CN109101778B - Wiener process parameter estimation method based on performance degradation data and life data fusion - Google Patents
Wiener process parameter estimation method based on performance degradation data and life data fusion Download PDFInfo
- Publication number
- CN109101778B CN109101778B CN201811395335.2A CN201811395335A CN109101778B CN 109101778 B CN109101778 B CN 109101778B CN 201811395335 A CN201811395335 A CN 201811395335A CN 109101778 B CN109101778 B CN 109101778B
- Authority
- CN
- China
- Prior art keywords
- data
- product
- parameter
- wiener
- performance degradation
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/04—Ageing analysis or optimisation against ageing
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Computational Mathematics (AREA)
- Mathematical Physics (AREA)
- Pure & Applied Mathematics (AREA)
- Mathematical Optimization (AREA)
- Mathematical Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Biology (AREA)
- Computer Hardware Design (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Operations Research (AREA)
- Probability & Statistics with Applications (AREA)
- Evolutionary Computation (AREA)
- Algebra (AREA)
- Geometry (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- Complex Calculations (AREA)
Abstract
The invention belongs to the technical field of reliability statistics and discloses a Wiener process parameter estimation method based on fusion of performance degradation data and life data. The method well solves the problem of Wiener process parameter estimation based on the fusion of performance degradation data and life data through the steps. Compared with the existing method, the algorithm provided by the invention is simple and effective in calculation process. The following embodiments show that the algorithm provided by the invention has high calculation precision.
Description
Technical field
Present invention relates generally to reliability statistics fields, refer in particular to a kind of based on Performance Degradation Data and lifetime data fusion
Wiener procedure parameter estimation method.
Background technique
In Reliability Engineering, the reliability estimation method based on lifetime data is traditional technology.This method is main
It is to be distributed using mathematical statistics using the specific service life, assesses the reliable of product by analyzing the lifetime data of product
Property.But with the progress of science and technology, the reliability of the raising of manufacturing process, product is higher and higher, has been difficult to be collected into a large amount of
Lifetime data, cause traditional reliability estimation method based on lifetime data to be difficult to be applicable in.
For this purpose, the reliability estimation method based on Performance Degradation Data has obtained broad development.This method is mainly logical
The excessively degenerative process of analysis properties of product data degenerates to the time of specific threshold by analysis properties of product to assess product
Reliability.
Wiener process is widely used in the reliability estimation method based on Performance Degradation Data.Remember the performance number of product
According to forIf it meets following property:
(1) momentTo the momentBetween increment Normal Distribution, i.e.,；
(2) to the disjoint time interval of any two,,, incrementWithIndependently of each other；
(3)And?Continuously.
ThenUnitary Wiener process is obeyed, and claims parameterFor drift parameter,It is diffusion parameter [with reference to text
It offers: Reliability Modeling research National University of Defense technology doctoral thesis of the Peng Baohua based on Wiener process, 2010.].
If the degradation of product is, i.e., when the performance of product degenerates to for the first timeWhen, product will fail, and at this time may be used
Push away to obtain life of productBe distributed as dead wind area, and life of productDistribution function be
(1)
WhereinIt is standardized normal distribution,,。
If existing simultaneously Performance Degradation Data and lifetime data, can carry out based on Performance Degradation Data and lifetime data
The Reliability Assessment method of fusion, key among these are Wiener procedure parametersWithEstimation.For both
The fusion of data, current is mostly that Wiener procedure parameter is estimated by Bayes theory, such as by Performance Degradation Data and service life number
Get up tectonic syntaxis likelihood function according to fusion, or converts prior distribation for Performance Degradation Data and merged with lifetime data, or general
Lifetime data is converted into prior distribation and merges with Performance Degradation Data.But these method for parameter estimation are directed to complicated number
Student movement is calculated, and engineering practice is not easy to.
Summary of the invention
The technical problem to be solved in the present invention is that: when the one-parameter Performance Degradation Data and service life for being collected simultaneously product
When data, and the Performance Degradation Data obedience parameter of product isWithUnitary Wiener process, the service life number of product at this time
According to the dead wind area obeyed in formula (1).It now needs to assess product by the Performance Degradation Data and lifetime data of fusion product
Reliability, key therein are how to estimate drift parameterAnd diffusion parameter.In order to solve the above-mentioned technical problem, of the invention
It is proposed the Wiener procedure parameter estimation method merged based on Performance Degradation Data and lifetime data.This method can simplify entirely
Mathematical operation in calculating process, and can guarantee the precision of calculated result.
The technical scheme is that
The Wiener procedure parameter estimation method merged based on Performance Degradation Data and lifetime data, comprising the following steps:
(1) utility degraded data estimates drift parameter according to maximum-likelihood methodAnd diffusion parameterInitial value:
If sharedA sample carries out performance degradation experiment to collect Performance Degradation Data.To sample, initial time
Performance measurement is, exist respectivelyMoment measures samplePerformance obtain its performance measurement and be respectively。
NoteIt is sampleAt the momentWithBetween performance degradation amount, according to
Known to the property of Wiener process
Wherein,,。
Then according to Maximum Likelihood Estimation Method, Performance Degradation DataLikelihood function be
Parameter can be acquired by likelihood functionWithMaximum-likelihood estimation be
(2)
It acquiresWithAs parameterWithInitial value.
(2) failure probability of life expectancy data.
If the degradation of product is, i.e., when the performance of product degenerates to for the first timeWhen, product will fail.In the service life
The performance measurement of each sample product is measured in test, cannot be worked on if a certain moment starts a certain sample product and is worked as production
The performance measurement of product degenerates to for the first timeWhen, then the moment is lifetime data namely its out-of-service time of the sample product.
If shared(M and the n in step (1) are not related, i.e., M is likely larger than n, it is also possible to be less than n, it is also possible to be equal to
N, should be subject to the data volume being actually collected into) a sample products carry out life test to collect lifetime data；If being collected into
A lifetime data isIt the out-of-service time of a sample products, is denoted as, and set；Then utilize step
(1) solution obtains inWith, life expectancy data as the following formulaFailure probability:
Wherein,,。
(3) the failure probability estimation procedure parameter of lifetime data is utilized.
According to all pointsIt is fitted product life distribution curve, wherein, that is, it is based on curve matching
Thought, utilize trusted zones reflect (trust region reflective) algorithm, solving optimization model
Obtain inverse Gauss service life distribution parameterWithEstimation, be denoted asWith, then Wiener procedure parameter can be obtained
WithBe estimated as
(3)
As described above, the present invention is asked first with Performance Degradation Data according to maximum-likelihood method according to step shown in FIG. 1
Wiener procedure parameter initial value, recycle the failure probability of the initial estimate lifetime data, finally be fitted failure probability and
Lifetime data life expectancy distribution parameter, and it is converted into the end value of Wiener procedure parameter.The present invention is through the above steps very
The good Wiener procedure parameter estimation problem solved based on Performance Degradation Data and lifetime data fusion, compares existing side
Method, algorithm calculating process proposed by the invention are simple and effective.By following specific embodiments it is found that proposed by the invention
Algorithm computational accuracy is very high.
Detailed description of the invention
Fig. 1 is flow chart of the invention,
Fig. 2 is the performance measurement of present example.
Specific embodiment
The present invention is described in further details below with reference to example and attached drawing.
This example discloses a kind of Wiener procedure parameter estimation side merged based on Performance Degradation Data and lifetime data
Method.In this example, it is assumed that have 6 samples, and the performance degradation amount of each sample obeys parameterWith's
Wiener process.6 groups of random numbers are generated based on the Wiener process, each group respectively there are 10001 numerical value, as 6 samples
Performance measurement, as shown in Figure 2.In addition, generating 6 random numbers based on dead wind area corresponding to the Wiener process and making
It is 1.4257,1.5118,1.5365,1.5641,2.2672 and 2.4917 for the lifetime data of 6 samples.Of the invention
Specific algorithm is as follows:
Firstly, based on the performance measurement in Fig. 2, acquiring Wiener procedure parameter using formula (2) according to maximum-likelihood method
Initial value beWith。
Then the service life distribution parameter initial value converted according to Wiener procedure parameter initial value, estimates 6 lifetime datas
Failure probability can obtain 0.2106,0.2631,0.2787,0.2962,0.7029 and 0.7895.
It is finally fitted 6 groups of lifetime datas and failure probability point estimation, inverse Gauss service life distribution parameter is acquired, is then converted to
The end value of Wiener procedure parameter can obtainWith。
By above example it is found that utilizing the Wiener for merging Performance Degradation Data and lifetime data proposed by the invention
Procedure parameter estimation method, the Wiener procedure parameter estimated value and true value acquired is very close, this demonstrate that the present invention is mentioned
The accuracy of method out.
Claims (1)
1. the Wiener procedure parameter estimation method merged based on Performance Degradation Data and lifetime data, it is characterised in that: including
Following steps:
(1) utility degraded data estimates the initial value of drift parameter μ and diffusion parameter σ according to maximum-likelihood method；
If shared n sample products carry out performance degradation experiment to collect Performance Degradation Data；For sample products i, when initial
Carve ti0Performance measurement is Xi0=0, exist respectivelyThe performance of moment measurement sample products i obtains its performance measurement point
It is not
Remember Δ xij=Xij-Xi(j-1)It is sample products i in moment ti(j-1)And tijBetween performance degradation amount, according to Wiener mistake
Known to the property of journey
Δxij~N (μ Δ tij,σ2Δtij)
Wherein Δ tij=tij-ti(j-1), i=1 ..., n, j=1 ..., mi；
Then according to Maximum Likelihood Estimation Method, Performance Degradation Data Δ xijLikelihood function be
It is by the Maximum-likelihood estimation that likelihood function can acquire parameter μ and σ
It acquiresWithInitial value as parameter μ and σ；
(2) failure probability of life expectancy data；
If the degradation of product is l, i.e., when the performance measurement of product degenerates to l for the first time, product will fail, and product loses
Its out-of-service time of corresponding lifetime data when effect；
If shared M sample products carry out life test to collect lifetime data；If being collected into M lifetime data i.e. M sample
The out-of-service time of product, it is denoted as t1,…,tM, and set t1≤…≤tM；Then obtained using solution in step (1)WithIt presses
The failure probability of following formula life expectancy data tk:
Wherein k=1 ..., M,
(3) Wiener procedure parameter is estimated using the failure probability of lifetime data；
According to all pointsIt is fitted product life distribution curve, i.e., based on the thought of curve matching, is reflected and is calculated using trusted zones
Method, solving optimization model
S.t α > 0, λ > 0
The estimation of inverse Gauss service life profile parameter and λ is obtained, is denoted asWithThe estimation of Wiener procedure parameter μ and σ can then be obtained
For
(4) according to the estimated value of the Wiener procedure parameter acquired in formula (3)WithThe reliablity estimation of t moment product can be obtained
Value are as follows:
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811395335.2A CN109101778B (en) | 2018-11-22 | 2018-11-22 | Wiener process parameter estimation method based on performance degradation data and life data fusion |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811395335.2A CN109101778B (en) | 2018-11-22 | 2018-11-22 | Wiener process parameter estimation method based on performance degradation data and life data fusion |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109101778A CN109101778A (en) | 2018-12-28 |
| CN109101778B true CN109101778B (en) | 2019-04-02 |
Family
ID=64870226
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811395335.2A Active CN109101778B (en) | 2018-11-22 | 2018-11-22 | Wiener process parameter estimation method based on performance degradation data and life data fusion |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109101778B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110568337A (en) * | 2019-09-16 | 2019-12-13 | 湖南银杏可靠性技术研究所有限公司 | Metal packaging power MOSFET storage reliability detection technology based on hidden degradation |
| CN111523251B (en) * | 2020-06-09 | 2023-04-21 | 江苏科技大学 | Method for rapidly evaluating service life of product under random environmental stress |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103488881A (en) * | 2013-09-06 | 2014-01-01 | 中国人民解放军第二炮兵工程大学 | Equipment residual service life prediction method under the condition of uncertain degradation measured data |
| CN104635155A (en) * | 2015-03-11 | 2015-05-20 | 哈尔滨工业大学 | Relay reliability estimation method based on Wiener process |
| CN107862134A (en) * | 2017-11-06 | 2018-03-30 | 河南科技大学 | A kind of Wiener process reliability analysis methods for considering autocorrelation measurement error |
| CN108520152A (en) * | 2018-04-13 | 2018-09-11 | 中国人民解放军火箭军工程大学 | A kind of the service life distribution determination method and system of engineering equipment |
| CN108647897A (en) * | 2018-05-16 | 2018-10-12 | 中南林业科技大学 | A kind of method and system of product reliability analysis |
-
2018
- 2018-11-22 CN CN201811395335.2A patent/CN109101778B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103488881A (en) * | 2013-09-06 | 2014-01-01 | 中国人民解放军第二炮兵工程大学 | Equipment residual service life prediction method under the condition of uncertain degradation measured data |
| CN104635155A (en) * | 2015-03-11 | 2015-05-20 | 哈尔滨工业大学 | Relay reliability estimation method based on Wiener process |
| CN107862134A (en) * | 2017-11-06 | 2018-03-30 | 河南科技大学 | A kind of Wiener process reliability analysis methods for considering autocorrelation measurement error |
| CN108520152A (en) * | 2018-04-13 | 2018-09-11 | 中国人民解放军火箭军工程大学 | A kind of the service life distribution determination method and system of engineering equipment |
| CN108647897A (en) * | 2018-05-16 | 2018-10-12 | 中南林业科技大学 | A kind of method and system of product reliability analysis |
Non-Patent Citations (2)
| Title |
|---|
| 基于Wiener过程的可靠性建模方法研究;彭宝华;《中国博士学位论文全文数据库工程科技Ⅱ辑》;20120415;C028-2 |
| 融合多源信息的维纳过程性能退化产品的可靠性评估;王小林 等;《电子学报》;20120531;第40卷(第5期);第977-982页 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109101778A (en) | 2018-12-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111046564B (en) | Residual life prediction method for two-stage degraded product | |
| CN117093879A (en) | Intelligent operation management method and system for data center | |
| CN106950945B (en) | A kind of fault detection method based on dimension changeable type independent component analysis model | |
| CN111414703B (en) | Method and device for predicting residual life of rolling bearing | |
| CN103761748A (en) | Method and device for detecting abnormal behaviors | |
| CN104376231A (en) | Damage identification method based on improved similar Bayesian calculation | |
| CN107249000B (en) | Method for detecting abnormal behaviors of mobile user | |
| CN105307200A (en) | Method for detecting abnormal value of multidimensional data of wireless sensor network based on trajectory | |
| CN109101778B (en) | Wiener process parameter estimation method based on performance degradation data and life data fusion | |
| CN108520267B (en) | Hydrological telemetering data anomaly detection method based on space-time characteristics | |
| CN117078048B (en) | Digital twinning-based intelligent city resource management method and system | |
| CN104462184A (en) | Large-scale data abnormity recognition method based on bidirectional sampling combination | |
| CN105468907A (en) | Accelerated degradation data validity testing and model selection method | |
| CN105843733A (en) | Big data platform performance detection method and device | |
| CN103425994A (en) | Feature selecting method for pattern classification | |
| CN108959188A (en) | Granger Causality based on quantization minimal error entropy criterion sentences the method for distinguishing | |
| CN105426647B (en) | Cold stand-by systems reliablity estimation method based on the fusion of reliability prior information | |
| CN103134433B (en) | Method for identifying slip factor caused by slope instability by utilizing displacement monitoring | |
| Tian et al. | An outliers detection method of time series data for soft sensor modeling | |
| CN105488598A (en) | Medium-and-long time electric power load prediction method based on fuzzy clustering | |
| CN117708625B (en) | Dam monitoring historical data filling method under spent data background | |
| Cheng et al. | Multipoint deformation safety monitoring model for concrete arch dams based on Bayesian model selection and averaging | |
| CN103559414A (en) | Environmental factor estimation method based on Arrhenius model | |
| CN105718733A (en) | Fault predicting method based on fuzzy nearness and particle filter | |
| Dupuis et al. | Detecting change-points in extremes |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |