual samples |
Also Published As
Publication number | Publication date |
---|---|
CN110595811B (en) | 2021-04-09 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Wang et al. | A deep learning method for bearing fault diagnosis based on time-frequency image | |
Gu et al. | Bearing fault diagnosis with varying conditions using angular domain resampling technology, SDP and DCNN | |
CN106124212B (en) | Fault Diagnosis of Roller Bearings based on sparse coding device and support vector machines | |
CN110702411B (en) | Residual error network rolling bearing fault diagnosis method based on time-frequency analysis | |
CN107702922B (en) | Rolling bearing fault diagnosis method based on LCD and stacking automatic encoder | |
CN113780412A (en) | Fault diagnosis model training method and system and fault diagnosis model training method and system | |
CN112633245A (en) | Planetary gear box fault diagnosis method based on deep reinforcement learning model | |
Liang et al. | Intelligent fault diagnosis of rolling element bearing based on convolutional neural network and frequency spectrograms | |
CN107024271A (en) | Mechanical oscillation signal compression reconfiguration method and system | |
Montechiesi et al. | Artificial immune system via Euclidean Distance Minimization for anomaly detection in bearings | |
CN116465623B (en) | Gearbox service life prediction method based on sparse converter | |
Wang et al. | An end-to-end fault diagnostics method based on convolutional neural network for rotating machinery with multiple case studies | |
CN107966287B (en) | Weak fault feature extraction method for self-adaptive electromechanical equipment | |
CN110595811B (en) | Method for constructing health state characteristic diagram of mechanical equipment | |
CN117708574B (en) | CNN variable-speed rolling bearing fault diagnosis method embedded with physical information | |
CN114462446A (en) | Rolling bearing fault diagnosis method based on vibration signal and electronic equipment | |
CN111610024A (en) | Gear crack fault diagnosis method based on vibration signals | |
CN113408578A (en) | Harmonic reducer fault diagnosis method and system based on generative countermeasure network | |
CN117330314A (en) | Bearing fault diagnosis method based on two-dimensional gray level diagram and Swin-transducer | |
Zhong et al. | Parallel learning attention-guided CNN for signal denoising and mechanical fault diagnosis | |
CN116720442A (en) | Wear prediction method based on improved Fourier neural operator | |
CN115808312A (en) | Rolling bearing fault diagnosis method based on deformable volume and Transformer | |
Shen et al. | Gear fault diagnosis based on short-time fourier transform and deep residual network under multiple operation conditions | |
Liu et al. | Sinc-based multiplication-convolution network for equipment intelligent edge diagnosis under small samples | |
Xin-yu et al. | NSP-CNN rolling bearing fault diagnosis method |
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 |