CN106596149A - Method for monitoring and diagnosing flat wheel state of rail vehicle based on VMD - Google Patents
Method for monitoring and diagnosing flat wheel state of rail vehicle based on VMD Download PDFInfo
- Publication number
- CN106596149A CN106596149A CN201611228519.0A CN201611228519A CN106596149A CN 106596149 A CN106596149 A CN 106596149A CN 201611228519 A CN201611228519 A CN 201611228519A CN 106596149 A CN106596149 A CN 106596149A
- Authority
- CN
- China
- Prior art keywords
- envelope spectrum
- vmd
- spectrum entropy
- rail vehicle
- wheel
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
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
- G01M17/00—Testing of vehicles
- G01M17/08—Railway vehicles
- G01M17/10—Suspensions, axles or wheels
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
Abstract
The invention discloses a method for monitoring and diagnosing a flat wheel state of a rail vehicle based on VMD. The method comprises the following steps of acquiring a vibration signal of a flat wheel through a sensor which is mounted in a rail; performing VMD variation modal decomposition on the vibration signal and obtaining a plurality of modals in different frequencies; constructing an envelope spectrum entropy characteristic for the plurality of modals; selecting an envelope spectrum entropy value with high consistency with the vibration signal from the envelope spectrum entropy characteristic; and inputting a plurality of selected envelope spectrum entropy values into a training support vector machine for performing identification, thereby identifying the working condition of the flat wheel. The method for monitoring and diagnosing the flat wheel state of the rail vehicle based on VMD has an advantage of high flat wheel working condition identification accuracy.
Description
Technical field
The invention belongs to flat wheel malfunction monitoring technical field, and in particular to a kind of flat wheel state of the rail vehicle based on VMD
Monitoring and diagnostic method.
Background technology
Wheel is the critical component of rail vehicle, under the background that train frequently raises speed, load-carrying is continued to increase, car body failure
Become complicated various, flat wheel failure is exactly one of them.The generation of flat wheel can cause vehicle bearing to damage, axle temperature is raised, rail
The problems such as ripple grinds, the operating mode for needing monitoring includes four kinds, and specially tread is normal, flat sliding, shelled tread and circumference mill
Consumption.Realizing the noncontact on-line monitoring method of flat wheel failure has a lot, and both at home and abroad common method has displacement detecting method, the signal of telecommunication
Detection method, vibration analysis method etc..Because vibration analysis law technology is more ripe, low cost extensively applies the online reality of height speed
When test and analyze.
Wherein empirical mode decomposition (EMD) is a kind of more inconvenient vibration analysis method, and it is a kind of self-adapting signal
Processing method, is widely applied in mechanical fault diagnosis,《Vibration engineering journal》In deliver " relevant rolling bearing therefore
Barrier EMD diagnostic method researchs ", are separated the high-frequency am signal that rolling bearing local damage is produced using EMD, using Hi
Lbert transformation calculations envelope spectrums, extract the Internal and external cycle fault characteristic frequency of rolling bearing.To realize flat wheel fault diagnosis,《Vibration
With impact》It is to realize stepping on reference to improved EMD methods and wavelet package transforms with regard to " wheel tread flat recognition methodss " in the book
Face abrades the quantitative identification of degree.There is the fault diagnosis that EEMD and approximate entropy are applied to train bogie critical component simultaneously
In, obtain preferable failure modes effect.But EMD easily causes modal overlap, so as to cause to omit important feature letter
Breath.
The content of the invention
In order to solve the above problems, it is an object of the invention to:A kind of rail vehicle based on VMD is provided and equals wheel state
Monitoring and diagnostic method, with the flat wheel operating mode accuracy rate of identification it is high the characteristics of.
For achieving the above object, the present invention is achieved by technical scheme below:
Rail vehicle based on VMD of the present invention puts down monitoring and the diagnostic method of wheel state, comprises the steps:
The vibration signal of wheel is put down by the sensor acquisition being installed in track;
VMD variation mode decompositions are carried out to the vibration signal, the mode under multiple different frequencies is obtained;
Envelope spectrum entropy feature is built to multiple mode;
The envelope spectrum entropy strong with the vibration signal concordance is chosen in envelope spectrum entropy feature;
To be identified in the multiple envelope spectrum entropy input Training Support Vector Machines chosen, the operating mode of the flat wheel of identification.
Further, VMD variations mode decomposition is carried out to the vibration signal using multiplication operator direction method, obtains multiple
Mode u under different frequencyk。
Further, the multiplication operator direction method is specially:
Modal components to the vibration signalMid frequencyOperatorInitialized;
Update uk、ωkAnd λ, concrete formula is:
Frequency centered on wherein ω, α be secondary multiplication factor, λ
For Lagrange multiplier operator;
Determine judgement precision e, according to formula is judged, during for meeting the judgement formula, stopping iteration output and decomposing
Mode uk。
Further, the judgement formula is specially:Wherein take e=1 × 10-6。
Further, the step of envelope spectrum entropy feature being built to multiple mode, specifically:
To mode ukCarry out Hilbert transform H [Uk(t)], envelope signal Z (t) is asked for, concrete formula is:
Fourier transformation is carried out to envelope signal Z (t), envelope spectrum Q is asked fori, concrete formula is Hi(w)=FFT (z
(t))、Qi=| Hi(w) |, wherein Hi (w) carries out the function of Fourier transformation to envelope signal Z (t);
The envelope spectrum is normalized, envelope spectrum entropy R is asked for, concrete formula is:
Wherein EiFor i-th (i=1,2,3,4....) the shared ratio in overall envelope spectrum of individual envelope spectrum, the envelope spectrum entropy R's
Collection is combined into envelope spectrum entropy feature.
Further, the envelope spectrum entropy strong with the vibration signal concordance, specially obtains Pearson came related
The big envelope spectrum entropy of coefficient.
Further, the big envelope spectrum entropy of the Pearson's correlation coefficient is located at the front position of envelope spectrum entropy feature.
Further, the Training Support Vector Machines are by by existing multiple known corresponding envelope spectrum entropy of operating mode
Input, then separates the envelope spectrum entropy under each operating mode, the reference model of the fixed mindset of formation.
Further, twice of the detection zone length of sensor place track more than flat wheel girth.
Further, the sensor includes at least 8 vibration acceleration sensors and at least 3 vehicle wheel speed sensings
Device;The vibration acceleration sensor is respectively symmetrically located at two inner sides of track;Two vehicle-wheel speed sensors are located at and shake
The front of dynamic acceleration transducer, remaining 1 vehicle-wheel speed sensor is located at the rear of vibration acceleration sensor, and
Three vehicle-wheel speed sensors are all located at the same side of track.
Compared with prior art, the invention has the beneficial effects as follows:
(1) using variation mode decomposition VMD and the flat wheel failure of envelope spectrum entropy analysis vehicle, the spy that can effectively extract failure
Levy frequency;
(2) after using variation mode decomposition VMD, according to component and the dependency of primary signal, suitable mode is chosen
Component, so as to improve the relatedness of feature;
(3) failure is identified using support vector machines, using VMD- envelope spectrums entropy as feature is extracted, is recognized
Accuracy rate realizes the accurate differentiation of flat wheel failure up to 100%.
Description of the drawings
The specific embodiment of the present invention is described in further detail below in conjunction with the accompanying drawings, wherein:
Fig. 1 is that the flow process of monitoring and diagnostic method that the rail vehicle based on VMD of the present invention equals wheel state is illustrated
Figure;
Fig. 2 is that the rail vehicle based on VMD of the present invention equals the monitoring of wheel state and multiplication operator in diagnostic method
The schematic flow sheet of direction method;
Fig. 3 is that the rail vehicle based on VMD of the present invention puts down the monitoring of wheel state and component envelope in diagnostic method
The schematic flow sheet of spectrum entropy feature;
Fig. 4 is that the rail vehicle based on VMD of the present invention equals the monitoring of wheel state and installation sensing in diagnostic method
The structural representation of device;
Fig. 5 is that the rail vehicle based on VMD of the present invention puts down adopt in the monitoring of wheel state and diagnostic method imitative
The discomposing effect figure of true signal and VMD;
Fig. 6 be adopt carry out emulation signal and the discomposing effect figure that EMD empirical mode decomposition modes are adopted;
Fig. 7 is that the rail vehicle based on VMD of the present invention puts down adopt in the monitoring of wheel state and diagnostic method imitative
U in true signal VMD decomposition1Spectrogram;
Fig. 8 is the spectrogram for carrying out the IMF1 in EMD empirical mode decompositions for adopting;
Fig. 9 is that the rail vehicle based on VMD of the present invention equals the reality adopted in the monitoring of wheel state and diagnostic method
Survey the discomposing effect figure of signal and VMD;
Figure 10 rail vehicles based on VMD of the present invention are put down during the monitoring of wheel state decomposes with VMD in diagnostic method
U1Envelope spectrogram;
Figure 11 is the monitoring operating mode different from diagnostic method that the rail vehicle based on VMD of the present invention equals wheel state
The scattergram of lower VMD- envelope spectrum entropy;
Figure 12 is the scattergram of the EMD- envelope spectrum entropy obtained using EMD empirical modal modes;
Figure 13 is the described monitoring and test set classification in diagnostic method that wheel state is equalled based on the rail vehicle of VMD of invention
Design sketch.
Specific embodiment
The preferred embodiments of the present invention are illustrated below in conjunction with accompanying drawing, it will be appreciated that preferred reality described herein
Apply example and be merely to illustrate and explain the present invention, be not intended to limit the present invention.
As shown in Fig. 1~Figure 13, the rail vehicle based on VMD of the present invention puts down the monitoring of wheel state and diagnosis side
Method, by the method using variation mode decomposition VMD, it is a kind of new mode estimation method, is asked by iterative search variation
Topic optimal solution, so as to the center frequency and bandwidth of decomposed component.Simultaneously variation mode decomposition VMD is applied to into flat wheel collision failure
In diagnosis, by the relative analyses with EMD, superioritys of the variation mode decomposition VMD in the fault type is diagnosed is drawn.This
It is bright variation mode decomposition VMD is incorporated in the flat wheel fault diagnosis of the vehicles such as subway first, the results show, the method energy
The flat wheel fault diagnosis of railcar is realized well.
As shown in figure 1, the rail vehicle based on VMD of the present invention puts down monitoring and the diagnostic method of wheel state, specifically
Comprise the steps:
S1:The vibration signal of wheel is put down by the sensor acquisition being installed in track;
Wherein, the sensor includes at least 8 vibration acceleration sensors and at least 3 vehicle-wheel speed sensors;Institute
State vibration acceleration sensor to be respectively symmetrically located at two inner sides of track;Two vehicle-wheel speed sensors accelerate located at vibration
The front of degree sensor, remaining 1 vehicle-wheel speed sensor is located at the rear of vibration acceleration sensor, and three institutes
State the same side that vehicle-wheel speed sensor is all located at track.
Meanwhile, it is to ensure that detection zone can at least collect the vibration signal for putting down wheel twice, need to make detection zone
Twice of the length more than flat wheel girth.
S2:VMD variation mode decompositions are carried out to the vibration signal, the mode under multiple different frequencies is obtained;
The step specifically adopts multiplication operator direction method, as shown in Fig. 2 obtaining mode u under multiple different frequenciesk;
The multiplication operator direction method is specially:
S21:Modal components to the vibration signalMid frequencyOperatorInitialized;
S22:Update uk、ωkAnd λ, concrete formula is:
Frequency centered on wherein ω, α be secondary multiplication factor, λ
For Lagrange multiplier operator;
S23:Judgement precision e is determined, according to judgement formulaWherein e > 0, in this enforcement
E=1 × 10 are taken in example-6, during for meeting the judgement formula, stopping iteration output and decomposing mode uk。
S3:Envelope spectrum entropy feature is built to multiple mode;As shown in Figure 3.
S31:To mode ukCarry out Hilbert transform H [Uk(t)], envelope signal Z (t) is asked for, concrete formula is:
S32:Fourier transformation is carried out to envelope signal Z (t), envelope spectrum Q is asked fori, concrete formula is Hi(w)=FFT
(z(t))、Qi=| Hi(w) |, wherein Hi (w) carries out the function of Fourier transformation to envelope signal Z (t);
S33:The envelope spectrum is normalized, envelope spectrum entropy R is asked for, concrete formula is:
Wherein EiFor the shared ratio in overall envelope spectrum of i-th envelope spectrum, it is special that the collection of the envelope spectrum entropy R is combined into envelope spectrum entropy
Levy.
S4:The envelope spectrum entropy strong with the vibration signal concordance is chosen in envelope spectrum entropy feature;
The envelope spectrum entropy strong with the vibration signal concordance, refers to many before envelope spectrum entropy feature
Individual envelope spectrum entropy.
Specifically, the Pearson's correlation coefficient of envelope spectrum entropy feature is obtained, wherein, Pearson's correlation coefficient is bigger, then table
The concordance for speaking frankly bright modal components and primary signal is better.The Pearson's correlation coefficient is phase between a kind of two variables of tolerance
The method of pass degree.It is a value between 1 and -1, wherein, 1 represents variable perfect positive correlation, and 0 represents unrelated, -1
Represent perfect negative correlation.
Meanwhile, by statistics, the envelope spectrum entropy positioned at front position is substantially than the envelope spectrum entropy positioned at back-end location
It is strong with the concordance of primary signal.
S5:To be identified in the multiple envelope spectrum entropy input Training Support Vector Machines chosen, the operating mode of the flat wheel of identification.
The Training Support Vector Machines are by existing multiple known floor data value inputs, then by under each operating mode
Data separate, the reference model of the fixed mindset of formation.
