CN114218980A - Method for predicting residual life of rolling bearing - Google Patents

Method for predicting residual life of rolling bearing Download PDF

Info

Publication number
CN114218980A
CN114218980A CN202111404253.1A CN202111404253A CN114218980A CN 114218980 A CN114218980 A CN 114218980A CN 202111404253 A CN202111404253 A CN 202111404253A CN 114218980 A CN114218980 A CN 114218980A
Authority
CN
China
Prior art keywords
rolling bearing
time domain
life
value
index
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202111404253.1A
Other languages
Chinese (zh)
Inventor
邵辰彤
刘莹
后麒麟
郭培培
杨乐
罗泽熙
王景霖
单添敏
曹亮
沈勇
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.)
AVIC Shanghai Aeronautical Measurement Controlling Research Institute
Original Assignee
AVIC Shanghai Aeronautical Measurement Controlling Research Institute
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 AVIC Shanghai Aeronautical Measurement Controlling Research Institute filed Critical AVIC Shanghai Aeronautical Measurement Controlling Research Institute
Priority to CN202111404253.1A priority Critical patent/CN114218980A/en
Publication of CN114218980A publication Critical patent/CN114218980A/en
Pending legal-status Critical Current

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
    • G06F2218/02—Preprocessing
    • G06F2218/04—Denoising
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
    • G01M13/00—Testing of machine parts
    • G01M13/04—Bearings
    • G01M13/045—Acoustic or vibration analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/044—Recurrent networks, e.g. Hopfield networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • 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

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Artificial Intelligence (AREA)
  • Software Systems (AREA)
  • Health & Medical Sciences (AREA)
  • Evolutionary Computation (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • Computational Linguistics (AREA)
  • Biophysics (AREA)
  • Mathematical Physics (AREA)
  • Biomedical Technology (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Strategic Management (AREA)
  • Human Resources & Organizations (AREA)
  • Data Mining & Analysis (AREA)
  • Economics (AREA)
  • Development Economics (AREA)
  • Game Theory and Decision Science (AREA)
  • Signal Processing (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Marketing (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Tourism & Hospitality (AREA)
  • General Business, Economics & Management (AREA)
  • Acoustics & Sound (AREA)
  • Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)

Abstract

The invention discloses a method for predicting the residual life of a rolling bearing, which comprises the following steps: firstly, collecting an original time domain vibration signal of the running of a rolling bearing; step two, carrying out noise reduction smoothing treatment on the original time domain vibration signal of the rolling bearing obtained in the step one to obtain a useful signal which can obviously represent the whole life cycle of the bearing; step three, extracting time domain characteristics and frequency domain characteristics of the bearing signals subjected to noise reduction in the step two; step four, carrying out correlation analysis on the time domain characteristics and the frequency domain characteristics extracted in the step three, and selecting characteristic indexes which are strongly related to the service life of the bearing; and step five, constructing a residual life prediction model, and extracting strong correlation characteristic indexes obtained in the step four of another bearing to predict the residual life of the bearing. The method for predicting the residual life of the rolling bearing is simple in input and output, high in prediction precision and good in prediction effect.

Description

Method for predicting residual life of rolling bearing
Technical Field
The invention relates to the field of bearing vibration signal processing and residual life prediction in mechanical equipment, in particular to a residual life prediction method of a rolling bearing based on MRSVD and LSTM.
Technical Field
Along with the deepening of the refinement, systematization, automation and large-scale degree of mechanical equipment, the sustainable requirement of the work is higher and higher, so that the health management requirement of the equipment is also improved. There is a need for a mechanical part characterized by a significant tendency to degrade, which is evaluated for remaining life, which ensures the working efficiency of the mechanical equipment and prevents secondary damage to the machine due to part failure. The rolling bearing is one of indispensable spare parts in most mechanical equipment, plays very important effect in mechanical system, simultaneously, because rolling bearing easily works under high temperature, high pressure environment, has also led to its life to reduce easily, becomes one of the easy machine parts that consume. Therefore, the current residual service life prediction of the rolling bearing becomes an important link in the health management of modern equipment.
In the process of signal acquisition and transmission, due to the influence of external environment interference and instruments per se, noise is mixed in the signal, and a great influence is generated on a signal analysis result, so that signal denoising is a basic work of signal analysis.
Disclosure of Invention
The invention aims to provide a method for predicting the residual life of a rolling bearing, which has the advantages of good prediction effect, high precision, no need of establishing a complex model and convenience in calculation.
The invention aims to be realized by the following technical scheme:
a method for predicting the remaining life of a rolling bearing, comprising:
step (1): collecting an original time domain vibration signal of the rolling bearing by using a vibration sensor;
step (2): carrying out noise reduction smoothing treatment on the original time domain vibration signal of the rolling bearing obtained in the step (1) to obtain a smooth time domain vibration signal after noise reduction;
and (3): extracting time domain characteristic indexes and frequency domain characteristic indexes of the smooth time domain vibration signals subjected to noise reduction in the step (2);
and (4): performing correlation analysis on the time domain characteristics and the frequency domain characteristics extracted in the step (3), and selecting characteristic indexes which are strongly correlated with the service life of the bearing;
and (5): and constructing a residual life prediction model by using characteristic indexes of the whole life cycle of the rolling bearing, which are strongly related to the service life of the bearing, wherein the residual life prediction model is used for predicting the residual life of the rolling bearing under the same condition.
Further, the noise reduction smoothing processing method in the step (2) is multi-resolution singular value decomposition.
Further, the number of decomposition layers of the multi-resolution singular value decomposition is: when the kurtosis value of the detail signal reaches the maximum, the decomposition layer number is determined as the current decomposition layer number plus 5.
Further, in the step (3), the time domain characteristic index includes dimensional characteristics and dimensionless characteristics, and the dimensional characteristics include average amplitude, variance, maximum value, square root mean value, minimum value, absolute average amplitude, kurtosis, root mean square, skewness and peak-to-peak value; the dimensionless features include a waveform index, a peak index, a pulse index, a margin index, a kurtosis index, and a skewness index.
Further, in the step (3), the frequency domain characteristic index includes a center of gravity frequency, an average frequency, a root mean square frequency, and a frequency standard deviation.
Further, the correlation analysis of step (4) is Pearson correlation analysis by the following formula:
Figure BDA0003372197480000031
where ρ isX,YIs the correlation coefficient of the variables X and Y; e is the mathematical expectation of the sample;
and selecting a time domain characteristic index and a frequency domain characteristic index with a correlation coefficient of 0.5-1 as characteristic indexes with strong correlation of bearing service life.
Further, step (5) constructing a residual life prediction model by using a long and short flag memory neural network;
carrying out min-max standardization processing on the data subjected to the correlation analysis in the step (4), and inputting the standardized data into the long-term and short-term memory neural network; the normalization process formula is:
xminmax＝(x-xmin)/(xmax-xmin)
wherein xminIs the minimum value in the sequence, xmaxIs the maximum value in the sequence and x is the current value of the sequence.
Further, the step (5) further comprises the step of obtaining the service life prediction precision of the current residual service life prediction model on the currently measured rolling bearing through the average error absolute percentage MAPE:
Figure BDA0003372197480000032
where y is the true value of the remaining life, yiThe predicted remaining life value is n, and the total predicted length is n.
The invention has the beneficial effects that: the vibration signals of the rolling bearing are collected through the vibration sensor, and the noise of the original vibration signals is eliminated through multi-resolution singular value decomposition. And extracting time domain characteristics and frequency domain characteristics from the denoised vibration signal to obtain characteristic parameters representing the running state of the bearing. And a data base for health management and fault diagnosis of the bearing is obtained. In the service life prediction, characteristic information strongly related to the service life of the bearing is obtained through Pearson correlation analysis, the LSTM neural network is trained by using data, the trained neural network is used, the data of the current state of the bearing to be detected is input into the network to predict the current service life, the residual service life is calculated, the service life of the rolling bearing can be predicted at any time before failure, and the prediction precision and the prediction effect are relatively high.
Drawings
In order to more clearly illustrate the technical solution of the embodiment of the present invention, the drawings used in the embodiment will be briefly described below. It is appreciated that the following drawings depict only certain embodiments of the invention and are therefore not to be considered limiting of its scope. For a person skilled in the art, it is possible to obtain other relevant figures from these figures without any inventive effort.
Fig. 1 is a schematic flow chart of a method for predicting the remaining life of a rolling bearing according to an embodiment of the present invention.
FIG. 2 is a schematic diagram of the processing flow of the multi-resolution singular value decomposition on the original vibration signal in the present invention.
Detailed Description
The following detailed description of embodiments of the invention refers to the accompanying drawings.
Referring to fig. 1, the present embodiment provides a method for predicting the remaining life of a rolling bearing, and the specific steps of the method will be described below.
Step (1): and acquiring a raw time domain vibration signal of the rolling bearing by using a vibration sensor. At this time, the signal collected by the vibration sensor includes data of bearing degradation information and noise data.
Step (2): and (3) carrying out noise reduction smoothing treatment on the original time domain vibration signal of the rolling bearing obtained in the step (1) to obtain a noise-reduced smooth time domain vibration signal, wherein the smooth time domain vibration signal can obviously represent a useful signal of the whole life cycle of the bearing.
In this embodiment, the noise reduction smoothing processing method is multi-resolution singular value decomposition (MRSVD), and obtains a smoothed time domain vibration signal after noise reduction. The multi-resolution singular value decomposition is a signal decomposition method for decomposing signals into different levels of subspaces by combining a matrix binary recursive structure principle and referring to a wavelet multi-resolution analysis idea on the basis of singular value decomposition. The multi-resolution singular value decomposition can keep the phase of the separated signal in the original signal unchanged and does not influence the phase of the retained signal, so that the method is a zero phase shift denoising method which is difficult to realize by a common filter method and simultaneously eliminates the boundary error in the conventional wavelet transform denoising.
For a signal A0＝(x1,x2,...,xN) Constructing a Hankel matrix with the number of rows 2
Figure BDA0003372197480000051
Obtain singular value of sigma1,σ2Then by σ1,σ2To obtain A1,D1Wherein A is1Is a relatively large contribution of, and is the main component of the signal, D1The contribution amount is small, and the composition is a detailed composition. Constructing a matrix of the main components of the signals, and obtaining a series of SVD signals by recursion, wherein the main signals are marked as AjDetail signal is Dj。
For any noisy discrete digital signal x (i), it can be expressed as
x(i)＝s(i)+ξ(i)i＝1,2,...,N
Where s (i) is a normal signal, ξ (i) is a noise signal, and N is a signal length, the Hankel matrix a constructed by x (i) can be expressed as:
A＝As+Aξ
wherein A issAnd AξHankel matrices constructed for s (i) and ξ (i), respectively.
Because of the irregularity of the noise signals, singular values after xi (i) decomposition are mutually uniform, so that each pair of x (i) and x (i) is subjected to singular value decomposition, most of normal signals are retained, the noise signals can be halved, and the noise signals after second decomposition only remain 1/4, so that the noise signals after N times of decomposition only remain 1/2 of the original noise signalsnMost of the useful signals are retained, and good effects can be obtained after a certain number of layers of decomposition.
For the determination of the decomposition layer number, when the kurtosis value of the detail signal reaches the maximum, namely the kurtosis value of the detail signal of the next layer obtained by continuing decomposition of the current similar signal has a descending trend, the decomposition layer number is determined as the current decomposition layer number plus 5.
And (3): and (3) extracting the time domain characteristic index and the frequency domain characteristic index of the smooth time domain vibration signal subjected to noise reduction in the step (2). The time domain characteristic indexes comprise dimensional characteristics (detailed in a table 1), dimensionless characteristics (detailed in a table 2) and frequency domain characteristic indexes (detailed in a table 3), and the multidimensional characteristic set of the vibration signal parameters is constructed.
TABLE 1
Figure BDA0003372197480000061
TABLE 2
Figure BDA0003372197480000062
Figure BDA0003372197480000071
TABLE 3
Figure BDA0003372197480000072
And (4): and (4) carrying out correlation analysis on the time domain characteristics and the frequency domain characteristics extracted in the step (3), and selecting to obtain characteristic indexes which are strongly correlated with the service life of the bearing.
In this embodiment, correlation analysis is performed on each characteristic parameter after noise elimination of the rolling bearing by using a Pearson correlation analysis method for the extracted time domain characteristic index and the extracted frequency domain characteristic index, and a vibration signal characteristic parameter strongly correlated with the remaining service life of the bearing is screened out and used as an input signal of a neural network afterwards. The Pearson correlation coefficient calculation formula is as follows:
Figure BDA0003372197480000073
where ρ isX,YIs the correlation coefficient of the variables X and Y; e is the mathematical expectation of the sample.
When the two variables X and Y are independent, the Pearson correlation coefficient is 0, and if the variables X and Y have negative correlation, the correlation coefficient rho isX,YIf the variables X and Y have positive correlation between-1 and 0, the correlation coefficient rhoX,YBetween 0 and 1. The larger the absolute value of the correlation coefficient is, the stronger the correlation is, and the specific coefficient range and the correlation degree are shown in the following table:
Figure BDA0003372197480000081
and removing irrelevant and weakly relevant characteristic parameters to prevent the characteristic parameters from polluting the input of a neural network, and selecting vibration signal characteristic parameters with strong relevance of 0.5-1 to construct a new strong relevant characteristic set matrix.
And (5): and constructing a residual life prediction model by using characteristic indexes of the whole life cycle of the rolling bearing, which are strongly related to the service life of the bearing, wherein the residual life prediction model is used for predicting the residual life of the rolling bearing under the same condition.
In the present embodiment, a long and short time memory neural network (LSTM) is used for the residual life prediction model training. The long-short term memory neural network is used as an improved cyclic neural network, and can effectively solve the problem of gradient explosion or gradient disappearance when the previous cyclic neural network is too deep or has too many time sequences. Meanwhile, the method is very good in performance of time sequence prediction, so that the method has higher application value in the life prediction of the rotary machine, and an LSTM prediction method is selected in the life prediction process.
And (3) carrying out min-max standardization processing on the data subjected to the Pearson correlation analysis in the step (4), inputting the standardized data into the long-term and short-term memory neural network, and constructing an LSTM residual service life prediction model by taking the actual service life of the rolling bearing as a mapping result of the network. The normalization process formula is:
xminmax＝(x-xmin)/(xmax-xmin)
wherein xminIs the minimum value in the sequence, xmaxIs the maximum value in the sequence. x is the current value of the sequence.
Step 5.1: and inputting the standardized data into a long-term and short-term memory neural network for training. With respect to long-short term memory neural networks (LSTM), the model has three gates and a memory cell, which are an input gate (input gate), an output gate (output gate), a forgetting gate (forget gate) and a memory cell (memory cell) for keeping information state. Memory cell ctFrom an input gate itAnd forget door ftAnd the control is used for saving the information state at the current moment. The input gate determines how much valid information can be retained in the memory unit when the input is currently input, the forgetting gate determines how much information can be inherited to the current state of the memory unit at the previous moment, and the output gate otThe information of the memory unit at the current moment is processed and output.
The output data of the neuron at the last moment and the current moment input enter a first interaction layer, namely a forgetting gate layer, and are output f after being processed by the forgetting gatetThe value is a number of 0 to 1, and is in positive correlation with the importance degree, and the specific expression is shown as the following formula：
ft＝σ[Wf·(ht-1,xt)+bf]
The second layer of interaction is the input gate layer, and the data at the previous time and the current input are processed by the input gate to determine which information should be updated into the memory unit, and the expression is shown as follows:
it＝σ[Wi·(ht-1,xt)+bi]
after the input gate calculation is finished, a candidate value is created
Figure RE-GDA0003506506750000093
For updating the calculated value of the current state input gate into the whole sequence memory unit CtWherein the expression is as follows:
Figure RE-GDA0003506506750000091
then the memory unit is updated, and the updated value and symbol can be obtained by combining the calculation result of the forgetting gate and the calculation result of the input gate
Figure BDA0003372197480000094
Representing a multiplication by element, the expression is as follows:
Figure BDA0003372197480000093
the output gate generates the output value of the LSTM of the current state by the data of the last moment and the current input, and the expression is as follows:
ot＝σ[Wo·(ht-1,xt)+bo]
based on the memory unit updates, the final output value of the LSTM neurons is determined:
Figure BDA0003372197480000102
step 5.2: and determining hyper-parameters of the LSTM by a trial and error method, wherein the hyper-parameters comprise cycle times, neuron number, dropout layer proportion, target precision and the like, and determining the LSTM network after the network converges.
Step 5.3: regarding the application of the remaining life prediction model: after the LSTM network is determined in the steps 5.1 and 5.2, the same type of bearing with the same working condition as the training data is used, the characteristic parameter data determined in the steps (1) - (4) of the vibration signal collected in real time are input into the residual service life prediction model, and the residual service life of the current rolling bearing is obtained through the mapping relation in the residual service life prediction model.
Step 5.4: and obtaining the service life prediction precision of the current neural network to the rolling bearing through the absolute percentage of mean error (MAPE), wherein the MAPE has the following calculation formula:
Figure BDA0003372197480000101
where y is the true value of the remaining life, yiThe predicted remaining life value is n, and the total predicted length is n. The smaller the MAPE, the higher the prediction accuracy.
The invention solves the defects in the service life prediction algorithm of the existing rolling bearing, compared with the existing common wavelet noise reduction, the noise reduction part eliminates the phenomena of phase deviation and boundary error, and determines the characteristic parameters with strong correlation through Pearson correlation analysis, thereby preventing data pollution.
It will be understood by those skilled in the art that the present invention is not limited to the embodiments described above, which are given by way of illustration of the principles of the present invention, and that various changes and modifications may be made without departing from the spirit and scope of the invention as defined by the appended claims. The scope of the invention is defined by the appended claims and equivalents thereof.

Claims (8)

1. A method for predicting the remaining life of a rolling bearing, comprising:
step (1): collecting an original time domain vibration signal of the rolling bearing by using a vibration sensor;
step (2): carrying out noise reduction smoothing treatment on the original time domain vibration signal of the rolling bearing obtained in the step (1) to obtain a smooth time domain vibration signal after noise reduction;
and (3): extracting a time domain characteristic index and a frequency domain characteristic index of the smooth time domain vibration signal subjected to noise reduction in the step (2);
and (4): performing correlation analysis on the time domain characteristics and the frequency domain characteristics extracted in the step (3), and selecting characteristic indexes which are strongly correlated with the service life of the bearing;
and (5): and constructing a residual life prediction model by using the characteristic indexes of the whole life cycle of the rolling bearing, which are strongly related to the service life of the bearing, wherein the residual life prediction model is used for predicting the residual life of the rolling bearing under the same condition.
2. The method for predicting the residual life of a rolling bearing according to claim 1, wherein the noise reduction smoothing processing method of the step (2) is multi-resolution singular value decomposition.
3. The method for predicting the remaining life of a rolling bearing according to claim 1, wherein the number of decomposition layers of the multi-resolution singular value decomposition is: when the kurtosis value of the detail signal reaches the maximum, the decomposition layer number is determined as the current decomposition layer number plus 5.
4. The method for predicting the remaining life of a rolling bearing according to claim 1, wherein in the step (3), the time domain characteristic index comprises dimensional characteristics and dimensionless characteristics, and the dimensional characteristics comprise average amplitude, variance, maximum value, square root mean value, minimum value, absolute average amplitude, kurtosis, root mean square, skewness and peak-to-peak value; the dimensionless characteristic comprises a waveform index, a peak index, a pulse index, a margin index, a kurtosis index and a skewness index.
5. The method according to claim 1, wherein in step (3), the frequency domain characteristic index includes a center of gravity frequency, an average frequency, a root mean square frequency, and a standard deviation of frequency.
6. The residual life prediction method of a rolling bearing according to claim 1, characterized in that the correlation analysis of the step (4) is a Pearson correlation analysis by the following formula:
Figure FDA0003372197470000021
where ρ isX,YIs the correlation coefficient of the variables X and Y; e is the mathematical expectation of the sample;
and selecting a time domain characteristic index and a frequency domain characteristic index with a correlation coefficient of 0.5-1 as characteristic indexes with strong correlation of bearing service life.
7. The residual life prediction method of a rolling bearing according to claim 1, wherein the step (5) constructs a residual life prediction model using a long and short flag memory neural network;
carrying out min-max standardization processing on the data subjected to the correlation analysis in the step (4), and inputting the standardized data into the long-term and short-term memory neural network; the normalization process formula is:
xminmax＝(x-xmin)/(xmax-xmin)
wherein xminIs the minimum value in the sequence, xmaxIs the maximum value in the sequence and x is the current value of the sequence.
8. The method for predicting the remaining life of a rolling bearing according to claim 1, wherein the step (5) further comprises obtaining the accuracy of the prediction of the life of the rolling bearing currently measured by the current remaining life prediction model through the mean absolute error percentage MAPE:
Figure FDA0003372197470000022
where y is the true value of the remaining life, yiThe predicted value of the residual service life is n, and the predicted total length is n.
CN202111404253.1A 2021-11-24 2021-11-24 Method for predicting residual life of rolling bearing Pending CN114218980A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111404253.1A CN114218980A (en) 2021-11-24 2021-11-24 Method for predicting residual life of rolling bearing

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111404253.1A CN114218980A (en) 2021-11-24 2021-11-24 Method for predicting residual life of rolling bearing

Publications (1)

Publication Number Publication Date
CN114218980A true CN114218980A (en) 2022-03-22

Family

ID=80698343

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111404253.1A Pending CN114218980A (en) 2021-11-24 2021-11-24 Method for predicting residual life of rolling bearing

Country Status (1)

Country Link
CN (1) CN114218980A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116502049A (en) * 2023-06-25 2023-07-28 山东科技大学 Rolling bearing residual service life prediction method, system, equipment and storage medium

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116502049A (en) * 2023-06-25 2023-07-28 山东科技大学 Rolling bearing residual service life prediction method, system, equipment and storage medium
CN116502049B (en) * 2023-06-25 2023-09-08 山东科技大学 Rolling bearing residual service life prediction method, system, equipment and storage medium

Similar Documents

Publication Publication Date Title
CN112418277B (en) Method, system, medium and equipment for predicting residual life of rotating machine parts
CN111967688B (en) Power load prediction method based on Kalman filter and convolutional neural network
CN113344295B (en) Method, system and medium for predicting residual life of equipment based on industrial big data
CN110992354A (en) Abnormal region detection method for countering self-encoder based on introduction of automatic memory mechanism
CN116226646B (en) Method, system, equipment and medium for predicting health state and residual life of bearing
CN111397901A (en) Rolling bearing fault diagnosis method based on wavelet and improved PSO-RBF neural network
CN110987436B (en) Bearing fault diagnosis method based on excitation mechanism
CN108399434B (en) Analysis and prediction method of high-dimensional time series data based on feature extraction
CN106656357B (en) Power frequency communication channel state evaluation system and method
CN113176092A (en) Motor bearing fault diagnosis method based on data fusion and improved experience wavelet transform
CN111931983A (en) Precipitation prediction method and system
CN115452362A (en) Fault diagnosis method for gear box
CN111291918B (en) Rotating machine degradation trend prediction method based on stationary subspace exogenous vector autoregression
CN114091349A (en) Multi-source field self-adaption based rolling bearing service life prediction method
CN113642508A (en) Bearing fault diagnosis method based on parameter self-adaptive VMD and optimized SVM
CN114218980A (en) Method for predicting residual life of rolling bearing
Sun et al. Data-driven fault diagnosis method based on second-order time-reassigned multisynchrosqueezing transform and evenly mini-batch training
Shi et al. BP neural network based bearing fault diagnosis with differential evolution & EEMD denoise
CN114841412A (en) Method for predicting pH value of sea cucumber growing water
CN116167007A (en) Analog circuit detection method based on gating recursion unit self-coding neural network
CN114021275B (en) Rolling bearing fault diagnosis method based on deep convolution fuzzy system
CN115901255A (en) Bearing fault diagnosis method based on parameter optimization VMD-SVM
CN116933119A (en) Signal data trend removal method based on convolutional neural network
CN114755010A (en) Rotary machine vibration fault diagnosis method and system
CN114647904A (en) Self-adaptive sparse graph learning bearing life prediction method based on digital twin dictionary

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