CN108398267A - High-speed train rail edge motion parameter self-adaptive identification method - Google Patents
High-speed train rail edge motion parameter self-adaptive identification method Download PDFInfo
- Publication number
- CN108398267A CN108398267A CN201810161027.7A CN201810161027A CN108398267A CN 108398267 A CN108398267 A CN 108398267A CN 201810161027 A CN201810161027 A CN 201810161027A CN 108398267 A CN108398267 A CN 108398267A
- Authority
- CN
- China
- Prior art keywords
- doppler
- window
- frequency
- signal
- parameter
- 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
- 238000000034 method Methods 0.000 title claims abstract description 21
- 238000009826 distribution Methods 0.000 claims abstract description 21
- 238000010276 construction Methods 0.000 claims description 11
- 239000011159 matrix material Substances 0.000 claims description 9
- 238000005457 optimization Methods 0.000 claims description 6
- 230000008569 process Effects 0.000 claims description 6
- 238000009825 accumulation Methods 0.000 claims description 3
- 230000009286 beneficial effect Effects 0.000 claims description 2
- 239000012141 concentrate Substances 0.000 claims description 2
- 238000006073 displacement reaction Methods 0.000 claims description 2
- 238000000926 separation method Methods 0.000 claims description 2
- 230000002159 abnormal effect Effects 0.000 claims 1
- 230000003044 adaptive effect Effects 0.000 abstract description 8
- 238000001914 filtration Methods 0.000 abstract 1
- 238000005070 sampling Methods 0.000 abstract 1
- 238000010586 diagram Methods 0.000 description 10
- 238000012937 correction Methods 0.000 description 5
- 238000001228 spectrum Methods 0.000 description 4
- 238000012545 processing Methods 0.000 description 3
- 238000001514 detection method Methods 0.000 description 2
- 238000003745 diagnosis Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 238000012952 Resampling Methods 0.000 description 1
- 230000004888 barrier function Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 230000003862 health status Effects 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000005096 rolling process Methods 0.000 description 1
- 230000005236 sound signal Effects 0.000 description 1
- 238000010183 spectrum analysis Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M13/00—Testing of machine parts
- G01M13/04—Bearings
- G01M13/045—Acoustic or vibration analysis
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B61—RAILWAYS
- B61K—AUXILIARY EQUIPMENT SPECIALLY ADAPTED FOR RAILWAYS, NOT OTHERWISE PROVIDED FOR
- B61K9/00—Railway vehicle profile gauges; Detecting or indicating overheating of components; Apparatus on locomotives or cars to indicate bad track sections; General design of track recording vehicles
-
- 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
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Mechanical Engineering (AREA)
- Acoustics & Sound (AREA)
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
Abstract
The invention discloses a high-speed train rail edge motion parameter self-adaptive identification method, which comprises the following steps: (1) down-sampling and filtering a track side microphone acquisition signal X (t) to obtain x (t); (2) performing short-time Fourier transform (STFT) on x (t) to obtain time-frequency distribution STFTx(t, f); (3) initializing a set of side-of-track model parameters γ { v, r, f0}; (4) frequency shift formula f based on parameters in step (3) and track-side signalsk(t) constructing a Doppler window w conforming to the Doppler time-frequency variation lawγ(t, f); (5) let x0(t)＝wγ(t,f)*STFTx(t, f); (6) doppler constructed by sequentially calculating from low frequency to high frequency in the whole time-frequency domainSignal energy values E corresponding to different frequency bands in the window area; 7) repeating the steps (3) to (6) until the maximum energy E is obtainedmaxCorresponding to the gamma {, v, r, f {0And the train motion parameter identification result is used as the train motion parameter identification result. The method improves the anti-noise capability and the adaptive degree of parameter estimation, and can be used for detecting the acoustic signal fault of the train bearing.
