CN114647909B - Method for determining degradation point of rolling bearing based on spectral kurtosis characteristic maximum value - Google Patents

Method for determining degradation point of rolling bearing based on spectral kurtosis characteristic maximum value Download PDF

Info

Publication number
CN114647909B
CN114647909B CN202210341586.2A CN202210341586A CN114647909B CN 114647909 B CN114647909 B CN 114647909B CN 202210341586 A CN202210341586 A CN 202210341586A CN 114647909 B CN114647909 B CN 114647909B
Authority
CN
China
Prior art keywords
characteristic
value
time
degradation
rolling bearing
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.)
Active
Application number
CN202210341586.2A
Other languages
Chinese (zh)
Other versions
CN114647909A (en
Inventor
孙丽
赵俊杰
任小蝶
周宏根
李国超
毕豪
杨景珲
罗皓文
王臻斌
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.)
Jiangsu University of Science and Technology
Original Assignee
Jiangsu University of Science and 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 Jiangsu University of Science and Technology filed Critical Jiangsu University of Science and Technology
Priority to CN202210341586.2A priority Critical patent/CN114647909B/en
Publication of CN114647909A publication Critical patent/CN114647909A/en
Application granted granted Critical
Publication of CN114647909B publication Critical patent/CN114647909B/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
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/10—Geometric CAD
    • G06F30/17—Mechanical parametric or variational design
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
    • G06F2119/04—Ageing analysis or optimisation against ageing
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
    • G06F2119/12—Timing analysis or timing optimisation

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Geometry (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Evolutionary Computation (AREA)
  • Computer Hardware Design (AREA)
  • General Engineering & Computer Science (AREA)
  • Pure & Applied Mathematics (AREA)
  • Mathematical Optimization (AREA)
  • Mathematical Analysis (AREA)
  • Computational Mathematics (AREA)
  • Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)

Abstract

The invention discloses a method for determining a degradation point of a rolling bearing based on a spectral kurtosis characteristic maximum, which comprises the steps of sampling a vibration signal of the rolling bearing in a whole life cycle of a tested rolling bearing T times to form a time sequence W= [ S ] 1 ,S 2 ,…,S T ]The sampling signal of each time unit is S; calculating the spectral kurtosis based on the sampling signal S of each time unit, extracting M time domain characteristic values according to the spectral kurtosis of each time unit, and forming a characteristic matrix X T×M Wherein M is the number of extracted features, and the number of M is at least 5; smoothing each column of characteristic time sequence of the characteristic matrix; carrying out 0-giving treatment on 5% of characteristic data at the beginning and ending stages of each column of characteristic time sequence; and selecting the corresponding time t' with the highest occurrence frequency as a bearing degradation point. According to the method, from the aspect of determining the degradation starting time of the rolling bearing, the degradation degree of the bearing is evaluated, degradation stages are divided, and the usability of performance degradation indexes is improved.

Description

Method for determining degradation point of rolling bearing based on spectral kurtosis characteristic maximum value
Technical Field
The invention relates to the field of rolling bearing residual life prediction, in particular to a method for determining a rolling bearing degradation point based on a spectral kurtosis characteristic maximum value.
Background
The rolling bearing is one of the parts which are widely applied to the rotary mechanical equipment, and the health state of the bearing is closely related to whether the rotary mechanical equipment can normally operate, so that the degradation degree of the rolling bearing is evaluated, and the rolling bearing which can not meet the working requirement can be replaced in time according to the use requirement of the rotary mechanical equipment, so that the rolling bearing has important significance for ensuring the stable operation of the rotary mechanical equipment. The rolling bearing service life prediction method is to build a degradation model according to performance degradation indexes to predict. The traditional performance degradation index construction method is constructed according to the rolling bearing full life cycle vibration signal, and the link of how to determine the bearing degradation starting moment is not considered, so that the fitting degree of the constructed performance degradation index to a degradation model is not high, and the prediction accuracy is low.
Therefore, there is a need to solve the above-mentioned problems.
Disclosure of Invention
The invention aims to: the invention aims to provide a method for determining a degradation point of a rolling bearing based on a spectral kurtosis characteristic maximum value, wherein the degradation start time can be determined.
The technical scheme is as follows: in order to achieve the above object, the invention discloses a method for determining a degradation point of a rolling bearing based on a spectral kurtosis characteristic maximum, which comprises the following steps:
(1) T times sampling is carried out on the full life cycle vibration signal of the rolling bearing to be tested so as to form a time sequence W= [ S ] 1 ,S 2 ,…,S T ]The sampling signal per time unit is s= [ S ] 1 ,S 2 ,…,S j ]Where j is the number of sampling points;
(2) Calculating the spectral kurtosis based on the sampling signal S of each time unit, extracting M time domain characteristic values according to the spectral kurtosis of each time unit, and forming a characteristic matrix X T×M Wherein M is the number of extracted features, and the number of M is at least 5;
(3) Smoothing each column of characteristic time sequence of the characteristic matrix;
(4) Carrying out 0-giving treatment on 5% of characteristic data at the beginning and ending stages of each column of characteristic time sequence;
(5) And selecting the corresponding time t' with the highest occurrence frequency as a bearing degradation point.
The time domain feature value in the step (2) refers to a maximum value, a minimum value, a median, a quarter bit distance, an absolute error mean value, an absolute error median, a variance, an empirical distribution function percentile, an empirical distribution function slope, a mean square value, a mean square error, a root mean square error, a mean value, a standard value, a skewness, a kurtosis, a peak-to-peak value, a root mean square, a peak factor, a shape factor, a pulse factor or a margin factor.
Preferably, the specific step of smoothing each column of the feature sequence of the feature matrix in the step (3) is as follows:
if the current value in each row of characteristic time sequences is more than 5 values before, the characteristic time sequences of each row are taken
Figure BDA0003579554380000021
Figure BDA0003579554380000022
The current value of (2) and the average of five values preceding the current value are taken as new current values: />
Figure BDA0003579554380000023
Wherein d is n A characteristic value of the nth time unit;
if the current value in each sequence of characteristic time sequences is less than 5 values, taking the average value of the current value in each sequence of characteristic time sequences and the existing value before the current value as a new current value:
Figure BDA0003579554380000024
the specific steps of the 0-adding process in the step (4) are as follows:
for the initial stage d 1 ,d 2 ,…,d a Make 0 and at the same time make the last stage d T-b ,d T-b+1 ,…,d T Carrying out 0 assignment; wherein a= [0.05×t ]]，b＝[0.05×T]。
Further, the specific step of selecting the corresponding time t' with the highest current frequency as the bearing degradation point in the step (5) is as follows:
calculating the maximum value of each column of characteristic time sequence, marking corresponding time t ', and counting corresponding time data [ t ] of the maximum value of all characteristic time sequences' 1 ,t′ 2 ,…,t′ M ]The method comprises the steps of carrying out a first treatment on the surface of the Extracting all unique values g in the corresponding time data, and setting out the occurrence frequency p of the unique values in the data; after obtaining all the frequency data, selecting the maximum value p in the frequency data max The method comprises the steps of carrying out a first treatment on the surface of the And taking the unique value g with the highest occurrence frequency as the degradation starting time of the bearing.
The beneficial effects are that: compared with the prior art, the invention has the following remarkable advantages: firstly, considering the aspect of determining the degradation starting time of the rolling bearing, evaluating the degradation degree of the bearing, dividing degradation stages and improving the usability of performance degradation indexes; secondly, on the one hand, the invention can better track transient components in the signal, and can determine the initial moment of non-stationary signal degradation under a strong noise background; on the other hand, the degradation degree of the rolling bearing can be accurately estimated, and the dominant performance degradation index is constructed according to the degradation degree, so that the residual life prediction accuracy is improved.
Drawings
FIG. 1 is a flow chart of the method of the present invention;
FIG. 2 is a schematic diagram of a data smoothing process step in the present invention;
FIG. 3 is a graph of 6 sets of time domain characteristic trends of the bearing 1_5 horizontal vibration signal for 52 minutes in the invention;
FIG. 4 is a graph showing characteristic trends of kurtosis of 6 groups of vibration signals of the bearing 1_5 in the horizontal direction for 52 minutes in the invention.
Detailed Description
The technical scheme of the invention is further described below with reference to the accompanying drawings.
As shown in fig. 1, the method for determining the degradation point of the rolling bearing based on the spectral kurtosis characteristic maximum value comprises the following steps:
(1) T times sampling the vibration signal of the rolling bearing in the life cycle to form one rolling bearingTime series w= [ S ] 1 ,S 2 ,…,S T ]The sampling signal per time unit is s= [ S ] 1 ,S 2 ,…,S j ]Where j is the number of sampling points;
(2) Calculating the spectral kurtosis based on the sampling signal S of each time unit, extracting M time domain characteristic values according to the spectral kurtosis of each time unit, and forming a characteristic matrix X T×M Wherein M is the number of extracted features, and the number of M is at least 5; wherein the time domain characteristic value refers to a maximum value, a minimum value, a median, a quartile range, an absolute error mean value, an absolute error median, a variance, an empirical distribution function percentile, an empirical distribution function slope, a mean square value, a mean square error, a root mean square error, a mean value, a standard value, a skewness, a kurtosis, a peak-to-peak value, a root mean square, a peak factor, a shape factor, a pulse factor or a margin factor;
(3) Smoothing each column of characteristic time sequence of the characteristic matrix, wherein the specific steps of smoothing are as follows:
if the current value in each row of characteristic time sequences is more than 5 values before, the characteristic time sequences of each row are taken
Figure BDA0003579554380000031
Figure BDA0003579554380000032
The current value of (2) and the average of five values preceding the current value are taken as new current values:
Figure BDA0003579554380000033
wherein d is n A characteristic value of the nth time unit;
if the current value in each sequence of characteristic time sequences is less than 5 values, taking the average value of the current value in each sequence of characteristic time sequences and the existing value before the current value as a new current value:
Figure BDA0003579554380000034
(4) Carrying out 0-giving treatment on 5% of characteristic data at the beginning and ending stages of each column of characteristic time sequence; the specific steps of the 0-giving treatment are as follows:
for the initial stage d 1 ,d 2 ,…,d a Make 0 and at the same time make the last stage d T-b ,d T-b+1 ,…,d T Carrying out 0 assignment; wherein a= [0.05×t ]]，b＝[0.05×T]；
(5) The corresponding time t' with highest occurrence frequency is selected as a bearing degradation point, and the specific steps are as follows: calculating the maximum value of each column of characteristic time sequence, marking corresponding time t ', and counting corresponding time data [ t ] of the maximum value of all characteristic time sequences' 1 ,t′ 2 ,…,t′ M ]The method comprises the steps of carrying out a first treatment on the surface of the Extracting all unique values g in the corresponding time data, and setting out the occurrence frequency p of the unique values in the data; after obtaining all the frequency data, selecting the maximum value p in the frequency data max The method comprises the steps of carrying out a first treatment on the surface of the And taking the unique value g with the highest occurrence frequency as the degradation starting time of the bearing.
Examples
The data are derived from XJTU-SY rolling bearing accelerated life test data set, and the test bearing is LDK UER204 rolling bearing. According to the embodiment of the invention, the vibration signals of the 5 bearings in the horizontal direction under the first working condition of the data set are selected as analysis sources, the first working condition sets the rotation speed born by the bearings to be 2100r/min, and the radial force born by the bearings to be 12kN.
The embodiment of the method for determining the degradation point of the rolling bearing based on the spectral kurtosis characteristic maximum comprises the following steps of:
(1) T times sampling is carried out on the full life cycle vibration signal of the rolling bearing to be tested so as to form a time sequence W= [ S ] 1 ,S 2 ,…,S T ]The sampling signal per time unit is s= [ S ] 1 ,S 2 ,…,s j ]Where j is the number of sample points, i.e. the vibration signal sampled per minute is s= [ S ] 1 ,S 2 ,…,S j ]The method comprises the steps of carrying out a first treatment on the surface of the Wherein the single sampling frequency is 25.6kHz, and the sampling duration is 1.28s, wherein the interval between two adjacent sampling is 1min; in addition, in order to ensure safety, the amplitude 30g of the signal is regarded as a bearing fault threshold value in the embodiment;
(2) The method comprises the steps of sequentially extracting the characteristics of all bearings, namely firstly extracting 10 time domain characteristic values from a sampling signal S of each time unit of a single bearing to form a characteristic matrix XT multiplied by 10; the 10 time domain characteristic values are mean value, standard value, skewness, kurtosis, peak-to-peak value, root mean square, peak factor, shape factor, pulse factor and margin factor; then, short-time Fourier transform is carried out on the signals based on the fixed window length 256, and the kurtosis of each frequency spectral line is calculated so as to obtain a spectral kurtosis vector K= [ K ] of the time unit 1 ,K 2 ,…,K m ]Where m is the number of frequency samples. Accordingly, 10 features including the mean value, standard deviation, skewness and the like are extracted again, and the above 20 feature calculation methods are shown in table 1 in detail.
TABLE 1
Figure BDA0003579554380000041
Figure BDA0003579554380000051
/>
Figure BDA0003579554380000061
(3) Smoothing each column of characteristic time sequence of the characteristic matrix, wherein the specific steps of smoothing are as follows:
the time domain feature matrix Y of all the bearings can be obtained by the feature extraction step T×10 Sum spectrum kurtosis characteristic matrix X T×10 To enhance the observability of the curve trend, the two matrices are smoothed, as shown in fig. 2;
if the current value in each row of characteristic time sequences is more than 5 values before, the characteristic time sequences of each row are taken
Figure BDA0003579554380000062
Figure BDA0003579554380000063
The current value of (2) and the average of five values preceding the current value are taken as new current values:
Figure BDA0003579554380000064
wherein d is n A characteristic value of the nth time unit;
if the current value in each sequence of characteristic time sequences is less than 5 values, taking the average value of the current value in each sequence of characteristic time sequences and the existing value before the current value as a new current value:
Figure BDA0003579554380000071
(4) Considering that data shock occurs in the initial stage and the tail cutting stage of the full life cycle signal of the bearing, so as to interfere with the selection of degradation points, and carrying out 0-giving processing on 5% of characteristic data in the beginning and ending stages of each row of characteristic time sequence; the specific steps of the 0-giving treatment are as follows:
for the initial stage d 1 ,d 2 ,…,d a Make 0 and at the same time make the last stage d T-b ,d T-b+1 ,…,d T Carrying out 0 assignment; wherein a= [0.05×t ]]，b＝[0.05×T]；
Taking a fifth bearing 1_5 as an example under the first working condition of the XJTU data set, collecting vibration signals of 52 time units of the bearing to form a spectral kurtosis characteristic matrix X 52×10 The method comprises the steps of carrying out a first treatment on the surface of the For characteristic matrix X 52×10 Each column of characteristic data
Figure BDA0003579554380000072
D at the initial stage 1 ,d 2 ,…,d a 0 is given, and d at the last stage is also given 52-b ,d 52-b+1 ,…,d 52 0 is added, and a= [ 0.05x52 ]]＝3，b＝[0.05×52]＝3；
(5) The corresponding time t' with highest occurrence frequency is selected as a bearing degradation point, and the specific steps are as follows: calculating the maximum value of each column of characteristic time sequence, marking corresponding time t ', and counting corresponding time data [ t ] of the maximum value of all characteristic time sequences' 1 ,t′ 2 ,…,t′ M ]The method comprises the steps of carrying out a first treatment on the surface of the Extracting all unique values g in the corresponding time data, and setting out the occurrence frequency p of the unique values in the data; after obtaining all the frequency data, selecting the maximum value p in the frequency data max The method comprises the steps of carrying out a first treatment on the surface of the Then taking the unique value g with highest occurrence frequency as the degradation starting time of the bearing;
taking bearing 1_5 as an example, after the data 0 processing is completed, the matrix X is obtained T×10 Selecting the maximum value of each column of characteristic data and marking the corresponding time t'; then, corresponding time data t 'of all maximum value points are counted' 1 ,t′ 2 ,…,t′ 10 ]Extracting all unique values g=4, 21, 35, 47 in the data and listing the frequency p at which the unique values occur 4 ＝1，p 21 ＝1，p 35 ＝6，p 47 =2; maximum p in frequency data max ＝p 35 =6, the unique value 35 with the highest occurrence frequency is taken as the degradation start time of the bearing, namely, the degradation start time of the bearing 1_5 is 35 th minute here; taking 6 groups of time domain and spectral kurtosis characteristics of the bearing 1_5 respectively, as shown in fig. 3 and 4; as can be seen from fig. 3 and 4, the determined degradation points can more accurately divide the degradation phase of the bearing;
the degradation points of all the bearings were obtained according to the above method, and the degradation start timings of the respective bearings are shown in table 2.
TABLE 2
Figure BDA0003579554380000073
Figure BDA0003579554380000081
In order to prove the superiority of the invention, 10 time domain indexes of each bearing in the undivided degradation stage are subjected to monotonicity calculation, and the characteristic with the highest score in the time domain characteristics of each bearing is selected as the advantage index 1. And in addition, 10 time domain indexes of each bearing after the degradation stage is divided based on the method are processed in the same way, and the characteristic with the highest score is selected as the advantage index 2.
Monotonicity is an important measure of the degradation process of the device, for each column of characteristic data
Figure BDA0003579554380000082
The monotonicity score calculation process is shown in the following equation.
Figure BDA0003579554380000083
/>
Wherein M is the monotonicity score for the feature, d (i) represents the i-th feature of the feature sequence, sgn (·) function returns the sign of the parameter; if the calculated value is greater than 0, sgn (·) returns to 1; returning to 0 if the calculated value is equal to 0; otherwise, return to-1.
The dominance index 1 is compared to the dominance index 2 as shown in table 3.
TABLE 3 Table 3
Figure BDA0003579554380000084
As can be seen from table 3, the dominant index 2 of the bearing 1_1 is slightly smaller than the dominant index 1, the dominant index 2 of the other bearings is better than the dominant index 1, and the overall average value of the indexes is larger than the dominant index 1. The monotonicity score of the dominant index 2 of the bearing 1_1 is larger than 0.5, and the bearing still has good characteristic performance of the bearing degradation process. In summary, the performance degradation index constructed according to the method of the present invention is more advantageous in characterizing the performance of the rolling bearing degradation process.

Claims (4)

1. A method for determining a degradation point of a rolling bearing based on a spectral kurtosis characteristic maximum, comprising the steps of:
(1) T times sampling is carried out on the full life cycle vibration signal of the rolling bearing to be tested so as to form a time sequence W= [ S ] 1 ,S 2 ,…,S T ]The sampling signal per time unit is s= [ S ] 1 ,S 2 ,…,S j ]Where j is the number of sampling points;
(2) Calculating the spectral kurtosis based on the sampling signal S of each time unit, extracting M time domain characteristic values according to the spectral kurtosis of each time unit, and forming a characteristic matrix X T×M Wherein M is the number of extracted features, and the number of M is at least 5;
(3) Smoothing each column of characteristic time sequence of the characteristic matrix;
(4) Carrying out 0-giving treatment on 5% of characteristic data at the beginning and ending stages of each column of characteristic time sequence;
(5) Selecting a corresponding time t' with highest occurrence frequency as a bearing degradation point;
the specific step of selecting the corresponding time t' with the highest current frequency as the bearing degradation point in the step (5) is as follows:
calculating the maximum value of each column of characteristic time sequence, marking corresponding time t ', and counting corresponding time data [ t ] of the maximum value of all characteristic time sequences' 1 ,t′ 2 ,…,t′ M ]The method comprises the steps of carrying out a first treatment on the surface of the Extracting all unique values g in the corresponding time data, and setting out the occurrence frequency p of the unique values in the data; after obtaining all the frequency data, selecting the maximum value p in the frequency data max The method comprises the steps of carrying out a first treatment on the surface of the And taking the unique value g with the highest occurrence frequency as the degradation starting time of the bearing.
2. The method for determining the degradation point of the rolling bearing based on the maximum value of the spectral kurtosis characteristic according to claim 1, wherein: the time domain feature value in the step (2) refers to a maximum value, a minimum value, a median, a quarter bit distance, an absolute error mean value, an absolute error median, a variance, an empirical distribution function percentile, an empirical distribution function slope, a mean square value, a mean square error, a root mean square error, a mean value, a standard value, a skewness, a kurtosis, a peak value, a root mean square, a peak factor, a shape factor, a pulse factor or a margin factor.
3. The method for determining the degradation point of the rolling bearing based on the maximum value of the characteristic of the spectral kurtosis according to claim 2, wherein the specific step of smoothing each column of the characteristic sequence of the characteristic matrix in the step (3) is as follows:
if the current value in each row of characteristic time sequences is more than 5 values before, the characteristic time sequences of each row are taken
Figure FDA0004141126180000011
Figure FDA0004141126180000012
The current value of (2) and the average of five values preceding the current value are taken as new current values:
Figure FDA0004141126180000013
wherein d is n A characteristic value of the nth time unit;
if the current value in each sequence of characteristic time sequences is less than 5 values, taking the average value of the current value in each sequence of characteristic time sequences and the existing value before the current value as a new current value:
Figure FDA0004141126180000021
4. the method for determining the degradation point of the rolling bearing based on the maximum value of the spectral kurtosis characteristic according to claim 3, wherein the specific steps of the 0-giving process in the step (4) are as follows:
for the initial stage d 1 ,d 2 ,…,d a Make 0 and at the same time make the last stage d T-b ,d T-b+1 ,…,d T Carrying out 0 assignment; wherein a= [0.05×t ]]，b＝[0.05×T]。
CN202210341586.2A 2022-04-02 2022-04-02 Method for determining degradation point of rolling bearing based on spectral kurtosis characteristic maximum value Active CN114647909B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210341586.2A CN114647909B (en) 2022-04-02 2022-04-02 Method for determining degradation point of rolling bearing based on spectral kurtosis characteristic maximum value

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210341586.2A CN114647909B (en) 2022-04-02 2022-04-02 Method for determining degradation point of rolling bearing based on spectral kurtosis characteristic maximum value

Publications (2)

Publication Number Publication Date
CN114647909A CN114647909A (en) 2022-06-21
CN114647909B true CN114647909B (en) 2023-04-21

Family

ID=81995383

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210341586.2A Active CN114647909B (en) 2022-04-02 2022-04-02 Method for determining degradation point of rolling bearing based on spectral kurtosis characteristic maximum value

Country Status (1)

Country Link
CN (1) CN114647909B (en)

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108760266A (en) * 2018-05-31 2018-11-06 西安交通大学 The virtual degeneration index building method of mechanical key component based on learning distance metric
CN113670616A (en) * 2021-09-03 2021-11-19 苏州大学 Bearing performance degradation state detection method and system

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
SE534531C2 (en) * 2010-03-30 2011-09-20 Rubico Ab Method for error detection of rolling bearings by increasing statistical asymmetry
EP2835707B1 (en) * 2013-08-05 2016-07-27 ABB Technology AG A method for condition monitoring of distributed drive-trains
CN104568444B (en) * 2015-01-28 2017-02-22 北京邮电大学 Method for extracting fault characteristic frequencies of train rolling bearings with variable rotational speeds
CN108195587B (en) * 2018-02-12 2023-08-25 西安交通大学 Motor rolling bearing fault diagnosis method and system
CN112001314A (en) * 2020-08-25 2020-11-27 江苏师范大学 Early fault detection method for variable speed hoist
CN113449618A (en) * 2021-06-17 2021-09-28 南京航空航天大学 Method for carrying out deep learning rolling bearing fault diagnosis based on feature fusion and mixed enhancement

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108760266A (en) * 2018-05-31 2018-11-06 西安交通大学 The virtual degeneration index building method of mechanical key component based on learning distance metric
CN113670616A (en) * 2021-09-03 2021-11-19 苏州大学 Bearing performance degradation state detection method and system

Also Published As

Publication number Publication date
CN114647909A (en) 2022-06-21

Similar Documents

Publication Publication Date Title
CN104236905A (en) Bearing fault diagnosis method
CN101995437B (en) Method for extracting features of crack acoustic emission signal of drawing part
CN110006652B (en) Rolling bearing fault diagnosis method and system
CN113375939B (en) Mechanical part fault diagnosis method based on SVD and VMD
CN106771598B (en) A kind of Adaptive spectra kurtosis signal processing method
CN109270345A (en) A kind of detection method of mains by harmonics signal
CN104215459A (en) Bearing fault diagnosis method
CN114492538A (en) Local discharge signal denoising method for urban medium-voltage distribution cable
CN113567127B (en) Rolling bearing degradation index extraction method based on time-frequency feature separation
CN113391244B (en) VMD-based transformer closing vibration signal characteristic frequency calculation method
CN106419911A (en) Emotional detection method based on brain electric wave analysis
CN111581715B (en) Rapid compression method for accelerating load spectrum of tractor part
CN112364296A (en) P wave arrival time automatic picking method based on deep learning
CN114647909B (en) Method for determining degradation point of rolling bearing based on spectral kurtosis characteristic maximum value
CN103852525A (en) Acoustic emission signal identification method based on AR-HMM
CN108802175A (en) A kind of Pulsed eddy current testing signal de-noising method based on EMD
CN110412366B (en) Dynamic time algorithm-based transformer live-line detection method
CN108020761B (en) A kind of Denoising of Partial Discharge
CN108020741B (en) Method and device for identifying damping characteristics of double-frequency harmonic attenuation signal
CN114462452B (en) Asynchronous motor rotor broken bar fault diagnosis method using successive variable mode decomposition algorithm
CN113406386B (en) Signal frequency accurate estimation method based on digital down-conversion
CN111007369A (en) Ultrahigh frequency electromagnetic wave signal arrival time difference calculation method and device
CN111458605A (en) Dielectric breakdown characteristic analysis method under impulse voltage and related device
CN113113052B (en) Discrete point voice fundamental tone recognition device and computer storage medium
Xu et al. An improved pitch detection of speech combined with speech enhancement

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