CN105224792B  A kind of rolling bearing performance keeps the Forecasting Methodology of reliability  Google Patents
A kind of rolling bearing performance keeps the Forecasting Methodology of reliability Download PDFInfo
 Publication number
 CN105224792B CN105224792B CN201510603662.2A CN201510603662A CN105224792B CN 105224792 B CN105224792 B CN 105224792B CN 201510603662 A CN201510603662 A CN 201510603662A CN 105224792 B CN105224792 B CN 105224792B
 Authority
 CN
 China
 Prior art keywords
 performance
 mrow
 rolling bearing
 reliability
 msub
 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.)
 Expired  Fee Related
Links
 238000005096 rolling process Methods 0.000 title claims abstract description 256
 238000000034 method Methods 0.000 title claims abstract description 48
 230000014759 maintenance of location Effects 0.000 claims abstract description 28
 230000000875 corresponding Effects 0.000 claims description 9
 238000001514 detection method Methods 0.000 claims description 5
 230000001133 acceleration Effects 0.000 description 20
 230000015556 catabolic process Effects 0.000 description 8
 230000004059 degradation Effects 0.000 description 8
 238000006731 degradation reaction Methods 0.000 description 8
 238000011056 performance test Methods 0.000 description 7
 230000000694 effects Effects 0.000 description 6
 238000005259 measurement Methods 0.000 description 5
 230000000737 periodic Effects 0.000 description 5
 230000035772 mutation Effects 0.000 description 4
 241001269238 Data Species 0.000 description 3
 238000010586 diagram Methods 0.000 description 3
 238000009826 distribution Methods 0.000 description 3
 239000000463 material Substances 0.000 description 3
 230000002378 acidificating Effects 0.000 description 2
 238000005516 engineering process Methods 0.000 description 2
 238000005461 lubrication Methods 0.000 description 2
 241001489523 Coregonus artedi Species 0.000 description 1
 238000004458 analytical method Methods 0.000 description 1
 238000011161 development Methods 0.000 description 1
 235000013399 edible fruits Nutrition 0.000 description 1
 238000011156 evaluation Methods 0.000 description 1
 238000000605 extraction Methods 0.000 description 1
 238000009776 industrial production Methods 0.000 description 1
 239000000843 powder Substances 0.000 description 1
 238000011160 research Methods 0.000 description 1
 230000003319 supportive Effects 0.000 description 1
Abstract
The invention discloses the Forecasting Methodology that a kind of rolling bearing performance keeps reliability, it is included in rolling bearing runnability best period, obtains performance data, builds performance sample rate function；The confidential interval of confidence level and performance stochastic variable is obtained according to Little Probability Event Princiole；According to Poisson counting process, acquisition performance data falls the frequency and rolling bearing performance holding reliability outside performance stochastic variable confidential interval, and then property retention Relative Reliability of the rolling bearing in future time is obtained, predicted roll bearing keeps the failure degree of optimum performance situation in future time accordingly.Any prior information of this method without Properties: Density function, without performance threshold is previously set, the failure degree of future time rolling bearing optimum performance situation can be predicted, prediction accuracy is high, the hidden danger that fails can be found in advance, to take intervening measure in time, avoiding generation severe safety accident from providing decisionmaking.
Description
Technical field
The invention belongs to rolling bearing military service performance failure to assess and performance reliability electric powder prediction, and in particular to one
Kind rolling bearing performance keeps the Forecasting Methodology of reliability.
Background technology
Rolling bearing is that the sliding friction between axle and axle bed by operating is changed into rolling friction, so as to reduce friction loss
A kind of precision mechanical components.Rolling bearing is typically made up of inner ring, outer ring, rolling element and retainer etc., and the effect of inner ring is
Merge with axle matching and rotate together with the axis；The effect of outer ring is engaged with bearing block, is played a supportive role；Rolling element in inner ring and
Rolled between outer ring, bear and transmit load；Retainer can be uniformly distributed rolling element, prevent rolling element from coming off with mutually touching
Hit, the rotation of guiding rolling element and improvement Bearing inner lubrication.
Rolling bearing is one of most important part in machine driven system, is handed in AeroSpace, ship, automobile, track
Logical field suffers from being widely applied.The part that rolling bearing is easily damaged simultaneously and in mechanical system is, it is necessary to periodically tie up
Shield and replacing.The maintenance and replacing of rolling bearing usually require to dismount whole mechanical system, are consumed in disassembly process
Time, manpower and materials cost are typically hundreds and thousands of times of bearing itself cost.However, that safeguards and change may cause not in time
Whole system is failed to work because of bearing, cause bigger economic loss or even jeopardize the life peace of operating personnel
Entirely.Therefore, according to specific operating mode and bearing parameter Accurate Prediction service life, excessive maintenance can be greatly reduced, reduces it
Use cost and maintenance cost, there is highly important effect to industrial production and development in science and technology.
The performance of rolling bearing mainly includes vibration, noise, moment of friction, temperature rise, running accuracy etc., and these performances are to machine
The runnability of tool system has a major impact.The performance failure of rolling bearing refers in running that rolling bearing is because of inside zero
Part insufficient lubrication, friction and wear, damage, bonding, burn into deformation etc. failure and lose its runnability or cisco unity malfunction
Phenomenon.Rolling bearing keeps the operation of optimum performance situation, is the basis that mechanical system realizes the operation of optimum performance situation.According to
Theory of random processes, optimum performance situation reliability of operation is kept to change in future time rolling bearing, this can increase
The big possibility for endangering mechanical system safe and reliable operation.Therefore, research rolling bearing performance keeps reliability to have important
Application value.
In general, the performance failure experiment of rolling bearing refers to：According to some standard, such as national standard or industry mark
Standard, certain amount sample is randomly selected from a collection of rolling bearing, then the sample of extraction is placed under identical experimental enviroment
Reliable life complete trial is carried out, the lifespan of each failure sample is obtained, finally according to standard GB/T246072009 to examination
Test data to be handled, provide the reliability indexs such as form parameter b, L10, characteristics life and the average life span of rolling bearing, according to
This makes reliability evaluation or analysis to this batch of bearing.But due to the raising of bearing quality, in reliability test process
In to accomplish that each sample fails be unrealistic also unnecessary；The rolling bearing that some are expensive, quantity is few is finished
Complete test is unpractical.
Existing performance reliability assessment and Forecasting Methodology, with assume in advance performance density function, performance degradation track with
And performance failure threshold value is known as, according to performance reliability is obtained, having been achieved for certain effect.Existing reliability assessment side
In method,《Aerospace journal》3rd phase in 2006 has delivered the article of entitled " reliability assessment based on Performance Degradation Data ", this article
Chapter assumes the linear function that performance Degradation path is the time, considers performance degradation value Normal Distribution and given threshold value, can be with
Carry out the reliability assessment of aerospace systems performance degradation.In existing rolling bearing performance reliability estimation method,
CN104318043A discloses a kind of bearing vibration performance reliability mutation process detection method, and this method relies on time sequence
The counting process of row, the very small amount raw information for the variation intensity that bear vibration is shown is obtained in Short Interval；By
To the selfservice resampling of variation intensity raw information, a large amount of generation information of variation intensity are simulated；Handled with Grey Prediction Model
Information is generated, obtains variation intensity estimate；Reliability function, realtime estimate bear vibration dependable performance are characterized with Poisson process
The mutation process of property.Time series of this method based on vibration information, the selfservice principle of ash is incorporated into Poisson process, proposes that ash is selfservice
Poisson method, with the mutation process of predicted roll bearing performance reliability.But this method needs prior performance test availability
Can threshold value, the threshold value is obtained by testing, different according to the damage location of selection and sensor, and corresponding threshold value is not yet
Together.Therefore, it is unknown in Properties: Density function prior information and performance threshold is not previously set in existing reliability estimation method
When, the mutation process of rolling bearing performance reliability can not be predicted.
The content of the invention
It is an object of the invention to provide the Forecasting Methodology that a kind of rolling bearing performance keeps reliability, in Properties: Density function
Prior information is unknown and without in the case of performance threshold is previously set, solving the problems, such as rolling bearing performance holding reliability prediction.
In order to realize the above object the technical solution adopted in the present invention is：
A kind of rolling bearing performance keeps the Forecasting Methodology of reliability, comprises the following steps：
1) in rolling bearing runnability best period, rolling bearing performance is measured, obtains performance data；
2) according to principle of maximum entropy, performance data builds performance sample rate function obtained by step 1)；
3) confidence level is obtained according to Little Probability Event Princiole, the confidence area of performance stochastic variable is obtained with quantile method
Between；
4) frequency outside performance stochastic variable confidential interval is fallen according to Poisson counting process, acquisition performance data；
5) according to the nofailure probability of Poisson counting process, fallen by performance data outside performance stochastic variable confidential interval
Frequency obtain rolling bearing performance keep reliability；
6) according to the relative error concept of measure theory, it is relatively reliable in the property retention of future time to obtain rolling bearing
Degree, the failure degree of optimum performance situation is kept in future time according to the property retention Relative Reliability predicted roll bearing.
Performance data obtained by step 1) refers to the performance data that rolling bearing is run in assessment time interval, the assessment
The rolling bearing runningin period that time interval refers to terminate after a time interval, assess end time of time interval for it is current when
Between, t=1；Assess the time interval after time interval, referred to as predicted time section, t>1, each predicted time section is with commenting
Estimating time interval has identical time span, and the end time in predicted time section is the future time described in step 6).
Often increase by 1 predicted time section, future time t adds 1；Time t unit is consistent with the unit for assessing time interval.
The assessment time interval is in rolling bearing runnability best period, refers to the axis of rolling in the time interval
It is optimal to carry row performance situation；In the rolling bearing runnability best period, rolling bearing operation keeps optimum performance gesture
State, refer to almost without the possibility of performance failure；The period is usually located at rolling bearing runningin period and terminates the rear neighbouring time
Section.
Rolling bearing performance mainly includes vibration, noise, moment of friction, temperature rise, running accuracy etc..In step 1), rolling
Dynamic bearing runnability best period, by measuring system periodic measurement rolling bearing performance, and then predicted roll bearing should
Time history of the performance failure degree in future time.
The performance data for building performance sample rate function is K, and kth of performance data is x_{k}, k=1,2 ..., K；K≥
1000；
The performance sample rate function is p (x)：
In formula (1), x is the performance stochastic variable of description rolling bearing performance；M is highest moment of the orign order；I is moment of the orign
Order；λ_{0}, λ_{1}..., λ_{m}For Lagrange multiplier, and there is first Lagrange multiplier λ_{0}For：
In formula (2), x is the performance stochastic variable of description rolling bearing performance；S_{1}For under performance stochastic variable x feasible zones
Dividing value；S_{2}For the upper dividing value of performance stochastic variable x feasible zones；I is moment of the orign order；λ_{i}For ith of Lagrange multiplier, glug
Bright day multiplier λ_{1}, λ_{2}..., λ_{m}Obtained by m equation group of formula (3)：
In formula (3), x is the performance stochastic variable of description rolling bearing performance；x_{k}For kth of performance data, k represents performance
The sequence number of data, K are performance data number；S_{1}For the floor value of performance stochastic variable x feasible zones, S_{2}Can for performance stochastic variable x
The upper dividing value in row domain；I and j is moment of the orign order, and m is highest moment of the orign order, λ_{i}For ith of Lagrange multiplier.
In step 3), confidence level is to the overall intrinsic optimum performance situation probability of happening of rolling bearing performance
Characterize, its acquisition methods is：According to statistical Little Probability Event Princiole, significance can be using value as 0~0.2, such as
0.01st, 0.05,0.1 etc., corresponding confidence level is 1~0.8, such as 0.99,0.95,0.9；The value of confidence level is with small probability thing
Part principle is foundation, and prior result of performance test determines, symbolizes the overall intrinsic optimum performance of rolling bearing performance
The probability of happening of situation.
The confidence level P is using Little Probability Event Princiole as foundation, and prior result of performance test determines, under specific method includes
Row step：
I) select same rolling bearing to make a service test in advance, enter in rolling bearing runnability in best period
Row detection, obtains K >=1000 performance data, K is performance data number；Performance sample rate letter is built with K performance data
Number p (x)；Select reliability estimating value P_{q}Be followed successively by 1,0.999,0.99,0.95,0.9,0.85,0.8 etc. 7 value respectively, with point
Digit method is obtained corresponding to P_{q}Qth of performance stochastic variable confidential interval [X_{Lq}, X_{Uq}], being recorded in K performance data has
How many individual data fall in performance stochastic variable confidential interval [X_{Lq}, X_{Uq}] outside, and thus obtain performance data fall it is random in performance
Variable confidential interval [X_{Lq}, X_{Uq}] outside qth of frequency values λ_{q}, X here_{Lq}With X_{Uq}It is floor value and upper dividing value respectively, sequence number q
=1,2,3 ..., 7；
Ii rolling bearing performance experiment and detection) are continued, W when performance failure, acquisition performance failure >=
1000 performance fail datas, W are performance failure data amount check；Or obtained in rolling bearing runnability in best period
After obtaining K performance data, pause experiment, the rolling bearing is taken out, when the rolling surface of its raceway constructs performance failure
Failure, simulate failure during performance failure, then the rolling bearing of failure detects during to there is simulated performance failure, obtains
W performance fail data during performance failure；
Iii) it is recorded in that how many data in W performance fail data fall in performance stochastic variable confidential interval [X_{Lq},
X_{Uq}] outside, and thus obtain performance fail data and fall in performance stochastic variable confidential interval [X_{Lq}, X_{Uq}] outside qth of frequency
Value β_{q}, X here_{Lq}With X_{Uq}It is floor value and upper dividing value respectively, sequence number q=1,2,3 ..., 7；
Iv formula d) is pressed_{q}={ [exp ( β_{q})exp(λ_{q})]/exp(λ_{q}) × 100% calculate performance failure when rolling
Bearing performance keeps Relative Reliability d_{q}, obtain qth of d_{q}It is worth, here λ_{q}Fall for performance data in performance stochastic variable confidence area
Between [X_{Lq}, X_{Uq}] outside qth of frequency values, β_{q}Fall for performance failure data in performance stochastic variable confidential interval [X_{Lq}, X_{Uq}] it
Qth outer of frequency values, sequence number q=1,2,3 ..., 7；From 7 d_{q}Choose in value be less than and near 10% that, mark
Its subscript q is q*, corresponding reliability estimating value P_{q*}It is exactly the prior result of performance test using Little Probability Event Princiole as foundation
The confidence level P of determination.
In the above method, X in step i)_{Lq}With X_{Uq}Computational methods same X respectively_{L}And X_{U}；Step ii) in, in rolling surface
Fault method when constructing performance failure be with acidic materials on the rolling surface of raceway along the circumferential direction with 120 degree of interval
Angle corrode 3 small and macroscopic spot altogether, simulate failure during performance failure.
In step 3), the confidential interval of performance stochastic variable is [X_{L}, X_{U}], floor value X_{L}Obtained with formula (4)：
Upper dividing value X_{U}Obtained with formula (5)：
In formula (4), formula (5), x is the performance stochastic variable of description rolling bearing performance；S_{1}It is feasible for performance stochastic variable x
The floor value in domain；S_{2}For the upper dividing value of performance stochastic variable x feasible zones；[X_{L}, X_{U}] be performance stochastic variable x confidential interval；p
(x) it is performance sample rate function；P is confidence level.
According to Poisson counting process, it is recorded in that how many data in K performance data fall putting in performance stochastic variable x
Believe section [X_{L}, X_{U}] outside, and thus obtain performance data and fall in performance stochastic variable x confidential intervals [X_{L}, X_{U}] outside frequency
λ。
In step 4), the frequency that performance data falls outside performance stochastic variable confidential interval is λ：
In formula (6), λ is that performance data falls in performance stochastic variable x confidential intervals [X_{L}, X_{U}] outside frequency, n is performance
Data fall in performance stochastic variable x confidential intervals [X_{L}, X_{U}] outside number, K is performance data number.
In the Forecasting Methodology of the present invention, rolling bearing performance keeps reliability, refers in experiment and rolls during one's term of military service
Bearing runs the possibility that can keep optimum performance situation.Property retention reliability shows as a function, the tool of the function
Body value is referred to as property retention reliability.I.e. step 5) is the nofailure probability according to Poisson counting process, is fallen by performance data
Frequency outside performance stochastic variable confidential interval obtains rolling bearing performance and keeps reliability function, and the specific of the function takes
It is worth for property retention reliability.
In step 5), it is R (t) that rolling bearing performance, which keeps reliability,：
R (t)=exp ( λ t)；T >=1 (7),
In formula (7), t is the time, t >=1；Rolling bearing performance keeps reliability when R (t) is time t, for characterizing the time
Rolling bearing operation can keep the possibility of optimum performance situation during t；λ is that performance data falls in performance stochastic variable x confidences
Section [X_{L}, X_{U}] outside frequency.
In step 6), rolling bearing is d (t) in the property retention Relative Reliability of future time：
In formula (8), rolling bearing performance keeps reliability when R (1) is current time t=1；T is future time, t>1；R
(t) rolling bearing performance keeps reliability when being future time t；D (t) is that rolling bearing performance keeps Relative Reliability, is used for
Rolling bearing operation keeps the failure degree of optimum performance situation when characterizing future time t.
In step 6), optimum performance situation is kept in future time according to property retention Relative Reliability predicted roll bearing
The method of failure degree include：According to significance tests principle and measure theory, rolling bearing runnability is classified；
It is classified according to rolling bearing runnability, the time history of predicted roll bearing optimum performance situation failure degree.
The general principle of rolling bearing runnability classification is as follows：
A) according to significance tests principle, if rolling bearing performance keeps Relative Reliability to be not less than 0%, institute is represented
The future time rolling bearing performance of prediction keeps reliability to keep reliability not less than current time rolling bearing performance, then not
Rolling bearing runnability can be refused and reached optimal situation；Otherwise, rolling bearing runnability can be refused to have reached
Optimal situation；
B) when rolling bearing performance keeps Relative Reliability to be less than 0%, according to measure theory, absolute relative error exists
(0%, 5%] between when measured value relative to true value error very little, absolute relative error (5%, 10%] between when survey
Value is becoming big relative to the error of true value, and measured value becomes relative to the error of true value when absolute relative error is more than 10%
Greatly.
Using abovementioned significance tests principle and measure theory as foundation, rolling bearing runnability classification is carried out.
The rolling bearing runnability classification refers to rolling bearing runnability being divided into S1, S2, S3, S4 totally 4 levels
Not：
S1：Rolling bearing performance keeps Relative Reliability d (t) >=0%, i.e. operation of the rolling bearing in future time t
Performance reaches optimal, and optimum performance situation is almost without the possibility of failure；
S2：Rolling bearing performance holding Relative Reliability d (t) ∈ [ 5%, 0%), i.e., rolling bearing is in future time t
Runnability it is normal, optimum performance situation failure possibility it is small；
S3：Rolling bearing performance holding Relative Reliability d (t) ∈ [ 10%, 5%), i.e., rolling bearing is in future time t
When runnability be deteriorated, optimum performance situation failure possibility increase；
S4：Rolling bearing performance keeps Relative Reliability d (t) ＜ 10%, i.e. fortune of the rolling bearing in future time t
Row degradation, the possibility of optimum performance situation failure become big.
4 grades being classified according to abovementioned rolling bearing runnability, the failure of predicted roll bearing optimum performance situation
The time history of degree.It is actually the optimum performance situation relative to current time that rolling bearing performance, which keeps Relative Reliability,
The attenuation degree of reliability is kept in future time rolling bearing performance, negative value represents decay, unattenuated on the occasion of expression.The axis of rolling
It is smaller to hold property retention Relative Reliability d (t), rolling bearing runnability becomes poorer, the possibility of optimum performance situation failure
Property becomes bigger.
Relative Reliability d (t)= 10% future time t is kept corresponding to rolling bearing performance, is rolling bearing performance
The crash time of variation, before the crash time arrives, intervening measure is taken, the rolling bearing is safeguarded or changed,
For avoiding that the severe safety accident brought by the failure of rolling bearing optimum performance situation occurs.
It is in state natural sciences fund that the rolling bearing performance of the present invention, which keeps the Forecasting Methodology of reliability,
(51475144) completed under subsidy.Confidence level that what the Forecasting Methodology included will be known as, confidential interval, property retention are reliable
Degree, and property retention Relative Reliability.Wherein, confidence level is to the overall intrinsic optimum performance of rolling bearing performance
The sign of situation probability of happening；Confidential interval, which can be characterized in experiment and the operation of rolling bearing during one's term of military service, can keep optimal
Performance situation；Property retention reliability represents that rolling bearing operation can keep the possibility of optimum performance situation；Property retention
Relative Reliability is used to characterize the failure degree that the operation of future time rolling bearing keeps optimum performance situation.
The rolling bearing performance of the present invention keeps the Forecasting Methodology of reliability, is when rolling bearing runnability is optimal
Phase, performance data is obtained, build performance sample rate function, obtain the confidential interval of confidence level and performance stochastic variable；According to
Poisson counting process, acquisition performance data fall the frequency and rolling bearing performance holding outside performance stochastic variable confidential interval
Reliability, and then rolling bearing is obtained in the property retention Relative Reliability of future time, predicted roll bearing will be in future accordingly
Time keeps the failure degree of optimum performance situation.Any prior information of this method without Properties: Density function, without thing
Performance threshold is first set, the failure degree of future time rolling bearing optimum performance situation, prediction accuracy can be predicted
Height, the hidden danger that fails can be found in advance, to take intervening measure in time, avoiding generation severe safety accident from providing decisionmaking.This method
Be a kind of Properties: Density function prior information it is unknown and need not be previously set performance threshold rolling bearing experiment with during one's term of military service
Property retention Reliability Prediction Method；Method according to the invention it is possible in the possibility of rolling bearing optimum performance situation failure
Property become big before take intervening measure, rolling bearing is safeguarded or changed, avoids that serious security incident occurs.
Brief description of the drawings
Fig. 1 is Frictional Moment for Rolling Bearings data profile in embodiment 1；
Fig. 2 is Frictional Moment for Rolling Bearings sample rate function curve diagram in embodiment 1；
Fig. 3 is the time history diagram that Frictional Moment for Rolling Bearings keeps Relative Reliability in embodiment 1；
Fig. 4 is bearing vibration acceleration information distribution map in embodiment 2；
Fig. 5 is bearing vibration acceleration samples density function curve figure in embodiment 2；
Fig. 6 is the time history diagram that bearing vibration acceleration keeps Relative Reliability in embodiment 2.
Embodiment
With reference to embodiment, the present invention is further illustrated.
In embodiment, the time of periodic measurement rolling bearing performance is after rolling bearing runningin period terminates
Carried out in one time interval；Rolling bearing runnability situation is optimal in the time interval, and the time interval is the assessment time
Section, the end time for assessing time interval are current time, t=1.Future time during prediction refers to after assessment time interval
Time interval, referred to as predicted time section often increases by 1 predicted time section, and t adds 1.Time t unit is with assessing the time
The unit in section is consistent.For example, periodic measurement rolling bearing performance since 1 day January in 2015, was tied on June 30th, 2015
Beam, to predict on January 1st, 2017 to the possibility of the rolling bearing optimum performance situation failure between June 30.Here, comment
It is 0.5 year (since 1 day January in 2015, to end of day June 30 in 2015) to estimate time interval, current time t=1, future
Time is t=1+4=5, experienced 4 predicted time sections altogether, each predicted time section is 0.5 year, when predicting the 5th
Between section (on January 1st, 2017 is between June 30) when the failure of rolling bearing optimum performance situation possibility.Here, 5
Time interval amounts to 2.5 years, and time t unit is 0.5 year.
Embodiment 1
The rolling bearing performance of the present embodiment 1 keeps the Forecasting Methodology of reliability, comprises the following steps：
1) in rolling bearing runnability best period, the rolling in time interval is assessed by measuring system periodic measurement
The moment of friction of bearing, in current time t=1, obtain K performance data of moment of friction, K=20000, kth of performance number
According to for x_{k}, k=1,2 ..., 20000, data unit Nm；Frictional Moment for Rolling Bearings data profile such as Fig. 1 institutes of acquisition
Show；
In the present embodiment 1, it is 1 year to assess time interval, and each predicted time section is 1 year, and time t unit is
Year；The Frictional Moment for Rolling Bearings performance is predicted using the Forecasting Methodology of the rolling bearing performance holding reliability of the present embodiment
Time history of the failure degree in future time；
2) according to principle of maximum entropy, performance data builds performance sample rate function p (x) obtained by step 1)：
In formula (1), x is the performance stochastic variable of description rolling bearing performance；M is highest moment of the orign order；I is moment of the orign
Order；λ_{0}, λ_{1}..., λ_{m}For Lagrange multiplier, and there is first Lagrange multiplier λ_{0}For：
In formula (2), x is the performance stochastic variable of description rolling bearing performance；S_{1}For under performance stochastic variable x feasible zones
Dividing value；S_{2}For the upper dividing value of performance stochastic variable x feasible zones；I is moment of the orign order；λ_{i}For ith of Lagrange multiplier, glug
Bright day multiplier λ_{1}, λ_{2}..., λ_{m}Obtained by m equation group of formula (3)：
In formula (3), x is the performance stochastic variable of description rolling bearing performance；x_{k}For kth of performance data, k represents performance
The sequence number of data, K are performance data number；S_{1}For the floor value of performance stochastic variable x feasible zones, S_{2}Can for performance stochastic variable x
The upper dividing value in row domain；I and j is moment of the orign order, and m is highest moment of the orign order, λ_{i}For ith of Lagrange multiplier；
With 20000 moment of friction performance datas structure moment of friction sample rate function p (x) obtained by step 1), knot
Fruit is as shown in Figure 2；
3) confidence level is obtained according to Little Probability Event Princiole, the confidence area of performance stochastic variable is obtained with quantile method
Between；
According to Little Probability Event Princiole, obtain confidence level P=0.99 by testing in advance, can calculate moment of friction with
Machine variable x confidential intervals [X_{L}, X_{U}]=[233.329Nm, 251.897Nm]：
Floor value X_{L}Obtained with formula (4)：
Upper dividing value X_{U}Obtained with formula (5)：
In formula (4), formula (5), x is the performance stochastic variable of description rolling bearing performance；S_{1}It is feasible for performance stochastic variable x
The floor value in domain；S_{2}For the upper dividing value of performance stochastic variable x feasible zones；[X_{L}, X_{U}] be performance stochastic variable x confidential interval；p
(x) it is performance sample rate function；P is confidence level；
Wherein, the confidence level P is using Little Probability Event Princiole as foundation, and prior result of performance test determines, specific method
For：
I) select same rolling bearing to make a service test in advance, enter in rolling bearing runnability in best period
Row detection, obtains K >=1000 performance data, K is performance data number；Performance sample rate letter is built with K performance data
Number p (x)；Select reliability estimating value P_{q}Be followed successively by 1,0.999,0.99,0.95,0.9,0.85,0.8 etc. 7 value respectively, with point
Digit method is obtained corresponding to P_{q}Qth of performance stochastic variable confidential interval [X_{Lq}, X_{Uq}](X_{Lq}With X_{Uq}Computational methods difference
Same X_{L}And X_{U}), it is recorded in that how many data in K performance data fall in performance stochastic variable confidential interval [X_{Lq}, X_{Uq}] outside,
And thus obtain performance data and fall in performance stochastic variable confidential interval [X_{Lq}, X_{Uq}] outside qth of frequency values λ_{q}, X here_{Lq}
With X_{Uq}It is floor value and upper dividing value respectively, sequence number q=1,2,3 ..., 7；
Ii) after rolling bearing runnability is in best period K performance data of acquisition, pause experiment, taking out should
Rolling bearing, the failure when the rolling surface of its raceway constructs performance failure, i.e., with acidic materials raceway rolling table
Along the circumferential direction corroded altogether with the angle at 120 degree of interval on face and 3 small and macroscopic spot, simulate performance failure
When failure, then the rolling bearing of failure detects during to there is simulated performance failure, obtains W performance mistake during performance failure
Imitate data；
Iii) it is recorded in that how many data in W performance fail data fall in performance stochastic variable confidential interval [X_{Lq},
X_{Uq}] outside, and thus obtain performance fail data and fall in performance stochastic variable confidential interval [X_{Lq}, X_{Uq}] outside qth of frequency
Value β_{q}, X here_{Lq}With X_{Uq}It is floor value and upper dividing value respectively, sequence number q=1,2,3 ..., 7；
Iv formula d) is pressed_{q}={ [exp ( β_{q})exp(λ_{q})]/exp(λ_{q}) × 100% calculate performance failure when rolling
Bearing performance keeps Relative Reliability d_{q}, obtain qth of d_{q}It is worth, here λ_{q}Fall for performance data in performance stochastic variable confidence area
Between [X_{Lq}, X_{Uq}] outside qth of frequency values, β_{q}Fall for performance failure data in performance stochastic variable confidential interval [X_{Lq}, X_{Uq}] it
Qth outer of frequency values, sequence number q=1,2,3 ..., 7；From 7 d_{q}Choose in value be less than and near 10% that, mark
Its subscript q is q*, corresponding reliability estimating value P_{q*}It is exactly the prior result of performance test using Little Probability Event Princiole as foundation
The confidence level P of determination；Confidence level P=0.99 obtained by the present embodiment；
4) frequency outside performance stochastic variable confidential interval is fallen according to Poisson counting process, acquisition performance data；
It is recorded in that how many data in 20000 data fall in moment of friction stochastic variable x confidential intervals [X_{L}, X_{U}] it
Outside, moment of friction data are obtained to fall in moment of friction stochastic variable x confidential intervals [X_{L}, X_{U}] outside frequency lambda：
In formula (6), λ is that performance data falls in performance stochastic variable x confidential intervals [X_{L}, X_{U}] outside frequency, n is performance
Data fall in performance stochastic variable x confidential intervals [X_{L}, X_{U}] outside number, K is performance data number；
5) according to the nofailure probability of Poisson counting process, fallen by performance data outside performance stochastic variable confidential interval
Frequency obtain rolling bearing performance keep reliability, obtain Frictional Moment for Rolling Bearings property retention reliability R (t) time
Course；
It is R (t) that rolling bearing performance, which keeps reliability,：
R (t)=exp ( λ t)；T >=1 (7),
In formula (7), t is the time, t >=1；Rolling bearing performance keeps reliability when R (t) is time t, for characterizing the time
Rolling bearing operation can keep the possibility of optimum performance situation during t；λ is that performance data falls in performance stochastic variable x confidences
Section [X_{L}, X_{U}] outside frequency；
6) according to the relative error concept of measure theory, it is relatively reliable in the property retention of future time to obtain rolling bearing
Degree, obtain the time history that Frictional Moment for Rolling Bearings keeps Relative Reliability d (t)；
Rolling bearing is d (t) in the property retention Relative Reliability of future time：
In formula (8), rolling bearing performance keeps reliability when R (1) is current time t=1；T is future time, t>1；R
(t) rolling bearing performance keeps reliability when being future time t；D (t) is that rolling bearing performance keeps Relative Reliability, is used for
Rolling bearing operation keeps the failure degree of optimum performance situation when characterizing future time t；
Frictional Moment for Rolling Bearings keeps Relative Reliability d (t) time history as shown in Figure 3；
7) according to significance tests principle and measure theory, rolling bearing runnability is divided into S1, S2, S3, S4
Totally 4 ranks：
S1：Rolling bearing performance keeps Relative Reliability d (t) >=0%, i.e. operation of the rolling bearing in future time t
Performance reaches optimal, and optimum performance situation is almost without the possibility of failure；
S2：Rolling bearing performance holding Relative Reliability d (t) ∈ [ 5%, 0%), i.e., rolling bearing is in future time t
Runnability it is normal, optimum performance situation failure possibility it is small；
S3：Rolling bearing performance holding Relative Reliability d (t) ∈ [ 10%, 5%), i.e., rolling bearing is in future time t
When runnability be deteriorated, optimum performance situation failure possibility increase；
S4：Rolling bearing performance keeps Relative Reliability d (t) ＜ 10%, i.e. fortune of the rolling bearing in future time t
Row degradation, the possibility of optimum performance situation failure become big；
8) 4 grades being classified according to abovementioned rolling bearing runnability, predicted roll bearing optimum performance situation are lost
The time history of effect degree is as follows：
In figure 3, as t=6, and the ∈ of d (t)= 4.43% [ 5%, 0%), d (t) values are close to 5%；
As t=7, and the ∈ of d (t)= 5.29% [ 10%, 5%), d (t) is already less than 5%；
As t=12, and the ∈ of d (t)= 9.47% [ 10%, 5%), d (t) values are close to 10%；
As t=13, the ＜ 10% of d (t)= 10.29%, d (t) values are already less than 10%.
The failure degree of optimum performance situation is kept in future time according to the above predicted roll bearing：
Accordingly it is expected that to before the 6th year, the runnability of the rolling bearing is normal, moment of friction optimum performance gesture
The possibility of state failure is small；To before the 12nd year after the 7th year, the runnability of the rolling bearing is deteriorated, frictional force
The possibility of square optimum performance situation failure increases；Until the 13rd year, the runnability of the rolling bearing was deteriorated, frictional force
The possibility of square optimum performance situation failure is big.
According to abovementioned time history, between the 12nd year and the 13rd year, intervening measure should be taken, the rolling bearing is carried out
Safeguard or change, avoid that the severe safety accident brought by the failure of bearing frictional torque optimum performance situation occurs.
Embodiment 2
The rolling bearing performance of the present embodiment 2 keeps the Forecasting Methodology of reliability, comprises the following steps：
1) in rolling bearing runnability best period, the rolling in time interval is assessed by measuring system periodic measurement
The vibration acceleration of bearing, in current time t=1, obtain K performance data of vibration acceleration, K=20000, kth of property
Energy data are x_{k}, k=1,2 ..., 20000, data unit for μm/s^{2}；The bearing vibration acceleration information distribution map of acquisition
As shown in Figure 4；
In the present embodiment 2, it is 1 year to assess time interval, and each predicted time section is 1 year, and time t unit is
Year；The bearing vibration acceleration is predicted using the Forecasting Methodology of the rolling bearing performance holding reliability of the present embodiment
Time history of the energy failure degree in future time；
2) according to principle of maximum entropy, performance data builds performance sample rate function p (x) obtained by step 1)：
In formula (1), x is the performance stochastic variable of description rolling bearing performance；M is highest moment of the orign order；I is moment of the orign
Order；λ_{0}, λ_{1}..., λ_{m}For Lagrange multiplier, and there is first Lagrange multiplier λ_{0}For：
In formula (2), x is the performance stochastic variable of description rolling bearing performance；S_{1}For under performance stochastic variable x feasible zones
Dividing value；S_{2}For the upper dividing value of performance stochastic variable x feasible zones；I is moment of the orign order；λ_{i}For ith of Lagrange multiplier, glug
Bright day multiplier λ_{1}, λ_{2}..., λ_{m}Obtained by m equation group of formula (3)：
In formula (3), x is the performance stochastic variable of description rolling bearing performance；x_{k}For kth of performance data, k represents performance
The sequence number of data, K are performance data number；S_{1}For the floor value of performance stochastic variable x feasible zones, S_{2}Can for performance stochastic variable x
The upper dividing value in row domain；I and j is moment of the orign order, and m is highest moment of the orign order, λ_{i}For ith of Lagrange multiplier；
With 20000 vibration acceleration performance datas structure vibration acceleration sample rate function p obtained by step 1)
(x), as a result as shown in Figure 5；
3) confidence level is obtained according to Little Probability Event Princiole, the confidence area of performance stochastic variable is obtained with quantile method
Between；
According to Little Probability Event Princiole, confidence level P=0.99 is obtained by testing in advance, vibration acceleration can be calculated
Stochastic variable x confidential intervals [X_{L}, X_{U} 0.0549 μm of]=[/s^{2}, 0.0692 μm/s^{2}]：
Floor value X_{L}Obtained with formula (4)：
Upper dividing value X_{U}Obtained with formula (5)：
In formula (4), formula (5), x is the performance stochastic variable of description rolling bearing performance；S_{1}It is feasible for performance stochastic variable x
The floor value in domain；S_{2}For the upper dividing value of performance stochastic variable x feasible zones；[X_{L}, X_{U}] be performance stochastic variable x confidential interval；p
(x) it is performance sample rate function；P is confidence level；
Wherein, the confidence level P is using Little Probability Event Princiole as foundation, and prior result of performance test determines, specific method
With embodiment 1；Confidence level P=0.99 obtained by the present embodiment；
4) frequency outside performance stochastic variable confidential interval is fallen according to Poisson counting process, acquisition performance data；
It is recorded in that how many data in 20000 data fall in vibration acceleration stochastic variable x confidential intervals [X_{L}, X_{U}]
Outside, obtain vibration acceleration data and fall in vibration acceleration stochastic variable x confidential intervals [X_{L}, X_{U}] outside frequency lambda：
In formula (6), λ is that performance data falls in performance stochastic variable x confidential intervals [X_{L}, X_{U}] outside frequency, n is performance
Data fall in performance stochastic variable x confidential intervals [X_{L}, X_{U}] outside number, K is performance data number；
5) according to the nofailure probability of Poisson counting process, fallen by performance data outside performance stochastic variable confidential interval
Frequency obtain rolling bearing performance keep reliability, obtain bearing vibration acceleration property retention reliability R (t) when
Between course；
It is R (t) that rolling bearing performance, which keeps reliability,：
R (t)=exp ( λ t)；T >=1 (7),
In formula (7), t is the time, t >=1；Rolling bearing performance keeps reliability when R (t) is time t, for characterizing the time
Rolling bearing operation can keep the possibility of optimum performance situation during t；λ is that performance data falls in performance stochastic variable x confidences
Section [X_{L}, X_{U}] outside frequency；
6) according to the relative error concept of measure theory, it is relatively reliable in the property retention of future time to obtain rolling bearing
Degree, obtain the time history that bearing vibration acceleration keeps Relative Reliability d (t)；
Rolling bearing is d (t) in the property retention Relative Reliability of future time：
In formula (8), rolling bearing performance keeps reliability when R (1) is current time t=1；T is future time, t>1；R
(t) rolling bearing performance keeps reliability when being future time t；D (t) is that rolling bearing performance keeps Relative Reliability, is used for
Rolling bearing operation keeps the failure degree of optimum performance situation when characterizing future time t；
Bearing vibration acceleration keeps Relative Reliability d (t) time history as shown in Figure 6；
7) according to significance tests principle and measure theory, rolling bearing runnability is divided into S1, S2, S3, S4
Totally 4 ranks：
S1：Rolling bearing performance keeps Relative Reliability d (t) >=0%, i.e. operation of the rolling bearing in future time t
Performance reaches optimal, and optimum performance situation is almost without the possibility of failure；
S2：Rolling bearing performance holding Relative Reliability d (t) ∈ [ 5%, 0%), i.e., rolling bearing is in future time t
Runnability it is normal, optimum performance situation failure possibility it is small；
S3：Rolling bearing performance holding Relative Reliability d (t) ∈ [ 10%, 5%), i.e., rolling bearing is in future time t
When runnability be deteriorated, optimum performance situation failure possibility increase；
S4：Rolling bearing performance keeps Relative Reliability d (t) ＜ 10%, i.e. fortune of the rolling bearing in future time t
Row degradation, the possibility of optimum performance situation failure become big；
8) 4 grades being classified according to abovementioned rolling bearing runnability, predicted roll bearing optimum performance situation are lost
The time history of effect degree is as follows：
In figure 6, as t=6, and the ∈ of d (t)= 4.66% [ 5%, 0%), d (t) values are very close to 5%；
As t=7, and the ∈ of d (t)= 5.57% [ 10%, 5%), d (t) is already less than 5%；
As t=12, and the ∈ of d (t)= 9.97% [ 10%, 5%), d (t) values are very close to 10%；
As t=13, the ＜ 10% of d (t)= 10.83%, d (t) values are already less than 10%.
The failure degree of optimum performance situation is kept in future time according to the above predicted roll bearing：
Accordingly it is expected that to before the 6th year, the runnability of the rolling bearing is normal, vibration acceleration optimum performance
The possibility of situation failure is small；To before the 12nd year after the 7th year, the runnability of the rolling bearing is deteriorated, vibration
The possibility of acceleration optimum performance situation failure increases；Until the 13rd year, the runnability of the rolling bearing was deteriorated, and shakes
The possibility of dynamic acceleration optimum performance situation failure is big.
According to abovementioned time history, between the 12nd year and the 13rd year, intervening measure should be taken, the rolling bearing is carried out
Safeguard or change, avoid that the severe safety accident brought by the failure of bear vibration acceleration optimum performance situation occurs.
The performance of rolling bearing has vibration, noise, moment of friction, temperature rise, running accuracy etc., and all performances can adopt
Property retention reliability prediction is carried out with technical scheme.In other embodiments of the invention, using the present invention's
Rolling bearing performance keeps the Forecasting Methodology of reliability, relatively reliable according to property retentions such as noise, temperature rise, running accuracies respectively
Degree, predicted roll bearing keep the failure degree of optimum performance situation in future time；Concrete operation method is the same as embodiment 1.
When the prediction result according to technical solution of the present invention carries out intervening measure, if the dissimilarity to same rolling bearing
The crash time of the prediction result of energy is different, then should take intervening measure before the most short crash time.
Claims (10)
1. a kind of rolling bearing performance keeps the Forecasting Methodology of reliability, it is characterised in that：Comprise the following steps：
1) in rolling bearing runnability best period, rolling bearing performance is measured, obtains performance data；
2) according to principle of maximum entropy, performance data builds performance sample rate function obtained by step 1)；
3) confidence level is obtained according to Little Probability Event Princiole, the confidential interval of performance stochastic variable is obtained with quantile method；
4) frequency outside performance stochastic variable confidential interval is fallen according to Poisson counting process, acquisition performance data；
5) according to the nofailure probability of Poisson counting process, the frequency outside performance stochastic variable confidential interval is fallen by performance data
Rate obtains rolling bearing performance and keeps reliability；
6) according to the relative error concept of measure theory, property retention Relative Reliability of the rolling bearing in future time is obtained,
The failure degree of optimum performance situation is kept in future time according to the property retention Relative Reliability predicted roll bearing；
The determination of confidence level comprises the following steps in step 3)：
I) same rolling bearing is selected to make a service test in advance, being in best period in rolling bearing runnability is examined
Survey, obtain K >=1000 performance data, K is performance data number；Performance sample rate function p is built with K performance data
(x)；Select reliability estimating value P_{q}1,0.999,0.99,0.95,0.9,0.85,0.8 totally 7 values are followed successively by respectively, with dividing position
Counting method is obtained corresponding to P_{q}Qth of performance stochastic variable confidential interval [X_{Lq}, X_{Uq}], be recorded in K performance data have it is more
A few data fall in performance stochastic variable confidential interval [X_{Lq}, X_{Uq}] outside, and thus obtain performance data and fall and become at random in performance
Measure confidential interval [X_{Lq}, X_{Uq}] outside qth of frequency values λ_{q}, X here_{Lq}With X_{Uq}It is floor value and upper dividing value respectively, sequence number q=
1,2,3,…,7；
Ii) continue rolling bearing performance experiment and detection, until performance failure, obtain W >=1000 during performance failure
Individual performance fail data, W are performance failure data amount check；Or it is in best period in rolling bearing runnability and obtains K
After performance data, pause experiment, the rolling bearing is taken out, the event when the rolling surface of its raceway constructs performance failure
Barrier, simulates failure during performance failure, then the rolling bearing of failure detects during to there is simulated performance failure, obtains performance
W performance fail data during failure；
Iii) it is recorded in that how many data in W performance fail data fall in performance stochastic variable confidential interval [X_{Lq}, X_{Uq}] it
Outside, performance fail data and is thus obtained to fall in performance stochastic variable confidential interval [X_{Lq}, X_{Uq}] outside qth of frequency values β_{q},
Here X_{Lq}With X_{Uq}It is floor value and upper dividing value respectively, sequence number q=1,2,3 ..., 7；
Iv formula d) is pressed_{q}={ [exp ( β_{q})exp(λ_{q})]/exp(λ_{q}) × 100% calculate performance failure when rolling bearing
Property retention Relative Reliability d_{q}, obtain qth of d_{q}It is worth, here λ_{q}Fall for performance data in performance stochastic variable confidential interval
[X_{Lq}, X_{Uq}] outside qth of frequency values, β_{q}Fall for performance failure data in performance stochastic variable confidential interval [X_{Lq}, X_{Uq}] outside
Qth of frequency values, sequence number q=1,2,3 ..., 7；From 7 d_{q}Choose in value be less than and near 10% that, mark it
Subscript q is q*, corresponding reliability estimating value P_{q*}The confidence level being just to determine.
2. rolling bearing performance according to claim 1 keeps the Forecasting Methodology of reliability, it is characterised in that：Step 1) institute
Obtain performance data and refer to the performance data that rolling bearing is run in assessment time interval, the assessment time interval refers to roll
The bearing running phase terminate after a time interval, the end time for assessing time interval is current time；Assess time interval
Time interval afterwards, referred to as predicted time section, each predicted time section have the identical time with assessing time interval
Span, the end time in predicted time section are the future time described in step 6).
3. rolling bearing performance according to claim 1 or 2 keeps the Forecasting Methodology of reliability, it is characterised in that：Structure
The performance data of performance sample rate function is K, and kth of performance data is x_{k}, k=1,2 ..., K；K≥1000；
The performance sample rate function is p (x)：
<mrow>
<mi>p</mi>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mi>exp</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>&lambda;</mi>
<mn>0</mn>
</msub>
<mo>+</mo>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>m</mi>
</munderover>
<msub>
<mi>&lambda;</mi>
<mi>i</mi>
</msub>
<msup>
<mi>x</mi>
<mi>i</mi>
</msup>
<mo>)</mo>
</mrow>
<mo></mo>
<mo></mo>
<mo></mo>
<mrow>
<mo>(</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
<mo>,</mo>
</mrow>
In formula (1), x is the performance stochastic variable of description rolling bearing performance；M is highest moment of the orign order；I is moment of the orign rank
It is secondary；λ_{0}, λ_{1}..., λ_{m}For Lagrange multiplier, and there is first Lagrange multiplier λ_{0}For：
<mrow>
<msub>
<mi>&lambda;</mi>
<mn>0</mn>
</msub>
<mo>=</mo>
<mo></mo>
<mi>l</mi>
<mi>n</mi>
<mrow>
<mo>(</mo>
<msubsup>
<mo>&Integral;</mo>
<msub>
<mi>S</mi>
<mn>1</mn>
</msub>
<msub>
<mi>S</mi>
<mn>2</mn>
</msub>
</msubsup>
<mrow>
<mi>exp</mi>
<mrow>
<mo>(</mo>
<mrow>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>m</mi>
</munderover>
<msub>
<mi>&lambda;</mi>
<mi>i</mi>
</msub>
<msup>
<mi>x</mi>
<mi>i</mi>
</msup>
</mrow>
<mo>)</mo>
</mrow>
<mi>d</mi>
<mi>x</mi>
</mrow>
<mo>)</mo>
</mrow>
<mo></mo>
<mo></mo>
<mo></mo>
<mrow>
<mo>(</mo>
<mn>2</mn>
<mo>)</mo>
</mrow>
<mo>,</mo>
</mrow>
In formula (2), x is the performance stochastic variable of description rolling bearing performance；S_{1}For the floor value of performance stochastic variable x feasible zones；
S_{2}For the upper dividing value of performance stochastic variable x feasible zones；I is moment of the orign order；λ_{i}For ith of Lagrange multiplier, Lagrange multiplies
Sub λ_{1}, λ_{2}..., λ_{m}Obtained by m equation group of formula (3)：
<mrow>
<mn>1</mn>
<mo></mo>
<mfrac>
<mrow>
<msubsup>
<mo>&Integral;</mo>
<msub>
<mi>S</mi>
<mn>1</mn>
</msub>
<msub>
<mi>S</mi>
<mn>2</mn>
</msub>
</msubsup>
<msup>
<mi>x</mi>
<mi>j</mi>
</msup>
<mi>exp</mi>
<mrow>
<mo>(</mo>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>m</mi>
</munderover>
<msub>
<mi>&lambda;</mi>
<mi>i</mi>
</msub>
<msup>
<mi>x</mi>
<mi>i</mi>
</msup>
<mo>)</mo>
</mrow>
<mi>d</mi>
<mi>x</mi>
</mrow>
<mrow>
<mfrac>
<mn>1</mn>
<mi>K</mi>
</mfrac>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>k</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>K</mi>
</munderover>
<msubsup>
<mi>x</mi>
<mi>k</mi>
<mi>j</mi>
</msubsup>
<msubsup>
<mo>&Integral;</mo>
<msub>
<mi>S</mi>
<mn>1</mn>
</msub>
<msub>
<mi>S</mi>
<mn>2</mn>
</msub>
</msubsup>
<mi>exp</mi>
<mrow>
<mo>(</mo>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>m</mi>
</munderover>
<msub>
<mi>&lambda;</mi>
<mi>i</mi>
</msub>
<msup>
<mi>x</mi>
<mi>i</mi>
</msup>
<mo>)</mo>
</mrow>
<mi>d</mi>
<mi>x</mi>
</mrow>
</mfrac>
<mo>=</mo>
<mn>0</mn>
<mo>;</mo>
<mi>j</mi>
<mo>=</mo>
<mn>1</mn>
<mo>,</mo>
<mn>2</mn>
<mo>,</mo>
<mo>...</mo>
<mo>,</mo>
<mi>m</mi>
<mo></mo>
<mo></mo>
<mo></mo>
<mrow>
<mo>(</mo>
<mn>3</mn>
<mo>)</mo>
</mrow>
<mo>,</mo>
</mrow>
In formula (3), x is the performance stochastic variable of description rolling bearing performance；x_{k}For kth of performance data, k represents performance data
Sequence number, K is performance data number；S_{1}For the floor value of performance stochastic variable x feasible zones, S_{2}For performance stochastic variable x feasible zones
Upper dividing value；I and j is moment of the orign order, and m is highest moment of the orign order, λ_{i}For ith of Lagrange multiplier.
4. rolling bearing performance according to claim 1 keeps the Forecasting Methodology of reliability, it is characterised in that：Step 3)
In, the confidential interval of performance stochastic variable is [X_{L}, X_{U}], floor value X_{L}Obtained with formula (4)：
<mrow>
<mfrac>
<mn>1</mn>
<mn>2</mn>
</mfrac>
<mrow>
<mo>(</mo>
<mn>1</mn>
<mo></mo>
<mi>P</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<msubsup>
<mo>&Integral;</mo>
<msub>
<mi>S</mi>
<mn>1</mn>
</msub>
<msub>
<mi>X</mi>
<mi>L</mi>
</msub>
</msubsup>
<mi>p</mi>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>)</mo>
</mrow>
<mi>d</mi>
<mi>x</mi>
<mo></mo>
<mo></mo>
<mo></mo>
<mrow>
<mo>(</mo>
<mn>4</mn>
<mo>)</mo>
</mrow>
<mo>,</mo>
</mrow>
Upper dividing value X_{U}Obtained with formula (5)：
<mrow>
<mfrac>
<mn>1</mn>
<mn>2</mn>
</mfrac>
<mrow>
<mo>(</mo>
<mn>1</mn>
<mo></mo>
<mi>P</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<msubsup>
<mo>&Integral;</mo>
<msub>
<mi>X</mi>
<mi>U</mi>
</msub>
<msub>
<mi>S</mi>
<mn>2</mn>
</msub>
</msubsup>
<mi>p</mi>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>)</mo>
</mrow>
<mi>d</mi>
<mi>x</mi>
<mo></mo>
<mo></mo>
<mo></mo>
<mrow>
<mo>(</mo>
<mn>5</mn>
<mo>)</mo>
</mrow>
<mo>,</mo>
</mrow>
In formula (4), formula (5), x is the performance stochastic variable of description rolling bearing performance；S_{1}For performance stochastic variable x feasible zones
Floor value；S_{2}For the upper dividing value of performance stochastic variable x feasible zones；[X_{L}, X_{U}] be performance stochastic variable x confidential interval；P (x) is
Performance sample rate function；P is confidence level.
5. the rolling bearing performance according to claim 1 or 4 keeps the Forecasting Methodology of reliability, it is characterised in that：Step
4) in, the frequency that performance data falls outside performance stochastic variable confidential interval is λ：
<mrow>
<mi>&lambda;</mi>
<mo>=</mo>
<mfrac>
<mi>n</mi>
<mi>K</mi>
</mfrac>
<mo></mo>
<mo></mo>
<mo></mo>
<mrow>
<mo>(</mo>
<mn>6</mn>
<mo>)</mo>
</mrow>
<mo>,</mo>
</mrow>
In formula (6), λ is that performance data falls in performance stochastic variable x confidential intervals [X_{L}, X_{U}] outside frequency, n is performance data
Fall in performance stochastic variable x confidential intervals [X_{L}, X_{U}] outside number, K is performance data number.
6. rolling bearing performance according to claim 5 keeps the Forecasting Methodology of reliability, it is characterised in that：Step 5)
In, it is R (t) that rolling bearing performance, which keeps reliability,：
R (t)=exp ( λ t)；T >=1 (7),
In formula (7), t is the time, t >=1；Rolling bearing performance keeps reliability when R (t) is time t, during for characterizing time t
Rolling bearing runs the possibility that can keep optimum performance situation；λ is that performance data falls in performance stochastic variable x confidential intervals
[X_{L}, X_{U}] outside frequency.
7. rolling bearing performance according to claim 6 keeps the Forecasting Methodology of reliability, it is characterised in that：Step 6)
In, rolling bearing is d (t) in the property retention Relative Reliability of future time：
<mrow>
<mi>d</mi>
<mrow>
<mo>(</mo>
<mi>t</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mfrac>
<mrow>
<mi>R</mi>
<mrow>
<mo>(</mo>
<mi>t</mi>
<mo>)</mo>
</mrow>
<mo></mo>
<mi>R</mi>
<mrow>
<mo>(</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
</mrow>
<mrow>
<mi>R</mi>
<mrow>
<mo>(</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
</mrow>
</mfrac>
<mo>&times;</mo>
<mn>100</mn>
<mi>%</mi>
<mo>;</mo>
<mi>t</mi>
<mo>></mo>
<mn>1</mn>
<mo></mo>
<mo></mo>
<mo></mo>
<mrow>
<mo>(</mo>
<mn>8</mn>
<mo>)</mo>
</mrow>
<mo>,</mo>
</mrow>
In formula (8), rolling bearing performance keeps reliability when R (1) is current time t=1；T is future time, t>1；R (t) is
Rolling bearing performance keeps reliability during future time t；D (t) is that rolling bearing performance keeps Relative Reliability, for characterizing not
Carry out the failure degree that rolling bearing operation during time t keeps optimum performance situation.
8. rolling bearing performance according to claim 7 keeps the Forecasting Methodology of reliability, it is characterised in that：Step 6)
In, according to property retention Relative Reliability predicted roll bearing in the side of the failure degree of future time holding optimum performance situation
Method includes：According to significance tests principle and measure theory, rolling bearing runnability is classified；Carried according to the axis of rolling
Row grading performance, the time history of predicted roll bearing optimum performance situation failure degree.
9. rolling bearing performance according to claim 8 keeps the Forecasting Methodology of reliability, it is characterised in that：The rolling
The classification of bearing runnability refers to rolling bearing runnability being divided into S1, S2, S3, S4 totally 4 ranks：
S1：Rolling bearing performance keeps Relative Reliability d (t) >=0%, i.e. runnability of the rolling bearing in future time t
Reach optimal, optimum performance situation is almost without the possibility of failure；
S2：Rolling bearing performance keep Relative Reliability d (t) ∈ [ 5%, 0%), i.e. fortune of the rolling bearing in future time t
Row performance is normal, and the possibility of optimum performance situation failure is small；
S3：Rolling bearing performance keep Relative Reliability d (t) ∈ [ 10%, 5%), i.e. rolling bearing is in future time t
Runnability is deteriorated, and the possibility of optimum performance situation failure increases；
S4：Rolling bearing performance keeps Relative Reliability d (t) ＜ 10%, i.e. maneuverability of the rolling bearing in future time t
It can be deteriorated, the possibility of optimum performance situation failure becomes big.
10. rolling bearing performance according to claim 9 keeps the Forecasting Methodology of reliability, it is characterised in that：Correspond to
Rolling bearing performance keep Relative Reliability d (t)= 10% future time t, be rolling bearing performance be deteriorated it is critical when
Between, before the crash time arrives, intervening measure is taken, for avoiding occurring because of rolling bearing optimum performance situation failure band
The severe safety accident come.
Priority Applications (1)
Application Number  Priority Date  Filing Date  Title 

CN201510603662.2A CN105224792B (en)  20150921  20150921  A kind of rolling bearing performance keeps the Forecasting Methodology of reliability 
Applications Claiming Priority (1)
Application Number  Priority Date  Filing Date  Title 

CN201510603662.2A CN105224792B (en)  20150921  20150921  A kind of rolling bearing performance keeps the Forecasting Methodology of reliability 
Publications (2)
Publication Number  Publication Date 

CN105224792A CN105224792A (en)  20160106 
CN105224792B true CN105224792B (en)  20180406 
Family
ID=54993758
Family Applications (1)
Application Number  Title  Priority Date  Filing Date 

CN201510603662.2A Expired  Fee Related CN105224792B (en)  20150921  20150921  A kind of rolling bearing performance keeps the Forecasting Methodology of reliability 
Country Status (1)
Country  Link 

CN (1)  CN105224792B (en) 
Families Citing this family (2)
Publication number  Priority date  Publication date  Assignee  Title 

CN106644481B (en) *  20161227  20181030  哈尔滨理工大学  Rolling bearing Predicting Reliability method based on mathematical morphology and IFOASVR 
CN109855875B (en) *  20190115  20201222  沈阳化工大学  Rolling bearing operation reliability prediction method 
Citations (5)
Publication number  Priority date  Publication date  Assignee  Title 

CN102331348A (en) *  20110609  20120125  河南科技大学  Evaluation method for characteristic variation of starting friction torque of slewing bearing 
CN102831325A (en) *  20120904  20121219  北京航空航天大学  Method for predicting bearing fault based on Gaussian process regression 
CN103246803A (en) *  20130407  20130814  河南科技大学  Significance testing method for rolling bearing performance variation process 
CN104318043A (en) *  20140220  20150128  河南科技大学  Rolling bearing vibration performance reliability variation process detection method and rolling bearing vibration performance reliability variation process detection device 
CN104598734A (en) *  20150122  20150506  西安交通大学  Life prediction model of rolling bearing integrated expectation maximization and particle filter 

2015
 20150921 CN CN201510603662.2A patent/CN105224792B/en not_active Expired  Fee Related
Patent Citations (5)
Publication number  Priority date  Publication date  Assignee  Title 

CN102331348A (en) *  20110609  20120125  河南科技大学  Evaluation method for characteristic variation of starting friction torque of slewing bearing 
CN102831325A (en) *  20120904  20121219  北京航空航天大学  Method for predicting bearing fault based on Gaussian process regression 
CN103246803A (en) *  20130407  20130814  河南科技大学  Significance testing method for rolling bearing performance variation process 
CN104318043A (en) *  20140220  20150128  河南科技大学  Rolling bearing vibration performance reliability variation process detection method and rolling bearing vibration performance reliability variation process detection device 
CN104598734A (en) *  20150122  20150506  西安交通大学  Life prediction model of rolling bearing integrated expectation maximization and particle filter 
NonPatent Citations (3)
Title 

Evaluatlon for optimum technical plan of rolling bearing vibration using grey system theory;XinTao Xia 等;《Journal of Grey System》;20061231;第9卷(第1期);914 * 
滚动轴承振动速度的乏信息真值估计;栗永非，时保吉;《轴承》;20150331(第3期);3942, 53 * 
航天轴承摩擦力矩的最大熵概率分布与bootstrap推断;夏新涛 等;《宇航学报》;20070930;第28卷(第5期);13951400 * 
Also Published As
Publication number  Publication date 

CN105224792A (en)  20160106 
Similar Documents
Publication  Publication Date  Title 

Sobie et al.  Simulationdriven machine learning: Bearing fault classification  
Cui et al.  A novel switching unscented Kalman filter method for remaining useful life prediction of rolling bearing  
Yang et al.  Wind turbine condition monitoring: technical and commercial challenges  
Singh et al.  An extensive review of vibration modelling of rolling element bearings with localised and extended defects  
Iliopoulos et al.  A modal decomposition and expansion approach for prediction of dynamic responses on a monopile offshore wind turbine using a limited number of vibration sensors  
Amarnath et al.  Exploiting sound signals for fault diagnosis of bearings using decision tree  
Cui et al.  HVSRMS localization formula and localization law: Localization diagnosis of a ball bearing outer ring fault  
Wilkinson et al.  Comparison of methods for wind turbine condition monitoring with SCADA data  
Liao et al.  A framework for predicting the remaining useful life of a single unit under timevarying operating conditions  
CN103955750B (en)  Rolling bearing remaining life prediction method based on feature fusion and particle filtering  
Dolenc et al.  Distributed bearing fault diagnosis based on vibration analysis  
Villa et al.  Statistical fault diagnosis based on vibration analysis for gear testbench under nonstationary conditions of speed and load  
US8290718B2 (en)  Bridge monitoring and safety evaluation method using a vibration technique  
US10337958B2 (en)  Bearing device vibration analysis method, bearing device vibration analyzer, and rolling bearing condition monitoring system  
Kankar et al.  Rolling element bearing fault diagnosis using wavelet transform  
Gray et al.  Physics of failure approach to wind turbine condition based maintenance  
CN101835974B (en)  Method for determining fatigue damage in a power train of a wind turbine  
Mathew et al.  Failure mechanisms based prognostics  
Ahmad et al.  A reliable technique for remaining useful life estimation of rolling element bearings using dynamic regression models  
Fischer et al.  Condition monitoring of wind turbines: state of the art, user experience and recommendations  
JP6216242B2 (en)  Anomaly detection method and apparatus  
Farrar et al.  An introduction to structural health monitoring  
Sun et al.  Mechanical systems hazard estimation using condition monitoring  
Azarbayejani et al.  A probabilistic approach for optimal sensor allocation in structural health monitoring  
EP2484900B1 (en)  Method for lubricating at least one blade pitch bearing of a wind turbine 
Legal Events
Date  Code  Title  Description 

PB01  Publication  
C06  Publication  
SE01  Entry into force of request for substantive examination  
C10  Entry into substantive examination  
GR01  Patent grant  
GR01  Patent grant  
CF01  Termination of patent right due to nonpayment of annual fee 
Granted publication date: 20180406 Termination date: 20180921 

CF01  Termination of patent right due to nonpayment of annual fee 