Description
Technical field
The present invention relates to the technical fields of bullet train wheel set bearing rail side Acoustic Based Diagnosis, and in particular to a kind of high speed
Train rail side kinematic parameter self-adaptive identification method.
Background technology
Contain in the train voice signal that wheel set bearing is sent out in high-speed cruising closely related with its health status
Information, in-orbit side install microphone collected sound signal and can carry out effective event to wheel set bearing by signal processing means
Barrier diagnosis has the advantages that contactless monitor, is at low cost and can find initial failure.However, running at high speed due to train,
Serious Doppler's time-frequency distortion can occur for acoustic signal by collected road, this meeting subsequent fault message of severe jamming carries
It takes, so distorted signal must be corrected, and the premise of distortion correction is the acquisition of train movement parameters.It is most of at present
The algorithm of train movement parameters extraction to there is adaptive degree mostly inadequate, the problem of depending on human intervention unduly.
The present invention is directed to improve the adaptive degree of algorithm, realization does not depend on the adaptive identification parameter of human intervention.
The estimation of the train movement parameters based on signal itself may be implemented in method provided by the invention, without relying on additional ranging, surveying
Fast sensor.
Invention content
The technical problem to be solved in the present invention is：Overcome the shortage of prior art, by time frequency analysis, instantaneous frequency distilling,
Doppler's window is built and function-fitting method, may be implemented to get parms based on collected rail side acoustic signal is adaptive, can
Apply to distortion correction.The structure of Doppler's window proposed by the present invention may also reach up the filter effect of variable frequency range, in this way may be used
It is final to improve the acoustic signal fault message extraction of rail side to eliminate some strong background noises close with major frequency components signal frequency
Effect, improve practicability.
The present invention solve the technical solution that uses of above-mentioned technical problem for：A kind of bullet train rail side kinematic parameter is adaptive
Recognition methods, using the microphone acquisition train high speed mounted on rail on both sides by when the knocking noise message that sends out of wheel set bearing
Number, as detection signal X (t), the processing step to the detection signal is：
Acoustic signal Doppler center of distortion frequency f by step (1-1), roadk(t) it builds；
Step (1-2), the window w γ (t, f) for meeting Doppler's Distortion Law based on above-mentioned parameter convergent movement Construction of A Model,
And the centre frequency f for Doppler's window that each group of parametric configuration comes out is moved successively from low to high in signal time-frequency distributions
It is dynamic, a Δ f is often moved so just and will produce new Doppler's window.
Step (1-3), the local signal energy value E for calculating all Doppler's windows region successively find ceiling capacity
Value Emax, the corresponding one group of parameter of Energy maximum value is exactly optimized parameter.
In the step (1-1), centre frequency fk(t) the step of building is as follows：
Step (2-1) sound source motion parameter set parameter set γ (v, r, f here0) in do not have include sound source motion model cross
To distance.This is all highly beneficial to the high efficiency and stability of algorithm.During algorithm actual match, as long as calculating
Go out the duration T of signal, so that it may which the lateral distance s to define motion model is：
S=vT/2
Step (2-2) develops the motion model of sound-source signal to obtain new motion model function.
Here c/v is exactly above-mentioned Mach number M, fkIt is the f in motion parameter set γ0One group of parameter.It can define
For fk={ f1...fk, k=1...Z }.
In the step (1-2), the construction step of Doppler's window is as follows：
Step (3-1) combines motion parameter set γ (v, r, f0) harmony source movement pattern function structure meet train sound letter
Doppler's window of number Distortion Law：
V={ v1...vi, i=1...L }
R={ r1...rj, j=1...M }
F={ f1...fk, k=1...Z }
Here L, M and Z are the fit range length of the parameter of these three movements, so the match cognization in total of this algorithm
Number is L*M*Z.Each group of kinematic parameter can construct Doppler's distortion curve.
Step (3-2) and then the Doppler's distortion curve constructed according to this reset one in time-frequency matrix
The section a Δs f moved up and down.Each group of Doppler's window can be expressed as in this way：
Boundary=fk(t) ± a Δ f (k=1...Z, t=1...N/Fs)
One time-frequency Doppler's window w can be constructed by this group of boundaryγ(t,f).Here Δ f is defined as the resolution of frequency shift
The width of rate, each group in this way Doppler's window constructed is 2a Δs f.Then each group of Doppler's window meeting for being constructed out
It is moved to high frequency treatment from the low frequency of time-frequency matrix.Around this principle, the entire of signal can be distributed in there are many Doppler's window
In time-frequency domain, each Doppler's window can be chosen in the time-frequency distributions of signal and correspond to signal distributions.
In the step (1-3), the search of local signal Energy maximum value and the step of Parameter optimization, are as follows：
The energy accumulation of each pixel of entire time-frequency matrix is got up to indicate the gross energy of signal by step (4-1).This
In use x1(t) indicate that the local signal chosen by Doppler's window, the principle that local signal is chosen can be summarized as：
Here x1(t) it can be expressed as：
x1(t)=x (t) S ' (t, f)
x1(t) it is the local signal chosen by Doppler's window, the energy value of this local signal can be in entire time-frequency square
It is calculated in battle array.
Step (4-2) is by local signal x1(t) each element point, which add up, can be obtained by x1(t) energy and,
As follows,
The signal energy for including in the Doppler's window constructed in this way is bigger, and the energy value being calculated is bigger.According to
The motion model of sound source can show that the energy value for the signal in Doppler's window that only optimal parameter is constructed is maximum
's.
Step (4-3) parameter optimization in time-frequency distributions, brighter region indicates that signal energy is higher, so more accurate
Region where Doppler's window of sound source kinematic parameter construction more concentrates on the high region of signal energy.So here according to all
The signal energy value being calculated carries out optimizing and can be obtained by optimal sound source kinematic parameter.The searching process is expressed as：
E(i,j,k)=max (max (max (E(i,j,k))))
V=vi
R=rj
F=fk
The advantages of the present invention over the prior art are that：Believed using structure Doppler's auto-adapted fitting rail side train acoustics
Number kinematic parameter method, can realize the identification parameter of monitoring system self-adaption, and examine independent of additional sensor
It surveys.Not only make the installation of detecting system easier in this way, also reduces the cost of system, and the parameter identified includes movement
Geometric parameter and frequency parameter；The structure of Doppler's window is realized compared with conventional method with interior de-noising, is the equal of that one kind can
Become the bandpass filter of frequency band；This gradually structure Doppler's window goes the method for calculating local signal energy extreme value ensureing parameter
In the case of precision, the adaptive degree of algorithm is greatly improved so that algorithm has good practicability.
Description of the drawings
Fig. 1 is the program flow diagram of sound source kinematic parameter adaptive recognition algorithm；
Fig. 2 is train sound source motion model schematic diagram；
Fig. 3 is signal energy schematic diagram in Doppler's window principle and window, wherein Fig. 3 (a) is that Doppler's window shows
It is intended to, Fig. 3 (b) is window self-energy schematic diagram, and Fig. 3 (c) is window distribution schematic diagram in time-frequency matrix；
Fig. 4 is inner ring local signal optimizing schematic diagram, wherein Fig. 4 (a) is to owe excellent tectonic window area schematic, Fig. 4 (b)
For optimal construction window area schematic, Fig. 4 (c) was excellent tectonic window area schematic；
Fig. 5 is inner ring signal window self-energy with sound source kinematic parameter distribution map, wherein Fig. 5 (a) is fore-and-aft distance r optimizing
Result schematic diagram, Fig. 5 (b) are speed v optimizing result schematic diagrames；
Fig. 6 is inner ring signal processing results schematic diagram, wherein Fig. 6 (a) is correcting signal time-frequency figure, and Fig. 6 (b) is correction
Signal Time Domain Amplitude is composed, and Fig. 6 (c) is correcting signal frequency spectrum, and Fig. 6 (d) is correcting signal envelope spectrum.
Specific implementation mode
Below in conjunction with the accompanying drawings and specific implementation case further illustrates the present invention.
Here it is verified using the inner ring Single Point of Faliure signal of true wheel set bearing, model China active service of bearing
Cylinder rolling bearing NJ (P) 3226X1 that lorry uses.It is real that the vibration signal of collected bearing tester is subjected to rail side
It tests, then acquires the rail side acoustic signal of bearing by microphone, the motion model of sound source experimental signal is as shown in Figure 2.Experimental signal
Sample frequency 50kHz, down-sampled is 10kHz.Bearing sound-source signal time domain waveform is as shown in figure 3, harmonic signal centre frequency is
Interior ring signal of the position 1800Hz, Fig. 4 based on this experiment carries out the local signal optimizing of time-frequency domain, obtains sound source in Figure 5 most
Excellent kinematic parameter, finally it can be seen that the time-frequency of signal distorts and corrected and in envelope spectrum in time-frequency distributions in figure 6
In extracted faint fault message.
It is as follows：
Step (1-1) obtains the signal x (t) of ambient noise based on time-frequency distributions to signal X (t) progress mean filters；
Acoustic signal Doppler center of distortion frequency f by step (1-2), roadk(t) it builds；
Step (1-3), the window w for meeting Doppler's Distortion Law based on above-mentioned parameter convergent movement Construction of A Modelγ(t, f),
And the centre frequency f for Doppler's window that each group of parametric configuration comes out is moved successively from low to high in signal time-frequency distributions
It is dynamic, a Δ f is often moved so just and will produce new Doppler's window, and detailed process is referring to Fig. 3 (a) and Fig. 3 (c) institutes
Show；
Step (1-4), the local signal energy value E for calculating all Doppler's windows region successively find ceiling capacity
Value Emax, the corresponding one group of parameter of Energy maximum value is exactly optimized parameter, as shown in Fig. 3 (b)；
Step (1-5) corrects distorted signal according to the optimal sound source kinematic parameter of identification, identifies fault characteristic information.
It is as follows that the step (1-1) is based on the step of time-frequency distributions carry out mean filter to signal X (t)：
Step (2-1) passes through the time-frequency distributions of Short Time Fourier Transform STFT signal Analysis first：
Here it is a kind of nonlinear distortion varying signal that the reason of selection STFT carries out signal time frequency analysis, which is rail side acoustic signal,
Frequency displacement can occur over time for the frequency of signal.STFT can time domain and frequency domain all keep good resolution ratio, and
Observe the rule that signal frequency distorts at any time.
Step (2-2) carries out time-frequency figure using the mean filter method in image procossing in above-mentioned time-frequency distributions global
Threshold process.
S'(t, f)=S (t, f) >=δ
Wherein, S (t, f) indicates that the time-frequency distributions of signal, S ' (t, f) indicate the signal time-frequency distributions after mean filter, δ
It is defined as the energy threshold of time-frequency image, can be defined as：
Wherein, K is a weights, is set as 15 based on experience value.I is the length of the time arrow of time-frequency matrix, and j is frequency
The length of rate vector；
In the step (1-2), center of distortion frequency fk(t) the step of building is as follows：
Step (2-1) is in sound source motion parameter set parameter set γ (v, r, f0) in, sound source movement velocity v is according to practical feelings
Condition progress is preset, and parameter section is the interval 0.5m/s from 10m/s to 60m/s.Fore-and-aft distance r is according to experimental provision
It is set according to situation, section is divided into 0.1m from 1.9m to 2.1m.Frequency f0Setting section is 1000Hz to 2000Hz,
Interval is 50Hz.Here the lateral distance s of motion model is defined according to the duration T of movement velocity and signal.
S=vT/2
Step (2-2) develops the motion model of sound-source signal to obtain new motion model function.
