Relative reliability quantitative evaluation method and equipment based on interference theory
Through the HALT test ultimate stress model and K-S test method based on interference theory, the problems of stress distribution type selection and reliability evaluation accuracy in HALT test are solved, and effective quantitative evaluation of product reliability is achieved, suitable for small samples occasion.
Patent Information
- Application Number
- CN202410136838.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-01-31
- Publication Date
- 2024-11-15
- Estimated Expiration
- 2044-01-31
AI Technical Summary
The lack of effective reliability quantitative evaluation methods in the prior art to utilize limited data from HALT tests, resulting in the inability to accurately evaluate the reliability index of the product, and the selection of stress distribution types is unreasonable. The quantitative evaluation method of accelerated tests is directly applied to HALT tests. The accuracy of reliability assessment cannot be guaranteed.
Based on the interference theory, an interference model for the ultimate stress tolerance of old and new products is constructed. The stress distribution type is determined by the K-S test method, and the interference model is used for quantitative evaluation to obtain the relative reliability of the product.
It effectively solves the problem of insufficient HALT test data, provides a quantitative reliability evaluation method suitable for small sample occasions, improves the accuracy and universal applicability of the evaluation, and eliminates the reliability evaluation difficulties caused by small sample size.
Smart Images

Figure CN117910268B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of reliability evaluation, and in particular to a relative reliability quantitative evaluation method and equipment based on interference theory highly accelerated life test. Background Art
[0002] HALT test is Highly Accelerated Life Testing (HALT), which refers to the test sample being subjected to low temperature step stress test, high temperature step stress test, rapid temperature change cycle stress test, vibration step stress test, and comprehensive environmental stress test in sequence, and the test stress applied to the test sample is gradually increased to quickly expose potential design defects and determine the product's stress tolerance limit. HALT test is also known as reliability enhancement test. In recent years, it has been widely used in the reliability test of electronic equipment internationally, especially in the fields of communications, energy, aviation, and aerospace. Internationally, HALT test technology is relatively mature, and some standards on how to conduct HALT tests have been formed; however, the research on HALT test technology in China started late, and HALT test is in the initial application state of following foreign countries and typical products.
[0003] The HALT test can quickly stimulate product defects in a short period of time, improve the product's tolerance in the use environment, increase the product's working limit and damage limit, and shorten the product's development time. However, it is precisely because the HALT test time is short and the degradation test data obtained is limited that it is difficult to directly derive the product's reliability indicators by relying on a large amount of test data for analysis like traditional reliability tests. A major advantage of the HALT test is that it shortens the product development time. At the beginning of product design or when conducting tests, a large number of samples are often not invested in testing, and the small sample size is also a difficulty in reliability evaluation. So far, there is no mature HALT test reliability quantitative evaluation method at home and abroad that can directly derive product reliability indicators.
[0004] At present, there are few studies on the reliability quantitative evaluation method of HALT test, and more studies on the test method. There is no mature reliability quantitative evaluation method, and no reliability index for HALT test is given. However, there are many studies on the reliability quantitative evaluation method of products, and the mainstream direction is the reliability quantitative evaluation research of accelerated life test and accelerated degradation test. Some scholars also use stress intensity interference model to conduct quantitative analysis of reliability strengthening test, or directly use the accelerated test quantitative evaluation method in HALT test evaluation, but there are the following shortcomings:
[0005] ① Some studies have not considered that the distribution type of the stress tolerance in the HALT test may be different for different products, which makes the selection of stress distribution type in the evaluation method unreasonable.
[0006] ② The reliability assessment direction is not appropriate. The reliability assessment method used in other accelerated tests is used to directly conduct a quantitative assessment of the reliability of the HALT test. The mechanism of the HALT test is not consistent with that of the accelerated test, and the data of the HALT test is also relatively small. The quantitative assessment method of the accelerated test is directly applied to the HALT test assessment, and the accuracy of the reliability index cannot be judged.
[0007] Therefore, how to utilize the limited test data of HALT test, select appropriate test indicators, and consider the universal applicability of stress distribution to conduct quantitative evaluation of product HALT test reliability is a key issue that needs to be solved urgently. Summary of the invention
[0008] The purpose of the present invention is to provide a relative reliability quantitative evaluation method and device based on interference theory.
[0009] To achieve the above-mentioned object of the invention, the present invention provides a relative reliability quantitative evaluation method based on interference theory, comprising:
[0010] S1. Based on the interference theory of product HALT test tolerance limit stress, an interference model is constructed to characterize the tolerance limit stress of old products and the tolerance limit stress of new products;
[0011] S2. Determine the distribution type to which the product HALT test tolerance limit stress is subject, and obtain a distribution function corresponding to the distribution type;
[0012] S3. Substitute the distribution function into the interference model and conduct a HALT test to quantitatively evaluate the relative reliability of the product based on the interference theory.
[0013] According to one aspect of the present invention, in step S1, the interference theory based on the product HALT test tolerance limit stress is: the probability density function curve of the HALT test tolerance limit stress of the old product that is first subjected to the HALT test for performance improvement and the probability density function curve of the HALT test tolerance limit stress of the new product that is subjected to the HALT test again for performance improvement effect are cross-overlapped interference;
[0014] In step S1, in the step of constructing an interference model for characterizing the HALT test tolerance limit stress of the old product and the HALT test tolerance limit stress of the new product, the interference model is expressed as:
[0015]
[0016] Wherein, P represents the probability of product performance improvement in the entire stress range, δ represents the ultimate stress tolerance of the old product HALT test, S represents the ultimate stress tolerance of the new product HALT test, h(δ) represents the probability density function of the ultimate stress tolerance of the old product HALT test, and f(S) represents the probability density function of the ultimate stress tolerance of the new product HALT test.
[0017] According to one aspect of the present invention, in step S2, the step of determining the distribution type obeyed by the HALT test tolerance limit stress of the product is used to obtain the distribution type obeyed by the HALT test tolerance limit stress of the old product and the new product; wherein, based on the small sample test method, the distribution hypothesis and hypothesis test are performed on the HALT test tolerance limit stress of the old product samples and the new product samples respectively, so as to determine the distribution type of the old product and the new product as a whole.
[0018] According to one aspect of the present invention, in step S2, in the step of determining the distribution type obeyed by the HALT test tolerance limit stress of the product, the small sample inspection method adopts the KS inspection method, which includes:
[0019] S21. Assume the distribution type of the HALT test tolerance limit stress of the old product sample and the new product sample respectively, wherein the distribution type is one of normal distribution, lognormal distribution, exponential distribution, and Weibull distribution;
[0020] S22. Based on the assumed distribution type, obtain a first hypothesis H0 and a second hypothesis H1 that obey the distribution type;
[0021] S23. Calculate the test statistic D for the preset number of old product samples and new product samples respectively n , and based on the test statistic D n The first hypothesis H0 or the second hypothesis H1 is selected to complete the determination of the distribution type of the HALT test tolerance limit stress of the old product sample and the new product sample, wherein if D n ≤d n,α , then accept the first hypothesis H0, otherwise, accept the second hypothesis H1, d n,α It represents the critical value obtained by looking up the table under the preset confidence level α;
[0022] S24. Based on the determined distribution types of the HALT test tolerance limit stress of the old product samples and the new product samples, parameter estimation of the sample population is performed respectively, wherein the parameter estimation includes characteristic parameters such as the population mean and the population mean square error.
[0023] According to one aspect of the present invention, in step S3, the distribution function is introduced into the interference model, and a HALT test is performed to quantitatively evaluate the relative reliability of the product based on the interference theory. The relative reliability of the interference theory is expressed as:
[0024]
[0025] Wherein, R represents the relative reliability of the new product after performance improvement relative to the old product before HALT test, P represents the probability of product performance improvement in the entire stress range, δ represents the ultimate stress tolerance of the old product in HALT test, S represents the ultimate stress tolerance of the new product in HALT test, h(δ) represents the probability density function of the ultimate stress tolerance of the old product in HALT test, and f(S) represents the probability density function of the ultimate stress tolerance of the new product in HALT test.
[0026] To achieve the above-mentioned object of the invention, the present invention provides a device applied to the above-mentioned relative reliability quantitative evaluation method, comprising at least one processor, at least one memory and a data bus;
[0027] The processor and the memory communicate with each other via the data bus;
[0028] The memory stores program instructions that can be executed by the processor, and the processor calls the program instructions to execute the relative reliability quantitative evaluation method.
[0029] According to a scheme of the present invention, the present invention aims at the tolerance limit stress of old products and new products during the HALT test, and based on the interference theory, establishes a tolerance limit stress model of old products HALT test - new products HALT test, uses distribution hypothesis and hypothesis testing method to make the distribution of product tolerance limit stress more universally applicable, and conducts quantitative evaluation of the relative reliability of HALT test of new products relative to old products.
[0030] According to a solution of the present invention, the method performs distribution hypothesis test based on the KS test method, and the established relative reliability quantitative evaluation method is suitable for small sample occasions, which effectively eliminates the disadvantage that the reliability evaluation cannot be performed due to the small sample size of the HALT test.
[0031] According to one solution of the present invention, the relative reliability evaluation method established by analyzing the distribution of tolerance limit stress during the HALT test solves the problem of applicability and effectiveness of traditional reliability evaluation methods in the HALT test evaluation process.
[0032] According to a solution of the present invention, the present invention effectively solves the problem that the HALT test cannot be directly quantitatively evaluated due to insufficient test data, and the present invention can effectively evaluate the reliability of the product HALT test without the need for a large number of samples and a large amount of test data. BRIEF DESCRIPTION OF THE DRAWINGS
[0033] Figure 1 is a flowchart schematically showing a relative reliability quantitative evaluation method according to an embodiment of the present invention;
[0034] Figure 2 is a diagram schematically showing a change process of the tolerance limit stress of a product according to an embodiment of the present invention after a HALT test;
[0035] Figure 3 It is a schematic diagram of an interference model of an old product HALT test tolerance limit stress-new product HALT test tolerance limit stress according to an embodiment of the present invention. DETAILED DESCRIPTION
[0036] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.
[0037] When describing the embodiments of the present invention, the orientation or positional relationship expressed by the terms "longitudinal", "lateral", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside" and "outside" are based on the orientation or positional relationship shown in the relevant drawings and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the referred device or element must have a specific orientation, be constructed and operate in a specific orientation. Therefore, the above terms should not be understood as limiting the present invention.
[0038] The present invention will be described in detail below in conjunction with the accompanying drawings and specific embodiments. The embodiments cannot be described one by one here, but the embodiments of the present invention are not therefore limited to the following embodiments.
[0039] like Figure 1 As shown, according to one embodiment of the present invention, a relative reliability quantitative evaluation method based on interference theory of the present invention includes:
[0040] S1. Based on the interference theory of product HALT test tolerance limit stress, an interference model is constructed to characterize the tolerance limit stress of old products and the tolerance limit stress of new products;
[0041] S2. Determine the distribution type of the product HALT test tolerance limit stress and obtain the distribution function corresponding to the distribution type;
[0042] S3. Substitute the distribution function into the interference model and conduct HALT test to quantitatively evaluate the product to obtain the relative reliability of the product based on interference theory.
[0043] A HALT test is performed on a group of old products to obtain new products that meet the preset design goals. The new products are subjected to a HALT test again to verify whether they meet the design goals. The tolerance limit stress of the product during the HALT test is collected. Assume that the tolerance limit stress collected from m product samples is X i , i=1,2,…,n.
[0044] After the product undergoes the HALT test, the tolerance limit stress is increased and a new product is obtained. The change process is as follows: Figure 2 As shown in the figure. The probability density function curve f(S1) of the HALT test tolerance limit stress of the old product at a certain moment is as follows Figure 2 As shown in the left figure, after the HALT test, a new product is obtained. The tolerance limit stress of the new product will increase. The probability density function curve of the new and old products is as follows: Figure 2 As shown on the right, the new product probability density curve f(S2) is as follows Figure 2 Shown in the upper right portion.
[0045] The condition for improving product performance is that the tolerance limit stress of the new product after the HALT test (set as S) is greater than the tolerance limit stress of the old product (set as δ). However, even if the mean value of the working limit stress S of the new product is greater than the mean value of the tolerance limit stress δ of the old product, it is possible that the working limit stress S of individual products is less than the tolerance limit stress δ of the old product, that is, interference superposition occurs and product performance is not improved.
[0046] Furthermore, in step S1, the interference theory based on the product HALT test tolerance limit stress is: the probability density function curve of the HALT test tolerance limit stress of the old product that is subjected to the HALT test for the first time for performance improvement (i.e., the old product HALT test tolerance limit stress) and the probability density function curve of the HALT test tolerance limit stress of the new product that is subjected to the HALT test again for verification of the performance improvement effect (i.e., the new product HALT test tolerance limit stress) are cross-overlapped interference; wherein, the old product HALT test tolerance limit stress - new product HALT test tolerance limit stress interference model schematic diagram is as follows Figure 3 shown.
[0047] Further, such as Figure 3 As shown in the figure, the HALT test tolerance limit stress δ1 of the old product falls on a width of d δThe probability of the interval is equal to the unit area A1 determined by the interval, that is:
[0048]
[0049] Among them, h(δ) represents the probability density function of the endurance limit stress of the old product HALT test;
[0050] Then, the probability that the HALT test tolerance limit stress S of the new product is greater than the HALT test tolerance limit stress δ1 of the old product is:
[0051]
[0052] Where, f(S) represents the probability density curve of the HALT test tolerance limit stress of the new product, and A2 represents the area enclosed by the probability density curve of the HALT test tolerance limit stress of the product;
[0053] Furthermore, the tolerance limit stress of the old product and the tolerance limit stress of the new product are two independent random events. According to the "multiplication theorem" of the probability of two independent events, the probability of simultaneous occurrence is equal to the product of the probabilities of the two events occurring separately. This probability is the probability that the product can work normally, that is:
[0054]
[0055] Wherein, h(δ1) represents the value of the probability density function of the endurance limit stress of the old product HALT test at δ1.
[0056] In the entire HALT test tolerance limit stress δ interval of the old product, the probability of product performance improvement is the relative reliability under the interference model of the HALT test tolerance limit stress of the old product - the HALT test tolerance limit stress of the new product. Therefore, in step S1, in the step of constructing an interference model for characterizing the HALT test tolerance limit stress of the old product - the HALT test tolerance limit stress of the new product, the interference model is expressed as:
[0057]
[0058] Wherein, P represents the probability of product performance improvement in the entire stress range, δ represents the HALT test tolerance limit stress of the old product, S represents the HALT test tolerance limit stress of the new product, h(δ) represents the probability density function of the HALT test tolerance limit stress of the old product, and f(S) represents the probability density function of the HALT test tolerance limit working stress of the new product.
[0059] According to one embodiment of the present invention, in step S2, the step of determining the distribution type obeyed by the product HALT test tolerance limit stress is used to obtain the distribution type obeyed by the old product and the new product HALT test tolerance limit stress; wherein, based on the small sample test method, the distribution hypothesis and hypothesis test are respectively performed on the HALT test tolerance limit stress of the old product and the new product samples to determine the distribution type of the old product and the new product as a whole.
[0060] According to an embodiment of the present invention, in step S2, in the step of determining the distribution type obeyed by the product HALT test tolerance limit stress, the small sample inspection method adopts the KS inspection method, which includes:
[0061] S21. Make distribution assumptions for the distribution types of the HALT test tolerance limit stress of old product samples and new product samples, respectively, where the distribution type of the samples is one of the normal distribution, lognormal distribution, exponential distribution, Weibull distribution, etc.; specifically,
[0062] 1) Normal distribution
[0063] When the HALT test tolerance limit stress of old product samples and new product samples obeys the normal distribution X~N(μ,σ), the probability density function is:
[0064]
[0065] Where μ is the mean of the normal distribution and σ is the standard deviation of the normal distribution.
[0066] 2) Lognormal distribution
[0067] The HALT test tolerance limit stress of old product samples and new product samples follows the logarithmic normal distribution lnX~N(μ,σ 2 ), its probability density function is:
[0068]
[0069] Where μ is the mean of the lognormal distribution and σ is the standard deviation of the lognormal distribution.
[0070] 3) Exponential distribution
[0071] When the HALT test tolerance limit stress of old product samples and new product samples obeys exponential distribution, the probability density function is:
[0072]
[0073] Where λ is the exponential parameter of the exponential distribution.
[0074] 4) Weibull distribution
[0075] When the HALT test tolerance limit stress of old product samples and new product samples obeys Weibull distribution, the probability density function is:
[0076]
[0077] Where γ is the scale parameter of the Weibull distribution and k is the shape parameter of the Weibull distribution.
[0078] S22. Based on the assumed distribution type, obtain the first hypothesis H0 and the second hypothesis H1 that obey the distribution type; specifically, based on the assumed distribution types of the old product and new product samples, obtain the first hypothesis H0 and the second hypothesis H1 that the old product and the new product population obey the distribution type respectively; in this embodiment, the KS test method (Kolmogorov-Smirnov test method), also known as the D test method, is used to test the fit between the sample assumed distribution of the collected data and the actual distribution of the population, which includes:
[0079] Assume that the distribution function of the simple random sample x (i.e. the HALT test tolerance limit stress of the old product sample and the new product sample) from the population X is F n (x), yes, arranged from small to large according to the sample data:
[0080] x (1) ≤x (2) ≤…≤x (n) ;
[0081] Furthermore, we can get the distribution function F n (x):
[0082]
[0083] Thus, the first hypothesis H0 and the second hypothesis H1 of the type of distribution of the population can be obtained based on the assumptions;
[0084] First hypothesis H0: The distribution function F(x) of the HALT test tolerance limit stress of the overall product of this type determined in step S2 is different from the distribution function F(x) of the HALT test tolerance limit stress of the sample used for the test. n (x) are equal, and F n (x) is also equal to the distribution function F0(x) of a hypothetical continuous distribution (such as normal distribution, lognormal distribution, exponential distribution, Weibull distribution, etc.), that is, the distribution type of the sample hypothesis is the same as the distribution type of the population. The first hypothesis H0 can be expressed as:
[0085] H0:F(x)=F n (x) = F0(x)
[0086] Second hypothesis H1: The distribution function F(x) of the HALT test tolerance limit stress of the overall product of this type determined in step S2 is different from the distribution function F(x) of the HALT test tolerance limit stress of the sample used for the test. n (x) are equal, but F n (x) is not equal to the distribution function F0(x) of a certain assumed continuous distribution (such as normal distribution, lognormal distribution, exponential distribution, Weibull distribution, etc.). The distribution type of the sample assumption is different from the distribution type of the population. The second assumption H1 can be expressed as:
[0087] H1:F(x)=F n (x)≠F0(x)
[0088] S23. Calculate the test statistic D for the preset number of old product samples and new product samples respectively n , and based on the test statistic D n The first hypothesis distribution H0 or the second hypothesis distribution H1 is selected to complete the determination of the distribution type of the HALT test tolerance limit stress of the old product sample and the new product sample. In this embodiment, the statistic D n It is expressed as:
[0089]
[0090] Furthermore, when the first hypothesis H0 is established, for a given number of stresses n (i.e., the number of HALT test tolerance limit stresses of the product used for the test), the statistic D can be obtained: n The exact distribution of and the limiting distribution when n→∞. Then:
[0091] D n =max{|F0(x i )-F n (x i )|,|F0(x i )-F n (x i-1 )|}, (i=1,…,n)
[0092] Among them, F0(x i ) represents the HALT test tolerance limit stress of the old product samples and the new product samples used for the test under a certain assumed continuous distribution x i The function value at n (x i ) represents the HALT test tolerance limit stress x of the old product samples and the new product samples i The distribution function F n (x) function value;
[0093] Therefore, it can be further expressed as:
[0094]
[0095] Furthermore, based on the statistic D n Test the results. Specifically, if D n ≤d n,α , then accept the first hypothesis H0, otherwise, accept the second hypothesis H1, d n,α It represents the critical value obtained by looking up the table under the preset confidence level α;
[0096] S24. Based on the distribution types of the HALT test tolerance limit stress of the old product samples and the new product samples, the parameters of the sample population are estimated respectively, wherein the parameters include characteristic parameters such as the population mean and the population mean square error. In this embodiment, if the distribution types of the old product samples and the new product samples determined in the above steps are different, the parameter estimation methods used are not completely the same. Generally, the parameter estimation methods such as moment method estimation, least squares method estimation, likelihood estimation, Bayesian estimation, etc. can be used.
[0097] In this embodiment, the maximum likelihood estimation (MLE) can be used to estimate the general mean point of the population mean μ To estimate, accordingly, the minimum variance unbiased estimation (Minimum Variance Unbiased Estimation: MVUE) can be used to estimate the mean square error of the population mean square error σ Make an estimate; then, the point estimate of the mean of the population mean μ is:
[0098]
[0099] in, is the sample mean, which means that in the maximum likelihood estimation, the population mean is equal to the sample mean.
[0100] The point estimate of the population mean square error σ is:
[0101]
[0102] in, is the mean square error of the sample, which means that the population mean square error is equal to the sample mean square error.
[0103] like Figure 1 As shown, according to an embodiment of the present invention, in step S3, the step of bringing the distribution function into the interference model and performing a HALT test to quantitatively evaluate and obtain the relative reliability of the product based on the interference theory includes:
[0104] S31. Based on the interference model of the HALT test tolerance limit stress of old products and the HALT test tolerance limit stress of new products, a general relative reliability formula for new products relative to old products is constructed. The general relative reliability formula is expressed as:
[0105]
[0106] Wherein, R represents the relative reliability of the new product after performance improvement relative to the old product before HALT test, P represents the probability of product performance improvement in the entire stress range, δ represents the ultimate stress tolerance of the old product in HALT test, S represents the ultimate stress tolerance of the new product in HALT test, h(δ) represents the probability density function of the ultimate stress tolerance of the old product in HALT test, and f(S) represents the probability density function of the ultimate stress tolerance of the new product in HALT test.
[0107] S32. Based on the general relative reliability formula, the distribution functions are respectively substituted to obtain the relative reliability of the new product relative to the old product under the distribution type.
[0108] (1) In this embodiment, if the distribution type of the HALT test tolerance limit stress of the old product and the new product determined in the previous step is a normal distribution, then the sum of the normal distribution is still a normal distribution. Let δ~N(μ δ , σ δ 2 )、S~N(μ S , σ S 2 ), the growth margin of the product HALT test is Z, then:
[0109] Z=S-δ>0
[0110] Moreover, the growth margin Z also conforms to the normal distribution Z~N(μ Z , σ Z 2 ); the probability density function of Z is:
[0111]
[0112] Among them, σ Z 2 =σ δ 2 +σ S 2 , μ Z =μ S -μ δ ;
[0113] Thus, the expression of the interference model that obeys the general normal distribution can be obtained, which is expressed as:
[0114]
[0115] The relative reliability formula of the new product relative to the old product based on the interference model of the old product HALT test tolerance limit stress and the new product HALT test tolerance limit stress can be expressed as:
[0116]
[0117] Among them, σ Z 2 =σ δ 2 +σ S 2 , μ Z =μ S -μ δ ; σ δ is the standard deviation of the ultimate stress of the HALT test of the old product samples, μ δ is the mean value of the ultimate stress of the HALT test of the old product samples; σ S is the standard deviation of the ultimate stress of the HALT test of new product samples, μ S It is the average value of the ultimate stress of the HALT test of new product samples.
[0118] (2) In this embodiment, if the distribution type of the HALT test tolerance limit stress of the old product and the new product determined in the above steps is a lognormal distribution, an expression of the interference model that obeys the general lognormal distribution can be obtained, which is expressed as:
[0119]
[0120] σ δ is the standard deviation of the ultimate stress of the HALT test of the old product samples, μ δ is the mean value of the ultimate stress of the HALT test of the old product samples; σ S is the standard deviation of the ultimate stress of the HALT test of new product samples, μ S It is the average value of the ultimate stress of the HALT test of new product samples.
[0121] The relative reliability formula of the new product relative to the old product based on the interference model of the old product HALT test tolerance limit stress and the new product HALT test tolerance limit stress can be expressed as:
[0122]
[0123] (3) In this embodiment, if the distribution type of the HALT test tolerance limit stress of the old product and the new product determined in the above steps is an exponential distribution, the expression of the exponential distribution of the interference model can be obtained, which is expressed as:
[0124]
[0125] λ1 is the exponential parameter of the ultimate stress tolerance of the HALT test of the old product sample, and λ2 is the exponential parameter of the ultimate stress tolerance of the HALT test of the new product sample.
[0126] The relative reliability formula of the new product relative to the old product based on the interference model of the old product HALT test tolerance limit stress and the new product HALT test tolerance limit stress can be expressed as:
[0127]
[0128] (4) In this embodiment, if the distribution type of the HALT test tolerance limit stress of the old product and the new product determined in the above steps is Weibull distribution, the expression of Weibull distribution of the interference model can be obtained, which is expressed as:
[0129]
[0130] γ1 is the proportional parameter of the ultimate stress tolerance of the HALT test of the old product sample, and k1 is the shape parameter of the ultimate stress tolerance of the HALT test of the old product sample; γ2 is the proportional parameter of the ultimate stress tolerance of the HALT test of the new product sample, and k2 is the shape parameter of the ultimate stress tolerance of the HALT test of the new product sample.
[0131] The relative reliability formula of the new product relative to the old product based on the interference model of the old product HALT test tolerance limit stress and the new product HALT test tolerance limit stress can be expressed as:
[0132]
[0133] According to one embodiment of the present invention, the present invention provides a device applied to the aforementioned relative reliability quantitative evaluation method, including at least one processor, at least one memory and a data bus; in this embodiment, the processor and the memory communicate with each other through the data bus; the memory stores program instructions that can be executed by the processor, and the processor calls the program instructions to execute the relative reliability quantitative evaluation method.
[0134] In this embodiment, the memory can be, but is not limited to, a random access memory (RAM), a read only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable read-only memory (EEPROM), etc.
[0135] In this embodiment, the processor can be an integrated circuit chip with signal processing capabilities. The processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.
[0136] The above contents are merely examples of specific solutions of the present invention. For devices and structures not described in detail therein, it should be understood that they can be implemented by adopting general devices and general methods available in the art.
[0137] The above is only one solution of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.
Claims
1. A relative reliability quantitative evaluation method based on interference theory, characterized in that: include: S1. Based on the interference theory of the product HALT test tolerance limit stress, an interference model is constructed to characterize the tolerance limit stress of the old product and the tolerance limit stress of the new product; wherein the interference model is expressed as: Wherein, P represents the probability of product performance improvement in the entire stress range, δ represents the HALT test tolerance limit stress of the old product, S represents the HALT test tolerance limit stress of the new product, h(δ) represents the probability density function of the HALT test tolerance limit stress of the old product, and f(S) represents the probability density function of the HALT test tolerance limit stress of the new product; S2. Determine the distribution type obeyed by the HALT test tolerance limit stress of the product, and obtain the distribution function corresponding to the distribution type; wherein, it is used to obtain the distribution type obeyed by the HALT test tolerance limit stress of the old product and the new product; wherein, based on the small sample test method, the distribution hypothesis and hypothesis test are respectively performed on the HALT test tolerance limit stress of the old product sample and the new product sample to determine the distribution type of the old product and the new product as a whole; the small sample test method adopts the KS test method, and step S2 includes: S21. Assume the distribution type of the HALT test tolerance limit stress of the old product sample and the new product sample respectively, wherein the distribution type is one of normal distribution, lognormal distribution, exponential distribution, and Weibull distribution; S22. Based on the assumed distribution type, obtain a first hypothesis H0 and a second hypothesis H1 that obey the distribution type; S23. Calculate the test statistic D for the preset number of old product samples and new product samples respectively n , and based on the test statistic D n The first hypothesis H0 or the second hypothesis H1 is selected to complete the determination of the distribution type of the HALT test tolerance limit stress of the old product sample and the new product sample, wherein if D n ≤d n,α , then accept the first hypothesis H0, otherwise, accept the second hypothesis H1, d n,α It represents the critical value obtained by looking up the table under the preset confidence level α; among them, when the first hypothesis H0 is established, for the given number n of HALT test tolerance limit stress of the product used for the test, the statistic D can be obtained n The exact distribution of and the limiting distribution when n→∞, then: D n =max{|F0(x i )-F n (x i )|,|F0(x i )-F n (x i-1 )|},(i=1,…,n), or, Among them, F0(x i ) represents the HALT test tolerance limit stress of the old product samples and the new product samples used for the test under a certain assumed continuous distribution x i The function value at F n (x i ) represents the HALT test tolerance limit stress x of the old product samples and the new product samples i The distribution function F n (x) function value; S24. Based on the determined distribution types of the HALT test tolerance limit stress of the old product sample and the new product sample, parameter estimation of the sample population is performed respectively, wherein the parameter estimation includes: a population mean and a population mean square error; S3. Substitute the distribution function into the interference model and conduct a HALT test to quantitatively evaluate the relative reliability of the product based on the interference theory.
2. The relative reliability quantitative evaluation method according to claim 1, characterized in that: In step S1, the interference theory based on the product HALT test tolerance limit stress is: the probability density function curve of the HALT test tolerance limit stress of the old product that is subjected to the HALT test for the first time for performance improvement and the probability density function curve of the HALT test tolerance limit stress of the new product that is subjected to the HALT test again for verification of the performance improvement effect are cross-overlapping interference.
3. The relative reliability quantitative evaluation method according to claim 2, characterized in that: In step S3, the distribution function is introduced into the interference model, and a HALT test is performed to quantitatively evaluate the relative reliability of the product based on the interference theory. The relative reliability of the interference theory is expressed as: Wherein, R represents the relative reliability of the new product after performance improvement relative to the old product before HALT test, P represents the probability of product performance improvement in the entire stress range, δ represents the ultimate stress tolerance of the old product in HALT test, S represents the ultimate stress tolerance of the new product in HALT test, h(δ) represents the probability density function of the ultimate stress tolerance of the old product in HALT test, and f(S) represents the probability density function of the ultimate stress tolerance of the new product in HALT test.
4. A device applied to the relative reliability quantitative evaluation method according to any one of claims 1 to 3, characterized in that: comprising at least one processor, at least one memory and a data bus; The processor and the memory communicate with each other via the data bus; The memory stores program instructions that can be executed by the processor, and the processor calls the program instructions to execute the relative reliability quantitative evaluation method.
