CN117664577A - Vehicle-mounted train bearing fault diagnosis method, system and equipment - Google Patents
Vehicle-mounted train bearing fault diagnosis method, system and equipment Download PDFInfo
- Publication number
- CN117664577A CN117664577A CN202311694019.6A CN202311694019A CN117664577A CN 117664577 A CN117664577 A CN 117664577A CN 202311694019 A CN202311694019 A CN 202311694019A CN 117664577 A CN117664577 A CN 117664577A
- Authority
- CN
- China
- Prior art keywords
- signal
- sound source
- target bearing
- noise
- bearing sound
- 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
- 238000003745 diagnosis Methods 0.000 title claims abstract description 55
- 238000000034 method Methods 0.000 title claims abstract description 42
- 238000012545 processing Methods 0.000 claims abstract description 34
- 230000007613 environmental effect Effects 0.000 claims abstract description 29
- 230000006870 function Effects 0.000 claims abstract description 29
- 238000007781 pre-processing Methods 0.000 claims abstract description 13
- 239000011159 matrix material Substances 0.000 claims description 21
- 238000012549 training Methods 0.000 claims description 21
- 239000013598 vector Substances 0.000 claims description 18
- 230000009467 reduction Effects 0.000 claims description 16
- 238000011068 loading method Methods 0.000 claims description 14
- 230000001629 suppression Effects 0.000 claims description 11
- 238000004590 computer program Methods 0.000 claims description 10
- 238000002405 diagnostic procedure Methods 0.000 claims description 8
- 238000005457 optimization Methods 0.000 claims description 8
- 230000001133 acceleration Effects 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 6
- 238000005259 measurement Methods 0.000 claims description 5
- 238000007476 Maximum Likelihood Methods 0.000 claims description 3
- 230000003044 adaptive effect Effects 0.000 claims description 3
- 125000004122 cyclic group Chemical group 0.000 claims description 3
- 230000002452 interceptive effect Effects 0.000 claims description 3
- 230000003313 weakening effect Effects 0.000 claims description 3
- 238000012806 monitoring device Methods 0.000 abstract description 8
- 238000009434 installation Methods 0.000 abstract description 3
- 230000008030 elimination Effects 0.000 abstract description 2
- 238000003379 elimination reaction Methods 0.000 abstract description 2
- 238000004891 communication Methods 0.000 abstract 1
- 238000010586 diagram Methods 0.000 description 9
- 230000005236 sound signal Effects 0.000 description 8
- 230000000694 effects Effects 0.000 description 7
- 230000002159 abnormal effect Effects 0.000 description 6
- 238000004458 analytical method Methods 0.000 description 6
- 238000001514 detection method Methods 0.000 description 6
- 238000004422 calculation algorithm Methods 0.000 description 5
- 230000003993 interaction Effects 0.000 description 5
- 238000012544 monitoring process Methods 0.000 description 5
- 238000012360 testing method Methods 0.000 description 5
- 230000002829 reductive effect Effects 0.000 description 3
- 238000005070 sampling Methods 0.000 description 3
- PCTMTFRHKVHKIS-BMFZQQSSSA-N (1s,3r,4e,6e,8e,10e,12e,14e,16e,18s,19r,20r,21s,25r,27r,30r,31r,33s,35r,37s,38r)-3-[(2r,3s,4s,5s,6r)-4-amino-3,5-dihydroxy-6-methyloxan-2-yl]oxy-19,25,27,30,31,33,35,37-octahydroxy-18,20,21-trimethyl-23-oxo-22,39-dioxabicyclo[33.3.1]nonatriaconta-4,6,8,10 Chemical compound C1C=C2C[C@@H](OS(O)(=O)=O)CC[C@]2(C)[C@@H]2[C@@H]1[C@@H]1CC[C@H]([C@H](C)CCCC(C)C)[C@@]1(C)CC2.O[C@H]1[C@@H](N)[C@H](O)[C@@H](C)O[C@H]1O[C@H]1/C=C/C=C/C=C/C=C/C=C/C=C/C=C/[C@H](C)[C@@H](O)[C@@H](C)[C@H](C)OC(=O)C[C@H](O)C[C@H](O)CC[C@@H](O)[C@H](O)C[C@H](O)C[C@](O)(C[C@H](O)[C@H]2C(O)=O)O[C@H]2C1 PCTMTFRHKVHKIS-BMFZQQSSSA-N 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 238000001914 filtration Methods 0.000 description 2
- 230000000670 limiting effect Effects 0.000 description 2
- 230000004807 localization Effects 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000036961 partial effect Effects 0.000 description 2
- 230000002265 prevention Effects 0.000 description 2
- 238000005096 rolling process Methods 0.000 description 2
- 239000004065 semiconductor Substances 0.000 description 2
- 238000000926 separation method Methods 0.000 description 2
- WHXSMMKQMYFTQS-UHFFFAOYSA-N Lithium Chemical compound [Li] WHXSMMKQMYFTQS-UHFFFAOYSA-N 0.000 description 1
- 238000009825 accumulation Methods 0.000 description 1
- 230000006978 adaptation Effects 0.000 description 1
- XAGFODPZIPBFFR-UHFFFAOYSA-N aluminium Chemical compound [Al] XAGFODPZIPBFFR-UHFFFAOYSA-N 0.000 description 1
- 229910052782 aluminium Inorganic materials 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 239000000428 dust Substances 0.000 description 1
- 210000005069 ears Anatomy 0.000 description 1
- 238000002592 echocardiography Methods 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 239000000835 fiber Substances 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 229910052744 lithium Inorganic materials 0.000 description 1
- 230000007257 malfunction Effects 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 239000012528 membrane Substances 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000003032 molecular docking Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 238000001228 spectrum Methods 0.000 description 1
- 238000010183 spectrum analysis Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000001960 triggered effect Effects 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
-
- 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
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T90/00—Enabling technologies or technologies with a potential or indirect contribution to GHG emissions mitigation
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Acoustics & Sound (AREA)
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
Abstract
The invention belongs to the technical field of fault diagnosis, and discloses a vehicle-mounted train bearing fault diagnosis method, system and equipment, wherein the diagnosis method is based on a monitoring device for collecting acoustic signals, the device comprises a reference microphone and a double-ring microphone array, the system has the functions of collecting, storing, transmitting, network communication and fault diagnosis, and the diagnosis method comprises the following steps: acquiring multichannel data x (n) by adopting a double-ring microphone array, preprocessing to obtain d (n, k), and positioning a sound source; selecting D (n, k) according to the sound source positioning result and far field condition to perform signal enhancement processing to obtain Y (t), and eliminating the environmental noise in Y (t) by using the environmental noise signal D (n) acquired by the reference microphone to obtain a desired signal Y (n); and extracting time domain, frequency domain and wavelet domain features from Y (n) and establishing a diagnosis model. The invention creatively provides a double-ring microphone array and a reference microphone scheme, realizes accurate fault identification under the environment of a multi-target sound source through signal enhancement and environmental noise elimination, and has the advantages of strong installation adaptability, accurate diagnosis and the like.
Description
Technical Field
The invention belongs to the technical field of fault diagnosis, and particularly relates to a vehicle-mounted train bearing fault diagnosis method, system and equipment.
Background
In various rotary machines in the industrial field, bearings are one of the key components. The bearings are prone to failure due to the high speed, variable load conditions that are typical. Once a fault occurs, a serious accident may be caused. The on-line detection of the running state of the bearing can effectively avoid serious consequences caused by faults. Currently, there are several methods for monitoring and diagnosing bearings: vibration diagnostics, acoustic diagnostics, temperature diagnostics, wear debris analysis, bearing clearance measurement, oil film resistance, fiber optic monitoring techniques.
The vibration mode is most widely applied, a vibration sensor arranged on a bearing seat is used for collecting a bearing vibration signal, then fault diagnosis is realized through signal analysis, the prior Chinese patent with publication number of CN115326396A discloses a method and a device for diagnosing bearing faults, and particularly discloses a method for further processing by adopting a filtering reconstruction method on the basis of collecting the vibration signal, and fault diagnosis is performed by calculating envelope spectrum analysis. However, the quality of the vibration signal can directly influence the processing effect of the subsequent method, in actual use, the equipment often comprises a plurality of vibration sources, and the interference vibration sources can be transmitted to the vibration sensor on the bearing seat through the equipment body to interfere the vibration signal of the detected bearing, so that the difficulty of signal analysis is increased, and the diagnosis precision and reliability are reduced. Meanwhile, the vibration sensor needs to be installed on the bearing seat, equipment needs to be perforated and modified, and extra manpower and material resources are needed to be paid.
In the prior art, a non-contact acoustic measurement mode exists, and the equipment is not required to be modified. Meanwhile, compared with a vibration mode, the acoustic mode adopts the microphone to collect the sound signal opposite to the fault position, and the collection mode is more direct. Research shows that the fault sound signal can generate obvious characteristic change at the early stage of the fault, and the early warning capability is strong.
During operation of the motor vehicle, a number of acoustic signals are generated, some of which are generated by malfunctions of the running components of the motor vehicle, known as abnormal sounds. Through detection and identification of abnormal sounds of the motor car, a running part with faults of the motor car can be found in time, corresponding fault information is reported, powerful support is provided for subsequent fault processing, and running safety of the motor car is ensured. At present, detection and identification of abnormal sounds of a motor car mainly depend on a skilled worker to listen through ears, and then the type of the abnormal sounds is judged. The current method of manual listening to detect and identify abnormal sounds has great limitation: firstly, a train is usually provided with a worker, and the abnormal sound is heard by a walking back and forth mode, so that the mode is not efficient; secondly, the accuracy is limited by the hearing of workers at the time, is easy to be interfered by the environment, and cannot be ensured; again, the workers need long training and accumulation to master the "listen" technique, and the mode has long forming cycle and is not easy to popularize.
The environment in the train is complex and narrow, the reverberation effect is particularly obvious, and because the collected environment noise is more in variety, such as the sound of people talking, the vibration of walking, the mechanical noise of train components and the like, and the signals can generate stronger echoes and reverberation in the carriage, the signal collection of the target bearing sound source becomes difficult, and the method and the device for effectively monitoring the train bearing in real time are difficult. Conventional microphone array acquisition algorithms are typically applied to enhance speech, removing mechanical noise and stationary noise.
In summary, how to provide a non-contact fault diagnosis method based on multi-target bearing acoustic source acoustic signals is a problem that needs to be solved by the person skilled in the art.
Disclosure of Invention
The invention aims to provide a vehicle-mounted train bearing fault diagnosis method, system and equipment, which are used for solving the problems in the background technology.
The invention realizes the above purpose through the following technical scheme:
an on-board train bearing fault diagnosis method, comprising the steps of:
s1, when the running speed of a target bearing sound source meets a first condition, acquiring multichannel data of the target bearing sound source from a microphone array;
s2, preprocessing multichannel data, positioning a target bearing sound source signal, performing signal enhancement on the preprocessed signal based on far-field conditions and sound source distance of the target bearing sound source, generating a near-end signal and a far-end signal of the target bearing sound source, extracting environmental noise of the preprocessed signal, and taking the environmental noise as a reference signal for noise suppression so as to eliminate the environmental noise outside the target bearing sound source signal;
s3, extracting time domain, frequency domain and wavelet domain characteristics of the target bearing sound source signal, performing feature dimension reduction by adopting a dimension reduction method to generate a training set, inputting the training set into a pre-constructed diagnosis model for training, inputting a signal to be diagnosed into the trained diagnosis model for classification, and determining whether the target bearing sound source has faults.
As a further optimization scheme of the present invention, in step S1, the first condition is: the running speed is determined by satellite positioning speed measurement, a triaxial acceleration sensor and a band-pass filter in combination, so that the vibration of the target bearing sound source carrier is prevented from interfering acceleration acquisition.
As a further optimized scheme of the invention, the diagnostic method acquisition device comprises an annular microphone array, wherein the annular microphone array comprises an inner ring microphone for acquiring a near-end target bearing sound source, an outer ring microphone for acquiring a far-end target bearing sound source and a top microphone for acquiring environmental noise.
As a further optimization scheme of the present invention, in step S2, specifically:
s201, preprocessing the multichannel data x, weakening reverberation and echo components in signals, and processing by adopting short-time Fourier transform to obtain x 1 (n, k) outputting a preprocessing desired signal d (n, k) as:
wherein n is the channel number, k is the signal frame number, τ is the signal frame time interval, g (k) is the filter coefficient, input x (n, k) is the observed signal, the time-varying Gaussian model and the maximum likelihood estimation rule are used for cyclic iteration to estimate the variance lambda and the weight g,
through iterative loop optimization, variance and weight are obtained, and the variance and weight are substituted into expected signals to obtain expected signals d { d1, d2, d3...d 14, d15, d16};
s202, the far field condition calculation formula is as follows:
wherein L is the distance between the microphone array and the target bearing sound source, R is the radius of the circular array microphone array, and lambda is the signal frequency of the target bearing sound source;
s203, a signal enhancement flow is as follows: far-field signals are parallelly incident on the center of the microphone array at the azimuth angle theta, the number of delay units in the microphone array is L, the delay between adjacent delay units is T, and then the output of signal enhancement processing is y (T):
and solving through a Lagrangian function to obtain an optimal solution as follows:
s204, defining a matrix for diagonally loading and updating the sample covariance matrix as:
wherein R is -1 For the correlation matrix R -1 ＝αR -1 +λI，For the power of the white noise component in the noise, lambda and alpha are the diagonal loadings, LNR is the loading noise level, whereby the in-focus loading lambda is obtained from the noise power, and R is used according to the optimal weight vector solution xx Approximation xx H Instead of R -1 ＝αR xx -1 Carrying out +lambdaI, obtaining a new weight vector, and obtaining a one-dimensional signal y (t) in a specified direction according to signal enhancement processing;
s205, inputting the signal D (n) collected by the reference microphone into noise suppression to obtain the weight vector W of the noise signal k Setting an approximation factor beta according to the approximation degree of the interference sound source and the bearing sound source 1 ，β 2 ，β 3 ，β 4 ]；
Wherein the noise suppressed input signal vector X (n), W k Is a weight coefficient vector, and the output is Y (n) is：
The adaptive linear combiner is configured to calculate the mean square value ζ of the error signal as:
e(n)＝d(n)-y(n)；
ξ(n)＝E[e 2 (n)]；
according to the criterion of minimum mean square value of error signal, obtaining weight vector of noise, the desired signal is
Y(n)＝X(n)-β·W k ·X(n)。
As a further optimization scheme of the invention, in the feature dimension reduction in the step S3, an optimal approximation matrix of the source signal is obtained in advance according to the unmixed matrix B and the nonlinear function phi, namely:
y＝B F Φ(x)＝U T WΦ(x)；
the covariance matrix of the mapped feature space F is obtained as follows:
based on the kernel function, a kernel matrix K of NxN is generated:
K ij ＝(Φ(x i )·Φ(x j )>＝k(x i ，x j )；
wherein K is ij The Mercer condition is required to be satisfied, i.e., k=ff T The kernel function includes at least: gaussian kernel function:polynomial kernel function: /> sigmoid kernel function:
an on-board train bearing fault diagnosis system for use in performing any one of the above diagnostic methods, the system comprising:
the signal acquisition module is used for acquiring multichannel data of the target bearing sound source from the microphone array when the running speed of the target bearing sound source meets a first condition;
the signal processing module is used for preprocessing the multichannel data, positioning the target bearing sound source signal, performing signal enhancement processing on the preprocessed signal based on the far-field condition and the sound source distance of the target bearing sound source, generating a near-end signal and a far-end signal of the target bearing sound source, extracting the environmental noise of the preprocessed signal, and taking the environmental noise as a reference signal for noise suppression so as to eliminate the environmental noise outside the target bearing sound source signal;
the fault diagnosis module is used for extracting the time domain, the frequency domain and the wavelet domain characteristics of the target bearing sound source signal, adopting a dimension reduction method to perform characteristic dimension reduction so as to generate a training set, inputting the training set into a pre-constructed diagnosis model for training, inputting the signal to be diagnosed into the trained diagnosis model for classification so as to determine whether the target bearing sound source has faults.
An on-board train bearing fault diagnosis apparatus comprising a memory, a processor and a computer program stored in the memory and executable on the processor, the processor implementing a diagnosis method according to any one of the preceding claims when executing the computer program.
The invention has the beneficial effects that:
(1) In the prior art, a single microphone array is adopted, the array interval is smaller, the microphone array is used for a voice enhancement scene, and the microphone array is not suitable for a scene for collecting low-frequency sound source signals.
(2) According to the invention, the reference microphone at the top of the device is used for collecting environmental noise as a reference signal for noise suppression processing, so that the environmental noise signal in the target bearing sound source signal is eliminated, the steady state signal is enhanced, the unsteady state signals such as human voice and impact noise are eliminated, the purer bearing sound source signal is obtained, and the accuracy of the diagnosis model is improved.
(3) The monitoring device is only required to be placed at the bottom of the train saddle, mechanical disassembly and assembly are not required, non-contact acquisition is adopted, equipment is not required to be modified, a microphone is adopted to collect sound signals opposite to a fault position in combination with an acoustic mode, and the collection mode is more direct.
(4) The invention eliminates the non-steady state signal by enhancing the steady state signal, judges the health state of the train bearing by a signal processing and fault diagnosis method, has the edge computing capability, and can stably realize the acquisition, processing, algorithm realization and fault judgment of sound data in an on-line/off-line state. And the audio data is uploaded in a timing synchronization way, so that multi-terminal information synchronization is realized, and subsequent deep learning training and diagnosis are facilitated.
Drawings
FIG. 1 is a schematic illustration of the performance of a diagnostic method of the present invention;
FIG. 2 is a flow chart of a diagnostic method of the present invention;
FIGS. 3 and 4 are diagrams of installation scenarios of an acoustic monitoring device of the present invention;
FIG. 5 is a physical diagram of an acoustic monitoring device according to the present invention;
FIG. 6 is a diagram of the internal structure of an acoustic monitoring apparatus according to the present invention;
FIG. 7 is a side view of the interaction zone within an acoustic monitoring apparatus of the present invention;
FIG. 8 is a schematic diagram of a portion of a bearing test stand according to an embodiment of the present invention;
FIGS. 9, 10, 11 and 12 are diagrams showing failure comparisons of bearing outer rings at different speeds in the example of the present invention;
FIGS. 13, 14, 15 and 16 are graphs showing fault comparisons of different speeds of inner rings of partial bearings according to the embodiment of the present invention;
FIG. 17 is a diagram of a case-portion directed acquisition experimental setup of an embodiment of the present invention;
FIG. 18 is a diagram showing the sound source localization effect of a case-portion directional acquisition experiment according to the embodiment of the present invention;
fig. 19 is a partial effect comparison chart of a case of the embodiment of the present invention.
In fig. 5, 6 and 7: 1. a battery; 2. a microphone array board; 3. an array case housing; 4. an array acquisition board; 5. a core plate; 6. an interaction zone; 7. a status light; 8. reference microphone.
Detailed Description
The following detailed description of the present application is provided in conjunction with the accompanying drawings, and it is to be understood that the following detailed description is merely illustrative of the application and is not to be construed as limiting the scope of the application, since numerous insubstantial modifications and adaptations of the application will be to those skilled in the art in light of the foregoing disclosure.
Example 1
As shown in fig. 1-2, the present embodiment provides a method for diagnosing a bearing failure of an on-board train, including the steps of:
s1, when the running speed of a target bearing sound source meets a first condition, acquiring multichannel data of the target bearing sound source from a microphone array;
s2, preprocessing multichannel data, positioning a target bearing sound source signal, performing signal enhancement processing on the preprocessed signal based on far-field conditions and sound source distance of the target bearing sound source, generating a near-end signal and a far-end signal of the target bearing sound source, extracting environmental noise of the preprocessed signal, and taking the environmental noise as a reference signal for noise suppression so as to eliminate the environmental noise outside the target bearing sound source signal;
s3, extracting time domain, frequency domain and wavelet domain characteristics of the target bearing sound source signal, performing feature dimension reduction by adopting a dimension reduction method to generate a training set, inputting the training set into a pre-constructed diagnosis model for training, and inputting a signal to be diagnosed into the trained diagnosis model for classification so as to determine whether the target bearing sound source has faults.
In this embodiment, in step S1, the first condition is: the running speed is determined by satellite positioning speed measurement, a triaxial acceleration sensor and a band-pass filter combination, so that the vibration of the target bearing sound source carrier is prevented from interfering acceleration acquisition.
In this embodiment, the application execution diagnostic method acquisition device includes an annular microphone array including an inner ring microphone for acquiring a near-end target bearing sound source, an outer ring microphone for acquiring a far-end target bearing sound source, and a top microphone for acquiring environmental noise.
In this embodiment, step S2 specifically includes:
s201, preprocessing the multichannel data x, weakening reverberation and echo components in signals, and processing by adopting short-time Fourier transform to obtain x 1 (n, k) outputting a preprocessing desired signal d (n, k) as:
wherein n is the channel number, k is the signal frame number, τ is the signal frame time interval, g (k) is the filter coefficient, input x (n, k) is the observed signal, the time-varying Gaussian model and the maximum likelihood estimation rule are used for cyclic iteration to estimate the variance lambda and the weight g,
through iterative loop optimization, variance and weight are obtained, and the variance and weight are substituted into expected signals to obtain expected signals d { d1, d2, d3...d 14, d15, d16};
s202, a far-field condition calculation formula is as follows:
wherein L is the distance between the microphone array and the target bearing sound source, R is the radius of the circular array microphone array, and lambda is the signal frequency of the target bearing sound source;
s203, the signal enhancement processing flow is as follows: far-field signals are parallelly incident on the center of the microphone array at the azimuth angle theta, the number of delay units in the microphone array is L, the delay between adjacent delay units is T, and then the output y (T) of signal enhancement processing is:
and solving through a Lagrangian function to obtain an optimal solution as follows:
s204, defining a matrix for diagonally loading and updating the sample covariance matrix as:
wherein R is -1 For the correlation matrix R -1 ＝αR -1 +λI，For the power of the white noise component in the noise, lambda and alpha are the diagonal loadings, LNR is the loading noise level, whereby the in-focus loading lambda is obtained from the noise power, and R is used according to the optimal weight vector solution xx Approximation xx H Instead of R -1 ＝αR xx -1 Carrying out +lambdaI, obtaining a new weight vector, and obtaining a one-dimensional signal y (t) in a specified direction according to signal enhancement processing;
s205, inputting the interference signal and the environmental noise as reference signals D (n) into noise suppression to obtain weight vectors W of the noise signals k Setting an approximation factor beta according to the approximation degree of the interference sound source and the bearing sound source 1 ，β 2 ，β 3 ，β 4 ]；
Wherein the noise suppressed input signal vector X (n),W k the weight coefficient vector, the output Y (n) is:
the adaptive linear combiner is configured to calculate the mean square value ζ of the error signal as:
e(n)＝d(n)-y(n)；
ξ(n)＝E[e 2 (n)]；
according to the criterion of minimum mean square value of error signal, obtaining weight vector of noise, the desired signal is
Y(n)＝X(n)-β·W k ·X(n)。
In this embodiment, in the feature dimension reduction in step S3, the optimal approximation matrix of the source signal is obtained in advance according to the unmixed matrix B and the nonlinear function Φ, that is:
y＝B F Φ(x)＝U T WΦ(x)；
the covariance matrix of the mapped feature space F is obtained as follows:
based on the kernel function, a kernel matrix K of NxN is generated:
K ij ＝<Φ(x i )·Φ(X j )>＝k(x i ，x j )；
wherein K is ij The Mercer condition is required to be satisfied, i.e., k=ff T The kernel function includes at least: gaussian kernel function:polynomial kernel function: /> sigmoid kernel function:
the embodiment also provides a vehicle-mounted train bearing fault diagnosis system, which is applied to executing any one of the diagnosis methods, and comprises the following steps:
the signal acquisition module is used for acquiring multichannel data of the target bearing sound source from the microphone array when the running speed of the target bearing sound source meets a first condition;
the signal processing module is used for preprocessing the multichannel data, positioning the target bearing sound source signal, performing signal enhancement processing on the preprocessed signal based on the far-field condition and the sound source distance of the target bearing sound source, generating a near-end signal and a far-end signal of the target bearing sound source, extracting the environmental noise of the preprocessed signal, and taking the environmental noise as a reference signal for noise suppression so as to eliminate the environmental noise outside the target bearing sound source signal;
the fault diagnosis module is used for extracting the time domain, the frequency domain and the wavelet domain characteristics of the target bearing sound source signal, adopting a dimension reduction method to perform characteristic dimension reduction so as to generate a training set, inputting the training set into a pre-constructed diagnosis model for training, inputting the signal to be diagnosed into the trained diagnosis model for classification so as to determine whether the target bearing sound source has faults.
An on-board train bearing failure device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, the processor implementing any one of the diagnostic methods described above when the computer program is executed.
The above diagnostic method is further described in connection with the actual processing routine:
firstly, acoustic signals of a multi-target bearing sound source are collected under multi-noise strong reverberations and reverberations environments through the arrangement of an acoustic monitoring device, the monitoring device collects sound signals of a train bearing through a signal collection array box, the device is composed of a battery 1, a microphone array plate 2, an array box shell 3, an array collection plate 4, a calculation core template 5, an interaction area 6 and a vehicle body, 8 holes on the side of the array box shell 3 are connected with the vehicle body, a buffer pad is arranged at the bottom of the shell, the influence of vibration of the train body on the collection box can be reduced, an LED state lamp 8 is arranged at the top of the shell to facilitate rapid judgment of the state of the device, and a reference microphone 8 at the top can collect environmental noise outside the box body and is used for subsequent optimization processing.
The microphone array board comprises a battery 1, a microphone array board 2, an array box shell 3 and an array acquisition board 4, wherein the core module 5 and an interaction area 6 are calculated. The battery 1 is a detachable lithium battery, when the use environment can not provide power, the battery is automatically powered, the microphone array plate 2 is a double-ring circular array microphone array, the pickup range of the microphone array can be enlarged by dividing the microphone array into an outer ring and an inner ring, the microphone array is installed on the bottom surface of the array case housing 3, and the microphone faces downwards. The bottom of the array case shell 3 is provided with 16 annular holes, and microporous sound-transmitting filter membranes made of TPU materials are respectively and adhered to the annular holes for dust prevention and static prevention. The array acquisition board 4 is arranged at the side of the shell, and an opening is arranged on the shell, so that the interface of the acquisition board can extend out of the shell, and the acquisition board is integrated with the docking interface of the calculation core board 5, so that the disassembly and the replacement are convenient. The interaction area 6 integrates a power switch with a power input interface. The LED status light arranged at the top of the case with status light 7 may show the operational status of the case. The top reference microphone 8 may collect ambient noise outside the tank for subsequent optimization.
The function introduction is as follows, and the fault diagnosis method provided by the invention is used for processing and analyzing signals, so as to realize nondestructive detection of the train bearing. The information processing technology is utilized to realize the multidimensional information management of the train, the detection result of the system is fed back in real time, the degree of automation of the fault detection of the vehicle is improved, the manual work amount is reduced, and the running safety of the train is effectively ensured. The system mainly comprises an array box body, a signal acquisition system, a signal processing and fault diagnosis method and other auxiliary facilities, and the main functions of each part are as follows:
(1) The device is arranged below the upper seat of the train as shown in fig. 3 and 4 (the physical structure and the electrical structure of the train are not changed, the safe and stable running of the train is not influenced), and the device has the function of directing and collecting sound signals of four bearing movements of the running part of the train. When the speed of the train is stable, the collection is triggered, the array box synchronously collects the sound signals of the bearing at high speed in a short time, and data are quickly transmitted to the signal analysis system.
(2) The main functions of the signal processing and fault diagnosis method are to carry out filtering noise reduction processing, fault information mining, diagnosis decision, fault early warning and the like on the collected sound signals, and the method mainly comprises sound source localization, reverberation elimination, signal enhancement and noise suppression, wherein the algorithms are core functional parts of the whole system.
(3) The equipment terminal has the edge computing capability, and can be used for collecting and processing sound data and stably realizing algorithm realization and fault judgment in an on-line/off-line state. And the audio data is uploaded in a timing synchronization way, so that multi-terminal information synchronization is realized, and subsequent deep learning training and diagnosis are facilitated.
(4) The device is designed with various power supply modes, is compatible with 24V/110V DC input and realizes a long-endurance function by a self-contained battery.
Fig. 3 and 4 are the installation scene and the simple principle of the acoustic monitoring device, the signals of the bearing are directionally collected, fig. 5 is a physical diagram of the whole device, fig. 7 is the internal structure diagram of the device, the model of the train wheel set bearing in China is NJ (P) 3226X1, and the bearing test bed shown in fig. 8 is used for simulating the bearing load of 0.75 ton.
The outer ring fault frequency f has the following calculation formula:
wherein: z represents the number of rolling elements, D represents the bearing pitch diameter, D represents the rolling element diameter, and fn represents the inner ring rotation frequency. The outer ring fault frequencies corresponding to the bearing at the rotation speeds of 198.8r/min (train speed of 31.46 km/h) and 511.20r/min (train speed of 80.9 km/h) are respectively 17.56Hz and 50.84Hz, the center frequency of a spectrogram is about 1100Hz, the bandwidth is about 1000Hz, the bearing is mounted on a bearing test bed for loading, the load is 0.75 ton, and the sampling analysis is carried out on the bearing, so that the results are shown in figures 9, 10, 11 and 12.
The corresponding inner ring fault frequencies of the bearing at the rotation speeds of 198.8r/min (train speed of 31.46 km/h) and 511.20r/min (train speed of 80.9 km/h) are respectively 26.69Hz and 69Hz, the center frequency of a spectrogram is about 1100Hz, the bandwidth is about 1000Hz, the bearing is mounted on a bearing test bed for loading, the load is 0.75 ton, the sampling analysis is carried out on the bearing test bed, the results are shown in figures 13, 14, 15 and 16,
in order to more obviously show the separation and extraction effects of the invention in a multi-sound source environment, four fault bearing simulation signals ABCD with different fault frequencies are arranged, the bandwidth of an A sound source is 1000hz, the central frequency is 1800hz, and the fault frequency is 35hz; the bandwidth of the B sound source is 1000hz, the center frequency is 2100hz, and the fault frequency is 44hz; c sound source bandwidth 1000hz center frequency 2400hz, fault frequency 56hz; the D sound source bandwidth is 1000hz at a center frequency 2800hz and the failure frequency is 62hz.
The monitoring device is arranged on an aluminum profile frame shown in fig. 17, the sampling frequency of signals is 32KHz, four speakers are used for playing four sound sources of ABCD, the sound source positioning effect is shown in fig. 18, the comparison effect is shown in fig. 19, the sound source fault components before processing are disordered, the sound source fault components in four directions after processing are clearly shown, and the diagnosis method has good effects on the separation and extraction of envelope spectrum sources before and after the sound signal enhancement processing in a multi-sound source environment.
The above embodiments may be implemented in whole or in part by software, hardware, firmware, or any other combination. When implemented in software, the above-described embodiments may be implemented in whole or in part in the form of a computer program product. The computer program product comprises one or more computer instructions or computer programs. When the computer instructions or computer program are loaded or executed on a computer, the processes or functions described in accordance with embodiments of the present invention are produced in whole or in part. The computer may be a general purpose computer, a special purpose computer, a network of computers, or other programmable devices. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium, for example, the computer instructions may be transmitted from one website site, computer, server, or data center to another website site, computer, server, or data center by wired (e.g., infrared, wireless, microwave, etc.). The computer readable storage medium may be any available medium that can be accessed by a computer or a data storage device such as a server, data center, etc. that contains one or more sets of available media. The usable medium may be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. The semiconductor medium may be a solid state disk.
Those of ordinary skill in the art will appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, or combinations of computer software and electronic hardware. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present application.
In addition, each functional module in each embodiment of the present application may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit.
The functions, if implemented in the form of software functional modules and sold or used as a stand-alone product, may be stored in a computer-readable storage medium. Based on such understanding, the technical solution of the present application may be embodied essentially or in a part contributing to the prior art or in a part of the technical solution, in the form of a software product stored in a storage medium, including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a read-only memory (ROM), a random access memory (random access memory, RAM), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
The above embodiments are only for illustrating the technical solution of the present application, and are not limiting thereof; although the present application has been described in detail with reference to the foregoing embodiments, one of ordinary skill in the art will appreciate that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not drive the essence of the corresponding technical solutions to depart from the spirit and scope of the technical solutions of the embodiments of the present application.
Claims (7)
1. An on-board train bearing fault diagnosis method is characterized by comprising the following steps:
s1, when the running speed of a train meets a first condition, acquiring multichannel data of a target bearing sound source from a microphone array in the train;
s2, preprocessing multichannel data, positioning a target bearing sound source signal based on a far-field condition and a sound source distance of the target bearing sound source, performing signal enhancement on the preprocessed signal to generate a near-end signal and a far-end signal of the target bearing sound source, extracting environmental noise of the preprocessed signal, and taking the environmental noise as a reference signal for noise suppression processing to eliminate the environmental noise outside the target bearing sound source signal;
s3, extracting time domain, frequency domain and wavelet domain characteristics of the target bearing sound source signal, performing feature dimension reduction by adopting a dimension reduction method to generate a training set, inputting the training set into a pre-constructed diagnosis model for training, inputting a signal to be diagnosed into the trained diagnosis model for classification, and determining whether the target bearing sound source has faults.
2. The on-board train bearing fault diagnosis method according to claim 1, characterized by: in step S1, the first condition is: the running speed is determined by satellite positioning speed measurement, a triaxial acceleration sensor and a band-pass filter in combination, so that the vibration of the target bearing sound source carrier is prevented from interfering acceleration acquisition.
3. The on-board train bearing fault diagnosis method according to claim 1, characterized by: the diagnostic method acquisition device comprises an annular microphone array, wherein the annular microphone array comprises an inner ring microphone for acquiring a near-end target bearing sound source, an outer ring microphone for acquiring a far-end target bearing sound source and a reference microphone for acquiring environmental noise.
4. The on-board train bearing fault diagnosis method according to claim 1, characterized by: in step S2, specifically:
s201, preprocessing the multichannel data x, weakening reverberation and echo components in signals, and processing by adopting short-time Fourier transform to obtain x 1 (n, k) preprocessing the desired signal d (n, k) as:
wherein n is the channel number, k is the signal frame number, τ is the signal frame time interval, g (k) is the filter coefficient, input x (n, k) is the observed signal, the time-varying Gaussian model and the maximum likelihood estimation rule are used for cyclic iteration to estimate the variance lambda and the weight g,
through iterative loop optimization, variance and weight are obtained, and the variance and weight are substituted into expected signals to obtain expected signals d { d1, d2, d3...d 14, d15, d16};
s202, the far field condition calculation formula is as follows:
wherein L is the distance between the microphone array and the target bearing sound source, R is the radius of the circular array microphone array, and lambda is the signal frequency of the target bearing sound source;
s203, a signal enhancement flow is as follows: far-field signals are parallelly incident on the center of the microphone array at the azimuth angle theta, the number of delay units in the microphone array is L, the delay between adjacent delay units is T, and then the signals are enhanced to output y (T):
and solving through a Lagrangian function to obtain an optimal solution as follows:
s204, defining a matrix for diagonally loading and updating the sample covariance matrix as:
wherein R is -1 For the correlation matrix R -1 ＝αR -1 +λI，For the power of the white noise component in the noise, lambda and alpha are the diagonal loadings, LNR is the loading noise level, whereby the in-focus loading lambda is obtained from the noise power, and R is used according to the optimal weight vector solution xx Approximation xx H Instead of R -1 ＝αR -1 Carrying out +lambdaI, namely obtaining a new weight vector, and obtaining a one-dimensional signal y (t) in a specified direction according to signal enhancement processing;
s205, inputting the interference signal and the environmental noise as reference signals D (n) into noise suppression to obtain weight vectors W of the noise signals k Setting an approximation factor beta according to the approximation degree of the interference sound source and the bearing sound source 1 ，β 2 ，β 3 ，β 4 ]；
Wherein the noise suppressed input signal vector X (n), W k The weight coefficient vector, the output Y (n) is:
the adaptive linear combiner is configured to, in terms of the mean square value of the error signal:
e(n)＝d(n)-y(n)；
ξ(n)＝E[e 2 (n)]；
according to the criterion of minimum mean square value of error signal, obtaining weight vector of noise, the desired signal is
Y(n)＝X(n)-β·W k ·X(n)。
5. The on-board train bearing fault diagnosis method according to claim 4, wherein: in the step S3, the feature dimension reduction includes obtaining an optimal approximation matrix of the source signal in advance according to the unmixed matrix B and the nonlinear function Φ, that is:
y＝B F Φ(x)＝U T WΦ(x)；
the covariance matrix of the mapped feature space F is obtained as follows:
based on the kernel function, a kernel matrix K of NxN is generated:
K ij ＝<Φ(x i )·Φ(x j )>＝k(x i ，x j )；
wherein K is ij The Mercer condition is required to be satisfied, i.e., k=ff T The kernel function includes at least: gaussian kernel function:polynomial kernel function: /> sigmoid kernel function:
6. an on-board train bearing fault diagnosis system for use in performing the diagnosis method of any one of claims 1 to 5, the system comprising:
the signal acquisition module is used for acquiring multichannel data of the target bearing sound source from the microphone array when the running speed of the target bearing sound source meets a first condition;
the signal processing module is used for preprocessing the multichannel data, positioning a target bearing sound source signal, carrying out signal enhancement on the preprocessed signal based on the far-field condition and the sound source distance of the target bearing sound source, generating a near-end signal and a far-end signal of the target bearing sound source, extracting the environmental noise of the preprocessed signal, and taking the environmental noise as a reference signal for noise suppression so as to eliminate the environmental noise outside the target bearing sound source signal;
the fault diagnosis module is used for extracting the time domain, the frequency domain and the wavelet domain characteristics of the target bearing sound source signal, adopting a dimension reduction method to perform characteristic dimension reduction so as to generate a training set, inputting the training set into a pre-constructed diagnosis model for training, inputting the signal to be diagnosed into the trained diagnosis model for classification so as to determine whether the target bearing sound source has faults.
7. An on-board train bearing fault diagnosis device comprising a memory, a processor and a computer program stored in the memory and operable on the processor, wherein the processor implements the diagnosis method according to any one of claims 1-5 when executing the computer program.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311694019.6A CN117664577A (en) | 2023-12-08 | 2023-12-08 | Vehicle-mounted train bearing fault diagnosis method, system and equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311694019.6A CN117664577A (en) | 2023-12-08 | 2023-12-08 | Vehicle-mounted train bearing fault diagnosis method, system and equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117664577A true CN117664577A (en) | 2024-03-08 |
Family
ID=90084283
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311694019.6A Pending CN117664577A (en) | 2023-12-08 | 2023-12-08 | Vehicle-mounted train bearing fault diagnosis method, system and equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117664577A (en) |
-
2023
- 2023-12-08 CN CN202311694019.6A patent/CN117664577A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110307994B (en) | Abnormal sound detection device and abnormal sound detection method | |
| US11688415B2 (en) | System and methods for machine anomaly detection based on sound spectrogram images and neural networks | |
| CN110472587B (en) | Micro vibration motor defect identification method and device based on CNN and sound time-frequency characteristic diagram | |
| US8271200B2 (en) | System and method for acoustic signature extraction, detection, discrimination, and localization | |
| Nakadai et al. | Real-time sound source localization and separation for robot audition. | |
| KR102272622B1 (en) | Deep learning based abnormal symptom detection system using mechanical sound of mechanical equipment | |
| US20220217468A1 (en) | Method, device and electronic device for controlling audio playback of multiple loudspeakers | |
| MX2010010905A (en) | Method and device for the classification of sound-generating processes. | |
| KR102509699B1 (en) | Sound-based Multi-device Operation Monitoring Method and System for the Same | |
| CN117309038A (en) | Mining electromechanical equipment fault diagnosis system and method based on multi-mode sensing | |
| CN113567162A (en) | Fan fault intelligent diagnosis device and method based on acoustic sensor | |
| Pan et al. | Cognitive acoustic analytics service for Internet of Things | |
| CN109934136B (en) | Rolling bearing fault diagnosis method based on Duffing vibrator and eigen mode component | |
| JP2013014286A (en) | Vehicle approach notification device and program | |
| CN112052712A (en) | Power equipment state monitoring and fault identification method and system | |
| JP2018189522A (en) | Method for diagnosing sign of facility failure | |
| CN117664577A (en) | Vehicle-mounted train bearing fault diagnosis method, system and equipment | |
| CN116907029A (en) | Method for detecting abnormality of fan in outdoor unit, control device and air conditioner outdoor unit | |
| Ooi et al. | Non-intrusive operation status tracking for legacy machines via sound recognition | |
| Zhang et al. | Sound event localization and classification using WASN in Outdoor Environment | |
| Hioka et al. | Clean audio recording using unmanned aerial vehicles | |
| KR102457642B1 (en) | Machine sound recording apparatus and abnormality detection system of machine equipment comprising the same | |
| JP2008096305A (en) | Abnormality monitor | |
| CN115389186A (en) | Anomaly analysis method, device, terminal and storage medium | |
| CN114719961B (en) | Machine abnormal sound identification method and device, computer equipment and readable storage medium |
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 |