Here c/v is exactly above-mentioned Mach number M, fkIt is the f in motion parameter set γ0One group of parameter.It can define
For fk={ f1...fk, k=1...Z }, t=[1/Fs...N/Fs], corresponding each moment.
In the step (1-3), the construction step of Doppler's window is as follows：
Step (3-1) combines motion parameter set γ (v, r, f0) harmony source movement pattern function structure meet train sound letter
Doppler's window of number Distortion Law：
V={ v1...vi, i=1...L }
R={ r1...rj, j=1...M }
F={ f1...fk, k=1...Z }
Here L, M and Z are respectively 100,10 and 20, so the match cognization number in total of this algorithm is 20000.It is each
Group kinematic parameter can construct Doppler's distortion curve.
Step (3-2) and then the Doppler's distortion curve constructed according to this reset one in time-frequency matrix
The interval Δ f moved up and down.Each group of Doppler's window can be expressed as in this way：
Boundary=fk(t) ± a Δ f (k=1...Z, t=1...N/Fs)
Here Δ f is defined as the resolution ratio of frequency shift, and the width of each group of Doppler's window constructed is determined by Δ f,
It is noted herein that Δ f have to it is smaller than frequency separation interval, this experimental setup be 10Hz.Here it is considered that algorithm tune
It is 3 that a scale a, which is arranged, in the convenience of examination.The building process of this step Doppler's window can be with parameter schematic diagram shown in Fig. 3.
In the step (1-4), the search of local signal Energy maximum value and the step of Parameter optimization, are as follows：
The energy accumulation of each pixel of entire time-frequency matrix is got up to indicate the gross energy of signal by step (4-1).This
In use x1(t) indicate that the local signal chosen by Doppler's window, the principle that local signal is chosen can be summarized as：
Here x1(t) it can be expressed as：
x1(t)=x (t) S ' (t, f)
The energy of signal is calculated generally in entire time-frequency distributions can be expressed as E=sum (S (t, f)).
Step (4-2) is by local signal x1(t) each element point, which add up, can be obtained by x1(t) energy and,
As follows,
The signal energy for including in the Doppler's window constructed in this way is bigger, and the energy value being calculated is bigger.According to
The motion model of sound source can show that the energy value for the signal in Doppler's window that only optimal parameter is constructed is maximum
's.
Step (4-3) parameter optimization in time-frequency distributions, as shown in figure 4, brighter region indicates that signal energy is higher, institute
The high region of signal energy is more concentrated on the region where Doppler's window of more accurate sound source kinematic parameter construction.Fig. 4 a),
(b) and in (c) result calculated is 384.75,321.51 and 406.00 respectively.Here according to all signal energy being calculated
Magnitude carries out optimizing and can be obtained by optimal sound source kinematic parameter.The searching process is expressed as：
E(i,j,k)=max (max (max (E(i,j,k))))
V=vi
R=rj
F=fk
In the step (1-5), correction distorted signal and identification fault characteristic information the step of it is as follows：
Step (5-1) derives interpolation time sequence { t according to movement relationr(i),1,2,…,N0, N0Be resampling it
Preceding signal length is 13333.
Step (5-2) calculates amplitude demodulation sequence { xd(i), 1,2 ..., N0, utilize { xd(i) } amplitude can be adjusted
The information of system is demodulated.{ t is used againr(i),1,2,…,N0The distortion that time domain interpolation eliminates time domain is carried out to demodulated signal.
Step (5-3) carries out envelope spectrum analysis to the signal after correction, extracts fault message, envelope as shown in FIG. 6
It can be seen that there is the failure-frequency of inner ring fault-signal at 195Hz in spectrum.
Above-described embodiment is provided just for the sake of the description purpose of the present invention, is not intended to limit the scope of the present invention.This hair
Bright range is defined by the following claims.It does not depart from spirit and principles of the present invention and the various equivalent replacements made and repaiies
Change, should all cover within the scope of the present invention.
Claims (4)
1. a kind of bullet train rail side kinematic parameter self-adaptive identification method, which is characterized in that realize that steps are as follows：
Acoustic signal Doppler center of distortion frequency f by step (1-1), roadk(t) it builds；
Step (1-2), the window w for meeting Doppler's Distortion Law based on above-mentioned parameter collection and motion model constructionγ(t, f), and will
The centre frequency f for Doppler's window that each group of parametric configuration comes out is moved successively from low to high in signal time-frequency distributions,
A Δ f is often moved so just and will produce new Doppler's window, and the construction of Doppler's window is as follows：
wγ(t, f)=fk(t) ± a Δ f (k=1...Z, t=1...N/Fs)
Wherein,
v,r,f0It is train movement parameters collection γ { v, r, f0Inner parameter, the train operation of the motion model of train is indicated respectively
The centre frequency of speed, fore-and-aft distance and signal, M are Mach numbers, are defined as the ratio of the speed and the theoretical velocity of sound of train, and s is
The lateral distance of motion model；
Step (1-3), the local signal energy value E for calculating all Doppler's windows region successively find maximum energy value
Emax。
2. bullet train rail according to claim 1 side kinematic parameter self-adaptive identification method, it is characterised in that：The step
Suddenly in (1-1), center of distortion frequency fk(t) the step of building is as follows：
Step (2-1) initializes sound source motion parameter set γ (v, r, f0), do not have in γ include sound source motion model lateral distance
S, this is all highly beneficial to the high efficiency and stability of raising algorithm, during algorithm actual match, as long as calculating
Go out the duration T of signal, so that it may which the lateral distance s to define motion model is：
S=vT/2
Step (2-2) develops original sound source motion model to obtain the sound source motion model function of variable number reduction,
Here c/v is exactly above-mentioned Mach number M, fkIt is the f in motion parameter set γ0One group of parameter, f can be defined ask
={ f1...fk, k=1...Z }.
3. bullet train rail according to claim 1 side kinematic parameter self-adaptive identification method, it is characterised in that：The step
Suddenly in (1-2), the construction step of Doppler's window is as follows：
Step (3-1) combines motion parameter set γ (v, r, f0) to meet train voice signal abnormal for harmony source movement pattern function structure
Become Doppler's window of rule：
V={ v1...vi, i=1...L }
R={ r1...rj, j=1...M }
F={ f1...fk, k=1...Z }
Here L, M and Z are the fit range length of the parameter of these three movements, and match cognization number in total is L*M*Z, each
Group kinematic parameter can construct Doppler's distortion curve fk(t)；
Step (3-2) sets Doppler's window frequency axial extent as 2a Δ f, and the up-and-down boundary of each group in this way Doppler's window is：
Boundary=fk(t) ± a Δ f (k=1...Z, t=1...N/Fs)
Time-frequency Doppler's window w can be constructed by this group of boundaryγ(t,f)；
Here Δ f is defined as the resolution ratio of frequency shift, and the width of each group of Doppler's window constructed is determined by Δ f, how general
The coboundary for strangling window is fkThe lower boundary of+a Δ f, Doppler's window are fk- a Δ f, so each group of Doppler's window constructed
Width is 2a Δ f, and then each group of Doppler's window for being constructed out can move from the low frequency of time-frequency matrix to high frequency treatment, root
According to this principle, can be distributed in the entire time-frequency domain of signal there are many Doppler's window, each Doppler's window can be in signal
In time-frequency distributions choose correspond to signal distributions, here why before frequency displacement separation delta f be arranged multiplier factor a, be in order to
The convenience design of algorithm debugging.
4. bullet train rail according to claim 1 or 2 side kinematic parameter self-adaptive identification method, it is characterised in that：Institute
It states in step (1-3), the search of local signal Energy maximum value and the step of Parameter optimization are as follows：
The energy accumulation of each pixel of entire time-frequency matrix gets up to indicate the gross energy of signal by step (4-1), uses x here1
(t) indicate that the local signal chosen by Doppler's window, the principle that local signal is chosen can be summarized as：
Here x1(t) it can be expressed as：
x1(t)=x (t) S ' (t, f)
x1(t) what is indicated is the signal for the local segments chosen by Doppler's window；
Step (4-2) is by local signal x1(t) each element point, which add up, can be obtained by x1(t) energy and, it is as follows
It is shown：
The signal energy for including in the Doppler's window constructed in this way is bigger, and the energy value being calculated is bigger, according to sound source
Motion model can show that the energy value for the signal in Doppler's window that only optimal parameter is constructed is the largest；
Step (4-3) parameter optimization in time-frequency distributions, brighter region indicates that signal energy is higher, so more accurate sound source
Region where Doppler's window of kinematic parameter construction more concentrates on the high region of signal energy, so here according to all calculating
Obtained signal energy value carries out optimizing and can be obtained by optimal sound source kinematic parameter, and the searching process is as follows, if,
E(i,j,k)=max (max (max (E(i,j,k))))
So,
V=vi
R=rj
F=fk。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810161027.7A CN108398267B (en) | 2018-02-27 | 2018-02-27 | High-speed train rail edge motion parameter self-adaptive identification method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810161027.7A CN108398267B (en) | 2018-02-27 | 2018-02-27 | High-speed train rail edge motion parameter self-adaptive identification method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108398267A true CN108398267A (en) | 2018-08-14 |
| CN108398267B CN108398267B (en) | 2019-12-03 |
Family
ID=63096670
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810161027.7A Active CN108398267B (en) | 2018-02-27 | 2018-02-27 | High-speed train rail edge motion parameter self-adaptive identification method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108398267B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109406147A (en) * | 2018-10-29 | 2019-03-01 | 安徽大学 | Train bearing rail side acoustic diagnosis method under variable speed working condition |
| CN112857767A (en) * | 2021-01-18 | 2021-05-28 | 中国长江三峡集团有限公司 | Hydro-turbo generator set rotor fault acoustic discrimination method based on convolutional neural network |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104568444A (en) * | 2015-01-28 | 2015-04-29 | 北京邮电大学 | Method for extracting fault characteristic frequencies of train rolling bearings with variable rotational speeds |
| CN205262744U (en) * | 2015-11-17 | 2016-05-25 | 苏州大学 | Train wheel pair bearing trouble transient state characteristic detection device based on parameterization doppler transient state model |
| CN107402131A (en) * | 2017-08-03 | 2017-11-28 | 安徽大学 | High-speed train motion parameter identification method based on rail-side acoustic signal time-frequency ridge line |
-
2018
- 2018-02-27 CN CN201810161027.7A patent/CN108398267B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104568444A (en) * | 2015-01-28 | 2015-04-29 | 北京邮电大学 | Method for extracting fault characteristic frequencies of train rolling bearings with variable rotational speeds |
| CN205262744U (en) * | 2015-11-17 | 2016-05-25 | 苏州大学 | Train wheel pair bearing trouble transient state characteristic detection device based on parameterization doppler transient state model |
| CN107402131A (en) * | 2017-08-03 | 2017-11-28 | 安徽大学 | High-speed train motion parameter identification method based on rail-side acoustic signal time-frequency ridge line |
Non-Patent Citations (2)
| Title |
|---|
| 刘星辰: "基于时频脊线旋转匹配的多普勒矫正方法研究", 《振动与冲击》 * |
| 张翱 等: "基于能量重心法的列车轴承多普勒畸变故障声信号校正诊断研究", 《振动与冲击》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109406147A (en) * | 2018-10-29 | 2019-03-01 | 安徽大学 | Train bearing rail side acoustic diagnosis method under variable speed working condition |
| CN109406147B (en) * | 2018-10-29 | 2020-11-13 | 安徽大学 | Train bearing rail side acoustic diagnosis method under variable speed working condition |
| CN112857767A (en) * | 2021-01-18 | 2021-05-28 | 中国长江三峡集团有限公司 | Hydro-turbo generator set rotor fault acoustic discrimination method based on convolutional neural network |
| CN112857767B (en) * | 2021-01-18 | 2022-03-11 | 中国长江三峡集团有限公司 | Hydro-turbo generator set rotor fault acoustic discrimination method based on convolutional neural network |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108398267B (en) | 2019-12-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108426713B (en) | Rolling bearing weak fault diagnosis method based on wavelet transformation and deep learning | |
| CN107356432B (en) | Fault Diagnosis of Roller Bearings based on frequency domain window experience small echo resonance and demodulation | |
| CN107402131A (en) | High-speed train motion parameter identification method based on rail-side acoustic signal time-frequency ridge line | |
| CN103424258A (en) | Fault diagnosis method for rolling bearing | |
| CN103699513B (en) | A kind of stochastic resonance method adjusted based on multiple dimensioned noise | |
| CN105424366A (en) | Bearing fault diagnosis method based on EEMD adaptive denoising | |
| CN108398267B (en) | High-speed train rail edge motion parameter self-adaptive identification method | |
| CN106872171A (en) | A kind of adaptive learning bearing calibration of Doppler's acoustic signal | |
| Li et al. | Adaptive cost function ridge estimation for rolling bearing fault diagnosis under variable speed conditions | |
| Shuuji et al. | Low-speed bearing fault diagnosis based on improved statistical filtering and convolutional neural network | |
| CN112904282A (en) | Radar interference signal identification method based on PWVD and convolutional neural network | |
| CN111504640B (en) | Weighted sliding window second-order synchronous compression S transformation bearing fault diagnosis method | |
| CN108844741A (en) | High-speed train bearing fault diagnosis method based on microphone uniform area array filtering | |
| CN109696906A (en) | Underwater robot propeller method for diagnosing faults based on small echo amendment Bayes's convolution energy | |
| CN114707558A (en) | Method and medium for extracting, classifying and identifying ice collapse infrasound characteristic | |
| CN103994820A (en) | Moving target identification method based on micro-aperture microphone array | |
| CN116522074A (en) | Rolling bearing signal noise reduction method based on adaptive window long time-frequency peak filtering | |
| CN113188797B (en) | Bearing fault diagnosis method based on microphone array | |
| CN109612730A (en) | A kind of rolling bearing fault localization method based on prewhitening analysis | |
| CN109670459A (en) | Helicopter Main Reducer fault sample generation method | |
| CN103714542B (en) | Extraction method for target highlight in low-resolution high-frequency sonar image | |
| CN109614887A (en) | A kind of vehicle whistle classification method based on support vector machines | |
| CN108061653A (en) | Train wheel set bearing rail edge sound signal separation method based on harmonic wave-impact Doppler modulation composite dictionary | |
| CN106125148B (en) | A kind of noise-reduction method and device for active cycle electromagnetic signal | |
| CN109738212A (en) | It is a kind of using frequency spectrum kurtosis as the adaptive Doppler antidote of optimizing index |
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 | ||
| CB03 | Change of inventor or designer information | ||
| CB03 | Change of inventor or designer information |
Inventor after: Liu Fang Inventor after: Liu Jiaqing Inventor after: Qian Qiang Inventor after: Fu Yangyang Inventor after: Gu Kangkang Inventor after: Liu Yongbin Inventor after: Lu Siliang Inventor after: Ju Bin Inventor before: Liu Fang Inventor before: Qian Qiang Inventor before: Fu Yangyang Inventor before: Gu Kangkang Inventor before: Liu Yongbin Inventor before: Lu Siliang Inventor before: Ju Bin |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |