CN108846241A - The rolling bearing life prediction technique to be linked based on Fu Leixie apart from dynamic digital-to-analogue - Google Patents
The rolling bearing life prediction technique to be linked based on Fu Leixie apart from dynamic digital-to-analogue Download PDFInfo
- Publication number
- CN108846241A CN108846241A CN201810725813.5A CN201810725813A CN108846241A CN 108846241 A CN108846241 A CN 108846241A CN 201810725813 A CN201810725813 A CN 201810725813A CN 108846241 A CN108846241 A CN 108846241A
- Authority
- CN
- China
- Prior art keywords
- health indicator
- rolling bearing
- leixie
- moment
- vector
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
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
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/04—Ageing analysis or optimisation against ageing
Landscapes
- Physics & Mathematics (AREA)
- Geometry (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Pure & Applied Mathematics (AREA)
- Mathematical Optimization (AREA)
- Mathematical Analysis (AREA)
- Computer Hardware Design (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Computational Mathematics (AREA)
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
Abstract
Based on the rolling bearing life prediction technique that Fu Leixie links apart from dynamic digital-to-analogue, regression analysis is carried out to rolling bearing health indicator using the Method Using Relevance Vector Machine with different nuclear parameters first, then utilization index models fitting associated vector；Dynamic evaluation is carried out to fitting degenerated curve by the Fu Leixie distance between digital simulation degenerated curve and smooth rear health indicator sequence, obtain optimal degenerated curve and its corresponding degradation model parameter, then recursion health indicator is until it reaches failure threshold, to realize based on the prediction under model and data-driven method dynamic linkage to rolling bearing remaining life, the invention avoids make a priori assumption to degradation model parameter, get rid of dependence of the prediction technique to a large amount of historical datas, improve the precision of rolling bearing life prediction, with preferable engineering application value.
Description
Technical field
The invention belongs to rolling bearing predicting residual useful life technical fields, and in particular to based on Fu Leixie apart from dynamic digital-to-analogue
The rolling bearing life prediction technique of linkage.
Background technique
Rolling bearing directly affects the reliable of equipment whether healthy as one of components most common in mechanical equipment
Property and comprehensive benefit, it is therefore desirable to formulate effective maintenance scheme to ensure equipment safety stable operation.Due to rolling bearing
It degenerates generally after long-term slowly early stage catagen phase and short-term violent serious catagen phase, according to traditional regular dimension
Shield scheme not only results in manpower and material resources waste, it is also possible to lead to equipment downtime due to replacing rolling bearing not in time, even
Cause the tragedy of fatal crass.Predictive maintenance scheme can then work out corresponding standby according to rolling bearing predicting residual useful life result
Part and maintenance plan maximize the reliability and comprehensive benefit of equipment.As the premise and crucial foundation for carrying out predictive maintenance,
How extensive concern and height weight of the Accurate Prediction rolling bearing remaining life just by domestic and international research institution and manufacturing enterprise
Depending on.
Existing rolling bearing method for predicting residual useful life can be roughly divided into two classes：Based on model method and data-driven side
Method.Based on model method by theory analysis to bearing degradation process or summary of experience, exponential model, Paris- are used
Erdogan model etc. describes its degradation trend, and carries out predicting residual useful life accordingly；And data-driven method then utilizes artificial intelligence
It can model (support vector machines, Method Using Relevance Vector Machine, artificial neural network etc.) excavation runing time, health status from historical data
Mapping relations between remaining life, to realize the predicting residual useful life of rolling bearing.But both the above method all exists
Obvious drawback：It needs researcher to analyse in depth rolling bearing degradation mechanism based on model method, while also needing to pre-suppose that mould
The prior distribution of shape parameter is easy to bring serious prediction error because of false supposition；The precision of prediction of data-driven method is not only
Depending on the quantity of historical data, its quality is relied more on, but is often difficult to obtain enough high quality in practice in engineering
Bearing history data.Therefore, above two life-span prediction method cannot effectively meet engineering actual demand at this stage.
Summary of the invention
In order to overcome the disadvantages of the above prior art, the purpose of the present invention is to provide be based on Fu Leixie apart from dynamic digital-to-analogue
The rolling bearing life prediction technique of linkage makes full use of the advantage based on two methods of model and data-driven, and abandons two
The deficiency of person, to improve the precision of rolling bearing predicting residual useful life.
In order to achieve the above object, the technical scheme adopted by the invention is as follows：
Based on the rolling bearing life prediction technique that Fu Leixie links apart from dynamic digital-to-analogue, include the following steps：
1) bearing vibration signal is acquired in real time using acceleration transducer, asinh is extracted from vibration signal
Health indicator of the index as rolling bearing：
Wherein, s (t) is the asinh index value of t moment, xtFor the vibration signal segment that t moment samples, σ
() is variance formula；
2) in tkThe health indicator obtained is input in the Method Using Relevance Vector Machine with different nuclear parameters by the moment, is executed back
Return analysis and obtains corresponding associated vector；Specific step is as follows：
2.1) rolling bearing is obtained from initial time t0To current time tkHealth indicator sequence Sk, wherein Sk={ s
(t0),...,s(tk)}；
2.2) kernel function that Method Using Relevance Vector Machine is arranged is gaussian kernel function, i.e.,：
K(tk,ti)=exp (- γ | | ti-tk||2) (2)
K () indicates kernel function, t in formulaiRepresent i-th of monitoring moment, i=0,1 ..., k, tk={ t0,t1,...,
tk, it represents from initial time t0To current time tkMonitoring moment sequence, γ is nuclear parameter；
2.3) it takes several values to carry out relevance vector regression nuclear parameter γ, obtains several associated vectorsWhereinN-th of associated vector is represented in tmThe estimated value of moment health indicator, n=1 ..., N, N are phase
Close vector sum；M=1 ..., M, M are each associated vector dimension, and further obtain the health indicator that confidence level is 95%
Estimated value confidence interval bound：
In above formula,For the confidence interval upper bound,For confidence interval lower bound, tM=(t1,t2,...,tM)T, generation
The time series of table associated vector,And Σn,MPFor the optimal hyper parameter of n-th of associated vector, the sparse pattra leaves of sequence is used
This learning algorithm obtains；
3) non-linear least square is carried out to several associated vectors obtained in step 2) using index degradation model to intend
It closes, obtains being fitted degenerated curve under different nuclear parameters：
gn(t)=an exp(bnt)+cn exp(dnt) (5)
Wherein, gn() indicates the corresponding fitting degenerated curve of n-th of associated vector, an, bn, cn, dnIt is that the curve is corresponding
Index degradation model parameter；
4) according to Fu Leixie apart from appraisal procedure 3) obtained in be respectively fitted degenerated curve, obtain optimal degenerated curve, have
Steps are as follows for body：
4.1) by rolling bearing from initial time t0To current time tkHealth indicator sequence SkIt is smoothed, structure
At smoothed out health indicator sequenceWhereinIt is smooth rear tiThe health indicator at moment
Value, i=1 ..., k；
4.2) calculate each associated vector fitting degenerated curve obtained in smoothed out health indicator sequence and step 3) it
Between Fu Leixie distance：
Wherein ξ () and ψ () is to meet ξ (0)=ψ (0)=t1, ξ (1)=ψ (1)=tkAnd domain is appointing for [0,1]
Meaning monotone nondecreasing function, D () are that Euclidean distance calculates function；
4.3) it finds the Fu Leixie between smoothed out health indicator sequence and remembers that this is quasi- apart from the smallest fitting degenerated curve
Conjunction curve is optimal degenerated curve, and correspondence model parameter is optimal degradation model parameter；
5) according to optimal degradation model parameter, using formula (5) recursion health indicator, remember the recursion value of health indicator reach or
It is t at the time of being for the first time more than default failure thresholdend, then the remaining life RUL (t of current time rolling bearingk) be expressed as：
RUL(tk)=tend-tk (7)。
Beneficial effects of the present invention：The present invention uses the Method Using Relevance Vector Machine with different nuclear parameters at each monitoring moment to rolling
Dynamic bearing health indicator carries out regression analysis, then by Exponential Model associated vector；It is obtained according to Fu Leixie distance screening
It can effectively reflect the optimal fitting degenerated curve of rolling bearing degradation trend, recursion health indicator, realizes based on model accordingly
With under data-driven method dynamic linkage to the predicting residual useful life of rolling bearing.The present invention overcomes needed based on model method
The dependence of the drawbacks of pre-supposing that model parameter prior distribution and data-driven method to a large amount of historical datas, improves the axis of rolling
Hold the accuracy of predicting residual useful life.
Detailed description of the invention
Fig. 1 is flow chart of the present invention.
Fig. 2 is embodiment PRONOSTIA experimental bench structure chart.
Fig. 3 is two test rolling bearing life cycle management internal vibration signals of embodiment.
Fig. 4 is two test rolling bearing degradation trend prediction result figures of embodiment.
Fig. 5 is two test rolling bearing predicting residual useful life result figures of embodiment.
Specific embodiment
The invention will be described in further detail with reference to the accompanying drawings and examples.
As shown in Figure 1, based on the rolling bearing life prediction technique that Fu Leixie links apart from dynamic digital-to-analogue, including following step
Suddenly：
1) bearing vibration signal is acquired in real time using acceleration transducer, asinh is extracted from vibration signal
Health indicator of the index as rolling bearing：
Wherein, s (t) is the asinh index value of t moment, xtFor the vibration signal segment that t moment samples, σ
() is variance formula；
2) in tkThe health indicator obtained is input in the Method Using Relevance Vector Machine with different nuclear parameters by the moment, is executed back
Return analysis and obtains corresponding associated vector；Specific step is as follows：
2.1) rolling bearing is obtained from initial time t0To current time tkHealth indicator sequence Sk, wherein Sk={ s
(t0),...,s(tk)}；
2.2) kernel function that Method Using Relevance Vector Machine is arranged is gaussian kernel function, i.e.,：
K(tk,ti)=exp (- γ | | ti-tk||2) (2)
K () indicates kernel function, t in formulaiRepresent i-th of monitoring moment, i=0,1 ..., k, tk={ t0,t1,...,
tk, it represents from initial time t0To current time tkMonitoring moment sequence, γ is nuclear parameter；
2.3) it takes several values to carry out relevance vector regression nuclear parameter γ, obtains several associated vectorsWhereinN-th of associated vector is represented in tmThe estimated value of moment health indicator, n=1 ..., N, N are phase
Close vector sum；M=1 ..., M, M are each associated vector dimension, and further obtain the health indicator that confidence level is 95%
Estimated value confidence interval bound：
In above formula,For the confidence interval upper bound,For confidence interval lower bound, tM=(t1,t2,...,tM)T,
The time series of associated vector is represented,And Σn,MPFor the optimal hyper parameter of n-th of associated vector, it can be used sequence sparse
Bayesian learning algorithm obtains；
3) non-linear least square is carried out to several associated vectors obtained in step 2) using index degradation model to intend
It closes, obtains being fitted degenerated curve under different nuclear parameters：
gn(t)=an exp(bnt)+cn exp(dnt) (5)
Wherein, gn() indicates the corresponding fitting degenerated curve of n-th of associated vector, an, bn, cn, dnIt is that the curve is corresponding
Index degradation model parameter；
4) according to Fu Leixie apart from appraisal procedure 3) obtained in be respectively fitted degenerated curve, obtain optimal degenerated curve, have
Steps are as follows for body：
4.1) by rolling bearing from initial time t0To current time tkHealth indicator sequence SkIt is smoothed, structure
At smoothed out health indicator sequenceWhereinIt is smooth rear tiThe health indicator at moment
Value, i=1 ..., k；
4.2) calculate each associated vector fitting degenerated curve obtained in smoothed out health indicator sequence and step 3) it
Between Fu Leixie distance：
Wherein ξ () and ψ () is to meet ξ (0)=ψ (0)=t1, ξ (1)=ψ (1)=tkAnd domain is appointing for [0,1]
Meaning monotone nondecreasing function, D () are that Euclidean distance calculates function；
4.3) it finds the Fu Leixie between smoothed out health indicator sequence and remembers that this is quasi- apart from the smallest fitting degenerated curve
Conjunction curve is optimal degenerated curve, and correspondence model parameter is optimal degradation model parameter；
5) according to optimal degradation model parameter, using formula (5) recursion health indicator, remember the recursion value of health indicator reach or
It is t at the time of being for the first time more than default failure thresholdend, then the remaining life RUL (tk) of current time rolling bearing is expressed as：
RUL(tk)=tend-tk (7)
Embodiment：Using the rolling bearing accelerated life test data on PRONOSTIA experimental bench to the present invention into
Row verifying.Fig. 2 is PRONOSTIA experimental bench structure chart, and the experimental bench is by transmission system, loading system and data collection system three
A subsystem composition, can carry out rolling bearing and accelerate experiment of degenerating.During the experiment, setting rolling bearing revolving speed is
1800rpm, load 4kN, using acceleration transducer acquire bearing vibration signal, sample frequency 25.6kHz, every time
Sample duration is 0.1s, sampling interval 10s.When bearing vibration amplitude reach or for the first time more than 20g when, it is believed that rolling
Dynamic bearing is entirely ineffective and stops testing.Two groups of rolling bearing life-cycle vibration signals of gained are as shown in Figure 3.
Before carrying out rolling bearing life prediction, a rolling bearing failure threshold need to be set.The present embodiment is according to rolling
The stop condition of dynamic bearing life-cycle experiment, it is believed that fail, and set when the asinh index of rolling bearing is more than 1.2
Setting the value is failure threshold.Simultaneously as the method for the present invention carries out associated vector using several nuclear parameters in each prediction time
It returns, also needs to predefine nuclear parameter before starting prediction.In the present embodiment, nuclear parameter γ=0.5n is set, wherein n=
1,2,...,100。
Choose T as shown in table 11、T2And T3When moment is as rolling bearing degeneration early stage, mid-term and the typical case in advanced stage
Carve and carry out degradation trend prediction, acquired results as shown in figure 4, as seen from the figure, although the degradation trend predicted value at moment early stage with
True value deviation is larger, but over time, the predicted value of degradation trend moves closer to true value, this shows energy of the present invention
It is enough to excavate rolling bearing degradation modes effectively from monitoring data and carry out predicting residual useful life.It is square for the present invention is furture elucidated
The superiority of method introduces the method based on Method Using Relevance Vector Machine and the method based on particle filter method as a comparison, uses respectively
Control methods and the method for the present invention carry out predicting residual useful life, obtained prediction result such as Fig. 5 to two above sample.Wherein,
The life prediction result obtained using the method based on Method Using Relevance Vector Machine is abbreviated as Method Using Relevance Vector Machine predicted value, using based on particle
The life prediction result that the method for filtering obtains is abbreviated as particle filter predicted value, and proposition method of the present invention will be used to obtain
Life prediction result is abbreviated as the method for the present invention predicted value.As shown, in life prediction early period, the method for the present invention and two kinds it is right
There is certain deviation in the prediction result and true value of ratio method.But as time goes by, compared to two kinds control methods, this hair
Bright obtained predicting residual useful life result has higher precision.
1 degradation trend prediction time of table
To find out its cause, on the one hand the method for the present invention can make full use of the prior information of degradation trend, data drive is overcome
Dynamic dependence of the method to a large amount of monitoring data, thus prediction error when reducing monitoring data deficiency early period；On the other hand, originally
Inventive method screens optimal degenerated curve by Fu Leixie distance dynamic and determines therefrom that index degradation model parameter, refers to tradition
Exponential model is compared, and without preassigning the prior distribution of parameter, is avoided since parameter Estimation inaccuracy bring predicts error.
The method of the present invention cannot be only used for the predicting residual useful life of rolling bearing, it may also be used for the residue of other electronic products
Life prediction.Implementer can suitably be adjusted the corresponding steps in the method for the present invention according to the degradation characteristics of object to be predicted
It is whole, to can be applied in the predicting residual useful life of different type product.Further it is proposed that it is a kind of based on away from
From assessment technology by data-driven and new approaches based on model prediction method dynamic linkage, before not departing from present inventive concept
It puts, the replacement that index, parameter or model etc. are made is carried out to the present invention and is modified, also should be regarded as protection scope of the present invention.
Claims (1)
1. the rolling bearing life prediction technique to be linked based on Fu Leixie apart from dynamic digital-to-analogue, which is characterized in that including following step
Suddenly：
1) bearing vibration signal is acquired in real time using acceleration transducer, asinh index is extracted from vibration signal
Health indicator as rolling bearing：
Wherein, s (t) is the asinh index value of t moment, xtFor the vibration signal segment that t moment samples, σ () is
Variance formula；
2) in tkThe health indicator obtained is input in the Method Using Relevance Vector Machine with different nuclear parameters by the moment, is executed to return and be divided
It analyses and obtains corresponding associated vector；Specific step is as follows：
2.1) rolling bearing is obtained from initial time t0To current time tkHealth indicator sequence Sk, wherein Sk={ s
(t0),...,s(tk)}；
2.2) kernel function that Method Using Relevance Vector Machine is arranged is gaussian kernel function, i.e.,：
K(tk,ti)=exp (- γ | | ti-tk||2) (2)
K () indicates kernel function, t in formulaiRepresent i-th of monitoring moment, i=0,1 ..., k, tk={ t0,t1,...,tk,
It represents from initial time t0To current time tkMonitoring moment sequence, γ is nuclear parameter；
2.3) it takes several values to carry out relevance vector regression nuclear parameter γ, obtains several associated vectorsIts
InN-th of associated vector is represented in tmThe estimated value of moment health indicator, n=1 ..., N, N are associated vector sum；M=
1 ..., M, M are each associated vector dimension, and further obtain the health indicator estimated value confidence interval that confidence level is 95%
Bound：
In above formula,For the confidence interval upper bound,For confidence interval lower bound, tM=(t1,t2,...,tM)T, represent phase
The time series of vector is closed,And Σn,MPFor the optimal hyper parameter of n-th of associated vector, sequence sparse Bayesian is used
Algorithm is practised to obtain；
3) nonlinear least square fitting is carried out to several associated vectors obtained in step 2) using index degradation model, obtained
Degenerated curve is fitted under to different nuclear parameters：
gn(t)=anexp(bnt)+cnexp(dnt) (5)
Wherein, gn() indicates the corresponding fitting degenerated curve of n-th of associated vector, an, bn, cn, dnIt is the corresponding finger of the curve
Number degradation model parameter；
4) according to Fu Leixie apart from appraisal procedure 3) obtained in be respectively fitted degenerated curve, obtain optimal degenerated curve, it is specific to walk
It is rapid as follows：
4.1) by rolling bearing from initial time t0To current time tkHealth indicator sequence SkIt is smoothed, constitutes flat
Health indicator sequence after cunningWhereinIt is smooth rear tiThe health indicator value at moment, i=
1,...,k；
4.2) it calculates between each associated vector fitting degenerated curve obtained in smoothed out health indicator sequence and step 3)
Fu Leixie distance：
Wherein ξ () and ψ () is to meet ξ (0)=ψ (0)=t1, ξ (1)=ψ (1)=tkAnd domain is any list of [0,1]
Nondecreasing function is adjusted, D () is that Euclidean distance calculates function；
4.3) it finds the Fu Leixie between smoothed out health indicator sequence and remembers fitting song apart from the smallest fitting degenerated curve
Line is optimal degenerated curve, and correspondence model parameter is optimal degradation model parameter；
5) according to optimal degradation model parameter, using formula (5) recursion health indicator, remember that the recursion value of health indicator reaches or for the first time
It is t at the time of more than default failure thresholdend, then the remaining life RUL (t of current time rolling bearingk) be expressed as：
RUL(tk)=tend-tk (7)。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810725813.5A CN108846241A (en) | 2018-07-04 | 2018-07-04 | The rolling bearing life prediction technique to be linked based on Fu Leixie apart from dynamic digital-to-analogue |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810725813.5A CN108846241A (en) | 2018-07-04 | 2018-07-04 | The rolling bearing life prediction technique to be linked based on Fu Leixie apart from dynamic digital-to-analogue |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108846241A true CN108846241A (en) | 2018-11-20 |
Family
ID=64200087
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810725813.5A Pending CN108846241A (en) | 2018-07-04 | 2018-07-04 | The rolling bearing life prediction technique to be linked based on Fu Leixie apart from dynamic digital-to-analogue |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108846241A (en) |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109726524A (en) * | 2019-03-01 | 2019-05-07 | 哈尔滨理工大学 | A kind of rolling bearing remaining life prediction technique based on CNN and LSTM |
| CN110967188A (en) * | 2019-11-08 | 2020-04-07 | 珠海格力电器股份有限公司 | Rolling bearing residual life prediction method and system based on iterative correlation vector machine |
| CN111241633A (en) * | 2020-01-20 | 2020-06-05 | 中国人民解放军国防科技大学 | Chopper residual life prediction method based on principal component analysis and double-exponential model |
| CN111783242A (en) * | 2020-06-17 | 2020-10-16 | 河南科技大学 | RVM-KF-based rolling bearing residual life prediction method and device |
| CN112417701A (en) * | 2020-12-01 | 2021-02-26 | 江苏南高智能装备创新中心有限公司 | Numerical control machine tool residual life prediction method and device based on autoregressive summation moving average model and network side server |
| CN112711832A (en) * | 2020-12-08 | 2021-04-27 | 重庆理工大学 | Method and system for early warning of temperature and fault identification of stator winding of synchronous generator |
| CN112906157A (en) * | 2021-02-20 | 2021-06-04 | 南京航空航天大学 | Method and device for evaluating health state of main shaft bearing and predicting residual life |
| CN113468720A (en) * | 2021-06-02 | 2021-10-01 | 中国人民解放***箭军工程大学 | Service life prediction method for digital-analog linked random degradation equipment |
| CN113487085A (en) * | 2021-07-06 | 2021-10-08 | 新智数字科技有限公司 | Joint learning framework-based equipment service life prediction method and device, computer equipment and computer-readable storage medium |
| CN113609685A (en) * | 2021-08-09 | 2021-11-05 | 电子科技大学 | Bearing residual life prediction method based on optimized RVM and mixed degradation model |
| CN113656910A (en) * | 2021-08-06 | 2021-11-16 | 电子科技大学 | Rolling bearing health index curve construction method based on AFF-AAKR fusion |
| CN113886920A (en) * | 2021-10-08 | 2022-01-04 | 中国矿业大学 | Bridge vibration response data prediction method based on sparse Bayesian learning |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102072829A (en) * | 2010-11-04 | 2011-05-25 | 同济大学 | Iron and steel continuous casting equipment oriented method and device for forecasting faults |
| CN104899608A (en) * | 2015-06-19 | 2015-09-09 | 西安交通大学 | Weight fusion relevant vector machine model for prediction of residual life of antifriction bearing |
| CN106909756A (en) * | 2017-03-29 | 2017-06-30 | 电子科技大学 | A kind of rolling bearing method for predicting residual useful life |
| RU2016137238A (en) * | 2016-09-16 | 2018-03-19 | Федеральное государственное бюджетное военное образовательное учреждение высшего образования "Черноморское высшее военно-морское ордена Красной Звезды училище имени П.С. Нахимова" Министерства обороны Российской Федерации | A method for predicting the residual life of rolling bearings according to their vibration activity |
-
2018
- 2018-07-04 CN CN201810725813.5A patent/CN108846241A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102072829A (en) * | 2010-11-04 | 2011-05-25 | 同济大学 | Iron and steel continuous casting equipment oriented method and device for forecasting faults |
| CN104899608A (en) * | 2015-06-19 | 2015-09-09 | 西安交通大学 | Weight fusion relevant vector machine model for prediction of residual life of antifriction bearing |
| RU2016137238A (en) * | 2016-09-16 | 2018-03-19 | Федеральное государственное бюджетное военное образовательное учреждение высшего образования "Черноморское высшее военно-морское ордена Красной Звезды училище имени П.С. Нахимова" Министерства обороны Российской Федерации | A method for predicting the residual life of rolling bearings according to their vibration activity |
| CN106909756A (en) * | 2017-03-29 | 2017-06-30 | 电子科技大学 | A kind of rolling bearing method for predicting residual useful life |
Non-Patent Citations (4)
| Title |
|---|
| WANG,BIAO,ETC: "An Improved Fusion Prognostics Method for Remaining Useful Life Prediction of Bearings", 《2017 IEEE INTERNATIONAL CONFERENCE ON PROGNOSTICS AND HEALTH MANAGEMENT (ICPHM)》 * |
| YAGUO LEI, ETC: "A Model-Based Method for Remaining Useful Life Prediction of Machinery", 《IEEE TRANSACTIONS ON RELIABILITY》 * |
| 燕晨耀: "基于多特征量的滚动轴承退化状态和剩余寿命预测方法研究", 《全国优秀硕士论文全文数据库 工程科技II辑》 * |
| 雷亚国，等: "基于粒子滤波的滚动轴承寿命预测方法", 《2014年全国机械行业可靠性技术学术交流会论文集》 * |
Cited By (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109726524A (en) * | 2019-03-01 | 2019-05-07 | 哈尔滨理工大学 | A kind of rolling bearing remaining life prediction technique based on CNN and LSTM |
| CN110967188A (en) * | 2019-11-08 | 2020-04-07 | 珠海格力电器股份有限公司 | Rolling bearing residual life prediction method and system based on iterative correlation vector machine |
| CN111241633A (en) * | 2020-01-20 | 2020-06-05 | 中国人民解放军国防科技大学 | Chopper residual life prediction method based on principal component analysis and double-exponential model |
| CN111241633B (en) * | 2020-01-20 | 2024-05-31 | 中国人民解放军国防科技大学 | Chopper residual life prediction method based on principal component analysis and double-index model |
| CN111783242A (en) * | 2020-06-17 | 2020-10-16 | 河南科技大学 | RVM-KF-based rolling bearing residual life prediction method and device |
| CN111783242B (en) * | 2020-06-17 | 2024-05-28 | 河南科技大学 | RVM-KF-based rolling bearing residual life prediction method and device |
| CN112417701A (en) * | 2020-12-01 | 2021-02-26 | 江苏南高智能装备创新中心有限公司 | Numerical control machine tool residual life prediction method and device based on autoregressive summation moving average model and network side server |
| CN112417701B (en) * | 2020-12-01 | 2024-01-23 | 江苏南高智能装备创新中心有限公司 | Method and device for predicting residual life of numerical control machine tool and network side server |
| CN112711832A (en) * | 2020-12-08 | 2021-04-27 | 重庆理工大学 | Method and system for early warning of temperature and fault identification of stator winding of synchronous generator |
| CN112711832B (en) * | 2020-12-08 | 2022-09-30 | 重庆理工大学 | Method and system for early warning of temperature and fault identification of stator winding of synchronous generator |
| CN112906157A (en) * | 2021-02-20 | 2021-06-04 | 南京航空航天大学 | Method and device for evaluating health state of main shaft bearing and predicting residual life |
| CN113468720B (en) * | 2021-06-02 | 2022-08-19 | 中国人民解放***箭军工程大学 | Service life prediction method for digital-analog linked random degradation equipment |
| CN113468720A (en) * | 2021-06-02 | 2021-10-01 | 中国人民解放***箭军工程大学 | Service life prediction method for digital-analog linked random degradation equipment |
| CN113487085A (en) * | 2021-07-06 | 2021-10-08 | 新智数字科技有限公司 | Joint learning framework-based equipment service life prediction method and device, computer equipment and computer-readable storage medium |
| CN113656910A (en) * | 2021-08-06 | 2021-11-16 | 电子科技大学 | Rolling bearing health index curve construction method based on AFF-AAKR fusion |
| CN113656910B (en) * | 2021-08-06 | 2023-04-07 | 电子科技大学 | Rolling bearing health index curve construction method based on AFF-AAKR fusion |
| CN113609685A (en) * | 2021-08-09 | 2021-11-05 | 电子科技大学 | Bearing residual life prediction method based on optimized RVM and mixed degradation model |
| CN113886920A (en) * | 2021-10-08 | 2022-01-04 | 中国矿业大学 | Bridge vibration response data prediction method based on sparse Bayesian learning |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108846241A (en) | The rolling bearing life prediction technique to be linked based on Fu Leixie apart from dynamic digital-to-analogue | |
| Lindemann et al. | Anomaly detection and prediction in discrete manufacturing based on cooperative LSTM networks | |
| CN110781592B (en) | Method and system for predicting residual life of equipment | |
| CN112131212A (en) | Hybrid cloud scene-oriented time sequence data anomaly prediction method based on ensemble learning technology | |
| CN107730127B (en) | Relay storage degradation data prediction method based on output characteristic initial distribution | |
| CN110378504B (en) | Photovoltaic power generation climbing event probability prediction method based on high-dimensional Copula technology | |
| CN111143973B (en) | Valve-regulated lead-acid storage battery degradation trend prediction method based on Gauss process regression | |
| CN112232386B (en) | Voltage sag severity prediction method based on support vector machine | |
| CN111832703B (en) | Irregular sampling dynamic sequence modeling method for process manufacturing industry | |
| CN108959498A (en) | A kind of big data processing platform and its design method for health monitoring | |
| CN110766236A (en) | Power equipment state trend prediction method based on statistical analysis and deep learning | |
| CN111310963A (en) | Power generation data prediction method and device for power station, computer equipment and storage medium | |
| Hu et al. | A hybrid model for predicting the degradation trend of hydropower units based on deep learning | |
| CN116050665A (en) | Heat supply equipment fault prediction method | |
| CN117077327A (en) | Bearing life prediction method and system based on digital twin | |
| CN115329669A (en) | Power system transient stability evaluation method based on deep Bayes active learning | |
| CN113554229A (en) | Three-phase voltage unbalance abnormality detection method and device | |
| CN109960146A (en) | The method for improving soft measuring instrument model prediction accuracy | |
| CN112132324A (en) | Ultrasonic water meter data restoration method based on deep learning model | |
| CN112700050A (en) | Method and system for predicting ultra-short-term 1 st point power of photovoltaic power station | |
| CN116644358A (en) | Power system transient stability evaluation method based on Bayesian convolutional neural network | |
| CN105741184A (en) | Transformer state evaluation method and apparatus | |
| CN116859255A (en) | Method, device, equipment and medium for predicting state of health of energy storage battery | |
| CN116662925A (en) | Industrial process soft measurement method based on weighted sparse neural network | |
| CN114298413A (en) | Hydroelectric generating set runout trend prediction method |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20181120 |
|
| WD01 | Invention patent application deemed withdrawn after publication |