CN109472042A - A kind of reliability sampling test method based on acceleration degraded data - Google Patents
A kind of reliability sampling test method based on acceleration degraded data Download PDFInfo
- Publication number
- CN109472042A CN109472042A CN201811076300.2A CN201811076300A CN109472042A CN 109472042 A CN109472042 A CN 109472042A CN 201811076300 A CN201811076300 A CN 201811076300A CN 109472042 A CN109472042 A CN 109472042A
- Authority
- CN
- China
- Prior art keywords
- test
- product
- reliability
- risk
- sampling
- 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]
- G06F30/20—Design optimisation, verification or simulation
Abstract
The present invention relates to reliability test technical fields, more particularly to a kind of based on the reliability sampling test method for accelerating degraded data, the present invention obtains the reliability function of product by Maximum Likelihood Estimation Method according to the data of the accelerated degradation test of product, and according to reliability function, to accelerate the test variance minimum of acceptance test to establish the Optimized model and constraint condition of acceptance test, and the parameter of Optimized model is optimized, it is difficult to solve the problems, such as that high reliability long life product batches are checked and accepted for the optimal acceptance scheme for obtaining accelerating acceptance test.The shortcomings that acceptance scheme based on the guarantee test time can not obtain the failure parameter of product by statistical modeling method is overcome simultaneously.
Description
Technical field
The present invention relates to reliability test technical fields, and in particular to a kind of based on the reliability sampling for accelerating degraded data
Test method.
Background technique
There are two types of currently used Demonstration Reliability Acceptance Test appraisal procedures: first is that the examination based on the guarantee test time, i.e.,
A certain amount of reliability test is carried out to test sample, sees whether product is able to take in the experimental condition and test period not
Exposure is out of order.The advantages of this method is that experimentation cost is controllable, can the verifying production to a certain extent in determining test period
The reliability of product；But this method can not accurately estimate product achieved reliability parameter, can not be obtained by statistical modeling method
The failure parameter of product, product can not effectively be managed and can not reliability quality to different manufacturers batch sample carry out
It distinguishes.
Second is that the examination based on out-of-service time and the number that fails, i.e., carry out the reliability test of certain sample size to product,
Its out-of-service time or failure quantity are collected, statistical model modeling is carried out to product by obtained data.The advantages of this method
It is that can judge whether product can receive or refuse by obtained parameter；But for high reliability long life product and
Speech, experimental condition not enough will cause test products and no-failure occurs causing test failure, and excessively high experimental condition can be made
It is excessively high at the uncertain and experimentation cost of test failure mode, so that production and user can not carry out test.
Summary of the invention
To solve the above-mentioned problems, the present invention provides a kind of based on the reliability sampling test side for accelerating degraded data
Method, specific technical solution are as follows:
It is a kind of based on accelerate degraded data reliability sampling test method the following steps are included:
(1) the acceleration degradation information of product is obtained by carrying out accelerated test to product, records moving back for accelerated degradation test
Change experimental condition, and the test data D being recorded under different testing time pointsiAnd testing time point Ti, wherein i=1,
2 ..., n, n are the quantity of test data；
(2) it is established by acceleration degradation information obtained above and accelerates degradation model, obtain the reliability function of product；
(3) design parameter of sampling test is determined；
(4) according to step (1)-step (3) result obtain accelerate acceptance test test parameters, and to test parameters into
Row optimization obtains accelerated test stress, product sample number, the receivable quantity and total time on test length to fail of test.
Preferably, step (2) specific steps include:
(1) assume that the Degradation path of product obeys random Brownian motionWherein, D0It is initial
Degradation values,For accelerate deterioration velocity, σ be constant value drift amount, B (t) be moment t standard Brownian movement B (t)~N (0,
T), DtFor in the degradation values of moment t；Obtain the reliability function of product:
Wherein, DfFor the threshold values of product failure；
(2) assume that the acceleration in accelerated degradation test is used as accelerated stress using vibration, then under the conditions of single stress
Coffin-Manson accelerates relationship are as follows:
Wherein, NSAnd N0The vibration frequency respectively under acceleration and conventional environment, w is index parameters；
To have
μ (N is set herein0)=aN0+ b, wherein a, b are agreement constant；Then have:
(3) reliability function of this batch of product is obtained according to Maximum Likelihood Estimation Method by step (1) and step (2) are as follows:
Preferably, step (3) specific steps include:
(1) assumed condition is established:
Assuming that (1): test product is sampled from process, and bi-distribution is selected to be distributed as bottom；
Assuming that (2) test cell be it is independent, can be carried out in testing time τ I type truncation test；
Assuming that it is continuous measurable, negligible increment space that (3), which degenerate,；
Assuming that (4) both sides' risk is made of the model risk of sampling risk and modeling of degenerating；
According to above-mentioned it is assumed that sampling risk are as follows:
I=1....n indicates n product；Have for each vIndicate different time of measuring points, Ω
(τ) representative products are in the fitting risk of time τ point, wherein DijkWithThe jth of i-th of sample respectively under k stress condition
The measured value and estimated value of a amount of degradation；For the variance of amount of degradation estimated value,For the estimated value of error；C is examination
Test the quantity of acceptable failure；α0For Production venture, α1For product sampling risk, α2For the fitting risk of product；Pr is failure
Risk function；R indicates r-th of failure product；R(t；It N) is reliability of the t moment product in the case where stress is N；M is product sample
Number
(2) composite risk is not only influenced by sampling factor but also is influenced by degradation model, i.e. α=α1+α2；Setting accelerates to check and accept examination
Test design principle: so that accelerate the test variance of acceptance test minimum,
Constant single stress acceptance test is then carried out in testing time τ, tests variance are as follows:
Var(Dt)=μ(N)+σ2τ2N2；⑩
Therefore the Optimized model of acceptance test are as follows:
NL< N < NH
Wherein, α0The producer risk maximum value fixed for both sides before on-test；Wherein, Var (Dt) for acceptance test institute
There is the variance of amount of degradation；NLFor the lower limit of proof stress；NHFor the upper limit of proof stress；
Parameter to be optimized: Θ=[N, M, c, τ]T, wherein N is proof stress, and τ is test period.
Preferably, step (4) specific steps include:
(1) objective function and constraint condition of acceptance test are established
NL< N < NH
Wherein:
(2) to above-mentioned carry out numerical optimization, the stress (N) of available sampling test, sample size M, the acceptable mistake of test
The quantity c and total time on test length τ lost；
(3) test can be carried out according to the testing program of step (2) parameters obtained, if sampling M sample in the batch, into
Row stress is the accelerated degradation test of N, if not failing within the τ time, then it represents that the batch products are qualified.
The invention has the benefit that
The present invention obtains the reliability function of product according to the data of the accelerated degradation test of product, and according to reliability letter
Number, to accelerate the test variance minimum of acceptance test to establish the Optimized model and constraint condition of acceptance test, and to Optimized model
Parameter optimize, obtain accelerate acceptance test optimal acceptance scheme, solve high reliability long life product batches examination
Difficult problem.Product can not be obtained by statistical modeling method by overcoming the acceptance scheme based on the guarantee test time simultaneously
The shortcomings that failure parameter.
Specific embodiment
In order to better understand the present invention, the present invention is further explained in the light of specific embodiments:
It is a kind of based on accelerate degraded data reliability sampling test method the following steps are included:
(1) the acceleration degradation information of product is obtained by carrying out accelerated test to product, records moving back for accelerated degradation test
Change experimental condition, and the test data D being recorded under different testing time pointsiAnd testing time point Ti, wherein i=1,
2 ..., n, n are the quantity of test data；By taking intelligent electric meter as an example, annual grid company will adopt a large amount of intelligent electric meter
It purchases, after electric energy meter arrival, needs to be sampled intelligent electric meter detection, detection content includes elementary error, starting, shunt running etc.
Multinomial electric energy meter performance indicator.The present embodiment is tested using acceleration shock and carries out accelerated degradation test to intelligent electric meter, and records
Lower 18 pieces of intelligent electric meter error of time of day, intelligent electric meter number BE1-BE18, as shown in Table 1 and Table 2:
1 BE1-BE9 intelligent electric meter timing error detail list of table
2 BE10-BE18 intelligent electric meter timing error detail list of table
(2) it is established by acceleration degradation information obtained above and accelerates degradation model, obtain the reliability function of product:
1) assume that the Degradation path of product obeys random Brownian motionWherein, D0Initially to move back
Change value,For accelerate deterioration velocity, σ be constant value drift amount, B (t) be moment t standard Brownian movement B (t)~N (0,
T), DtFor in the degradation values of moment t；Obtain the reliability function of product:
Wherein, DfFor the threshold values of product failure；
2) assume that the acceleration in accelerated degradation test is used as accelerated stress using vibration, then under the conditions of single stress
Coffin-Manson accelerates relationship are as follows:
Wherein, NSAnd N0The vibration frequency respectively under acceleration and conventional environment, w is index parameters；
To have
μ (N is set herein0)=aN0+ b, wherein a, b are agreement constant；Then have:
(3) reliability function of this batch of product is obtained according to Maximum Likelihood Estimation Method by step (1) and step (2) are as follows:
Wherein, μ (N)=(0.02N0+ 0.05) * (N/200), σ=0.5, Df=5, D0=0.5.
There are many documents to have the method for how acquiring the function, this hair from degraded data using Maximum Likelihood Estimation Method
It is bright not repeat herein, then D is utilized to iti(i=1...n) and testing time point Ti(i=1...n) Maximum-likelihood estimation is carried out
Available above-mentioned reliability function.
(3) design parameter of sampling test is determined:
1) assumed condition is established:
Assuming that (1): test product is sampled from process, and bi-distribution is selected to be distributed as bottom；
Assuming that (2) test cell be it is independent, can be carried out in testing time τ I type truncation test；
Assuming that it is continuous measurable, negligible increment space that (3), which degenerate,；
Assuming that (4) both sides' risk is made of the model risk of sampling risk and modeling of degenerating；
According to above-mentioned it is assumed that sampling risk are as follows:
I=1....n indicates n product；Have for each vIndicate different time of measuring points；, Ω
(τ) representative products are in the fitting risk of time τ point, wherein DijkWithThe jth of i-th of sample respectively under k stress condition
The measured value and estimated value of a amount of degradation；For the variance of amount of degradation estimated value,For the estimated value of error；C is test
The quantity of acceptable failure；α0For Production venture, α1For product sampling risk, α2For the fitting risk of product.；Wherein T is to learn
Raw-t statistic, freedom degree m-n, wherein n is the number of parameter in degradation model,To indicate to survey
Magnitude and the unequal probability of true value, statistic T meet student t distribution, and r indicates r-th of failure product；R(t；It N) is t moment
Reliability of the product in the case where stress is N；M is product sample number.
2) composite risk is not only influenced by sampling factor but also is influenced by degradation model, i.e. α=α1+α2；Setting accelerates acceptance test
Design principle: so that accelerate the test variance of acceptance test minimum,
Constant single stress acceptance test is then carried out in testing time τ, tests variance are as follows:
Var(Dt)=μ (N)+σ2τ2N2；⑩
Therefore the Optimized model of acceptance test are as follows:
NL< N < NH
Wherein, α0The producer risk maximum value fixed for both sides before on-test；Wherein, Var (Dt) for acceptance test institute
There is the variance of amount of degradation；NLFor the lower limit of proof stress；NHFor the upper limit of proof stress；
Parameter to be optimized: Θ=[N, M, c, τ]T, wherein N is proof stress, and τ is test period, according to the deliberation of both sides,
Determine that producer risk maximum value is α0=0.2.
(4) according to step (1)-step (3) result obtain accelerate acceptance test test parameters, and to test parameters into
Row optimization obtains accelerated test stress, product sample number, the receivable quantity and total time on test length to fail of test:
1) objective function and constraint condition of acceptance test are established:
NL< N < NH
Wherein:
Wherein, NL=100, NH=1000,
(2) to above-mentioned carry out numerical optimization, available proof stress N, product sample number M, test are subjected to failure
Quantity c and test period τ；Optimum results are as shown in table 3:
3 optimum results detail of table
| τ (hour) | N (newton) | c | M(a) | |
| Optimal value | 0.813 | 261.1 | 0 | 4 |
(3) test can be carried out according to the testing program of step (2) parameters obtained, if sampling M sample in the batch, into
Row stress is the accelerated degradation test of N, if not failing within the τ time, then it represents that the batch products are qualified；Taken out in the batch
4 sample tables are taken, under the proof strength of proof stress N=261.1 newton, are not failed by 0.813 hour product, then table
Show that the batch products meet quality requirement.
The present invention is not limited to above-described specific embodiment, and the foregoing is merely preferable case study on implementation of the invention
, it is not intended to limit the invention, any modification done within the spirit and principles of the present invention and changes equivalent replacement
Into etc., it should all be included in the protection scope of the present invention.
Claims (4)
1. a kind of based on the reliability sampling test method for accelerating degraded data, it is characterised in that: the following steps are included:
(1) the acceleration degradation information of product is obtained by carrying out accelerated test to product, records the degeneration examination of accelerated degradation test
The test data D for testing condition, and being recorded under different testing time pointsiAnd testing time point Ti, wherein i=1,2 ..., n, n
For the quantity of test data；
(2) it is established by acceleration degradation information obtained above and accelerates degradation model, obtain the reliability function of product；
(3) design parameter of sampling test is determined；
(4) it obtains accelerating the test parameters of acceptance test according to step (1)-step (3) result, and test parameters is carried out excellent
Change, obtains accelerated test stress, product sample number, the receivable quantity and total time on test length to fail of test.
2. according to claim 1 a kind of based on the reliability sampling test method for accelerating degraded data, it is characterised in that:
Step (2) specific steps include:
(1) assume that the Degradation path of product obeys random Brownian motionWherein, D0Initially to degenerate
Value,To accelerate deterioration velocity, σ is constant value drift amount, and B (t) is standard Brownian movement B (t)~N (0, t) in moment t,
DtFor in the degradation values of moment t；Obtain the reliability function of product:
Wherein, DfFor the threshold values of product failure；
(2) assume that the acceleration in accelerated degradation test is used as accelerated stress using vibration, then the Coffin- under the conditions of single stress
Manson accelerates relationship are as follows:
Wherein, NSAnd N0The vibration frequency respectively under acceleration and conventional environment, w is index parameters；
To have
μ (N is set herein0)=aN0+ b, wherein a, b are agreement constant；Then have:
(3) reliability function of this batch of product is obtained according to Maximum Likelihood Estimation Method by step (1) and step (2) are as follows:
3. according to claim 1 a kind of based on the reliability sampling test method for accelerating degraded data, it is characterised in that:
Step (3) specific steps include:
(1) assumed condition is established:
Assuming that (1): test product is sampled from process, and bi-distribution is selected to be distributed as bottom；
Assuming that (2) test cell be it is independent, can be carried out in testing time τ I type truncation test；
Assuming that it is continuous measurable, negligible increment space that (3), which degenerate,；
Assuming that (4) both sides' risk is made of the model risk of sampling risk and modeling of degenerating；
According to above-mentioned it is assumed that sampling risk are as follows:
Wherein:
C is the quantity of the acceptable failure of test；α1For Production venture；Pr is failure risk function；R indicates that r-th of failure produces
Product；R(t；It N) is reliability of the t moment product in the case where stress is N；M is product sample number；α2,p(t；N) the meaning indicated.(2)
Composite risk is not only influenced by sampling factor but also is influenced by degradation model, i.e. α=α1+α2；Setting accelerates acceptance test design principle:
So that accelerate the test variance of acceptance test minimum,
Constant single stress acceptance test is then carried out in testing time τ, tests variance are as follows:
Var(Dt)=μ (N)+σ2τ2N2； ⑨
Therefore the Optimized model of acceptance test are as follows:
Wherein, α0The producer risk maximum value fixed for both sides before on-test；Wherein, Var (Dt) moved back for acceptance test is all
The variance of change amount；NLFor the lower limit of proof stress；NHFor the upper limit of proof stress；
Parameter to be optimized: Θ=[N, M, c, τ]T, wherein N is proof stress, and τ is test period.
4. according to claim 1 a kind of based on the reliability sampling test method for accelerating degraded data, it is characterised in that:
Step (4) specific steps include:
(1) objective function and constraint condition of acceptance test are established
Wherein:
I=1....n indicates n product；Have for each vIndicate different time of measuring points, Ω (τ) generation
Table product is in the fitting risk of time τ point, wherein DijkWithJ-th of degeneration of i-th of sample respectively under k stress condition
The measured value and estimated value of amount；For the variance of amount of degradation estimated value,For the estimated value of error；C is that test can connect
By the quantity of failure；α0For Production venture, α1For product sampling risk, α2For the fitting risk of product；Pr is failure risk letter
Number；R indicates r-th of failure product；R(t；It N) is reliability of the t moment product in the case where stress is N；M is product sample number.
(2) to above-mentioned carry out numerical optimization, the stress (N) of available sampling test, sample size M, test are subjected to failure
Quantity c and total time on test length τ；
(3) test can be carried out according to the testing program of step (2) parameters obtained, if sampling M sample in the batch, be answered
Power is the accelerated degradation test of N, if not failing within the τ time, then it represents that the batch products are qualified.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811076300.2A CN109472042A (en) | 2018-09-14 | 2018-09-14 | A kind of reliability sampling test method based on acceleration degraded data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811076300.2A CN109472042A (en) | 2018-09-14 | 2018-09-14 | A kind of reliability sampling test method based on acceleration degraded data |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109472042A true CN109472042A (en) | 2019-03-15 |
Family
ID=65664615
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811076300.2A Pending CN109472042A (en) | 2018-09-14 | 2018-09-14 | A kind of reliability sampling test method based on acceleration degraded data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109472042A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110826234A (en) * | 2019-11-08 | 2020-02-21 | 中国航天标准化研究所 | Simulation-based multi-stress accelerated life test scheme optimization method |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101666662A (en) * | 2009-09-25 | 2010-03-10 | 北京航空航天大学 | Accelerated degradation test prediction method based on fuzzy theory |
| CN101793927A (en) * | 2010-01-12 | 2010-08-04 | 北京航空航天大学 | Optimization design method of step-stress accelerated degradation test |
| JP2010186704A (en) * | 2009-02-13 | 2010-08-26 | Hitachi Maxell Ltd | Lifetime acceleration testing method of polymer electrolyte fuel cell |
| CN102253242A (en) * | 2011-04-27 | 2011-11-23 | 北京航空航天大学 | Method for determining stationary phase of accelerometer based on dual-parameter accelerated degradation data |
| CN102707255A (en) * | 2012-06-19 | 2012-10-03 | 华北电网有限公司计量中心 | Quick sampling and detecting method for intelligent ammeter |
| KR20150035183A (en) * | 2013-09-27 | 2015-04-06 | 한국전력공사 | Apparatus for accelerated degradation testing of cable |
| CN106372315A (en) * | 2016-08-31 | 2017-02-01 | 华南理工大学 | Accelerated degradation testing method based on improved Brown drift motion |
| CN106991074A (en) * | 2017-04-11 | 2017-07-28 | 中国人民解放军海军航空工程学院 | Accelerated degradation test Optimization Design based on accelerated factor principle of invariance |
-
2018
- 2018-09-14 CN CN201811076300.2A patent/CN109472042A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2010186704A (en) * | 2009-02-13 | 2010-08-26 | Hitachi Maxell Ltd | Lifetime acceleration testing method of polymer electrolyte fuel cell |
| CN101666662A (en) * | 2009-09-25 | 2010-03-10 | 北京航空航天大学 | Accelerated degradation test prediction method based on fuzzy theory |
| CN101793927A (en) * | 2010-01-12 | 2010-08-04 | 北京航空航天大学 | Optimization design method of step-stress accelerated degradation test |
| CN102253242A (en) * | 2011-04-27 | 2011-11-23 | 北京航空航天大学 | Method for determining stationary phase of accelerometer based on dual-parameter accelerated degradation data |
| CN102707255A (en) * | 2012-06-19 | 2012-10-03 | 华北电网有限公司计量中心 | Quick sampling and detecting method for intelligent ammeter |
| KR20150035183A (en) * | 2013-09-27 | 2015-04-06 | 한국전력공사 | Apparatus for accelerated degradation testing of cable |
| CN106372315A (en) * | 2016-08-31 | 2017-02-01 | 华南理工大学 | Accelerated degradation testing method based on improved Brown drift motion |
| CN106991074A (en) * | 2017-04-11 | 2017-07-28 | 中国人民解放军海军航空工程学院 | Accelerated degradation test Optimization Design based on accelerated factor principle of invariance |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110826234A (en) * | 2019-11-08 | 2020-02-21 | 中国航天标准化研究所 | Simulation-based multi-stress accelerated life test scheme optimization method |
| CN110826234B (en) * | 2019-11-08 | 2022-11-29 | 中国航天标准化研究所 | Simulation-based multi-stress accelerated life test scheme optimization method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10706332B2 (en) | Analog circuit fault mode classification method | |
| CN109409628B (en) | Acquisition terminal manufacturer evaluation method based on metering big data clustering model | |
| CN107478455B (en) | Timing truncation reliability test method suitable for Weibull distribution type product | |
| CN109389145B (en) | Electric energy meter manufacturer evaluation method based on metering big data clustering model | |
| CN112465040B (en) | Software defect prediction method based on class unbalance learning algorithm | |
| CN106201871A (en) | Based on the Software Defects Predict Methods that cost-sensitive is semi-supervised | |
| CN111178587B (en) | Spark framework-based short-term power load rapid prediction method | |
| CN103308723A (en) | Product service life rapid test method based on physical model | |
| CN110443117B (en) | Wind turbine generator fault diagnosis method | |
| CN111174370A (en) | Fault detection method and device, storage medium and electronic device | |
| CN106021928A (en) | Comprehensive stress acceleration test method | |
| CN107015875A (en) | A kind of complete electronic set storage life appraisal procedure and device | |
| CN110046789B (en) | Automatic generation method and system for student information literacy assessment test paper | |
| CN110135754A (en) | Examination question appraisal procedure and device | |
| CN101727441B (en) | Evaluating method and evaluating system targeting Chinese name identifying system | |
| CN111400505A (en) | Method and system for matching fault elimination scheme of power consumption information acquisition system | |
| CN109543720A (en) | A kind of wafer figure defect mode recognition methods generating network based on confrontation | |
| CN109472042A (en) | A kind of reliability sampling test method based on acceleration degraded data | |
| CN114662405A (en) | Rock burst prediction method based on few-sample measurement and ensemble learning | |
| CN113946920A (en) | Rolling bearing fault diagnosis method with unbalanced data and data set deviation | |
| CN109375143A (en) | A kind of method of determining intelligent electric energy meter remaining life | |
| CN110956112B (en) | Novel high-reliability slewing bearing service life assessment method | |
| CN116662800A (en) | Rolling bearing fault diagnosis method based on self-adaptive attention mechanism | |
| CN110569893A (en) | distribution equipment defect analysis management method and system | |
| CN111898314A (en) | Lake water body parameter detection method and device, electronic equipment and storage medium |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190315 |