Specifically, in order to make normal data and fault data as separate as possible, by the corresponding envelope spectrum of multiple known conditions
Entropy is input in support vector machine, is then separated the envelope spectrum entropy under each operating mode so that support vector machine shape
Into a kind of mindset, so that by a unknown envelope spectrum entropy of input, into the Training Support Vector Machines, seeing this
Envelope spectrum entropy is near with a distance from normal data or fault data, no if close to normal data, it is classified as normal data
Then, then it is fault data.Meanwhile, if fault data, and be close to any fault data, so as to judge the concrete of flat wheel
Operating mode.
In order to better illustrate monitoring and diagnostic method that the rail vehicle based on VMD of the present invention equals wheel state
Principle, describes with reference to following concrete contrast experiment:
Scheme of installation as shown in Figure 4, i.e., be provided with 8 vibration acceleration sensors in two inner sides of track, is used for
The size of impulsive force in collection vibration processes.Wherein, each 4 per side, side (left side) is L1, L2, L3, L4, and opposite side is (right
Side) it is R1, R2, R3, R4, while being provided with vehicle-wheel speed sensor on the track that right side is located, cross for gathering wheel
Speed during track, wherein having two W1, W2 to be distributed in the front of R1, remaining 1 W3 is distributed in the rear of R4.
For the number of above vibration acceleration sensor and vehicle-wheel speed sensor can be carried out according to the actual requirements
Adjustment, but the vibration acceleration sensor on the track of the left and right sides is symmetrical, while bus or train route velocity pick-up
Device is also to may be located on the track that left side is located, and number can also be according to actual adjustment.
Step one:Collection vibration signal;
As shown in Figure 4, the spacing between vibration acceleration sensor is 1.8m, between vehicle-wheel speed sensor W1 and W2
Spacing for 1m, W2 and R1 spacing for 5m, R4 and W3 spacing be 5m.For the spacing of each sensor in diagram, it is also
Can finely tune, purpose is intended merely to more effectively collect required vibration signal.
In the case of arrangement above, the sample frequency in test is set to 7992Hz, and selection speed is 40Km/h, each car
Take turns the discrete point number that corresponding signal length is exactly that wheel crosses 10m effective coverages, about 7193-9590 data point, for convenience
Signal processing unification takes 8192 points, that is, the vibration signal for obtaining.
Step 2:VMD variation mode decompositions
Emulation signal x (t) for wherein adopting is set to:X (t)=sin (80 π t)+0.6sin (160 π t)+randn (size
(t)), the sample frequency for emulating signal x (t) is 1000Hz, and the sampling time is 1s, and randn functions are analogue noise sources.Emulation knot
Fruit is as follows:Fig. 5 is the discomposing effect for emulating signal and VMD, and Fig. 7 is U in the VMD decomposition of emulation signal1Spectrogram.
Meanwhile, in order to better illustrate advantage of the present invention using the variation mode decomposition method of VMD, with reference to adopting EMD
Isolation be compared, its emulation signal and discomposing effect such as Fig. 6 emulates the spectrogram of IMF1 in signal EMD decomposition such as
Shown in Fig. 8.
By contrast, Fig. 5 and Fig. 6 is contrasted, it can be seen that be compared to the discomposing effect of EMD, VMD is to emulating signal decomposition
Go out three components, compare the composition for having reappeared original signal for completing, and EMD decomposites four components, occurs in that modal overlap
Phenomenon, wherein IMF3 are false mode.Contrast Fig. 7 and Fig. 8, it can be seen that relative to decomposition of the EMD to noisy emulation signal,
VMD decomposes affected by noise little, can preferably decomposite the signal frequency of 40Hz and 80Hz.
Subway equals four kinds of common operating modes of wheel:Tread is normal, flat sliding, shelled tread and circular wear.This literary grace
Decomposed with VMD and circular wear is analyzed, and the radio-frequency component to mode carries out envelope spectrum analysis, experimental result is as follows:Fig. 9
For measured signal and the discomposing effect of VMD, Figure 10 is U in measured signal VMD decomposition1Envelope spectrogram.
The flat wheel failure of subway can generally excite high frequency natural frequency, therefore respectively to the high frequency mode of two kinds of decomposition methods,
Carry out envelope spectrum analysis.From figure 7 it can be seen that in U1In have an obvious peak value, what corresponding circular wear failure caused
Wheel Rail Vibration frequency.
From the contrast of Fig. 9 and Figure 10, the flat wheel failure of subway can generally excite high frequency natural frequency, therefore respectively to two kinds points
The high frequency mode of solution method, carries out envelope spectrum analysis.From fig. 9, it can be seen that in U1In have an obvious peak value, corresponding circle
The Wheel Rail Vibration frequency that all wear-out failures cause.
Step 3:Build envelope spectrum entropy feature
Pearson's correlation coefficient is that its his value of the statistical indicator of related intimate degree is bigger between a kind of reflection variable, is said
Degree of association is bigger between bright two signals.Consider and screen each sample, decompose the larger mode of 5 correlation coefficienies for producing
Component UkFollow-up feature extraction is carried out, modal components U that original vibration signal and VMD decompose are calculatedkBetween Pearson came it is related
Coefficient, as shown in Table 1.
U modal components U after the VMD of table one decompositionkWith the correlation coefficient of primary signal
By table one it can be seen that first three U after VMD decomposeskComponent is high with the degree of association of primary signal, calculates first three
The envelope spectrum entropy of component, and normalized is done as characteristic vector, its result is as shown in Table 2.
First three U after the VMD of table two decompositionkThe envelope spectrum entropy of component
To contrast the feature extraction effect of VMD- envelope spectrum entropy, then EMD empirical modals point are carried out to Wheel Rail Vibration signal
Solution, constructs EMD- envelope spectrum entropy characteristic vectors, is contrasted.The each unification of four kinds of tread operating modes takes 40 samples, is visualization
The effect of two kinds of feature extractions, takes the entropy of first three dimension of characteristic vector, shows result as shown in FIG. 11 and 12.
According to structure laid out above, the sample that VMD- envelope spectrum entropy characteristic vector is represented, quasi-cohesion ratio are may indicate that
Preferably, class inner boundary is clear, can improve the accuracy rate of automatic identification.
Step 4:The selection of envelope spectrum entropy
To verify good result of the VMD- envelope spectrums entropy in track car fault diagnosis, according to the method described above in failure
Feature extraction phases calculate 3 layers of WAVELET PACKET DECOMPOSITION third layer, 8 signal band energy of original vibration signal as characteristic vector.
Step 5:Recognize in input support vector machine
According to entropy theory, the fault feature vector under the different operating modes of construction, every kind of operating mode takes 40 groups of data, is then made
Input for support vector machine classifier is trained, and the model of grader is obtained by the training to training set, to test set
Classification prediction is carried out, every kind of operating mode takes at random 10 groups of test set samples, classifying quality below figure, according to Figure 13 as can be seen that institute
There is test set all correctly to be classified, achieve preferable classifying quality.
By the 8 signal band energy that will be obtained in step 4, it is identified with support vector machine, the classification for obtaining is accurate
True rate, as shown in Table 3.
The classification accuracy of the different characteristic extracting method of table three
Comparative test result with VMD- envelope spectrums entropy it is found that when carrying out discriminant classification using support vector machine, made
Vector is characterized, more preferable classifying quality can be obtained.
Rail vehicle based on VMD of the present invention puts down monitoring and the diagnostic method of wheel state, by with traditional EMD
Empirical mode decomposition mode is compared, and can obviously obtain following features:
(1) the flat wheel fault vibration signal of subway has non-stationary, emulation and shows with the analysis result of measured signal, profit
With variation mode decomposition and the flat wheel failure of envelope spectrum entropy analysis subway, the characteristic frequency of failure can be effectively extracted.Should be noted
, after decomposing using VMD, it should according to component and the dependency of primary signal, suitable modal components are chosen, to improve
The relatedness of feature.
(2) EMD and VMD is respectively adopted herein carries out feature extraction, recycles SVM to be identified failure, and experiment shows,
The accurate differentiation of flat wheel failure is realized up to 100% as feature, recognition accuracy is extracted using VMD- envelope spectrums entropy.
The above, is only presently preferred embodiments of the present invention, and any pro forma restriction is not made to the present invention, therefore
Every any modification without departing from technical solution of the present invention content, above example made according to the technical spirit of the present invention,
Equivalent variations and modification, still fall within the range of technical solution of the present invention.
Claims (10)
1. a kind of rail vehicle based on VMD puts down monitoring and the diagnostic method of wheel state, it is characterised in that comprise the steps:
The vibration signal of wheel is put down by the sensor acquisition being installed in track;
VMD variation mode decompositions are carried out to the vibration signal, mode u under multiple different frequencies is obtainedk;
Envelope spectrum entropy feature is built to multiple mode;
The envelope spectrum entropy strong with the vibration signal concordance is chosen in envelope spectrum entropy feature;
To be identified in the multiple envelope spectrum entropy input Training Support Vector Machines chosen, the operating mode of the flat wheel of identification.
2. the rail vehicle based on VMD according to claim 1 puts down monitoring and the diagnostic method of wheel state, and its feature exists
In:
VMD variations mode decomposition is carried out to the vibration signal using multiplication operator direction method, under obtaining multiple different frequencies
Mode uk。
3. the rail vehicle based on VMD according to claim 2 puts down monitoring and the diagnostic method of wheel state, and its feature exists
In:
The multiplication operator direction method is specially:
Modal components to the vibration signalMid frequencyOperatorInitialized;
Update uk、ωkAnd λ, concrete formula is:
Frequency centered on wherein ω, α is secondary multiplication factor, and λ is drawing
Ge Lang multiplication operators;
Determine judgement precision e, according to formula is judged, during for meeting the judgement formula, stopping iteration output and decomposing mode
uk。
4. the rail vehicle based on VMD according to claim 3 puts down monitoring and the diagnostic method of wheel state, and its feature exists
In:
The judgement formula is specially:Wherein take e=1 × 10-6。
5. monitoring and the diagnostic method that wheel state is equalled based on the rail vehicle of VMD according to claim 3 or 4, its feature
It is:
The step of envelope spectrum entropy feature is built to multiple mode, specifically:
To mode ukCarry out Hilbert transform H [Uk(t)], envelope signal Z (t) is asked for, concrete formula is:
Fourier transformation is carried out to envelope signal Z (t), envelope spectrum Qi is asked for, concrete formula is Hi(w)=FFT (z (t)),
Qi=| HiW (), wherein Hi (w) carry out the function of Fourier transformation to envelope signal Z (t);
The envelope spectrum is normalized, envelope spectrum entropy R is asked for, concrete formula is:Its
Middle EiFor i-th (i=1,2,3,4 ... .) shared ratio in overall envelope spectrum of individual envelope spectrum, the collection of the envelope spectrum entropy R
It is combined into envelope spectrum entropy feature.
6. the rail vehicle based on VMD according to claim 5 puts down monitoring and the diagnostic method of wheel state, and its feature exists
In:
The envelope spectrum entropy strong with the vibration signal concordance, specially obtains the big envelope spectrum of Pearson's correlation coefficient
Entropy.
7. the rail vehicle based on VMD according to claim 6 puts down monitoring and the diagnostic method of wheel state, and its feature exists
In:
The big envelope spectrum entropy of the Pearson's correlation coefficient is located at the front position of envelope spectrum entropy feature.
8. the rail vehicle based on VMD according to claim 1 puts down monitoring and the diagnostic method of wheel state, and its feature exists
In:
The Training Support Vector Machines be by by the corresponding envelope spectrum entropy input of existing multiple known operating modes, then will be each
Envelope spectrum entropy under individual operating mode separates, the reference model of the fixed mindset of formation.
9. the rail vehicle based on VMD according to claim 1 puts down monitoring and the diagnostic method of wheel state, and its feature exists
In:
Twice of the detection zone length of sensor place track more than flat wheel girth.
10. monitoring and the diagnostic method that wheel state is equalled based on the rail vehicle of VMD according to claim 1 or 8, its feature
It is:
The sensor includes at least 8 vibration acceleration sensors and at least 3 vehicle-wheel speed sensors;
The vibration acceleration sensor is respectively symmetrically located at two inner sides of track;
Two vehicle-wheel speed sensors are located at the front of vibration acceleration sensor, 1 vehicle wheel speed sensing of residue
Device is located at the rear of vibration acceleration sensor, and three vehicle-wheel speed sensors are all located at the same side of track.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201611228519.0A CN106596149A (en) | 2016-12-27 | 2016-12-27 | Method for monitoring and diagnosing flat wheel state of rail vehicle based on VMD |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201611228519.0A CN106596149A (en) | 2016-12-27 | 2016-12-27 | Method for monitoring and diagnosing flat wheel state of rail vehicle based on VMD |
Publications (1)
Publication Number | Publication Date |
---|---|
CN106596149A true CN106596149A (en) | 2017-04-26 |
Family
ID=58604346
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201611228519.0A Pending CN106596149A (en) | 2016-12-27 | 2016-12-27 | Method for monitoring and diagnosing flat wheel state of rail vehicle based on VMD |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106596149A (en) |
Cited By (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107229795A (en) * | 2017-06-02 | 2017-10-03 | 东北大学 | A kind of milling parameter recognition methods based on variation mode decomposition and Energy-Entropy |
CN107563403A (en) * | 2017-07-17 | 2018-01-09 | 西南交通大学 | A kind of recognition methods of bullet train operating condition |
CN108760305A (en) * | 2018-06-13 | 2018-11-06 | 中车青岛四方机车车辆股份有限公司 | A kind of Bearing Fault Detection Method, device and equipment |
CN109059840A (en) * | 2018-05-29 | 2018-12-21 | 南京理工大学 | A kind of city rail vehicle wheel out of round is along detection method |
CN109059839A (en) * | 2018-05-23 | 2018-12-21 | 唐智科技湖南发展有限公司 | A kind of wheel tread loses diagnostic method, the apparatus and system of circle failure |
CN109556863A (en) * | 2018-06-13 | 2019-04-02 | 南京工业大学 | A kind of acquisition of large-scale turntable bearing Vibration Signal in Frequency Domain and processing method based on MSPAO-VMD |
CN109975044A (en) * | 2017-12-28 | 2019-07-05 | 中冶长天国际工程有限责任公司 | A kind of pallet fault detection method, apparatus and system |
CN110146294A (en) * | 2019-04-23 | 2019-08-20 | 莆田学院 | A kind of wind-driven generator vibrating failure diagnosis method and storage medium |
CN110726875A (en) * | 2019-12-02 | 2020-01-24 | 山东大学 | New energy flexible direct-current grid-connected transient harmonic detection method and system |
CN111238808A (en) * | 2020-02-04 | 2020-06-05 | 沈阳理工大学 | Compound fault diagnosis method for gearbox based on empirical mode decomposition and improved variational mode decomposition |
CN112200015A (en) * | 2020-09-16 | 2021-01-08 | 昆明理工大学 | Rolling bearing fault diagnosis method based on improved VMD |
CN112733692A (en) * | 2021-01-04 | 2021-04-30 | 润联智慧科技(西安)有限公司 | Fault prediction method and device based on integrated hybrid model and related equipment |
WO2021154172A1 (en) | 2020-01-28 | 2021-08-05 | Gokmen Sabri Haluk | Method for detection of flat wheel deformation by vibration measurement from rails |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105758644A (en) * | 2016-05-16 | 2016-07-13 | 上海电力学院 | Rolling bearing fault diagnosis method based on variation mode decomposition and permutation entropy |
CN106017926A (en) * | 2016-05-13 | 2016-10-12 | 山东理工大学 | Rolling bearing fault diagnosis method based on variational mode decomposition |
-
2016
- 2016-12-27 CN CN201611228519.0A patent/CN106596149A/en active Pending
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106017926A (en) * | 2016-05-13 | 2016-10-12 | 山东理工大学 | Rolling bearing fault diagnosis method based on variational mode decomposition |
CN105758644A (en) * | 2016-05-16 | 2016-07-13 | 上海电力学院 | Rolling bearing fault diagnosis method based on variation mode decomposition and permutation entropy |
Non-Patent Citations (9)
Title |
---|
FU WENLONG等: "Fault diagnosis for rolling element bearings with VMD time-frequency analysis and SVM", 《2015 FIFTH INTERNATIONAL CONFERENCE ON INSTRUMENTATION AND MEASUREMENT, COMPUTER, COMMUNICATION AND CONTROL》 * |
XIAOAN YAN等: "Compound fault diagnosis of rotating machinery based on OVMD and a 1.5-dimension envelope spectrum", 《MEASUREMENT SCIENCE AND TECHNOLOGY》 * |
孙洁娣等: "基于LMD包络谱熵及SVM的天然气管道微小泄漏孔径识别", 《机械工程学报》 * |
李志农: "基于局域均值分解包络谱和SVM的滚动轴承故障诊断方法研究", 《机械设计与制造》 * |
武英杰等: "变分模态分解在风电机组故障诊断中的应用", 《机械传动》 * |
潘玉娜等: "包络谱熵在滚动轴承性能退化评估中的应用", 《上海应用技术学院学报》 * |
石文磊等: "基于EMD和Hilbert包络谱的滚动轴承故障诊断", 《机械工程与自动化》 * |
程军圣等: "基于SVM 和EMD 包络谱的滚动轴承故障诊断方法", 《***工程理论与实践》 * |
郭翔: "城市轨道交通车辆平轮监测***设计与算法研究", 《万方学术期刊数据库》 * |
Cited By (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107229795A (en) * | 2017-06-02 | 2017-10-03 | 东北大学 | A kind of milling parameter recognition methods based on variation mode decomposition and Energy-Entropy |
CN107229795B (en) * | 2017-06-02 | 2019-07-19 | 东北大学 | A kind of milling parameter recognition methods based on variation mode decomposition and Energy-Entropy |
CN107563403A (en) * | 2017-07-17 | 2018-01-09 | 西南交通大学 | A kind of recognition methods of bullet train operating condition |
CN109975044A (en) * | 2017-12-28 | 2019-07-05 | 中冶长天国际工程有限责任公司 | A kind of pallet fault detection method, apparatus and system |
CN109059839A (en) * | 2018-05-23 | 2018-12-21 | 唐智科技湖南发展有限公司 | A kind of wheel tread loses diagnostic method, the apparatus and system of circle failure |
CN109059840A (en) * | 2018-05-29 | 2018-12-21 | 南京理工大学 | A kind of city rail vehicle wheel out of round is along detection method |
CN108760305A (en) * | 2018-06-13 | 2018-11-06 | 中车青岛四方机车车辆股份有限公司 | A kind of Bearing Fault Detection Method, device and equipment |
CN109556863A (en) * | 2018-06-13 | 2019-04-02 | 南京工业大学 | A kind of acquisition of large-scale turntable bearing Vibration Signal in Frequency Domain and processing method based on MSPAO-VMD |
CN110146294A (en) * | 2019-04-23 | 2019-08-20 | 莆田学院 | A kind of wind-driven generator vibrating failure diagnosis method and storage medium |
CN110726875A (en) * | 2019-12-02 | 2020-01-24 | 山东大学 | New energy flexible direct-current grid-connected transient harmonic detection method and system |
WO2021154172A1 (en) | 2020-01-28 | 2021-08-05 | Gokmen Sabri Haluk | Method for detection of flat wheel deformation by vibration measurement from rails |
EP4085241A4 (en) * | 2020-01-28 | 2023-07-05 | Gokmen, Sabri Haluk | Method for detection of flat wheel deformation by vibration measurement from rails |
CN111238808A (en) * | 2020-02-04 | 2020-06-05 | 沈阳理工大学 | Compound fault diagnosis method for gearbox based on empirical mode decomposition and improved variational mode decomposition |
CN111238808B (en) * | 2020-02-04 | 2021-08-17 | 沈阳理工大学 | Compound fault diagnosis method for gearbox based on empirical mode decomposition and improved variational mode decomposition |
CN112200015A (en) * | 2020-09-16 | 2021-01-08 | 昆明理工大学 | Rolling bearing fault diagnosis method based on improved VMD |
CN112733692A (en) * | 2021-01-04 | 2021-04-30 | 润联智慧科技(西安)有限公司 | Fault prediction method and device based on integrated hybrid model and related equipment |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106596149A (en) | Method for monitoring and diagnosing flat wheel state of rail vehicle based on VMD | |
Cui et al. | Quantitative trend fault diagnosis of a rolling bearing based on Sparsogram and Lempel-Ziv | |
CN100485342C (en) | Integrated supporting vector machine mixed intelligent diagnosing method for mechanical fault | |
Wang et al. | Optimization of segmentation fragments in empirical wavelet transform and its applications to extracting industrial bearing fault features | |
Ye et al. | Fault diagnosis of high-speed train suspension systems using multiscale permutation entropy and linear local tangent space alignment | |
CN104236911B (en) | A kind of train bogie bearing military service process monitoring and fault diagnosis system and method | |
CN109100143A (en) | Fault Diagnosis of Roller Bearings and equipment based on CEEMDAN and CFSFDP | |
Soualhi et al. | Pattern recognition method of fault diagnostics based on a new health indicator for smart manufacturing | |
Guo et al. | Rolling bearing fault classification based on envelope spectrum and support vector machine | |
CN110490071A (en) | A kind of substation's Abstraction of Sound Signal Characteristics based on MFCC | |
CN105275833A (en) | CEEMD (Complementary Empirical Mode Decomposition)-STFT (Short-Time Fourier Transform) time-frequency information entropy and multi-SVM (Support Vector Machine) based fault diagnosis method for centrifugal pump | |
CN103335617A (en) | Vibrational signal based railway track geometric deformation detection method | |
CN109827777A (en) | Rolling bearing fault prediction technique based on Partial Least Squares extreme learning machine | |
CN106796579A (en) | For the method and system of the defect in automatic detection rotary shaft | |
CN103048137A (en) | Fault diagnosis method of rolling bearing under variable working conditions | |
CN103499445A (en) | Time-frequency slice analysis-based rolling bearing fault diagnosis method | |
CN103674511A (en) | Mechanical wearing part performance assessment and prediction method based on EMD (empirical mode decomposition)-SVD (singular value decomposition) and MTS (Mahalanobis-Taguchi system) | |
EP3913453B1 (en) | Fault detection system and method for a vehicle | |
CN108254179A (en) | A kind of bullet train wheel set bearing method for diagnosing faults based on MEEMD arrangement entropys | |
CN106021789A (en) | Fuzzy-intelligence-based rail car suspension system fault classification method and system | |
CN114997218B (en) | Identification and detection method for polygonal abrasion of wheels of railway vehicles | |
CN105258789A (en) | Method and device for extracting vibration signal characteristic frequency band | |
CN110411766A (en) | The snakelike unstability detection method of train bogie, device, system and storage medium | |
CN107121285A (en) | A kind of bearing vibration signal fault feature extracting method | |
CN106383028A (en) | Gear case 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 | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170426 |