CN105203645A - Intelligent detection method of high-speed turnout crack damage based on vibration signal fusion - Google Patents
Intelligent detection method of high-speed turnout crack damage based on vibration signal fusion Download PDFInfo
- Publication number
- CN105203645A CN105203645A CN201510310524.5A CN201510310524A CN105203645A CN 105203645 A CN105203645 A CN 105203645A CN 201510310524 A CN201510310524 A CN 201510310524A CN 105203645 A CN105203645 A CN 105203645A
- Authority
- CN
- China
- Prior art keywords
- hurt
- vibration signal
- turnout
- track switch
- switch
- 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
Abstract
The invention discloses an intelligent detection method of high-speed turnout crack damage based on vibration signal fusion. The intelligent detection method comprises the following steps: firstly, respectively mounting vibration acceleration sensors at a turnout tip, a turnout center and a turnout tail of a high-speed turnout, so as to collect vibration signals, and respectively carrying out CEEMD adaptive decomposition on the vibration signals of the three detection points, so as to obtain a limit number of IMF components; then, screening out a main IMF component by using correlation analysis, and calculating singularity entropy eigenvalue of the main IMF component; fusing singularity entropy eigenvalue of the vibration signals of a plurality of the detection points into an eigenvector, and inputting the eigenvector into LSSVM for training and testing, so as to achieve judgement of the working condition and damage type of the turnout. According to the intelligent detection method, different crack damage of the turnout can be effectively distinguished, and the method has good realtime performance and noise resistance, and provides a novel technological means for intelligent diagnosis of turnout fault.
Description
Technical field
The present invention relates to a kind of high-speed switch crackle hurt intelligent detecting method based on vibration signal, belong to railway switch and detect and maintenance technology field.
Background technology
Track switch, as the important component part of railroad track, is the indispensable line facility of high-speed railway, is also the weak link on circuit simultaneously, with train running speed and security performance closely related.Train through turnout passing conversion line time, because wheel is to the powerful impact force action of track, track switch may be caused to occur contact fatigue, wearing and tearing, the crackle even hurt type such as track deformation.If detect not timely track switch hurt, process, As time goes on, track switch hurt worsens further, may cause the major accidents such as derail, and this constitutes serious threat to the safe operation of train.Therefore, the identification of research track switch hurt, Real-time Obtaining turnout work status information, to ensureing that train is efficient, safe operation is significant.
For track switch hurt monitoring problem, each state all conducts in-depth research and develops corresponding switch monitor system, as the TrackandTurnoutmonitoring monitoring system that French track switch adopts, the Roadmaster2000 monitoring system etc. that Germany's track switch adopts, above system can realize the monitoring to the conversion power of track switch communication apparatus, electric current, voltage, track circuit and goat state, each towing point etc., and the Real-Time Monitoring for turnout work status information provides strong means.But these systems are Shortcomings part also, such as lack the detection to hurts such as track switch abrasion, crackle, contact fatigues.At present, the track switch hurt of China detects the defectoscope mainly combined with traditional large-scale inspection car and small-sized defectoscope and is made as master, although this method of detection can detect track switch hurt to a certain extent, but have that detection efficiency is low, sensing range is limited, road occupying inspection affects the problems such as train operation efficiency, be difficult to the demand meeting China Express Railway fast development.
Publication number is " a kind of high ferro rail defects and failures detection method based on vibration signal and device " patented claim of CN102175768A, calculate the power spectrum of first IMF component after utilizing EMD decomposition vibration signal, and carry out hurt detection using this power spectrum as the distinguishing characteristic of rail defects and failures.In theory, the analysis of the modal overlap problems affect signal local feature of EMD and extraction.Say from technical standpoint, this patented claim only carries out hurt detection from a pick-up point collection signal, can not detect the position (along rail line direction) that hurt occurs, and does not also know from pick-up point how far hurt can detect.Because track circuit is oversize, carry out the detection of hurt position, need the sensor of laying enormous amount along the line and set up sensor network (see this patent application specification the 26th section), the consequence done like this, one is that data volume is difficult to too greatly process; Two is that vibration signal all has impact to multiple sensor after the rigidity of rail is propagated, and namely hurt information all can embody on the power spectrum characteristic of the vibration signal of multiple sensor, causes power spectrum characteristic differentiating method will be difficult to identify hurt.The measured result of Fig. 8 also demonstrates this power spectrum characteristic and cannot judge turnout work state and hurt type.
Summary of the invention
The object of the invention is the deficiency for existing detection technique, a kind of high-speed switch crackle hurt intelligent detecting method based on vibration signal is provided.
The concrete implementation step of the technical solution adopted in the present invention is as follows:
(1) according to the Changing Pattern of the single span free beam Mode Shape of track switch, in the switch blade, trouble of high-speed switch and trouble tail three measuring point places vibration acceleration sensors are installed;
(2) train is when turnout passing, gathers the track switch vibration signal at three measuring point places respectively.Without loss of generality, the vibration signal at note switch blade measuring point place is x (i), i=1 ..., N, N are sampling number.
(3) adopt CEEMD to carry out adaptive decomposition to x (i), obtain M the intrinsic modal components (IMF) comprising track switch hurt information, be designated as c
j(i), j=1,2 ..., M.
(4) IMF component c is calculated
jthe related coefficient of (i) and former vibration signal x (i)
Wherein,
with
represent x (i) and c respectively
jthe mean value of (i).
(5) η is chosen
jbe greater than the main IMF component of m IMF component as this measuring point vibration signal of threshold value T, be designated as
k=1,2 ..., m.
(6) main IMF component is calculated
singular entropy E
kand by this m singular entropy E
k(k=1,2 ..., m) as the track switch hurt eigenwert of switch blade measuring point, be designated as H
1={ E
1, E
2... E
m.
(7) similarly, in trouble, the vibration signal of trouble tail measuring point collection performs step (3)-step (6) respectively, obtains branching off the track switch hurt eigenwert H at neutralization trouble tail measuring point place
2, H
3.Then to H
1, H
2, H
3carry out being augmented Fusion Features, form the track switch hurt proper vector H={H of 3m dimension
1, H
2, H
3.
(8) hurt proper vector H is inputted Least square support vector (LSSVM) model, choose radial basis function as kernel function, utilize grid search and cross validation to carry out optimizing to LSSVM penalty factor and radial basis function parameter, and then realize the judgement of turnout work state and hurt type.
At above-mentioned a kind of high-speed switch crackle hurt intelligent detecting method based on vibration signal, it is characterized in that, the threshold value T=0.1 in described step (5).
In reality is implemented, the main IMF component of described step (6)
singular entropy E
kcalculation procedure is as follows:
(6.1) right
carry out phase space reconfiguration.Will
be embedded into (N-n+1) × n to tie up in phase space, obtain continuous wavelet transform track matrix X
Wherein n is Embedded dimensions, and N is that signal sampling is counted.
Svd is carried out to X, namely
X＝UΛV
T(3)
Wherein,
with
for orthogonal matrix,
for diagonal matrix, and q meets q=min (N-n+1, n).λ
j(j=1,2 ..., q) be called the singular value of matrix X.
(6.2) calculate
singular entropy E
k.From λ
j(j=1,2 ..., choose a front v maximum singular value q), and v meets
then singular entropy E
kfor
Wherein,
for the weight of a jth singular value in whole singular value.
In reality is implemented, the parameter optimization step of described step (8) is as follows:
(8.1) initialization penalty factor γ ∈ [e
-5, e
5], kernel functional parameter σ ∈ [e
-5, e
5], sizing grid gets 10 × 10, obtains 100 groups of parameters pair altogether.
(8.2) training sample data are divided into 10 groups, to (γ, σ), following operation are performed to each group parameter in grid:
(8.3) wherein one group of sample data is as test set in selection, and all the other 9 groups, as training set, obtain the predicated error δ of LSSVM.
(8.4) repeat step (8.3) to perform 10 times, all select different subset as training set at every turn, and the predicated error that 10 experiments obtain is averaged, obtain the predicated error of this group (γ, σ)
(8.5) parameter sets (γ, σ is changed
2), repeated execution of steps (8.3) and (8.4), obtain the predicated error of LSSVM under various combination parameter successively
using minimum for predicated error average one group of parameter as the optimization model parameter combinations in grid.
Compared with prior art, the invention has the beneficial effects as follows:
1) the present invention is based on vibration signal and carry out hurt detection to track switch, the signals collecting of the method is simple, and signal has contained abundant switch status information, can realize the real-time detection of track switch operating mode and need not take track equipment in a large number.
2) the noise residual problem of the CEEMD method modal overlap problem that effectively inhibits EMD to exist and EEMD, be suitable for processing track switch vibration signal that is non-linear, non-stationary, and singular entropy has svd and excavates the feature that the function of matrix modal characteristics and information entropy describe burst complicacy, the IMF singular entropy extracted better can reflect the hurt feature of track switch.
3) the present invention adopts LSSVM as sorter, artificially need not set decision threshold, can realize track switch hurt type automatic discrimination.Meanwhile, adopt grid search and cross validation to carry out optimizing to LSSVM parameter, reduce the blindness of Selecting parameter, improve the accuracy rate that hurt detects.
Accompanying drawing explanation
Fig. 1 is process flow diagram of the present invention;
Fig. 2 is track switch vibration transducer scheme of installation;
Fig. 3 is the time domain waveform of the most advanced and sophisticated different operating mode of track switch, wherein: (a) normal track switch, (b) crackle 0.5cm, (c) crackle 1.5cm;
Fig. 4 is normal track switch vibration signal CEEMD decomposition result;
Fig. 5 is the related coefficient of each IMF of different operating mode track switch vibration signal and original signal, wherein: (a) track switch is most advanced and sophisticated, in the middle part of (b) track switch, and (c) track switch tail end;
Fig. 6 A is the singular entropy distribution plan of the most advanced and sophisticated different operating mode of track switch;
Fig. 6 B is the singular entropy distribution plan of crackle 1.5cm diverse location;
Fig. 7 is grid search and cross validation parameter optimization result;
Fig. 8 is the power spectrum density of the first rank IMF under the most advanced and sophisticated different operating mode of track switch, wherein: (a) normal track switch, (b) crackle 0.5cm, (c) crackle 1.5cm;
Fig. 9 is the impact of different noise on hurt recognition result.
Embodiment
Below in conjunction with drawings and Examples, the present invention is described in detail.
As shown in Figure 1, the specific embodiment of the present invention is, a kind of high-speed switch crackle hurt intelligent detecting method based on vibration signal, the steps include:
(1) according to the Changing Pattern of the single span free beam Mode Shape of track switch, in the switch blade, trouble of high-speed switch and trouble tail three measuring point places vibration acceleration sensors are installed;
(2) train is when turnout passing, gathers the track switch vibration signal at three measuring point places respectively.Without loss of generality, the vibration signal at note switch blade measuring point place is x (i), i=1 ..., N, N are sampling number.
(3) adopt complete set empirical mode decomposition (CEEMD) to carry out adaptive decomposition to x (i), obtain M the intrinsic modal components (IMF) comprising track switch hurt information, be designated as c
j(i), j=1,2 ..., M.
(4) IMF component c is calculated
jthe related coefficient of (i) and former vibration signal
Wherein
with
represent x (i) and c respectively
jthe mean value of (i).Choose η
jbe greater than the IMF component of threshold value T=0.1 as main IMF.
(5) the singular entropy E of main IMF component is calculated
k, and by this m singular entropy E
k(k=1,2 ..., m) as the track switch hurt eigenwert that switch blade measuring point gathers, be designated as H
1={ E
1, E
2... E
m.The concrete calculation procedure of singular entropy is as follows:
(5.1) right
carry out phase space reconfiguration.Will
be embedded into (N-n+1) × n to tie up in phase space, obtain continuous wavelet transform track matrix X
Wherein n is Embedded dimensions, and N is that signal sampling is counted.
(5.2) svd is carried out to X, namely
X＝UΛV
T(7)
Wherein,
with
for orthogonal matrix,
for diagonal matrix, and q meets q=min (N-n+1, n).λ
j(j=1,2 ..., q) be called the singular value of matrix X.
(5.3) calculate
singular entropy E
k.From λ
j(j=1,2 ..., choose a front v maximum singular value q), and v meets
then singular entropy E
kfor
Wherein,
for the weight of a jth singular value in whole singular value.
(6) similarly, in trouble, the vibration signal of trouble tail measuring point collection performs step (3)-step (5) respectively, obtains branching off the track switch hurt eigenwert H at neutralization trouble tail measuring point place
2, H
3.Then to H
1, H
2, H
3carry out being augmented Fusion Features, form the track switch hurt proper vector H={H of 3m dimension
1, H
2, H
3.
(7) hurt proper vector H is inputted LSSVM model, choose radial basis function as kernel function, utilize grid search and cross validation to carry out optimizing to LSSVM penalty factor and radial basis function parameter, and then realize the judgement of turnout work state and hurt type.Parameter optimization step is as follows:
(7.1) initialization penalty factor γ ∈ [e
-5, e
5], kernel functional parameter σ ∈ [e
-5, e
5], sizing grid gets 10 × 10, obtains 100 groups of parameters pair altogether.
(7.2) training sample data are divided into 10 groups, to (γ, σ), following operation are performed to each group parameter in grid:
(7.3) wherein one group of sample data is as test set in selection, and all the other 9 groups, as training set, obtain the predicated error δ of LSSVM.
(7.4) repeat step (c) and perform 10 times, all select different subset as training set at every turn, and the predicated error that 10 experiments obtain is averaged, obtain the predicated error of this group (γ, σ)
(7.5) parameter sets (γ, σ is changed
2), repeated execution of steps (7.3) and (7.4), obtain the predicated error of LSSVM under various combination parameter successively
using minimum for predicated error average one group of parameter as the optimization model parameter combinations in grid.
Below effect of the present invention is verified.
(1) vibration signal at track switch different measuring points place is gathered by vibration transducer.As shown in Figure 2, wherein measuring point 1 is positioned at track switch tip to sensor mounting location, and measuring point 2 is positioned in the middle part of track switch, and measuring point 3 is positioned at track switch tail end.Fig. 3 is the time domain waveform of the different operating mode of switch blade.Adopt CEEMD to decompose signal, the decomposition result of normal track switch vibration signal as shown in Figure 4.Can find out, original signal is decomposed into several IMF by CEEMD, and each IMF contains the different characteristic information of signal.
(2) related coefficient of track switch different measuring points place each IMF component and former vibration signal is calculated, as shown in Figure 5, the correlativity of front 5 IMF components and original signal is larger, characterize the main hurt feature of original signal, and the correlativity of other high-order IMF and original signal is all below 0.1, visual to abandon for chaff component.Therefore choose the data source that front 5 IMF components calculate as singular entropy.
(3) singular entropy of main IMF component that obtains of calculation procedure (2).The distribution of part singular entropy is as shown in Fig. 6 A, Fig. 6 B.Can find out, normally different in the distribution of its singular entropy of same observation station with crackle hurt two kinds of operating modes, the singular entropy distribution of different crackle hurt degree is similar, but size there are differences.Track switch different measuring points is due to conduct vibrations path and different to the sensitivity of crackle hurt, and same operating also there are differences in the distribution of its singular entropy of diverse location, and namely the hurt information that comprises of diverse location is different.Therefore, the singular entropy feature of different measuring points can be carried out being augmented fusion and form hurt proper vector.More than analyze and show that singular entropy can characterize the hurt feature of the different operating mode of track switch diverse location preferably.
(4) the different operating mode vibration signal roads of track switch totally 172 groups of sample datas are gathered respectively, wherein, nominal situation 60 groups, each 56 groups of crackle 0.5cm and crackle 1.5cm.According to step (1) ~ (3)) obtain respective hurt proper vector and input LSSVM training and test.Adopt grid search and cross validation to carry out optimizing to LSSVM penalty factor and radial basis function parameter, obtain γ=1.63, σ=7.31.Grid search and cross validation parameter optimization result are as shown in Figure 7.
The test results of single measuring point and 3 fusions are as shown in table 1, can find out, based on the comprehensive analysis of multi-measuring point vibration signal, have and make full use of multi-sensor information redundancy and complementary advantage, the proper vector of single measuring point is after 3 fusion treatment, and the crackle hurt discrimination of test sample book reaches 91.25%.Show that this method can be effective to track switch hurt and detect.In addition, the hurt recognition time calculating the inventive method is 6.49s, and real-time is better.
The single measuring point of table 1 and 3 test results merged
As a comparison, adopt the patent No. for method described in CN102175768A is to the track switch vibration signal experiment Analysis of above-mentioned collection, result as shown in Figure 8.Can find out, the IMF Power Spectrum Distribution of the different operating mode of track switch is complicated, does not have obvious distinguishing characteristic, is difficult to according to method described in this patent document the identification realizing track switch operating mode and hurt type.
Noise testing:
Because actual track switch vibration signal is subject to Environmental Noise Influence in gatherer process, for assessing the noiseproof feature of the inventive method, the basis of original vibration signal is added white Gaussian noise, gauss heat source model, the impulsive noise of different signal to noise ratio (S/N ratio) respectively, and carries out emulation experiment.Wherein gauss heat source model by variance be 1 white Gaussian noise obtained by a fourth-order band-pass wave filter.Impulsive noise is by formula n'(k)=B (k) G (k) produces, wherein G (k) for average be 0, variance is the white Gaussian noise of 1; B (k) is Bernoulli process.
Under different signal to noise ratio (S/N ratio), as shown in Figure 9, can find out, the hurt discrimination of the inventive method reduces with the increase adding noise intensity the test result of the inventive method, and when signal to noise ratio (S/N ratio) is higher than 20dB, the method is little by three kinds of noise effects, good stability; When signal to noise ratio (S/N ratio) is lower than 20dB, noise is comparatively large to the interference of signal, and discrimination reduces to decline comparatively fast with signal to noise ratio (S/N ratio).Even if but when signal to noise ratio (S/N ratio) is 5dB, discrimination still can reach about 68%, show that the inventive method has good noise immunity.Field experiment shows, the signal to noise ratio (S/N ratio) of the vibration signal that train gathers when turnout passing is generally higher than 20dB.Therefore, the present invention can realize the automatic detection of track switch crackle hurt at Railway Site.
Claims (4)
1., based on a high-speed switch crackle hurt intelligent detecting method for vibration signal, it is characterized in that, comprise the steps:
(1) according to the Changing Pattern of the single span free beam Mode Shape of track switch, in the switch blade, trouble of high-speed switch and trouble tail three measuring point places vibration acceleration sensors are installed;
(2) train is when turnout passing, gathers the track switch vibration signal at three measuring point places respectively; Without loss of generality, the vibration signal at note switch blade measuring point place is x (i), i=1 ..., N, N are sampling number;
(3) adopt complete set empirical mode decomposition CEEMD to carry out adaptive decomposition to x (i), obtain M the intrinsic modal components IMF comprising track switch hurt information, be designated as c
j(i), j=1,2 ..., M;
(4) IMF component c is calculated
jthe related coefficient of (i) and former vibration signal x (i)
wherein,
with
represent x (i) and c respectively
jthe mean value of (i);
(5) η is chosen
jbe greater than the main IMF component of m IMF component as this measuring point vibration signal of threshold value T, be designated as
k=1,2 ..., m;
(6) main IMF component is calculated
singular entropy E
kand by this m singular entropy E
k(k=1,2 ..., m) as the track switch hurt eigenwert of switch blade measuring point, be designated as H
1={ E
1, E
2... E
m;
(7) in trouble, the vibration signal of trouble tail measuring point collection performs step (3)-step (6) respectively, obtains branching off the track switch hurt eigenwert H at neutralization trouble tail measuring point place
2, H
3; Then to H
1, H
2, H
3carry out being augmented Fusion Features, form the track switch hurt proper vector H={H of 3m dimension
1, H
2, H
3;
(8) hurt proper vector H is inputted Least square support vector (LSSVM) model, choose radial basis function as kernel function, utilize grid search and cross validation to carry out optimizing to LSSVM penalty factor and radial basis function parameter, and then realize the judgement of turnout work state and hurt type.
2. a kind of high-speed switch crackle hurt intelligent detecting method based on vibration signal as claimed in claim 1, is characterized in that, the threshold value T=0.1 in described step (5).
3. a kind of high-speed switch crackle hurt intelligent detecting method based on vibration signal as claimed in claim 1, is characterized in that, the main IMF component of described step (6)
singular entropy E
kcalculation procedure is as follows:
(6.1) right
carry out phase space reconfiguration; Will
be embedded into (N-n+1) × n to tie up in phase space, obtain continuous wavelet transform track matrix X
Wherein n is Embedded dimensions, and N is that signal sampling is counted;
Svd is carried out to X, namely
X=U Λ V
t(3) wherein,
with
for orthogonal matrix,
for diagonal matrix, and q meets q=min (N-n+1, n); λ
j(j=1,2 ..., q) be called the singular value of matrix X;
(6.2) calculate
singular entropy E
k; From λ
j(j=1,2 ..., choose a front v maximum singular value q), and v meets
then singular entropy E
kfor
wherein,
for the weight of a jth singular value in whole singular value.
4. a kind of high-speed switch crackle hurt intelligent detecting method based on vibration signal as claimed in claim 1, it is characterized in that, the parameter optimization step of described step (8) is as follows:
(8.1) initialization penalty factor γ ∈ [e
-5, e
5], kernel functional parameter σ ∈ [e
-5, e
5], sizing grid gets 10 × 10, obtains 100 groups of parameters pair altogether;
(8.2) training sample data are divided into 10 groups, to (γ, σ), following operation are performed to each group parameter in grid:
(8.3) wherein one group of sample data is as test set in selection, and all the other 9 groups, as training set, obtain the predicated error δ of LSSVM;
(8.4) repeat step (8.3) to perform 10 times, all select different subset as training set at every turn, and the predicated error that 10 experiments obtain is averaged, obtain the predicated error of this group (γ, σ)
(8.5) parameter sets (γ, σ is changed
2), repeated execution of steps (8.3) and (8.4), obtain the predicated error of LSSVM under various combination parameter successively
using minimum for predicated error average one group of parameter as the optimization model parameter combinations in grid.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510310524.5A CN105203645B (en) | 2015-06-08 | 2015-06-08 | A kind of high-speed switch crackle hurt intelligent detecting method based on vibration signal fusion |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510310524.5A CN105203645B (en) | 2015-06-08 | 2015-06-08 | A kind of high-speed switch crackle hurt intelligent detecting method based on vibration signal fusion |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105203645A true CN105203645A (en) | 2015-12-30 |
| CN105203645B CN105203645B (en) | 2018-03-13 |
Family
ID=54951436
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510310524.5A Expired - Fee Related CN105203645B (en) | 2015-06-08 | 2015-06-08 | A kind of high-speed switch crackle hurt intelligent detecting method based on vibration signal fusion |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105203645B (en) |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105740840A (en) * | 2016-02-29 | 2016-07-06 | 中南大学 | Nonlinear identification method for rock fracture signal and blasting vibration signal |
| CN105784936A (en) * | 2016-03-09 | 2016-07-20 | 西安交通大学 | Method and system for quickly detecting damage to composite material plate |
| CN106017879A (en) * | 2016-05-18 | 2016-10-12 | 河北工业大学 | Universal circuit breaker mechanical fault diagnosis method based on feature fusion of vibration and sound signals |
| CN106203382A (en) * | 2016-07-20 | 2016-12-07 | 河海大学 | A kind of excitation surge current based on kernel function extreme learning machine and fault current recognition methods |
| CN106556647A (en) * | 2016-11-18 | 2017-04-05 | 湘潭大学 | A kind of impact echo data processing method |
| CN108937847A (en) * | 2017-05-22 | 2018-12-07 | 北京大学 | A method of evaluation human body sports coordination |
| CN109085248A (en) * | 2018-07-03 | 2018-12-25 | 内蒙古科技大学 | Localization method, the apparatus and system of bearing pipe wall impulse source |
| CN109827082A (en) * | 2019-03-13 | 2019-05-31 | 常州大学 | A kind of pipeline multipoint leakage accurate positioning method |
| CN109975013A (en) * | 2019-04-30 | 2019-07-05 | 南京工程学院 | Gearbox of wind turbine fault signature extracting method based on IVMD-SE |
| CN110274957A (en) * | 2019-07-19 | 2019-09-24 | 西安交通大学 | One kind damping greatly laminated structural interface quality of adhesive detection method |
| CN111344210A (en) * | 2017-10-30 | 2020-06-26 | 科路实有限责任公司 | Method for determining element characteristics of a railway element |
| CN111721324A (en) * | 2020-05-15 | 2020-09-29 | 中铁第一勘察设计院集团有限公司 | Contact net dropper breakage detection method based on acceleration signals |
| CN113255577A (en) * | 2021-06-18 | 2021-08-13 | 中铁大桥科学研究院有限公司 | Active control intelligent data processing method for construction vibration parameters of cable-stayed bridge |
| CN113501028A (en) * | 2021-07-07 | 2021-10-15 | 中国铁道科学研究院集团有限公司 | Method and device for diagnosing poor welded joint of heavy-duty railway steel rail |
| CN114166943A (en) * | 2021-12-03 | 2022-03-11 | 四川西南交大铁路发展股份有限公司 | Active rail damage monitoring method and terminal based on nonlinear ultrasonic waves |
| WO2022134868A1 (en) * | 2020-12-25 | 2022-06-30 | 杭州浙达精益机电技术股份有限公司 | Turnout switch rail damage state monitoring method based on feature fusion |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| DE19912785A1 (en) * | 1999-03-12 | 2000-10-05 | Siemens Ag | Railway points rail defect detection device |
| CN102175768A (en) * | 2011-02-22 | 2011-09-07 | 哈尔滨工业大学 | Method and device for detecting defects and failures of high-speed rail based on vibration signals |
| CN103175897A (en) * | 2013-03-13 | 2013-06-26 | 西南交通大学 | High-speed turnout damage recognition method based on vibration signal endpoint detection |
| CN103197001A (en) * | 2013-03-13 | 2013-07-10 | 西南交通大学 | High speed turnout injury identification method based on vibration signal wavelet threshold value denoising |
-
2015
- 2015-06-08 CN CN201510310524.5A patent/CN105203645B/en not_active Expired - Fee Related
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| DE19912785A1 (en) * | 1999-03-12 | 2000-10-05 | Siemens Ag | Railway points rail defect detection device |
| CN102175768A (en) * | 2011-02-22 | 2011-09-07 | 哈尔滨工业大学 | Method and device for detecting defects and failures of high-speed rail based on vibration signals |
| CN103175897A (en) * | 2013-03-13 | 2013-06-26 | 西南交通大学 | High-speed turnout damage recognition method based on vibration signal endpoint detection |
| CN103197001A (en) * | 2013-03-13 | 2013-07-10 | 西南交通大学 | High speed turnout injury identification method based on vibration signal wavelet threshold value denoising |
Non-Patent Citations (3)
| Title |
|---|
| 张超 等: "奇异值熵和支持向量机的齿轮故障诊断", 《振动、测试与诊断》 * |
| 秦娜 等: "高速列车转向架故障的经验模态熵特征分析", 《交通运输工程学报》 * |
| 陆森林 等: "CEEMD-FFT在滚动轴承故障诊断中的应用", 《郑州大学学报（工学版）》 * |
Cited By (23)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105740840A (en) * | 2016-02-29 | 2016-07-06 | 中南大学 | Nonlinear identification method for rock fracture signal and blasting vibration signal |
| CN105784936A (en) * | 2016-03-09 | 2016-07-20 | 西安交通大学 | Method and system for quickly detecting damage to composite material plate |
| CN106017879A (en) * | 2016-05-18 | 2016-10-12 | 河北工业大学 | Universal circuit breaker mechanical fault diagnosis method based on feature fusion of vibration and sound signals |
| CN106017879B (en) * | 2016-05-18 | 2018-07-03 | 河北工业大学 | Omnipotent breaker mechanical failure diagnostic method based on acoustic signal Fusion Features |
| CN106203382A (en) * | 2016-07-20 | 2016-12-07 | 河海大学 | A kind of excitation surge current based on kernel function extreme learning machine and fault current recognition methods |
| CN106556647A (en) * | 2016-11-18 | 2017-04-05 | 湘潭大学 | A kind of impact echo data processing method |
| CN106556647B (en) * | 2016-11-18 | 2019-02-22 | 湘潭大学 | A kind of impact echo data processing method |
| CN108937847A (en) * | 2017-05-22 | 2018-12-07 | 北京大学 | A method of evaluation human body sports coordination |
| CN111344210A (en) * | 2017-10-30 | 2020-06-26 | 科路实有限责任公司 | Method for determining element characteristics of a railway element |
| US11597418B2 (en) | 2017-10-30 | 2023-03-07 | Konux Gmbh | Method for determining an element characteristic of a railroad element |
| CN109085248B (en) * | 2018-07-03 | 2021-04-06 | 内蒙古科技大学 | Method, device and system for positioning impact source of pressure-bearing pipeline wall |
| CN109085248A (en) * | 2018-07-03 | 2018-12-25 | 内蒙古科技大学 | Localization method, the apparatus and system of bearing pipe wall impulse source |
| CN109827082B (en) * | 2019-03-13 | 2020-10-09 | 常州大学 | Pipeline multi-point leakage accurate positioning method |
| CN109827082A (en) * | 2019-03-13 | 2019-05-31 | 常州大学 | A kind of pipeline multipoint leakage accurate positioning method |
| CN109975013A (en) * | 2019-04-30 | 2019-07-05 | 南京工程学院 | Gearbox of wind turbine fault signature extracting method based on IVMD-SE |
| CN110274957A (en) * | 2019-07-19 | 2019-09-24 | 西安交通大学 | One kind damping greatly laminated structural interface quality of adhesive detection method |
| CN111721324A (en) * | 2020-05-15 | 2020-09-29 | 中铁第一勘察设计院集团有限公司 | Contact net dropper breakage detection method based on acceleration signals |
| CN111721324B (en) * | 2020-05-15 | 2021-12-07 | 中铁第一勘察设计院集团有限公司 | Contact net dropper breakage detection method based on acceleration signals |
| WO2022134868A1 (en) * | 2020-12-25 | 2022-06-30 | 杭州浙达精益机电技术股份有限公司 | Turnout switch rail damage state monitoring method based on feature fusion |
| CN113255577A (en) * | 2021-06-18 | 2021-08-13 | 中铁大桥科学研究院有限公司 | Active control intelligent data processing method for construction vibration parameters of cable-stayed bridge |
| CN113501028A (en) * | 2021-07-07 | 2021-10-15 | 中国铁道科学研究院集团有限公司 | Method and device for diagnosing poor welded joint of heavy-duty railway steel rail |
| CN113501028B (en) * | 2021-07-07 | 2022-08-09 | 中国铁道科学研究院集团有限公司 | Method and device for diagnosing poor welded joint of heavy-duty railway steel rail |
| CN114166943A (en) * | 2021-12-03 | 2022-03-11 | 四川西南交大铁路发展股份有限公司 | Active rail damage monitoring method and terminal based on nonlinear ultrasonic waves |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105203645B (en) | 2018-03-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105203645A (en) | Intelligent detection method of high-speed turnout crack damage based on vibration signal fusion | |
| Liu et al. | Advances of research on high-speed railway catenary | |
| CN104898696B (en) | High pressure based on electric-field intensity rate of change is the same as tower list back transmission line unmanned plane inspection barrier-avoiding method | |
| CN101907437B (en) | Wavelet difference algorithm-based cable fault localization method | |
| CN107256635B (en) | Vehicle identification method based on distributed optical fiber sensing in intelligent traffic | |
| CN103144937B (en) | System and method for intelligently monitoring belt-type conveyer for coal mine steel wire rope core | |
| CN101900708B (en) | Vibration and audio signal-based high-speed train track defect detecting method | |
| CN106443335A (en) | Lightning stroke fault identification method and system | |
| CN103175897B (en) | A kind of high-speed switch hurt recognition methods based on vibration signal end-point detection | |
| CN105424364A (en) | Diagnostic method and device of train bearing failure | |
| CN102829974A (en) | LMD (local mean decomposition) and PCA (principal component analysis) based rolling bearing state identification method | |
| CN108830328B (en) | Microseismic signal SMOTE identification method and monitoring system fusing spatial knowledge | |
| CN110285877B (en) | Train real-time positioning tracking and speed calculating method based on Spark Streaming | |
| CN101975909A (en) | Method for judging failure properties of shielding failure and counterattack of transmission line | |
| CN108848571B (en) | Rail transit safety monitoring system and monitoring method based on MEMS sensor | |
| CN106651031B (en) | Lightning stroke flashover method for early warning and system based on historical information | |
| CN103712806B (en) | High-speed train bogie serpentine locomotion detection, analytic system and method thereof | |
| CN103076168A (en) | Diagnosis method for mechanical faults of circuit breaker | |
| CN109605127A (en) | A kind of cutting-tool wear state recognition methods and system | |
| CN108645634A (en) | A kind of rail vehicle trouble-shooter | |
| CN103808801A (en) | Real-time detection method for high-speed rail injury based on vibration and audio composite signals | |
| CN104820168A (en) | Lightning stroke fault determination method based on waveform difference degree and lightning stroke fault sample database | |
| CN107257351A (en) | One kind is based on grey LOF Traffic anomaly detections system and its detection method | |
| CN104297628A (en) | Method for detecting and positioning section faults of power distribution network containing DGs (distributed generators) | |
| CN105626502A (en) | Plunger pump health assessment method based on wavelet packet and Laplacian Eigenmap |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20180313 Termination date: 20180608 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |