CN111814301B - Reliability evaluation method, computer device, and computer-readable storage medium - Google Patents
Reliability evaluation method, computer device, and computer-readable storage medium Download PDFInfo
- Publication number
- CN111814301B CN111814301B CN202010468980.3A CN202010468980A CN111814301B CN 111814301 B CN111814301 B CN 111814301B CN 202010468980 A CN202010468980 A CN 202010468980A CN 111814301 B CN111814301 B CN 111814301B
- Authority
- CN
- China
- Prior art keywords
- function
- connection
- performance parameter
- functions
- product
- 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
Images
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
- G06F2111/00—Details relating to CAD techniques
- G06F2111/08—Probabilistic or stochastic CAD
-
- 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
- G06F2119/02—Reliability analysis or reliability optimisation; Failure analysis, e.g. worst case scenario performance, failure mode and effects analysis [FMEA]
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Hardware Design (AREA)
- Evolutionary Computation (AREA)
- Geometry (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Testing Or Measuring Of Semiconductors Or The Like (AREA)
Abstract
The present application relates to a reliability evaluation method, a computer device, and a computer-readable storage medium. The reliability evaluation method comprises the following steps: and judging whether the degradation process of each performance parameter of the product to be detected obeys the wiener process. And if the degradation process of each performance parameter of the product to be tested obeys the wiener process, obtaining the edge distribution function of the joint distribution function corresponding to each performance parameter by using the property of the wiener process. And obtaining parameter estimation values of at least two connection functions of the joint distribution function of each performance parameter according to the edge distribution function corresponding to each performance parameter. And determining an optimal connection function in the at least two connection functions according to the parameter estimation values of the at least two connection functions. And obtaining a reliability function of the product to be tested. The reliability evaluation method is a multivariate performance degradation product reliability evaluation method based on a wiener process and a connection function, so that the reliability evaluation process is more practical, and the evaluation result is more accurate.
Description
Technical Field
The present application relates to the field of reliability evaluation technologies, and in particular, to a reliability evaluation method, a computer device, and a computer-readable storage medium.
Background
The reliability evaluation means that the reliability evaluation value of the product under a certain specific condition is given by using the test or use information generated at each stage of the life cycle of the product and a probability statistical method. The product reliability evaluation is an important component of reliability engineering.
At present, most reliability evaluation methods only consider the random degradation process of single performance of products, so that the error between the existing reliability evaluation result and the actual result is large.
Disclosure of Invention
In view of the above, it is necessary to provide a reliability evaluation method, a computer device, and a computer-readable storage medium capable of improving evaluation accuracy in view of the above technical problems.
A reliability evaluation method, comprising:
judging whether the degradation process of each performance parameter of the product to be detected obeys the wiener process or not;
if the degradation process of each performance parameter of the product to be tested obeys the wiener process, obtaining the edge distribution function of the joint distribution function corresponding to each performance parameter by using the property of the wiener process;
obtaining parameter estimation values of at least two connection functions of the joint distribution function of each performance parameter according to the edge distribution function corresponding to each performance parameter;
determining an optimal connection function in the at least two connection functions according to the parameter estimation values of the at least two connection functions;
and obtaining a reliability function of the product to be tested according to the edge distribution function corresponding to each performance parameter, the parameter estimation value of the optimal connection function and the optimal connection function.
In one embodiment, the determining an optimal connection function among the at least two connection functions according to the parameter estimation values of the at least two connection functions includes:
obtaining likelihood functions of the at least two connection functions according to the parameter estimation values of the at least two connection functions;
calculating the Chi information values of the at least two connection functions according to the likelihood functions of the at least two connection functions;
and determining an optimal connection function according to the Chi information value.
In one embodiment, the determining an optimal connection function among the at least two connection functions according to the parameter estimation values of the at least two connection functions includes:
obtaining likelihood functions of the at least two connection functions according to the parameter estimation values of the at least two connection functions;
calculating a maximum of the likelihood functions of the at least two connection functions;
and determining an optimal connection function according to the maximum value of the likelihood function.
In one embodiment, before determining the optimal connection function among the at least two connection functions according to the parameter estimation values of the at least two connection functions, the method further includes:
calculating a correlation coefficient between the degradation amounts of the performance parameters;
and acquiring at least two connection functions of which the difference between the correlation coefficient and the degradation quantity of each performance parameter is within a preset range.
In one embodiment, the at least two join functions include a frank join function, a clement join function, and a gaussian join function.
In one embodiment, the determining whether the degradation process of each performance parameter of the product to be tested is compliant with the wiener process includes:
and obtaining a test result of multiple tests before failure of each performance parameter of a plurality of products to be tested according to a preset time interval.
In one embodiment, the determining whether the degradation process of each performance parameter of the product to be tested is compliant with the wiener process includes:
carrying out normal goodness-of-fit test on the increment of the degradation quantity of each performance parameter of the sample to be tested in each time interval through the Kolmogorov-Similov test;
and judging whether the degradation process of each performance parameter of the product to be tested is compliant with the wiener process according to the test result of the normal goodness of fit.
In one embodiment, the product to be tested is an insulated gate bipolar transistor, and the performance parameters of the product to be tested include collector-emitter saturation voltage drop and gate threshold voltage.
A computer device comprising a memory storing a computer program and a processor implementing the steps of the method described above when executing the computer program.
A computer-readable storage medium, on which a computer program is stored which, when being executed by a processor, carries out the steps of the above-mentioned method.
The reliability evaluation method is a multivariate performance degradation product reliability evaluation method based on a wiener process and a connection function. The method considers the randomness of the degradation process of a plurality of performance parameters of the product, simultaneously considers the coupling competition relationship of the plurality of performance parameters of the product, and provides a general reliability evaluation model of the multi-performance degraded product, so that the reliability evaluation process is more practical, the evaluation result is more accurate, and the application range is wider.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments or the conventional technologies of the present application, the drawings needed to be used in the description of the embodiments or the conventional technologies will be briefly introduced below, it is obvious that the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can be obtained according to the drawings without any creative efforts.
FIG. 1 is a schematic flow chart of a reliability evaluation method in one embodiment;
FIG. 2 is a graph illustrating degradation of collector-emitter saturation voltage drop performance in one embodiment;
FIG. 3 is a graph illustrating gate threshold voltage performance degradation in one embodiment;
FIG. 4 is a diagram illustrating the reliability evaluation results in different cases in another embodiment.
Detailed Description
To facilitate an understanding of the present application, the present application will now be described more fully with reference to the accompanying drawings. Embodiments of the present application are set forth in the accompanying drawings. This application may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used in the description of the present application herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the application.
As used herein, the singular forms "a", "an" and "the" may include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms "comprises/comprising," "includes" and "including," or "having," etc., specify the presence of stated features, integers, steps, operations, components, parts, or combinations thereof, but do not preclude the presence or addition of one or more other features, integers, steps, operations, components, parts, or combinations thereof.
At present, most of the traditional reliability evaluation methods based on a large number of failure samples do not consider the characteristics of high reliability, long service life, small samples, randomness in the performance degradation process, multi-performance parameter coupling competition and the like of products, and the requirements of reliability evaluation of multi-performance degradation products cannot be met. Recently, some scholars have proposed reliability assessment methods based on the wiener process. While such methods consider randomness in the process of performance degradation of highly reliable, long-lived products, most often only consider single-performance random degradation processes of products. Even if the random degradation process of the multivariate performance of the product is considered, the performance parameters are always assumed to be independent from each other at a small part. This makes the reliability evaluation result largely erroneous from the actual result. Therefore, in order to solve the above problems, the present application proposes a multivariate performance degradation product reliability evaluation method based on a wiener process and a connection function, a computer device, and a computer-readable storage medium.
The reliability evaluation method is a multivariate performance degradation product reliability evaluation method based on a wiener process and a connection function, and can realize reliability evaluation of products with high reliability, long service life, small samples, randomness in a performance degradation process and multi-performance parameter coupling competition.
In one embodiment, as shown in fig. 1, there is provided a reliability evaluation method including:
and step S1, judging whether the degradation process of each performance parameter of the product to be tested is subject to the wiener process.
Before the step of judging, firstly, test data of each performance parameter of the product to be tested is obtained, and then whether the degradation process of each performance parameter of the product to be tested obeys the wiener process or not is judged according to the test data of each performance parameter of the product to be tested.
Let x (t) be a standard wiener process variable with the following basic properties:
(1) x (0) ═ 0 and X (t) continues at t ═ 0.
(2) X (t) t ≧ 0 has stable independent increment, in any two disjoint time intervals [ t ≧ t1,t2]，[t3,t4]Inner X (t)4)-X(t3) And X (t)2)-X(t1) Are independent of each other.
(3) For each increment in the time interval, Δ X ═ X (t + Δ t) -X (t) obeys N (μ Δ t, σ)2Δt2) A normal distribution.
Suppose that a product to be tested has n performance parameters, wherein n is a positive integer greater than 1. Performance is continually degraded as product age increases. At the time of t, the degradation quantity X of the ith individual performance parameter of the product to be detectedi(t) is a random variable that obeys a certain distribution.
Due to the degradation X of the performance parameters of the product to be testedi(t) is a continuous variable. And at the time when t is 0, the amount of degradation is 0. Therefore, the degradation process of each performance parameter of the product to be tested meets the (1) th performance requirement of the wiener process variable. Meanwhile, X (t) t is more than or equal to 0 and has stable independent increment, and the increment is in any two non-intersected time intervals [ t ≧ t [1,t2]，[t3,t4]Inner X (t)4)-X(t3) And X (t)2)-X(t1) Are independent of each other. Therefore, the degradation process variables of the performance parameters of the product to be tested meet the (2) th performance requirement of the wiener process.
Therefore, the step determines whether the degradation process of each performance parameter of the product to be tested complies with the wiener process, that is, whether the increment of the degradation amount of each performance parameter in each time interval complies with the normal distribution. When the normal distribution is obeyed, the degradation process of each performance parameter of the product to be tested is obeyed the wiener process.
And step S2, if the degradation process of each performance parameter of the product to be tested obeys the wiener process, obtaining the edge distribution function of the combined distribution function corresponding to each performance parameter by using the wiener process.
When the degradation process of each performance parameter of the product to be tested obeys the wiener process, the degradation amount X of each performance of the product to be testedi(t) can be expressed as a wiener process variable, i.e.
Xi(t)＝μit+σiX0i(t)
In the formula: parameter(s)μiA drift parameter in the i-th performance parameter degradation amount; sigmaiFor diffusion parameters in the i-th performance parameter degradation quantity, X0i(t) is the standard wiener process for the ith performance parameter, and E [ X ]i(t)]＝μit， 
Suppose the amount of product performance degradation Xi(t) obeying some form of univariate or multivariate wiener process { Xi(t), t … 0}, and determiningThe false failure threshold is li(li>0) The product life T is n degradation amounts X by definition of degradation failurei(t) (i ═ 1,2, …, n) when the i-th performance parameter degradation reaches the failure threshold li first, that is, at the time corresponding to the failure threshold li
Ti＝inf{t|Xi(t)…li}
Considering a single performance parameter, its lifetime TiObey an inverse Gaussian distribution
TiRespectively as a distribution function and a probability density function of
The reliability of a single performance parameter of the product is expressed as:
a joint distribution is decomposed into k edge distributions and a join function that describes the correlation between variables. The join function is actually a type of function that joins the joint distribution functions with their respective edge distribution functions, and therefore, it is also referred to as a join function.
Let H be an n-dimensional distribution function whose edge distribution is F1(x1),F2(x2),…,Fn(xn) There is an n-dimensional linkage function C such that
H(x1,x2,…,xn)＝C(F1(x1),F2(x2),…,Fn(xn))
As long as a suitable connection function C is determined, the joint distribution function of the respective edge distribution functions can be solved with the connection function.
Lifetime T assuming degradation amount of ith individual performance parameter of product to be testediIndicates the life T1,T2,…,TnCan be expressed as H (t)1,t2,…,tn) According to the definition of the connection function, there is a unique connection function C, so that
H(t1,t2,…,tn)＝C(F1(t1),F2(t2),…,Fn(tn)；θ)
In which theta is the parameter estimate of the connection function, Fi(ti)＝Fi(T) lifetime T of i-th performance parameteriCorresponding edge distribution function.
Fi(t)＝1-Ri(t) as previously described, wherein Ri(t) is the reliability of a single performance parameter of the product, and the expression is as follows:
or, due to Fi(t) is the above-mentioned FTi(t) so that it may also pass
And (4) obtaining.
Setting a drift parameter mu in the i-th performance parameter degradation amount in the above expressioniDiffusion parameter sigma in ith performance parameter degradationiAre all edge function parameters.
At this time, specifically, the step may include:
and step S21, estimating the edge function parameters of each performance parameter according to the test data of each performance parameter by using the wiener process property to obtain the edge function parameter estimation value of each performance parameter.
Assuming that the total number of the products to be tested is M, and counting the performance of each productAccording to the K times of measurement, the time of each performance detection of the products in the same batch is the same, and the first detection is set as the initial time t0The subsequent detection time is tkK is 1,2, …, K. Wherein the ith performance parameter of the mth product is at the initial time t0The amount of degeneration of (A) is Xm,i(t0)＝0，Xm,i(tk) The ith performance parameter representing the mth product at time tkThe performance degradation amount of (c), the ith performance parameter of the mth product is in the time interval [ t ]k-1,tk]Increment of the amount of degradation of
ΔXm,i(tk)＝Xm,i(tk)-Xm,i(tk-1)
By nature of the wiener process
The probability density function of the increment of the performance degradation amount can be expressed as
Having a likelihood function of
Logarithm is taken on both sides of the above formula, and mu is respectively measuredi，σiCalculating the partial derivative, and estimating the parameters of the edge function
And step S22, obtaining the edge distribution function of the combined distribution function corresponding to each performance parameter according to the edge function parameter estimation value of each performance parameter.
After the edge function parameter estimation is performed on the test data of each performance parameter, the edge function parameter estimation value can be substituted into a correlation formula, so as to obtain an edge distribution function.
For example, the correlation formula is:
then, it is passed through Fi(t)＝1-Ri(t) the edge distribution function of the joint distribution function can be obtained.
Alternatively, the correlation formula is:
and Fi(t) is FTi(t) from which an edge distribution function of the joint distribution function can be derived.
And step S3, obtaining parameter estimated values of at least two connection functions of the joint distribution function of each performance parameter according to the edge distribution function corresponding to each performance parameter.
Using the edge distribution function F obtained in step S21(t1),F2(t2),…,Fn(tn) As an input in the connection function. c (F)1(t1),F2(t2),…,Fn(tn) (ii) a θ) is the probability density function of the multi-dimensional join function, and the likelihood function of the join function is:
the parameter theta of the connection function is obtained by derivation of the parameter theta
In this way, the parameter estimation values of at least two connection functions of the joint distribution function of each performance parameter can be obtained. The at least two join functions may include a frank join function, a cleton join function, a gaussian join function, and the like.
And step S4, determining the optimal connection function in the at least two connection functions according to the parameter estimation values of the at least two connection functions.
The determination of the optimal connection function is directly related to the accuracy of the correlation description of the multiple performance parameters. Through the determination of the optimal connection function, the reliability evaluation can be more accurately carried out.
And step S5, obtaining a reliability function of the product to be tested according to the edge distribution function corresponding to each performance parameter, the parameter estimation value of the optimal connection function and the optimal connection function.
Suppose that a product has n performance parameters and there is a certain correlation between the performance parameters. The track of product performance degradation at time t can be expressed as X (t) ═ X1(t),X2(t),…,Xn(t)), the corresponding degradation failure threshold is l ═ l (l)1,l2,…,ln). Meanwhile, assuming that n performance degradation amounts of the product can be described by a wiener process, the life of the product can be expressed as T ═ min (T ═ min)1,T2,…,Tn). The reliability of the product can thus be expressed as:
in the formula Cn(. cndot.) represents an n-dimensional connection function C, 2. ltoreq. k.ltoreq.n.
Substituting the edge distribution function corresponding to each performance parameter, the parameter estimation value of the optimal connection function and the optimal connection function to obtain the reliability function of the product to be tested.
The reliability evaluation method of the embodiment is a multivariate performance degradation product reliability evaluation method based on a wiener process and a connection function. The method considers the randomness of the degradation process of a plurality of performance parameters of the product, also considers the coupling competition relationship of the plurality of performance parameters of the product, and provides a general reliability evaluation model of the multi-performance degraded product, so that the reliability evaluation process is more practical, the evaluation result is more accurate, and the application range is wider.
In one embodiment, the step S4 (determining the optimal connection function among the at least two connection functions according to the parameter estimation values of the at least two connection functions) includes:
step S411, according to the parameter estimation values of at least two connection functions, obtaining the likelihood functions of at least two connection functions.
The likelihood function of the previous join function is:
estimating the parameter theta of each connection function
And substituting the expressions to obtain the likelihood function of each connection function.
Step S412, calculating the Chi information values of the at least two connection functions according to the likelihood functions of the at least two connection functions.
The expression of the erythropool information value a is as follows:
A＝-2lgL(θ)+2k
in the formula: k represents the number of parameters in the connection function, and lgL (θ) is the maximum value of the overall log-likelihood function.
And step S413, determining an optimal connection function according to the Chichi information value.
The Chi-cell information value A may effectively reflect the degree of fit of the model, indicating a higher degree of fit of the model if it is calculated to be smaller. Therefore, according to the method, the optimal connection function can be accurately determined, so that the reliability function of the product to be tested obtained in step S5 is more accurate.
Of course, the determination method of the optimal connection function is not limited in the present application, and may be obtained by other methods.
For example, in another embodiment, the step S4 (determining the optimal connection function among the at least two connection functions according to the parameter estimation values of the at least two connection functions) includes:
step S421, obtaining likelihood functions of at least two connection functions according to the parameter estimation values of the connection functions.
This step is the same as step S411 of the above embodiment.
Step S422, a maximum value of the likelihood functions of the at least two connection functions is calculated.
After the likelihood functions of the connection functions are obtained, the maximum value of the likelihood functions of the connection functions can be obtained through calculation.
Step S423, determining an optimal connection function according to the maximum value of the likelihood function.
The larger the maximum value of the likelihood function, the higher the degree of fitting of the model. Thus, an optimal connection function can be conveniently determined according to the method.
In one embodiment, before step S4 (determining the optimal connection function among the at least two connection functions according to the parameter estimation values of the at least two connection functions), the method further includes:
in step S01, a correlation coefficient between the degradation amounts of the respective performance parameters is calculated.
The solution formula of the correlation coefficient tau between the degradation amounts of the performance parameters is as follows:
in the formula: (x)i,yi) i is 1,2, …, K is a sample of the amount of degradation, and K is the number of measurements of the amount of degradation. sign (x) is a sign function when x>0, sign (x) 1; when x is 0, sign (x) is 0; when x is<0，sign(x)＝-1。
Step S02, at least two connection functions are obtained in which the difference between the correlation coefficient and the amount of degradation of each performance parameter is within a preset range.
There are at least tens of connection functions that connect the joint distribution function with their respective edge distribution functions. The calculation process is complicated by performing parameter estimation on all the connection functions to select the optimal connection function.
In this embodiment, at least two connection functions having a difference between the correlation coefficient and τ within a preset range are selected by comparing the correlation coefficient τ between the degradation amounts of the performance parameters with the correlation coefficients of the existing connection functions. The "preset range" is a range in which the correlation coefficient τ between the correlation coefficient of the connection function and the degradation amount of each performance parameter can reach a preset proximity, and may be set according to actual conditions.
In this embodiment, at least two connection functions that may be the optimal connection function are initially selected from the connection functions, so that the multiple connection functions can be comprehensively considered, and the selection and calculation process of the optimal connection function is not complicated.
Of course, the embodiments of the present application are not limited thereto, and in other embodiments, for example, the evaluator may directly select at least the connection function according to experience.
In one embodiment, the step S1 (determining whether the degradation process of each performance parameter of the product to be tested complies with the wiener process) includes: and obtaining a test result of multiple tests before failure of each performance parameter of a plurality of products to be tested according to a preset time interval.
The preset time interval can be set according to actual conditions. At the moment, the test data of each performance parameter of the product to be tested for reliability evaluation is more uniform and comprehensive, and the evaluation accuracy is further improved. The products to be tested can be products of the same batch, so that the performance consistency of the products to be tested is higher, and higher reliability performance evaluation accuracy is obtained. Of course, the present application is not limited thereto, and the plurality of products to be tested may not be the same batch of products.
In one embodiment, the step S1 (determining whether the degradation process of each performance parameter of the product to be tested complies with the wiener process) includes:
and step S11, carrying out normal goodness-of-fit test on the increment of the degradation quantity of each performance parameter of the sample to be tested in each time interval through the Kolmogorov-Similov test.
And step S12, judging whether the degradation process of each performance parameter of the product to be tested is compliant with the wiener process according to the test result of the normal goodness of fit.
The Kolmogorov-Schmilov test is mature data statistical software, and can conveniently obtain the normal fitting goodness test result of the increment of the degradation quantity of each performance parameter in each time interval, thereby conveniently judging whether the degradation process of each performance parameter of the product to be tested complies with the wiener process.
In one embodiment, the product to be tested is an Insulated Gate Bipolar Transistor (IGBT). The bipolar transistor can be a flexible direct current converter valve insulated gate bipolar transistor. Each performance parameter of the product to be tested comprises collector-emitter saturation voltage drop VCE(sat)And a gate threshold voltage VGE(th)。
Through analyzing the failure mode and failure mechanism of the IGBT, the main characteristic quantity of the IGBT is collector-emitter saturation voltage drop VCE(sat)And a gate threshold voltage VGE(th)The failure thresholds were 5% and 20%, respectively. Therefore, the present embodiment selects these two key performance parameters for performance degradation analysis.
The performance degradation amount of 20 IGBT samples (the specification is 1200V/75A) is tested according to the preset time interval, and the degradation trend of each performance parameter is respectively shown in FIG. 2 and FIG. 3.
Then, according to the characteristics of Gaussian increment in the wiener process, the collector-emitter saturation voltage drop V of the sample to be measuredCE(sat)And a gate threshold voltage VGE(th)And (4) carrying out normal fitting goodness test on the increment of the degradation quantity in each time interval so as to judge whether the degradation processes of the two performance parameters of the IGBT obey the wiener process.
This example uses Kolmogorov in the data statistics softwareThe test data of the performance parameters of the 20 samples were tested for a goodness of fit of normal with a 95% confidence by the smiloff test (K-S test). Collector-emitter saturation voltage drop V for all samplesCE(sat)And a gate threshold voltage VGE(th)And the data K-S inspection probability P values are all larger than 5%, and the inspection shows that the IGBT performance parameter degradation process can be modeled by using a wiener process.
And (3) respectively carrying out edge function parameter estimation on the two performance parameters by utilizing the properties of the wiener process to obtain edge function parameter estimation values of the two performance parameters, as shown in table 1.
TABLE 1
Substituting the obtained edge distribution parameter estimated value into a reliability expression of a single performance parameter of the product, and then passing through Fi(t)＝1-Ri(t) deriving an edge distribution function of the joint distribution function. Taking this as the input in the connection function, the parameter estimation value of the common connection function obtained by solving with the maximum likelihood estimation method is shown in table 2.
TABLE 2
Next, a frank join function is synthetically determined as an optimum join function by a join function determination method.
And obtaining the reliability function of the product to be tested according to the edge distribution function corresponding to each performance parameter, the parameter estimation value of the optimal connection function and the optimal connection function. Thus, a reliability function curve of the IGBT can be obtained as shown in fig. 4.
Fig. 4 shows a reliability function contrast curve of the evaluation result of the method of the present embodiment, the evaluation result of the single performance parameter, the evaluation result of the binary independent performance parameter, and the actual statistical condition. It can be seen that:
a) the reliability of the single performance parameter degradation model is significantly higher than that of the multi-element correlation performance degradation model (the model of the embodiment). For a complex product such as an IGBT, a plurality of monitored performance parameter indexes represent different degradation conditions, and a reliability model considering degradation of a single performance parameter may result in overestimation of reliability of the IGBT.
b) The reliability of the multivariate mutually independent degradation model is higher than that of the multivariate correlation performance degradation model (the model of the embodiment). Meanwhile, the reliability difference between the two increases along with the increase of time. And the deviation of the finally predicted reliability value from the actual value is larger along with the increase of time if the correlation between the two performance parameters is not considered in the use process of the IGBT.
c) The reliability evaluation result of the multivariate correlation performance degradation model (the model of the embodiment) is closest to the actual statistical result. The method of this embodiment evaluates the average number of cycles of the device before failure to be 76784, which is very small in error, only 2.77% from the actual statistical result of 74712 times.
In conclusion, the method provided by the embodiment is more reasonable and accurate.
It should be understood that, although the steps in the flowchart of fig. 1 are shown in order as indicated by the arrows, the steps are not necessarily performed in order as indicated by the arrows. The steps are not performed in a strict order unless explicitly stated herein, and may be performed in other orders. Moreover, at least a portion of the steps in fig. 1 may include multiple steps or multiple stages, which are not necessarily performed at the same time, but may be performed at different times, and the order of performing the steps or stages is not necessarily sequential, but may be performed alternately or in alternation with other steps or at least a portion of the steps or stages in other steps.
In one embodiment, a computer device is provided, comprising a memory and a processor, the memory having a computer program stored therein, the processor implementing the following steps when executing the computer program:
and step S1, judging whether the degradation process of each performance parameter of the product to be tested is subject to the wiener process.
And step S2, if the degradation process of each performance parameter of the product to be tested obeys the wiener process, obtaining the edge distribution function of the combined distribution function corresponding to each performance parameter by using the property of the wiener process.
And step S3, obtaining parameter estimated values of at least two connection functions of the joint distribution function of each performance parameter according to the edge distribution function corresponding to each performance parameter.
Step S4, determining an optimal connection function in the at least two connection functions according to the parameter estimation values of the at least two connection functions;
and step S5, obtaining a reliability function of the product to be tested according to the edge distribution function corresponding to each performance parameter, the parameter estimation value of the optimal connection function and the optimal connection function.
In one embodiment, the processor when executing step S4 includes implementing the steps of: step S411, according to the parameter estimation values of at least two connection functions, obtaining the likelihood functions of at least two connection functions. Step S412, calculating the Chi information values of the at least two connection functions according to the likelihood functions of the at least two connection functions. And step S413, determining an optimal connection function according to the Chichi information value.
In one embodiment, the processor when executing step S4 includes implementing the steps of: step S421, obtaining likelihood functions of at least two connection functions according to the parameter estimation values of the connection functions. Step S422 calculates the maximum value of the likelihood functions of the at least two connection functions. Step S423, determining an optimal connection function according to the maximum value of the likelihood function.
In one embodiment, a computer-readable storage medium is provided, having stored thereon a computer program which, when executed by a processor, performs the steps of:
and step S1, judging whether the degradation process of each performance parameter of the product to be tested is subject to the wiener process.
Step S2, if the degradation process of each performance parameter of the product to be tested obeys the wiener process, the edge distribution function of the combined distribution function corresponding to each performance parameter is obtained by using the property of the wiener process;
and step S3, obtaining parameter estimated values of at least two connection functions of the joint distribution function of each performance parameter according to the edge distribution function corresponding to each performance parameter.
Step S4, determining an optimal connection function in the at least two connection functions according to the parameter estimation values of the at least two connection functions;
and step S5, obtaining a reliability function of the product to be tested according to the edge distribution function corresponding to each performance parameter, the parameter estimation value of the optimal connection function and the optimal connection function.
In one embodiment, the processor when executing step S4 includes implementing the steps of: step S411, according to the parameter estimation values of at least two connection functions, obtaining the likelihood functions of at least two connection functions. Step S412, calculating the Chi information values of the at least two connection functions according to the likelihood functions of the at least two connection functions. And step S413, determining an optimal connection function according to the Chichi information value.
In one embodiment, the processor when executing step S4 includes implementing the steps of: step S421, obtaining likelihood functions of at least two connection functions according to the parameter estimation values of the connection functions. Step S422 calculates the maximum value of the likelihood functions of the at least two connection functions. Step S423, determining an optimal connection function according to the maximum value of the likelihood function.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by hardware related to instructions of a computer program, which can be stored in a non-volatile computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. Any reference to memory, storage, database, or other medium used in the embodiments provided herein can include at least one of non-volatile and volatile memory. Non-volatile Memory may include Read-Only Memory (ROM), magnetic tape, floppy disk, flash Memory, optical Memory, or the like. Volatile Memory can include Random Access Memory (RAM) or external cache Memory. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM), for example.
In the description herein, references to the description of "one embodiment," "another embodiment," or the like, mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, schematic descriptions of the above terms do not necessarily refer to the same embodiment or example.
The technical features of the above embodiments can be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the above embodiments are not described, but should be considered as the scope of the present specification as long as there is no contradiction between the combinations of the technical features.
The above examples only express several embodiments of the present application, and the description thereof is more specific and detailed, but not construed as limiting the scope of the invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the concept of the present application, which falls within the scope of protection of the present application. Therefore, the protection scope of the present patent shall be subject to the appended claims.
Claims (10)
1. A reliability evaluation method is characterized by comprising the following steps:
judging whether the degradation process of each performance parameter of the product to be detected obeys the wiener process or not;
if the degradation process of each performance parameter of the product to be tested obeys the wiener process, obtaining the edge distribution function of the joint distribution function corresponding to each performance parameter by using the property of the wiener process;
obtaining parameter estimation values of at least two connection functions of the joint distribution function of each performance parameter according to the edge distribution function corresponding to each performance parameter;
determining an optimal connection function in the at least two connection functions according to the parameter estimation values of the at least two connection functions;
obtaining a reliability function of a product to be tested according to the edge distribution function corresponding to each performance parameter, the parameter estimation value of the optimal connection function and the optimal connection function;
before determining the optimal connection function in the at least two connection functions according to the parameter estimation values of the at least two connection functions, the method further includes:
calculating a correlation coefficient between the degradation amounts of the performance parameters;
and acquiring at least two connection functions of which the difference between the correlation coefficient and the degradation quantity of each performance parameter is within a preset range.
2. The reliability evaluation method according to claim 1, wherein the determining an optimal connection function among the at least two connection functions according to the parameter estimation values of the at least two connection functions comprises:
obtaining likelihood functions of the at least two connection functions according to the parameter estimation values of the at least two connection functions;
calculating the Chi information values of the at least two connection functions according to the likelihood functions of the at least two connection functions;
and determining an optimal connection function according to the Chi information value.
3. The reliability evaluation method according to claim 1, wherein the determining an optimal connection function among the at least two connection functions according to the parameter estimation values of the at least two connection functions comprises:
obtaining likelihood functions of the at least two connection functions according to the parameter estimation values of the at least two connection functions;
calculating a maximum of the likelihood functions of the at least two connection functions;
and determining an optimal connection function according to the maximum value of the likelihood function.
4. The reliability evaluation method according to claim 1, wherein the obtaining of the edge distribution function of the joint distribution function corresponding to each performance parameter by using the wiener process property includes:
estimating the edge function parameters of each performance parameter according to the test data of each performance parameter by using the property of the wiener process to obtain the edge function parameter estimation value of each performance parameter;
and obtaining the edge distribution function of the joint distribution function corresponding to each performance parameter according to the edge function parameter estimation value of each performance parameter.
5. The reliability evaluation method according to claim 1, wherein the at least two connection functions include a frank connection function, a cleton connection function, and a gaussian connection function.
6. The reliability evaluation method according to claim 1, wherein before the determining whether the degradation process of each performance parameter of the product to be tested is subject to the wiener process, the method comprises:
and obtaining a test result of multiple tests before failure of each performance parameter of a plurality of products to be tested according to a preset time interval.
7. The reliability evaluation method according to claim 1, wherein the judging whether the degradation process of each performance parameter of the product to be tested complies with the wiener process comprises:
carrying out normal goodness-of-fit test on the increment of the degradation quantity of each performance parameter of the product to be tested in each time interval through the Kolmogorov-Similov test;
and judging whether the degradation process of each performance parameter of the product to be tested is compliant with the wiener process or not according to the test result of the normal goodness of fit.
8. The reliability evaluation method according to claim 1, wherein the product under test is an insulated gate bipolar transistor, and the performance parameters of the product under test include collector-emitter saturation voltage drop and gate threshold voltage.
9. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor, when executing the computer program, implements the steps of the method of any of claims 1 to 8.
10. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method of any one of claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010468980.3A CN111814301B (en) | 2020-05-28 | 2020-05-28 | Reliability evaluation method, computer device, and computer-readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010468980.3A CN111814301B (en) | 2020-05-28 | 2020-05-28 | Reliability evaluation method, computer device, and computer-readable storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111814301A CN111814301A (en) | 2020-10-23 |
| CN111814301B true CN111814301B (en) | 2021-07-27 |
Family
ID=72847804
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010468980.3A Active CN111814301B (en) | 2020-05-28 | 2020-05-28 | Reliability evaluation method, computer device, and computer-readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111814301B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115809569B (en) * | 2023-02-01 | 2023-06-20 | 中国电子产品可靠性与环境试验研究所（（工业和信息化部电子第五研究所）（中国赛宝实验室）） | Reliability evaluation method and device based on coupling competition failure model |
| CN115841046B (en) * | 2023-02-10 | 2023-06-20 | 中国电子产品可靠性与环境试验研究所（（工业和信息化部电子第五研究所）（中国赛宝实验室）） | Accelerated degradation test data processing method and device based on wiener process |
| CN116108697B (en) * | 2023-04-04 | 2023-08-04 | 中国电子产品可靠性与环境试验研究所（（工业和信息化部电子第五研究所）（中国赛宝实验室）） | Acceleration test data processing method, device and equipment based on multiple performance degradation |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107506337A (en) * | 2017-10-12 | 2017-12-22 | 中国人民解放军海军航空工程学院 | Reliability statistics estimating method based on polynary acceleration degraded data |
| 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 |
| CN108920341A (en) * | 2018-04-26 | 2018-11-30 | 航天东方红卫星有限公司 | A kind of moonlet and its constellation availability appraisal procedure based on Monte Carlo simulation |
| CN110895625A (en) * | 2018-09-11 | 2020-03-20 | 湖南银杏可靠性技术研究所有限公司 | Method for simulating reliability confidence interval estimation value of performance degradation product |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7203730B1 (en) * | 2001-02-13 | 2007-04-10 | Network Appliance, Inc. | Method and apparatus for identifying storage devices |
-
2020
- 2020-05-28 CN CN202010468980.3A patent/CN111814301B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107506337A (en) * | 2017-10-12 | 2017-12-22 | 中国人民解放军海军航空工程学院 | Reliability statistics estimating method based on polynary acceleration degraded data |
| CN108920341A (en) * | 2018-04-26 | 2018-11-30 | 航天东方红卫星有限公司 | A kind of moonlet and its constellation availability appraisal procedure based on Monte Carlo simulation |
| 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 |
| CN110895625A (en) * | 2018-09-11 | 2020-03-20 | 湖南银杏可靠性技术研究所有限公司 | Method for simulating reliability confidence interval estimation value of performance degradation product |
Non-Patent Citations (1)
| Title |
|---|
| 基于维纳过程和蒙特卡洛法的多元性能退化产品可靠性评估;潘广泽 等;《环境技术增刊》;20190930;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111814301A (en) | 2020-10-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111814301B (en) | Reliability evaluation method, computer device, and computer-readable storage medium | |
| US10496515B2 (en) | Abnormality detection apparatus, abnormality detection method, and non-transitory computer readable medium | |
| Carvalho et al. | The horseshoe estimator for sparse signals | |
| CN109657937B (en) | Product reliability evaluation and service life prediction method based on degradation data | |
| Nguyen et al. | Model selection for degradation modeling and prognosis with health monitoring data | |
| US20190261204A1 (en) | Method and system for abnormal value detection in lte network | |
| US20060074828A1 (en) | Methods and apparatus for detecting temporal process variation and for managing and predicting performance of automatic classifiers | |
| WO2006014509A2 (en) | Quantitative pcr data analysis system (qdas) | |
| US7373332B2 (en) | Methods and apparatus for detecting temporal process variation and for managing and predicting performance of automatic classifiers | |
| WO2006014464A2 (en) | Method for quantitative pcr data analysis system (qdas) | |
| US11099105B2 (en) | Data processing method, data processing device, and non-transitory computer-readable recording medium having recorded thereon data processing program | |
| Schubert et al. | Metadetect: Uncertainty quantification and prediction quality estimates for object detection | |
| CN113484813A (en) | Intelligent ammeter fault rate estimation method and system under multi-environment stress | |
| Fryzlewicz | Robust narrowest significance pursuit: Inference for multiple change-points in the median | |
| US20180300453A1 (en) | Estimation of Descriptive Parameters from a Sample | |
| Rancoita et al. | Bayesian DNA copy number analysis | |
| JP3681683B2 (en) | Insulating film lifetime estimation method and semiconductor device management method | |
| Tang et al. | Nonparametric goodness-of-fit tests for uniform stochastic ordering | |
| Evans et al. | Measuring statistical evidence and multiple testing | |
| Manna | Small Sample Estimation of Classification Metrics | |
| CN114020905A (en) | Text classification external distribution sample detection method, device, medium and equipment | |
| CN115964211A (en) | Root cause positioning method, device, equipment and readable medium | |
| CN112149833B (en) | Prediction method, device, equipment and storage medium based on machine learning | |
| US20060074827A1 (en) | Methods and apparatus for detecting temporal process variation and for managing and predicting performance of automatic classifiers | |
| Betta et al. | Uncertainty evaluation in algorithms with conditional statement |
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 |