CN112395550A - Rotary machine fault early warning method based on visual characteristic parameter matrix - Google Patents

Rotary machine fault early warning method based on visual characteristic parameter matrix Download PDF

Info

Publication number
CN112395550A
CN112395550A CN202011302101.6A CN202011302101A CN112395550A CN 112395550 A CN112395550 A CN 112395550A CN 202011302101 A CN202011302101 A CN 202011302101A CN 112395550 A CN112395550 A CN 112395550A
Authority
CN
China
Prior art keywords
characteristic parameter
trend
characteristic
sensor channel
frequency
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
CN202011302101.6A
Other languages
Chinese (zh)
Other versions
CN112395550B (en
Inventor
俞炅旻
章艺
符栋梁
钟焱
李国平
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.)
704th Research Institute of CSIC
Original Assignee
704th Research Institute of CSIC
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 704th Research Institute of CSIC filed Critical 704th Research Institute of CSIC
Priority to CN202011302101.6A priority Critical patent/CN112395550B/en
Publication of CN112395550A publication Critical patent/CN112395550A/en
Application granted granted Critical
Publication of CN112395550B publication Critical patent/CN112395550B/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
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/16—Matrix or vector computation, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization
    • 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

Landscapes

  • Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Computational Mathematics (AREA)
  • Mathematical Analysis (AREA)
  • Mathematical Optimization (AREA)
  • Pure & Applied Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Theoretical Computer Science (AREA)
  • Computing Systems (AREA)
  • Acoustics & Sound (AREA)
  • Algebra (AREA)
  • Databases & Information Systems (AREA)
  • Software Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)

Abstract

The invention provides a rotary machine fault early warning method based on a visual characteristic parameter matrix. According to the method, the characteristic frequency of the equipment is automatically calculated through the equipment process parameters, the frequency domain characteristics and the time domain characteristics are extracted, the comprehensive fault early warning of the multi-dimensional characteristic parameters is realized, and the problem of insufficient warning is reduced; according to the invention, after the system applying the method is deployed, the upper limit and the lower limit of the threshold can be adaptively adjusted based on data, so that the workload and uncertainty of manually adjusting the threshold are reduced, and the working efficiency and the early warning accuracy are effectively improved; according to the method, through l1 trend filtering, the filtered data trend reflects the state change trend of the equipment more accurately, and the probability of false alarm is reduced; according to the method, the thermodynamic diagram is used for displaying the difference matrix of the characteristic parameters and the threshold value, the multichannel comprehensive display of the equipment state is achieved, the information expression mode is optimized, and the fault information is clear at a glance.

Description

Rotary machine fault early warning method based on visual characteristic parameter matrix
Technical Field
The invention relates to a rotary machine fault early warning method based on a visual characteristic parameter matrix, and belongs to the technical field of state monitoring.
Background
The rotating machinery is widely applied to the fields of ships, aerospace, power plants, metallurgy and the like, and the safe and reliable operation of equipment not only relates to the economic benefit of enterprises, but also influences the safety of production. Therefore, the realization of the fault early warning of the rotating machinery equipment has important value for reducing the maintenance cost of the equipment and improving the utilization rate of the equipment in the life cycle. At present, a common mechanical equipment state monitoring system adopts a vibration total value trend curve based on a fixed threshold value to realize fault early warning, and the method has the following defects:
(1) due to the complex working condition of the equipment, the original data has larger disturbance and can not accurately reflect the change trend of the vibration total value;
(2) only the vibration total value is used for fault early warning, the characteristics are single, and the fault early warning accuracy rate is very limited;
(3) the trend map is adopted for early warning, and under the condition that more test channels and characteristic values exist, the excessive map and the poor information visualization effect are caused;
(4) the threshold is fixed and invariable, corresponding thresholds need to be determined for different working conditions of different equipment, the workload is large, and artificial interference is easily introduced, so that the early warning effect is poor;
disclosure of Invention
The technical problem to be solved by the invention is as follows: the multi-channel and multi-characteristic parameter comprehensive fault early warning false alarm rate is high, the visualization effect is poor, and the like.
In order to solve the technical problem, the technical scheme of the invention is to provide a rotary machine fault early warning method based on a visual characteristic parameter matrix, which is characterized by comprising the following steps of:
step 1, constructing a characteristic parameter matrix:
Figure BDA0002787155430000011
in formula (1), M is the total number of sensor channels tested on the device, and then:
Figure BDA0002787155430000021
for the inner ring vibration characteristic frequency of the mth sensor channel tested on the device,
Figure BDA0002787155430000022
n is the equipment rotating speed obtained by the mth sensor channel;
fi (m)for the outer ring vibration characteristic frequency of the mth sensor channel tested on the device,
Figure BDA0002787155430000023
Figure BDA0002787155430000024
Dmthe diameter of a bearing pitch circle is used, d is the diameter of a bearing rolling element and is troublesome supplement, alpha is a bearing contact angle, and Z is the number of gear teeth;
Figure BDA0002787155430000025
the characteristic frequency of vibration when the retainer of the mth sensor channel tested on the device touches the outer ring,
Figure BDA0002787155430000026
Figure BDA0002787155430000027
the characteristic frequency of vibration when the cage of the mth sensor channel tested on the device hits the inner ring,
Figure BDA0002787155430000028
Figure BDA0002787155430000029
for the rolling element vibration characteristic frequency of the mth sensor channel tested on the device,
Figure BDA00027871554300000210
Figure BDA00027871554300000211
Figure BDA00027871554300000212
for the gear mesh frequency of the mth sensor channel tested on the device,
Figure BDA00027871554300000213
Figure BDA00027871554300000214
for the rms value of the mth sensor channel tested on the device,
Figure BDA00027871554300000215
n is the sample sequence length, x (i) is the ith sample data;
Figure BDA00027871554300000216
for the margin factor of the mth sensor channel tested on the device,
Figure BDA00027871554300000217
Xpis the peak-to-peak value;
Figure BDA00027871554300000218
for the kurtosis factor of the mth sensor channel tested on the device,
Figure BDA00027871554300000219
x is sample data, E (X) is a sample mean value, and sigma is a sample variance;
Figure BDA00027871554300000220
for the pulse factor of the mth sensor channel tested on the device,
Figure BDA00027871554300000221
Figure BDA00027871554300000222
is the sample mean;
the inner ring vibration characteristic frequency, the outer ring vibration characteristic frequency, the vibration characteristic frequency when the retainer touches the outer ring, the vibration characteristic frequency when the retainer touches the inner ring, the rolling body vibration characteristic frequency, the gear meshing frequency, the root mean square value, the margin factor, the kurtosis factor and the pulse factor are all characteristic parameters;
step 2, after acquiring trend data of stable operation of equipment with a certain data volume by M sensor channels, calculating probability density distribution of each characteristic parameter in a characteristic parameter matrix shown in formula (1), calculating a beta distribution shape parameter most approximate to actual data by least square fitting to obtain a beta distribution probability density model beta (gamma, eta) of the trend data, calculating an adaptive threshold interval of each characteristic parameter by using the beta distribution probability density model beta (gamma, eta), wherein Th1 represents the upper threshold limit of the adaptive threshold interval of a certain characteristic parameter, and Th2 represents the lower threshold limit of the adaptive threshold interval of a certain characteristic parameter;
step 3, carrying out l1 trend filtering on the trend data to obtain a trend curve of each characteristic parameter in the characteristic parameter matrix shown in the formula (1), and estimating to obtain a new characteristic parameter matrix;
step 4, calculating a difference matrix of the feature parameter matrix obtained in step 3 in the upper threshold limit of the adaptive threshold interval of each feature parameter obtained in step 2, as shown in the following formula:
Figure BDA0002787155430000031
in the above formula, the superscript of Th1 indicates the sensor channel to which it belongs, and the subscript of Th1 indicates the specific characteristic parameter to which it belongs;
and 5, visually displaying the states of the characteristic parameters according to the predefined thermal grade based on the difference matrix obtained in the step 4.
Preferably, in step 2, after normalizing the trend data, calculating the probability density distribution of each characteristic parameter, and then fitting the shape parameters γ and η of the beta distribution by using a least square method to solve the beta distribution model β (γ and η), wherein the ith trend data X is subjected to the following formula (2)iNormalization:
Figure BDA0002787155430000032
in the formula (2), the reaction mixture is,
Figure BDA0002787155430000033
for normalized trend data, XminIs the minimum of all N trend data, XmaxIs the maximum of all N trend data.
Preferably, in step 2, after obtaining the beta distribution probability density model β (γ, η), the bilateral α quantile λ is calculated by using the following formulas (3) and (4)1、λ2：
Figure BDA0002787155430000034
Figure BDA0002787155430000035
In the formulas (3) and (4),
Figure BDA0002787155430000036
is a random variable and is used as a random variable,
Figure BDA0002787155430000037
for random variables less than λ1The probability of (a) of (b) being,
Figure BDA0002787155430000038
for random variables greater than λ2A is a quantile;
based on bilateral alpha quantiles lambda1、λ2Th1 and Th2 are calculated by the following formulas (5) and (6):
Th1＝λ1×(Xmax-Xmin)+Xmin (5)
Th2＝λ2×(Xmax-Xmin)+Xmin (6)
thereby obtaining the adaptive threshold interval of each characteristic parameter.
Preferably, in step 3, the l1 trend filtering is performed by using the following formula (7):
Figure BDA0002787155430000041
in equation (7), y is a time series of eigenvalues, λ is a non-negative parameter that controls the smoothness of the trend line and the size of the balance remainder to achieve a balance between the trend of the control estimation and the signal redundancy, and θ ═ θ (θ ═ θ1,θ2,…,θn) Are characteristic parameters (n is the number of characteristic parameters). A is the lower triangular matrix:
Figure BDA0002787155430000042
the solution yields the result of the problem shown in equation (7)
Figure BDA0002787155430000043
Then l1 trend filter results
Figure BDA0002787155430000044
Compared with the prior art, the invention has the following beneficial effects:
1) according to the method, the characteristic frequency of the equipment is automatically calculated through the equipment process parameters, the frequency domain characteristics and the time domain characteristics are extracted, the comprehensive fault early warning of the multi-dimensional characteristic parameters is realized, and the problem of insufficient warning is reduced;
2) according to the invention, after the system applying the method is deployed, the upper limit and the lower limit of the threshold can be adaptively adjusted based on data, so that the workload and uncertainty of manually adjusting the threshold are reduced, and the working efficiency and the early warning accuracy are effectively improved;
3) according to the method, through l1 trend filtering, the filtered data trend reflects the state change trend of the equipment more accurately, and the probability of false alarm is reduced;
4) according to the method, the thermodynamic diagram is used for displaying the difference matrix of the characteristic parameters and the threshold value, the multichannel comprehensive display of the equipment state is achieved, the information expression mode is optimized, the fault information is clear at a glance, the running state of the equipment is mastered in real time, and the technical support is provided for the fault early warning of the equipment.
Drawings
FIG. 1 is a flow chart of a fault early warning method provided by the present invention;
fig. 2 is a visual characteristic parameter matrix obtained by the present invention.
Detailed Description
The invention will be further illustrated with reference to the following specific examples. It should be understood that these examples are for illustrative purposes only and are not intended to limit the scope of the present invention. Further, it should be understood that various changes or modifications of the present invention may be made by those skilled in the art after reading the teaching of the present invention, and such equivalents may fall within the scope of the present invention as defined in the appended claims.
As shown in fig. 1, the invention relates to a rotary machine fault early warning method based on a visual characteristic parameter matrix, which comprises the following steps:
step 1, constructing a characteristic parameter matrix:
Figure BDA0002787155430000051
in formula (1), M is the total number of sensor channels tested on the device, and then:
Figure BDA0002787155430000052
for the inner ring vibration characteristic frequency of the mth sensor channel tested on the device,
Figure BDA0002787155430000053
n is the equipment rotating speed obtained by the mth sensor channel;
fi (m)for the outer ring vibration characteristic frequency of the mth sensor channel tested on the device,
Figure BDA0002787155430000054
Figure BDA0002787155430000055
Dmthe diameter of a bearing pitch circle, d is the diameter of a bearing rolling element, alpha is a bearing contact angle, and Z is the number of gear teeth;
Figure BDA0002787155430000056
the characteristic frequency of vibration when the retainer of the mth sensor channel tested on the device touches the outer ring,
Figure BDA0002787155430000057
Figure BDA0002787155430000058
the characteristic frequency of vibration when the cage of the mth sensor channel tested on the device hits the inner ring,
Figure BDA0002787155430000059
Figure BDA00027871554300000510
for the rolling element vibration characteristic frequency of the mth sensor channel tested on the device,
Figure BDA00027871554300000511
Figure BDA00027871554300000512
Figure BDA00027871554300000513
for the gear mesh frequency of the mth sensor channel tested on the device,
Figure BDA00027871554300000514
Figure BDA0002787155430000061
for the rms value of the mth sensor channel tested on the device,
Figure BDA0002787155430000062
n is the sample sequence length, x (i) is the ith sample data;
Figure BDA0002787155430000063
for the margin factor of the mth sensor channel tested on the device,
Figure BDA0002787155430000064
Xpis the peak-to-peak value;
Figure BDA0002787155430000065
for the kurtosis factor of the mth sensor channel tested on the device,
Figure BDA0002787155430000066
x is sample data, E (X) is a sample mean value, and sigma is a sample variance;
Figure BDA0002787155430000067
for the pulse factor of the mth sensor channel tested on the device,
Figure BDA0002787155430000068
Figure BDA0002787155430000069
is the sample mean;
the inner ring vibration characteristic frequency, the outer ring vibration characteristic frequency, the vibration characteristic frequency when the retainer touches the outer ring, the vibration characteristic frequency when the retainer touches the inner ring, the rolling body vibration characteristic frequency, the gear meshing frequency, the root mean square value, the margin factor, the kurtosis factor and the pulse factor are all characteristic parameters;
setting parameters such as a bearing, a gear, a rotating speed and the like of equipment, and extracting characteristic frequency; setting the arrangement of measuring points of the equipment for determining the number of rows of the characteristic parameter matrix
Step 2, calculating a characteristic parameter matrix according to the set parameters, recording real-time vibration data, and continuously enriching a trend database until the requirement of the data volume of threshold dynamic adjustment is met; after acquiring trend data of stable operation of equipment with a certain data volume by M sensor channels, calculating probability density distribution of each characteristic parameter in a characteristic parameter matrix shown in formula (1), calculating a beta distribution shape parameter most approximate to actual data by least square fitting to obtain a beta distribution probability density model beta (gamma, eta) of the trend data, calculating an adaptive threshold interval of each characteristic parameter by using the beta distribution probability density model beta (gamma, eta), wherein Th1 represents the upper threshold limit of the adaptive threshold interval of a certain characteristic parameter, and Th2 represents the lower threshold limit of the adaptive threshold interval of a certain characteristic parameter;
in the step, after the trend data is normalized, the probability density distribution of each characteristic parameter is calculated, then the least square method is used for fitting the shape parameters gamma and eta of the beta distribution, and the beta distribution model beta (gamma and eta) is solved, wherein the following formula (2) is adopted for the ith trend data XiNormalization:
Figure BDA00027871554300000610
in the formula (2), the reaction mixture is,
Figure BDA00027871554300000611
for normalized trend data, XminIs the minimum of all N trend data, XmaxThe maximum value of all the N trend data;
after a beta distribution probability density model beta (gamma, eta) is obtained, the alpha quantiles lambda at two sides of the model beta (gamma, eta) are calculated by using the following formulas (3) and (4)1、λ2：
Figure BDA0002787155430000071
Figure BDA0002787155430000072
In the formulas (3) and (4),
Figure BDA0002787155430000073
is a random variable and is used as a random variable,
Figure BDA0002787155430000074
for random variables less than λ1The probability of (a) of (b) being,
Figure BDA0002787155430000075
for random variables greater than λ2A is a quantile;
based on bilateral alpha quantiles lambda1、λ2Th1 and Th2 are calculated by the following formulas (5) and (6):
Th1＝λ1×(Xmax-Xmin)+Xmin (5)
Th2＝λ2×(Xmax-Xmin)+Xmin (6)
thereby obtaining the self-adaptive threshold interval of each characteristic parameter;
step 3, carrying out l1 trend filtering on the trend data to obtain a trend curve of each characteristic parameter in the characteristic parameter matrix shown in the formula (1), and estimating to obtain a new characteristic parameter matrix;
in this step, the l1 trend filtering is performed using the following equation (7):
Figure BDA0002787155430000076
in equation (7), y is a time series of eigenvalues, λ is a non-negative parameter that controls the smoothness of the trend line and the size of the balance remainder to achieve a balance between the trend of the control estimation and the signal redundancy, and θ ═ θ (θ ═ θ1,θ2,…,θn) Are characteristic parameters (n is the number of characteristic parameters). A is the lower triangular matrix:
Figure BDA0002787155430000077
the solution yields the result of the problem shown in equation (7)
Figure BDA0002787155430000078
Then l1 trend filter results
Figure BDA0002787155430000079
Step 4, calculating a difference matrix of the feature parameter matrix obtained in step 3 in the upper threshold limit of the adaptive threshold interval of each feature parameter obtained in step 2, as shown in the following formula:
Figure BDA00027871554300000710
in the above formula, the superscript of Th1 indicates the sensor channel to which it belongs, and the subscript of Th1 indicates the specific characteristic parameter to which it belongs;
and 5, visually displaying the states of the characteristic parameters according to the predefined thermal grade based on the difference matrix obtained in the step 4. As shown in fig. 2, the visual display of the device state is realized according to the colors of the parameters in the set thermodynamic diagram display matrix, wherein the thermodynamic diagram shows that the characteristic parameter of the position exceeds the upper threshold, and the deeper the color shows that the difference is larger, the device is about to fail.

Claims (4)

1. A rotary machine fault early warning method based on a visual characteristic parameter matrix is characterized by comprising the following steps:
step 1, constructing a characteristic parameter matrix:
Figure FDA0002787155420000011
in formula (1), M is the total number of sensor channels tested on the device, and then:
Figure FDA0002787155420000012
for the inner ring vibration characteristic frequency of the mth sensor channel tested on the device,
Figure FDA0002787155420000013
n is the equipment rotating speed obtained by the mth sensor channel;
fi (m)for the outer ring vibration characteristic frequency of the mth sensor channel tested on the device,
Figure FDA0002787155420000014
Figure FDA0002787155420000015
Dmthe diameter of a bearing pitch circle, d is the diameter of a bearing rolling element, alpha is a bearing contact angle, and Z is the number of gear teeth;
Figure FDA0002787155420000016
the characteristic frequency of vibration when the retainer of the mth sensor channel tested on the device touches the outer ring,
Figure FDA0002787155420000017
Figure FDA0002787155420000018
the characteristic frequency of vibration when the cage of the mth sensor channel tested on the device hits the inner ring,
Figure FDA0002787155420000019
Figure FDA00027871554200000110
for the rolling element vibration characteristic frequency of the mth sensor channel tested on the device,
Figure FDA00027871554200000111
Figure FDA00027871554200000112
Figure FDA00027871554200000113
for the gear mesh frequency of the mth sensor channel tested on the device,
Figure FDA00027871554200000114
Figure FDA00027871554200000115
for the rms value of the mth sensor channel tested on the device,
Figure FDA00027871554200000116
n is the sample sequence length, x (i) is the ith sample data;
Figure FDA00027871554200000117
for the margin factor of the mth sensor channel tested on the device,
Figure FDA00027871554200000118
Xpis the peak-to-peak value;
Figure FDA0002787155420000021
for the kurtosis factor of the mth sensor channel tested on the device,
Figure FDA0002787155420000022
x is sample data, E (X) is a sample mean value, and sigma is a sample variance;
Figure FDA0002787155420000023
for the pulse factor of the mth sensor channel tested on the device,
Figure FDA0002787155420000024
Figure FDA0002787155420000025
is the sample mean;
the inner ring vibration characteristic frequency, the outer ring vibration characteristic frequency, the vibration characteristic frequency when the retainer touches the outer ring, the vibration characteristic frequency when the retainer touches the inner ring, the rolling body vibration characteristic frequency, the gear meshing frequency, the root mean square value, the margin factor, the kurtosis factor and the pulse factor are all characteristic parameters;
step 2, after acquiring trend data of stable operation of equipment with a certain data volume by M sensor channels, calculating probability density distribution of each characteristic parameter in a characteristic parameter matrix shown in formula (1), calculating a beta distribution shape parameter most approximate to actual data by least square fitting to obtain a beta distribution probability density model beta (gamma, eta) of the trend data, calculating an adaptive threshold interval of each characteristic parameter by using the beta distribution probability density model beta (gamma, eta), wherein Th1 represents the upper threshold limit of the adaptive threshold interval of a certain characteristic parameter, and Th2 represents the lower threshold limit of the adaptive threshold interval of a certain characteristic parameter;
step 3, carrying out l1 trend filtering on the trend data to obtain a trend curve of each characteristic parameter in the characteristic parameter matrix shown in the formula (1), and estimating to obtain a new characteristic parameter matrix;
step 4, calculating a difference matrix of the feature parameter matrix obtained in step 3 in the upper threshold limit of the adaptive threshold interval of each feature parameter obtained in step 2, as shown in the following formula:
Figure FDA0002787155420000026
in the above formula, the superscript of Th1 indicates the sensor channel to which it belongs, and the subscript of Th1 indicates the specific characteristic parameter to which it belongs;
and 5, visually displaying the states of the characteristic parameters according to the predefined thermal grade based on the difference matrix obtained in the step 4.
2. The method as claimed in claim 1, wherein in step 2, after normalizing the trend data, calculating the probability density distribution of each characteristic parameter, and then fitting the shape parameters γ and η of the beta distribution by using the least square method to solve the beta distribution model β (γ, η), wherein the ith trend data X is obtained by using the following formula (2)iNormalization:
Figure FDA0002787155420000031
in the formula (2), the reaction mixture is,
Figure FDA0002787155420000032
for normalized trend data, XminIs the minimum of all N trend data, XmaxIs the maximum of all N trend data.
3. A process as claimed in claim 1The rotary machine fault early warning method based on the visual characteristic parameter matrix is characterized in that in the step 2, after a beta distribution probability density model beta (gamma, eta) is obtained, the following formulas (3) and (4) are utilized to calculate alpha quantiles lambda of two sides of the rotary machine1、λ2：
Figure FDA0002787155420000033
Figure FDA0002787155420000034
In the formulas (3) and (4),
Figure FDA0002787155420000035
is a random variable and is used as a random variable,
Figure FDA0002787155420000036
for random variables less than λ1The probability of (a) of (b) being,
Figure FDA0002787155420000037
for random variables greater than λ2A is a quantile;
based on bilateral alpha quantiles lambda1、λ2Th1 and Th2 are calculated by the following formulas (5) and (6):
Th1＝λ1×(Xmax-Xmin)+Xmin (5)
Th2＝λ2×(Xmax-Xmin)+Xmin (6)
thereby obtaining the adaptive threshold interval of each characteristic parameter.
4. The rotating machine fault early warning method based on the visualized characteristic parameter matrix as claimed in claim 1, wherein in step 3, the l1 trend filtering is performed by using the following formula (7):
Figure FDA0002787155420000038
in equation (7), y is a time series of eigenvalues, λ is a non-negative parameter that controls the smoothness of the trend line and the size of the balance remainder to achieve a balance between the trend of the control estimation and the signal redundancy, and θ ═ θ (θ ═ θ1,θ2,…,θn) Is a characteristic parameter, n is the number of the characteristic parameters, A is a lower triangular matrix:
Figure FDA0002787155420000039
the solution yields the result of the problem shown in equation (7)
Figure FDA00027871554200000310
Then l1 trend filter results
Figure FDA00027871554200000311
CN202011302101.6A 2020-11-19 2020-11-19 Rotary machine fault early warning method based on visual characteristic parameter matrix Active CN112395550B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011302101.6A CN112395550B (en) 2020-11-19 2020-11-19 Rotary machine fault early warning method based on visual characteristic parameter matrix

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011302101.6A CN112395550B (en) 2020-11-19 2020-11-19 Rotary machine fault early warning method based on visual characteristic parameter matrix

Publications (2)

Publication Number Publication Date
CN112395550A true CN112395550A (en) 2021-02-23
CN112395550B CN112395550B (en) 2023-02-03

Family

ID=74606560

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011302101.6A Active CN112395550B (en) 2020-11-19 2020-11-19 Rotary machine fault early warning method based on visual characteristic parameter matrix

Country Status (1)

Country Link
CN (1) CN112395550B (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114135477A (en) * 2021-10-11 2022-03-04 昆明嘉和科技股份有限公司 Pump equipment state monitoring dynamic threshold early warning method
CN116756490A (en) * 2023-06-15 2023-09-15 沈阳航空航天大学 Rolling bearing fault early warning method based on beta distribution and EEMD-CMSE

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110208019A (en) * 2019-05-14 2019-09-06 北京博华信智科技股份有限公司 A kind of dynamic equipment condition monitoring dynamic threshold method for early warning
CN111120348A (en) * 2019-12-25 2020-05-08 中国石化销售股份有限公司华南分公司 Centrifugal pump fault early warning method based on support vector machine probability density estimation
CN111222247A (en) * 2020-01-13 2020-06-02 北京化工大学 Early fault early warning method for rotary machine

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110208019A (en) * 2019-05-14 2019-09-06 北京博华信智科技股份有限公司 A kind of dynamic equipment condition monitoring dynamic threshold method for early warning
CN111120348A (en) * 2019-12-25 2020-05-08 中国石化销售股份有限公司华南分公司 Centrifugal pump fault early warning method based on support vector machine probability density estimation
CN111222247A (en) * 2020-01-13 2020-06-02 北京化工大学 Early fault early warning method for rotary machine

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
张明等: "基于动态自学习阈值和趋势滤波的机械故障智能预警方法", 《振动与冲击》 *
马波等: "变分自编码器在机械故障预警中的应用", 《计算机工程与应用》 *

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114135477A (en) * 2021-10-11 2022-03-04 昆明嘉和科技股份有限公司 Pump equipment state monitoring dynamic threshold early warning method
CN114135477B (en) * 2021-10-11 2024-04-02 昆明嘉和科技股份有限公司 Dynamic threshold early warning method for monitoring state of machine pump equipment
CN116756490A (en) * 2023-06-15 2023-09-15 沈阳航空航天大学 Rolling bearing fault early warning method based on beta distribution and EEMD-CMSE

Also Published As

Publication number Publication date
CN112395550B (en) 2023-02-03

Similar Documents

Publication Publication Date Title
CN112395550B (en) Rotary machine fault early warning method based on visual characteristic parameter matrix
CN109556863B (en) MSPAO-VMD-based large turntable bearing weak vibration signal acquisition and processing method
CN111366123B (en) Part surface roughness and cutter wear prediction method based on multi-task learning
CN112304613B (en) Wind turbine generator bearing early warning method based on feature fusion
CN110926809B (en) Big data analysis-based wind turbine generator transmission chain fault early warning method
CN110648480B (en) Single variable alarm system and method based on change rate
CN104794492B (en) Machine tool processing running status ONLINE RECOGNITION method based on power features model
CN112160761B (en) Hard rock TBM cutter disc hob abrasion real-time evaluation method based on field parameters
CN116400126B (en) Low-voltage power box with data processing system
CN116660672B (en) Power grid equipment fault diagnosis method and system based on big data
CN114813124B (en) Bearing fault monitoring method and device
CN116189407B (en) Intelligent early warning system based on data monitoring
CN114135477A (en) Pump equipment state monitoring dynamic threshold early warning method
CN114871850B (en) Tool wear state assessment method based on vibration signals and BP neural network
CN111444471A (en) Cable production quality abnormity detection method and system based on multivariate Gaussian distribution
CN112946469A (en) Monitoring method for self-adaptive dynamic alarm threshold of hydraulic generator
CN109211564B (en) Self-adaptive threshold detection method for health assessment of ball screw pair
CN111191950B (en) Method and device for analyzing abnormal oil temperature of gearbox of wind turbine generator
CN115758233A (en) Fault diagnosis method for main transmission gear box of high-speed precision strip steel cold rolling mill
CN116562623A (en) Real-time planting risk assessment system suitable for velvet mushroom is planted
CN114004059B (en) Health portrait method for hydroelectric generating set
CN110925182B (en) Early warning method for vibration degradation of descaling pump of hot rolling production line
CN112749473B (en) Energy efficiency safety diagnosis system based on real-time online analysis and simulated offline analysis
CN111125626B (en) Model order fixing method based on S-shaped function random subspace identification
CN109975702B (en) Direct-current gear reduction motor quality inspection method based on circulation network classification model

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