CN112183344A - Large unit friction fault analysis method and system based on waveform and dimensionless learning - Google Patents

Large unit friction fault analysis method and system based on waveform and dimensionless learning Download PDF

Info

Publication number
CN112183344A
CN112183344A CN202011041725.7A CN202011041725A CN112183344A CN 112183344 A CN112183344 A CN 112183344A CN 202011041725 A CN202011041725 A CN 202011041725A CN 112183344 A CN112183344 A CN 112183344A
Authority
CN
China
Prior art keywords
fault
dimensionless
signal
waveform
data
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
Application number
CN202011041725.7A
Other languages
Chinese (zh)
Other versions
CN112183344B (en
Inventor
荆晓远
陈润航
王许辉
张清华
成明康
孔晓辉
陈俊均
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Guangdong University of Petrochemical Technology
Original Assignee
Guangdong University of Petrochemical Technology
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Guangdong University of Petrochemical Technology filed Critical Guangdong University of Petrochemical Technology
Priority to CN202011041725.7A priority Critical patent/CN112183344B/en
Priority to PCT/CN2020/131616 priority patent/WO2022062161A1/en
Publication of CN112183344A publication Critical patent/CN112183344A/en
Application granted granted Critical
Publication of CN112183344B publication Critical patent/CN112183344B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
    • G06F2218/08—Feature extraction
    • G06F2218/10—Feature extraction by analysing the shape of a waveform, e.g. extracting parameters relating to peaks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/14—Fourier, Walsh or analogous domain transformations, e.g. Laplace, Hilbert, Karhunen-Loeve, transforms
    • G06F17/141—Discrete Fourier transforms
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/14—Fourier, Walsh or analogous domain transformations, e.g. Laplace, Hilbert, Karhunen-Loeve, transforms
    • G06F17/148—Wavelet transforms
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
    • G06F2218/12—Classification; Matching
    • G06F2218/14—Classification; Matching by matching peak patterns

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Data Mining & Analysis (AREA)
  • General Engineering & Computer Science (AREA)
  • Mathematical Optimization (AREA)
  • Pure & Applied Mathematics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Mathematical Analysis (AREA)
  • Software Systems (AREA)
  • Computational Mathematics (AREA)
  • Artificial Intelligence (AREA)
  • Algebra (AREA)
  • Databases & Information Systems (AREA)
  • Evolutionary Computation (AREA)
  • Evolutionary Biology (AREA)
  • Discrete Mathematics (AREA)
  • Computing Systems (AREA)
  • Medical Informatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Signal Processing (AREA)
  • Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
  • Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)

Abstract

The invention belongs to the technical field of fault detection, and discloses a large unit friction fault analysis method and system based on waveform and dimensionless learning, wherein a double probe is used for extracting a machine fault vibration signal and preprocessing data; simultaneously, extracting friction fault characteristics, and establishing a fault prediction model by using a machine learning method; predicting whether the unknown tag signal has a fault or not, and determining the type of the fault. In the process of preprocessing a machine fault vibration signal and data, two probe points are installed, and a vibration double-view signal of the large-scale sliding unit is acquired through the two probe points; after the probe collects data, aligning and performing discrete Fourier transform, and modifying the transformed Fourier value; and setting an adaptive threshold according to the signal condition, reducing the signal storage amount and accelerating transmission. In the process of diagnosing the friction fault of the large unit, the problem of difficult feature extraction can be effectively solved; effective characteristics can be extracted, and the problem of failure prediction is solved.

Description

Large unit friction fault analysis method and system based on waveform and dimensionless learning
Technical Field
The invention belongs to the technical field of fault detection, and particularly relates to a large unit friction fault analysis method and system based on waveform and dimensionless learning.
Background
At present, large-scale mechanized equipment has complex structure, perfect functions and close connection among internal parts of the equipment, so that high speed and large-scale production are achieved in the production process, the large-scale mechanized equipment breaks down to cause huge loss, and the difficulty of fault diagnosis of the large-scale mechanized equipment is increased. When an object and another object move along the tangent direction of the contact surfaces or have a relative motion tendency, a force resisting the relative motion between the contact surfaces of the two objects exists, and the force is called friction force. This phenomenon or characteristic between the contact surfaces is called "friction", and therefore, there is a great significance in detecting a fault in a large mechanized apparatus based on a friction vibration signal. The existing common method has the problems of difficult feature extraction, incomplete feature extraction and the like.
Through the above analysis, the problems and defects of the prior art are as follows: the existing common method has the problems of difficult feature extraction, incomplete feature extraction and the like. Causing difficulty in fault diagnosis in the operation of the large computerized equipment.
The difficulty in solving the above problems and defects is:
the problems of friction failure, such as difficult feature extraction, incomplete feature extraction and the like.
The significance of solving the problems and the defects is as follows:
the problem about friction fault detection can be solved well, and the friction fault detection precision is improved.
Disclosure of Invention
Aiming at the problems in the prior art, the invention provides a large unit friction fault analysis method and system based on waveform and dimensionless learning.
The invention is realized in such a way that a large unit friction fault analysis method based on waveform and dimensionless learning comprises the following steps:
extracting a machine fault vibration signal by using a double probe, and preprocessing data;
step two, extracting friction fault characteristics;
thirdly, establishing a fault prediction model by using a machine learning method;
and step four, predicting whether the unknown label signal has a fault or not, and determining the fault type.
Further, in the step one, the process of preprocessing the machine fault vibration signal and the data is as follows:
1) installing two probe points, acquiring a vibration double-view signal of the large-scale sliding unit through the acquisition of the two probe points, wherein the data acquisition is 32/rms, namely, 32 points are sampled every time the bearing rotates, and 32 circles of data are acquired;
2) after the probe collects data, performing discrete Fourier transform in an aligned mode, wherein the window size 32 x 32 is 1024 points, and modifying a Fourier value after the transform; and setting an adaptive threshold according to the signal condition, reducing the signal storage amount and accelerating transmission.
Further, the discrete fourier transform formula in 2) is specifically as follows:
Figure BDA0002706844540000021
where N is 0, …, N-1, N denotes the data length.
Further, in the second step, the friction fault feature extraction process is as follows:
(1) carrying out wavelet packet 2-layer decomposition transformation on the fault signal, wherein the wavelet packet is to analyze the detail part of an input signal by utilizing multi-iteration wavelet transformation to obtain wavelet coefficients under different scales, and setting the scale coefficient of an HH layer of the signal to be zero;
(2) calculating dimensionless characteristic waviness index SfTaking the volatility index as one of the extracted features, the specific calculation formula is as follows:
Figure BDA0002706844540000022
wherein
Figure BDA0002706844540000023
Represents the root mean square value of the waveform data,
Figure BDA0002706844540000024
represents the absolute average of the waveform data;
(3) calculating a dimensionless characteristic peak index, and taking the peak index as one of the extracted characteristics, wherein a specific calculation formula is as follows:
Figure BDA0002706844540000031
wherein xmaxWhich is indicative of the peak of the waveform,
Figure BDA0002706844540000032
the representation represents a root mean square value;
(4) calculating a dimensionless characteristic pulse index, and taking the pulse index as one of the extracted characteristics, wherein a specific calculation formula is as follows:
Figure BDA0002706844540000033
wherein xmaxWhich is indicative of the peak of the waveform,
Figure BDA0002706844540000034
represents the absolute average of the waveform data;
(5) calculating a non-dimensional feature kurtosis index to represent the height of the actual kurtosis relative to the normal kurtosis, wherein the kurtosis index reflects the impact feature in the vibration signal, and the non-dimensional feature kurtosis index is taken as one of the extracted features, and a specific calculation formula is as follows:
Figure BDA0002706844540000035
wherein
Figure BDA0002706844540000036
(6) Calculating a dimensionless characteristic margin index, which is generally used for detecting the abrasion condition of mechanical equipment; if the distortion index changes little, the ratio of the effective value to the average value is increased, which shows that the clearance is increased due to abrasion, so that the effective value of the energy index of vibration is increased faster than the average value, the margin index of the energy index of vibration is also increased, and the dimensionless feature margin index is taken as one of the extracted features, and the specific calculation formula is as follows:
Figure BDA0002706844540000037
wherein
Figure BDA0002706844540000038
(7) Calculating a dimensionless characteristic Teager energy operator, wherein the Teager energy operator is one of the extracted characteristics, and the specific calculation formula is as follows:
Figure BDA0002706844540000039
wherein t represents the data acquisition time,
Figure BDA0002706844540000041
alpha t is an offset angle before and after the time t;
(8) calculating a standard deviation characterizing the degree of dispersion of the data and the variability exhibited by a single statistic over multiple samples; it can be understood that the former is the variability of the data itself, and the latter characterizes the variability of the sampling behavior, and the specific calculation formula is as follows:
Figure BDA0002706844540000042
(9) calculating the standard deviation of the average, wherein the standard deviation of the average is a standard for measuring the dispersion degree of data distribution and is used for measuring the degree of the data value deviating from the arithmetic average; the smaller the standard deviation, the less the values deviate from the mean and vice versa; the magnitude of the standard deviation can be measured by the multiplying power relationship between the standard deviation and the average value, and the specific calculation formula is as follows:
Figure BDA0002706844540000043
(10) calculating a sample circle mean (circle _ mean) of the sample, wherein the sample circle mean is taken as one of the extracted features, and the specific calculation formula is as follows:
Figure BDA0002706844540000044
wherein X is a sample, sin is a sine function, cos is a cosine function, arctan2 is a tangent function, and pi is a circumferential ratio;
wherein
Figure BDA0002706844540000045
S＝∑isin(angle)C＝∑icos(angle)，res＝arctan2(S，C)。
Further, cca dimension reduction is performed by using the two view features extracted in the second step, the features of the two views after dimension reduction are spliced and used as input vectors, and a machine learning model is used for training.
Another object of the present invention is to provide a waveform and dimensionless learning-based mainframe friction fault analysis system for implementing the waveform and dimensionless learning-based mainframe friction fault analysis method, which includes:
the data acquisition module extracts a machine fault vibration signal by using the double probes and preprocesses data;
the characteristic extraction module is used for extracting the characteristics of the friction fault signal;
the prediction model building module is used for building a fault prediction model by utilizing a machine learning method;
and the fault prediction module predicts whether the unknown label signal has a fault or not and determines the type of the fault.
It is a further object of the invention to provide a computer device comprising a memory and a processor, the memory storing a computer program which, when executed by the processor, causes the processor to perform the steps of:
extracting a machine fault vibration signal by using a double probe, and preprocessing data;
extracting the characteristics of the friction fault signal;
establishing a fault prediction model by using a machine learning method;
predicting whether the unknown tag signal has a fault or not, and determining the type of the fault.
It is another object of the present invention to provide a computer-readable storage medium storing a computer program which, when executed by a processor, causes the processor to perform the steps of:
extracting a machine fault vibration signal by using a double probe, and preprocessing data;
extracting the characteristics of the friction fault signal;
establishing a fault prediction model by using a machine learning method;
predicting whether the unknown tag signal has a fault or not, and determining the type of the fault.
The invention also aims to provide a large mechanized device for implementing the large-unit friction fault analysis method based on waveform and dimensionless learning.
By combining all the technical schemes, the invention has the advantages and positive effects that:
in the process of diagnosing the friction fault of the large unit, the problem of difficult feature extraction can be effectively solved; meanwhile, effective features can be extracted, the problem of failure prediction is solved, and a new feature extraction method is provided. The invention obtains good results on the diagnosis problem of the friction fault of the large unit.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed to be used in the embodiments of the present application will be briefly described below, and it is obvious that the drawings described below are only some embodiments of the present application, and it is obvious for those skilled in the art that other drawings can be obtained from the drawings without creative efforts.
Fig. 1 is a flowchart of a large-scale unit friction fault analysis method based on waveform and dimensionless learning according to an embodiment of the present invention.
FIG. 2 is a schematic structural diagram of a large-scale unit friction fault analysis system based on waveform and dimensionless learning according to an embodiment of the present invention;
in the figure: 1. a data acquisition module; 2. a feature extraction module; 3. a prediction model construction module; 4. and a failure prediction module.
Fig. 3 is a schematic diagram of a wavelet packet 2 layer decomposition transformation structure according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail with reference to the following embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
Aiming at the problems in the prior art, the invention provides a large-unit friction fault analysis method and system based on waveform and dimensionless learning, and the invention is described in detail with reference to the attached drawings.
As shown in fig. 1, a large-unit friction fault analysis method based on waveform and dimensionless learning according to an embodiment of the present invention includes:
s101: and extracting a machine fault vibration signal by using the double probes, and preprocessing data.
S102: and (5) extracting the friction fault characteristics.
S103: and establishing a fault prediction model by using a machine learning method.
S104: predicting whether the unknown tag signal has a fault or not, and determining the type of the fault.
In S101 provided by the embodiment of the present invention, a process of preprocessing a machine fault vibration signal and data includes:
1) two probe points are installed, vibration double-view signals of the large-scale sliding unit are acquired through the two probe points, data acquisition is 32/rms, namely 32 points are sampled every time the bearing rotates one circle, and 32 circles of data are acquired.
2) After the probe collects data, performing discrete Fourier transform in an aligned mode, wherein the window size 32 x 32 is 1024 points, and modifying a Fourier value after the transform; and setting an adaptive threshold according to the signal condition, reducing the signal storage amount and accelerating transmission.
The discrete fourier transform formula is specifically as follows:
Figure BDA0002706844540000071
where N is 0, …, N-1, N denotes the data length.
In S102 provided by the embodiment of the present invention, the process of extracting the friction fault feature is as follows:
(1) the fault signal is processed by wavelet packet 2 layer decomposition transformation, and the wavelet packet is a detailed part of the input signal analyzed by multi-iteration wavelet transform, and its specific structure diagram is shown in fig. 2. Wavelet coefficients under different scales are obtained, and the scale coefficient of the HH layer of the signal is set to be zero.
(2) Calculating dimensionless characteristic waviness index SfWill waveThe index is taken as one of the extracted features, and the specific calculation formula is as follows:
Figure BDA0002706844540000072
wherein
Figure BDA0002706844540000073
Represents the root mean square value of the waveform data,
Figure BDA0002706844540000074
representing the absolute average of the waveform data.
(3) Calculating a dimensionless characteristic peak index, and taking the peak index as one of the extracted characteristics, wherein a specific calculation formula is as follows:
Figure BDA0002706844540000075
wherein xmaxWhich is indicative of the peak of the waveform,
Figure BDA0002706844540000076
the representation represents the root mean square value.
(4) Calculating a dimensionless characteristic pulse index, and taking the pulse index as one of the extracted characteristics, wherein a specific calculation formula is as follows:
Figure BDA0002706844540000077
wherein xmaxWhich is indicative of the peak of the waveform,
Figure BDA0002706844540000081
representing the absolute average of the waveform data.
(5) Calculating a non-dimensional feature kurtosis index to represent the height of the actual kurtosis relative to the normal kurtosis, wherein the kurtosis index reflects the impact feature in the vibration signal, and the non-dimensional feature kurtosis index is taken as one of the extracted features, and a specific calculation formula is as follows:
Figure BDA0002706844540000082
wherein
Figure BDA0002706844540000083
(6) And calculating a dimensionless characteristic margin index, which is generally used for detecting the abrasion condition of the mechanical equipment. If the distortion index changes little, the ratio of the effective value to the average value is increased, which shows that the clearance is increased due to abrasion, so that the effective value of the energy index of vibration is increased faster than the average value, the margin index of the energy index of vibration is also increased, and the dimensionless feature margin index is taken as one of the extracted features, and the specific calculation formula is as follows:
Figure BDA0002706844540000084
wherein
Figure BDA0002706844540000085
(7) Calculating a dimensionless characteristic Teager energy operator, wherein the Teager energy operator is one of the extracted characteristics, and the specific calculation formula is as follows:
Figure BDA0002706844540000086
wherein t represents the data acquisition time,
Figure BDA0002706844540000087
αtis the offset angle before and after time t.
(8) The standard deviation, which characterizes the degree of dispersion of the data, is calculated, and characterizes the variability exhibited by individual statistics over multiple samples. It can be understood that the former is the variability of the data itself, and the latter characterizes the variability of the sampling behavior, and the specific calculation formula is as follows:
Figure BDA0002706844540000091
(9) the standard deviation of the mean, which is a measure of the degree of dispersion of the data distribution, is calculated as the degree to which the data values deviate from the arithmetic mean. The smaller the standard deviation, the less the values deviate from the mean and vice versa. The magnitude of the standard deviation can be measured by the multiplying power relationship between the standard deviation and the average value, and the specific calculation formula is as follows:
Figure BDA0002706844540000092
(10) calculating a sample circle mean (circle _ mean) of the sample, wherein the sample circle mean is taken as one of the extracted features, and the specific calculation formula is as follows:
Figure BDA0002706844540000093
where X is the sample, sin is the sine function, cos is the cosine function, arctan2 is the tangent function, and π is the circumferential ratio.
Wherein
Figure BDA0002706844540000094
S＝∑isin(angle)C＝∑icos(angle)，res＝arctan2(S，C)。
In the embodiment of the invention, cca dimension reduction is performed by using the two view features extracted in the step S102 in the step S103, the features of the two views after dimension reduction are spliced to be used as an input vector, and a machine learning model is used for training.
As shown in fig. 2, the large-unit friction fault analysis system based on waveform and dimensionless learning according to the embodiment of the present invention includes:
and the data acquisition module 1 is used for extracting a machine fault vibration signal by using the double probes and preprocessing data.
And the characteristic extraction module 2 is used for extracting the characteristics of the friction fault signal.
The prediction model construction module 3 builds a fault prediction model by using a machine learning method.
And the fault prediction module 4 predicts whether the unknown label signal has a fault or not and determines the type of the fault.
The technical scheme of the invention is further described by taking the sliding mechanical data of a large unit as an example.
The large unit provided by the invention is provided with two probes for data acquisition, wherein the data is acquired at 32 points every time the mechanical bearing rotates for one circle, and then the period is 32 revolutions. One set of data is 1024 waveform points, which translates to a waveform length of 1024. Extracting a frequency spectrum vector and other dimensionless vectors from the data, and splicing the frequency spectrum vector and the other dimensionless vectors to form the characteristics of a view; performing feature dimension reduction on a single view through cca, splicing the features subjected to dimension reduction, and then bringing the features into a machine learning model for training; extracting the characteristics of the fault signal of the unknown label, and obtaining a fault prediction result through a trained model.
In the description of the present invention, "a plurality" means two or more unless otherwise specified; the terms "upper", "lower", "left", "right", "inner", "outer", "front", "rear", "head", "tail", and the like, indicate orientations or positional relationships based on the orientations or positional relationships shown in the drawings, are only for convenience in describing and simplifying the description, and do not indicate or imply that the device or element referred to must have a particular orientation, be constructed in a particular orientation, and be operated, and thus, should not be construed as limiting the invention. Furthermore, the terms "first," "second," "third," and the like are used for descriptive purposes only and are not to be construed as indicating or implying relative importance.
The above description is only for the purpose of illustrating the present invention and the appended claims are not to be construed as limiting the scope of the invention, which is intended to cover all modifications, equivalents and improvements that are within the spirit and scope of the invention as defined by the appended claims.

Claims (9)

1. A large unit friction fault analysis method based on waveform and dimensionless learning is characterized in that the large unit friction fault analysis method based on waveform and dimensionless learning comprises the following steps:
extracting a machine fault vibration signal by using a double probe, and preprocessing data;
extracting friction fault characteristics;
establishing a fault prediction model by using a machine learning method;
predicting whether the unknown tag signal has a fault or not, and determining the type of the fault.
2. The analysis method for large unit friction fault based on waveform and dimensionless learning of claim 1, wherein the pre-processing of the vibration signals and data of machine fault is as follows:
1) installing two probe points, acquiring a vibration double-view signal of the large-scale sliding unit through the acquisition of the two probe points, wherein the data acquisition is 32/rms, namely, 32 points are sampled every time the bearing rotates, and 32 circles of data are acquired;
2) after the probe collects data, performing discrete Fourier transform in an aligned mode, wherein the window size 32 x 32 is 1024 points, and modifying a Fourier value after the transform; and setting an adaptive threshold according to the signal condition, reducing the signal storage amount and accelerating transmission.
3. The waveform and dimensionless learning based analysis method for large-unit friction faults as claimed in claim 2, wherein the discrete fourier transform formula in 2) is specifically as follows:
Figure FDA0002706844530000011
where N is 0, …, N-1, N denotes the data length.
4. The analysis method for the large unit friction fault based on the waveform and the dimensionless learning of claim 1, wherein the process of extracting the friction fault features is as follows:
(1) carrying out wavelet packet 2-layer decomposition transformation on the fault signal, wherein the wavelet packet is to analyze the detail part of an input signal by utilizing multi-iteration wavelet transformation to obtain wavelet coefficients under different scales, and setting the scale coefficient of an HH layer of the signal to be zero;
(2) calculating dimensionless characteristic waviness index SfTaking the volatility index as one of the extracted features, the specific calculation formula is as follows:
Figure FDA0002706844530000012
wherein
Figure FDA0002706844530000021
Represents the root mean square value of the waveform data,
Figure FDA0002706844530000022
represents the absolute average of the waveform data;
(3) calculating a dimensionless characteristic peak index, and taking the peak index as one of the extracted characteristics, wherein a specific calculation formula is as follows:
Figure FDA0002706844530000023
wherein xmaxWhich is indicative of the peak of the waveform,
Figure FDA0002706844530000024
the representation represents a root mean square value;
(4) calculating a dimensionless characteristic pulse index, and taking the pulse index as one of the extracted characteristics, wherein a specific calculation formula is as follows:
Figure FDA0002706844530000025
wherein xmaxWhich is indicative of the peak of the waveform,
Figure FDA0002706844530000026
represents the absolute average of the waveform data;
(5) calculating a non-dimensional feature kurtosis index to represent the height of the actual kurtosis relative to the normal kurtosis, wherein the kurtosis index reflects the impact feature in the vibration signal, and the non-dimensional feature kurtosis index is taken as one of the extracted features, and a specific calculation formula is as follows:
Figure FDA0002706844530000027
wherein
Figure FDA0002706844530000028
(6) Calculating a dimensionless characteristic margin index, which is generally used for detecting the abrasion condition of mechanical equipment; if the distortion index changes little, the ratio of the effective value to the average value is increased, which shows that the clearance is increased due to abrasion, so that the effective value of the energy index of vibration is increased faster than the average value, the margin index of the energy index of vibration is also increased, and the dimensionless feature margin index is taken as one of the extracted features, and the specific calculation formula is as follows:
Figure FDA0002706844530000029
wherein
Figure FDA00027068445300000210
(7) Calculating a dimensionless characteristic Teager energy operator, wherein the Teager energy operator is one of the extracted characteristics, and the specific calculation formula is as follows:
Figure FDA0002706844530000031
wherein t represents the data acquisition time,
Figure FDA0002706844530000032
αtis the offset angle before and after the time t;
(8) calculating a standard deviation characterizing the degree of dispersion of the data and the variability exhibited by a single statistic over multiple samples; it can be understood that the former is the variability of the data itself, and the latter characterizes the variability of the sampling behavior, and the specific calculation formula is as follows:
Figure FDA0002706844530000033
(9) calculating the standard deviation of the average, wherein the standard deviation of the average is a standard for measuring the dispersion degree of data distribution and is used for measuring the degree of the data value deviating from the arithmetic average; the smaller the standard deviation, the less the values deviate from the mean and vice versa; the magnitude of the standard deviation can be measured by the multiplying power relationship between the standard deviation and the average value, and the specific calculation formula is as follows:
Figure FDA0002706844530000034
(10) calculating a sample circle mean (circle _ mean) of the sample, wherein the sample circle mean is taken as one of the extracted features, and the specific calculation formula is as follows:
Figure FDA0002706844530000035
wherein X is a sample, sin is a sine function, cos is a cosine function, arctan2 is a tangent function, and pi is a circumferential ratio;
wherein
Figure FDA0002706844530000036
S＝∑isin(angle)C＝∑icos(angle)，res＝arctan2(S,C)。
5. The method for analyzing large unit friction fault based on waveform and dimensionless learning of claim 1, wherein the cca dimension reduction is performed by using the extracted features of the two views, and the features of the two views after dimension reduction are spliced to be used as input vectors and trained by using a machine learning model.
6. A waveform and dimensionless learning-based mainframe friction fault analysis system implementing the waveform and dimensionless learning-based mainframe friction fault analysis method according to any of claims 1-5, wherein the waveform and dimensionless learning-based mainframe friction fault analysis system comprises:
the data acquisition module extracts a machine fault vibration signal by using the double probes and preprocesses data;
the characteristic extraction module is used for extracting the characteristics of the friction fault signal;
the prediction model building module is used for building a fault prediction model by utilizing a machine learning method;
and the fault prediction module predicts whether the unknown label signal has a fault or not and determines the type of the fault.
7. A computer device, characterized in that the computer device comprises a memory and a processor, the memory storing a computer program which, when executed by the processor, causes the processor to carry out the steps of:
extracting a machine fault vibration signal by using a double probe, and preprocessing data;
extracting the characteristics of the friction fault signal;
establishing a fault prediction model by using a machine learning method;
predicting whether the unknown tag signal has a fault or not, and determining the type of the fault.
8. A computer-readable storage medium storing a computer program which, when executed by a processor, causes the processor to perform the steps of:
extracting a machine fault vibration signal by using a double probe, and preprocessing data;
extracting the characteristics of the friction fault signal;
establishing a fault prediction model by using a machine learning method;
predicting whether the unknown tag signal has a fault or not, and determining the type of the fault.
9. A large mechanized apparatus for implementing a large unit friction fault analysis method based on waveform and dimensionless learning according to any of claims 1-5.
CN202011041725.7A 2020-09-28 2020-09-28 Large unit friction fault analysis method and system based on waveform and dimensionless learning Active CN112183344B (en)

Priority Applications (2)

Application Number Priority Date Filing Date Title
CN202011041725.7A CN112183344B (en) 2020-09-28 2020-09-28 Large unit friction fault analysis method and system based on waveform and dimensionless learning
PCT/CN2020/131616 WO2022062161A1 (en) 2020-09-28 2020-11-26 Large machine set friction fault analysis method and system based on waveform and dimensionless learning

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011041725.7A CN112183344B (en) 2020-09-28 2020-09-28 Large unit friction fault analysis method and system based on waveform and dimensionless learning

Publications (2)

Publication Number Publication Date
CN112183344A true CN112183344A (en) 2021-01-05
CN112183344B CN112183344B (en) 2021-06-01

Family

ID=73944550

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011041725.7A Active CN112183344B (en) 2020-09-28 2020-09-28 Large unit friction fault analysis method and system based on waveform and dimensionless learning

Country Status (2)

Country Link
CN (1) CN112183344B (en)
WO (1) WO2022062161A1 (en)

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114993677B (en) * 2022-05-11 2023-05-02 山东大学 Rolling bearing fault diagnosis method and system for unbalanced small sample data
CN114781467B (en) * 2022-06-22 2022-09-06 济南嘉宏科技有限责任公司 Fault detection method based on vibration similarity
CN115392317A (en) * 2022-08-31 2022-11-25 广东石油化工学院 Petrochemical unit rotating machinery fault diagnosis method, device, equipment and medium
CN116070154B (en) * 2023-03-29 2023-06-20 山东金帝精密机械科技股份有限公司 Fault detection method, equipment and medium for large bearing retainer
CN117272030B (en) * 2023-11-20 2024-01-26 南通市计量检定测试所 Method for sampling and processing dynamic signal packet

Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7269536B1 (en) * 2006-03-23 2007-09-11 Sun Microsystems, Inc. Method and apparatus for quantitatively determining severity of degradation in a signal
CN105760839A (en) * 2016-02-22 2016-07-13 重庆大学 Bearing fault diagnosis method based on multi-feature manifold learning and support vector machine
CN105910805A (en) * 2016-04-25 2016-08-31 电子科技大学 Wavelet local mean decomposition method used for rotor rub-impacting fault diagnosis
CN106323635A (en) * 2016-08-04 2017-01-11 中国航空综合技术研究所 Rolling bearing fault on-line detection and state assessment method
CN106932200A (en) * 2017-04-20 2017-07-07 北京信息科技大学 The Gaussian process multiple features intelligent diagnosing method of rolling bearing fault
CN107831012A (en) * 2017-10-11 2018-03-23 温州大学 A kind of Method for Bearing Fault Diagnosis based on Walsh conversion with Teager energy operators
CN109323860A (en) * 2018-10-31 2019-02-12 广东石油化工学院 A kind of rotating machinery gearbox fault data set optimization method
CN109598255A (en) * 2018-12-19 2019-04-09 北京化工大学 A kind of reciprocating mechanical vibration signal impact initial point self-adaptation extraction method based on energy operator k- gradient
CN109726893A (en) * 2018-11-27 2019-05-07 广东石油化工学院 A kind of determination and evaluation method of rotating machinery fault data set attribute weight
CN110160781A (en) * 2019-06-18 2019-08-23 广东石油化工学院 A kind of the test set reconstruct and prediction technique of rotating machinery fault classification

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108388860B (en) * 2018-02-12 2020-04-28 大连理工大学 Aero-engine rolling bearing fault diagnosis method based on power entropy spectrum-random forest
CN108760305B (en) * 2018-06-13 2021-03-09 中车青岛四方机车车辆股份有限公司 Bearing fault detection method, device and equipment
CN110108431B (en) * 2019-05-22 2021-07-16 西安因联信息科技有限公司 Mechanical equipment fault diagnosis method based on machine learning classification algorithm

Patent Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7269536B1 (en) * 2006-03-23 2007-09-11 Sun Microsystems, Inc. Method and apparatus for quantitatively determining severity of degradation in a signal
CN105760839A (en) * 2016-02-22 2016-07-13 重庆大学 Bearing fault diagnosis method based on multi-feature manifold learning and support vector machine
CN105910805A (en) * 2016-04-25 2016-08-31 电子科技大学 Wavelet local mean decomposition method used for rotor rub-impacting fault diagnosis
CN106323635A (en) * 2016-08-04 2017-01-11 中国航空综合技术研究所 Rolling bearing fault on-line detection and state assessment method
CN106932200A (en) * 2017-04-20 2017-07-07 北京信息科技大学 The Gaussian process multiple features intelligent diagnosing method of rolling bearing fault
CN107831012A (en) * 2017-10-11 2018-03-23 温州大学 A kind of Method for Bearing Fault Diagnosis based on Walsh conversion with Teager energy operators
CN109323860A (en) * 2018-10-31 2019-02-12 广东石油化工学院 A kind of rotating machinery gearbox fault data set optimization method
CN109726893A (en) * 2018-11-27 2019-05-07 广东石油化工学院 A kind of determination and evaluation method of rotating machinery fault data set attribute weight
CN109598255A (en) * 2018-12-19 2019-04-09 北京化工大学 A kind of reciprocating mechanical vibration signal impact initial point self-adaptation extraction method based on energy operator k- gradient
CN110160781A (en) * 2019-06-18 2019-08-23 广东石油化工学院 A kind of the test set reconstruct and prediction technique of rotating machinery fault classification

Non-Patent Citations (9)

* Cited by examiner, † Cited by third party
Title
DE ZHI LI ET AL.: "An Enhanced Bispectrum Technique With Auxiliary Frequency Injection for Induction Motor Health Condition Monitoring", 《IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT》 *
FARDIN DALVAND ET AL.: "Current Noise Cancellation for Bearing Fault Diagnosis Using Time Shifting", 《IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS》 *
QIN HU ET AL.: "A rotating machinery fault diagnosis method based on multi-scale dimensionless indicators and random forests", 《MECHANICAL SYSTEMS AND SIGNAL PROCESSING》 *
QIN HU ET AL.: "Machinery Fault Diagnosis Scheme Using Redefined Dimensionless Indicators and mRMR Feature Selection", 《IEEE ACCESS》 *
WEI LIU ET AL.: "Spectral Decomposition for Hydrocarbon Detection Based on VMD and Teager–Kaiser Energy", 《IEEE GEOSCIENCE AND REMOTE SENSING LETTERS》 *
李启华 等: "标准偏差和平均值的标准偏差探讨", 《金属材料与冶金工程》 *
李瑞莲 等: "基于小波包变换和极限学习机的滚动轴承故障诊断", 《太原理工大学学报》 *
覃爱淞 等: "基于量纲一指标和极限学习机的滚动轴承故障诊断方法", 《机床与液压》 *
陈仕琦 等: "基于小波包和Teager能量算子的齿轮故障诊断研究", 《机械传动》 *

Also Published As

Publication number Publication date
CN112183344B (en) 2021-06-01
WO2022062161A1 (en) 2022-03-31

Similar Documents

Publication Publication Date Title
CN112183344B (en) Large unit friction fault analysis method and system based on waveform and dimensionless learning
CN109765490B (en) Power battery fault detection method and system based on high-dimensional data diagnosis
CN109375060B (en) Method for calculating fault waveform similarity of power distribution network
CN111504647A (en) AR-MSET-based performance degradation evaluation method for rolling bearing
CN113702044B (en) Bearing fault detection method and system
CN113340598B (en) Rolling bearing intelligent fault diagnosis method based on regularized sparse model
CN114705432B (en) Method and system for evaluating health state of explosion-proof motor bearing
CN115758200A (en) Vibration signal fault identification method and system based on similarity measurement
Geng et al. Generalized broadband mode decomposition method and its application in fault diagnosis of variable speed spherical roller bearing
CN112232414B (en) Triple concurrency fault analysis method based on X and Y dual-measurement-point spectrum data
CN114154779A (en) Health assessment and fault diagnosis method and system based on EMD-SOM
CN114330096A (en) Deep learning-based rail transit motor bearing fault degree classification model construction method
US11809517B1 (en) Adaptive method of cleaning structural health monitoring data based on local outlier factor
CN110994802A (en) Method and device for monitoring running state of power transmission conductor
CN116226719A (en) Bearing fault diagnosis method based on multidimensional steady-state vibration characteristics and related components
CN112284721A (en) Double fault analysis method and system for friction and rotor imbalance of large unit
CN110160781B (en) Test set reconstruction and prediction method for rotary machine fault classification
CN112232212B (en) Triple concurrent fault analysis method and system, large unit device and storage medium
CN115659271A (en) Sensor abnormality detection method, model training method, system, device, and medium
CN115839788A (en) Fixed state detection system of feeder fixture based on stress measurement
CN114715354A (en) Health management device for marine equipment and equipment state detection and fault diagnosis method
CN111833905B (en) System and method for detecting quality of marked character based on audio analysis
CN111650436B (en) Subsynchronous oscillation identification method for high-proportion renewable energy power system
CN112270227A (en) Oil film whirl and friction concurrent fault analysis method and analysis system
CN117664420A (en) Multi-probe bolt shaft force monitoring method and device, electronic equipment and 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
GR01 Patent grant
GR01 Patent grant