A Design Method for Uncertain Acceptance Sampling Schemes Based on Performance Degradation

Through the uncertain acceptance sampling scheme based on performance degradation, the product's performance degradation process model and risk model are established, and the problem of low accuracy of small sample acceptance sampling scheme is solved, and high accuracy acceptance under small sample conditions is achieved, reducing the risk.

CN118780685BActive Publication Date: 2024-11-19BEIHANG UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411252296.6
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-09-09
Publication Date
2024-11-19
Estimated Expiration
2044-09-09

AI Technical Summary

Technical Problem

The existing small sample acceptance sampling plan cannot accurately measure cognitive uncertainty due to the probability theory, resulting in low accuracy of the acceptance sampling plan, which increases the risks of the producer and the user.

Method used

A uncertain acceptance sampling scheme based on performance degradation was designed. By establishing a product's performance degradation process model, the risks of the producer and user are obtained, and the uncertain acceptance sampling model is constructed, including acceptance index module, reception reliability module and both parties. Risk module, a multi-objective optimization algorithm is used to obtain the optimal acceptance sampling scheme.

Benefits of technology

Under the conditions of small samples, the accuracy of the acceptance sampling plan is improved, the risks of the producer and user are reduced, and the scientific acceptance of batches of products is achieved.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118780685B_ABST
    Figure CN118780685B_ABST
Patent Text Reader

Abstract

The present invention relates to a design method for an uncertain acceptance sampling scheme based on performance degradation, which belongs to the field of acceptance sampling technology and solves the problem of low accuracy of acceptance sampling schemes based on uncertain theoretical sampling schemes under small sample conditions in the prior art. The design method of the present invention specifically includes the following steps: Step 1, establishing a performance degradation process model of a product; Step 2, establishing an uncertain acceptance sampling model based on the performance degradation process model of the product in Step 1, and obtaining the production risk and the user risk; Step 3, determining a timed truncation uncertain acceptance sampling scheme based on the obtained production risk and user risk. Compared with the prior art, the present invention has at least the following beneficial effects: it can be used for acceptance sampling schemes under small sample conditions, thereby achieving the purpose of scientific acceptance of batch products with small sample characteristics.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention belongs to the technical field of acceptance sampling, and in particular relates to a design method for an uncertain acceptance sampling scheme based on performance degradation. Background Art

[0002] Acceptance sampling is a sampling inspection method used to determine whether the quality and reliability of products produced by the manufacturer meet the requirements of the user. At the same time, the user can decide whether to accept this batch of products based on the sampling results. In order to design small sample acceptance sampling plans, accurately measure cognitive uncertainty, and improve the accuracy of product batch identification, there are currently three main methods, namely Bayesian acceptance sampling, fuzzy acceptance sampling, and neutral acceptance sampling.

[0003] However, whether it is the Bayesian method, the fuzzy method, or the neutral statistical method, all of them are based on probability theory to design small sample acceptance sampling plans. However, the theoretical basis of probability theory is the law of large numbers, that is, the premise of large samples needs to be met, which is contrary to the actual situation of small samples. As a result, the cognitive uncertainty caused by small samples cannot be accurately measured, which reduces the accuracy of the acceptance sampling plan for product identification and increases the risks for both parties. Summary of the invention

[0004] In view of the above problems, the present invention provides a design method for an uncertain acceptance sampling scheme based on performance degradation, which solves the problem of low accuracy of acceptance sampling schemes based on uncertain theoretical sampling schemes under small sample conditions in the prior art.

[0005] The present invention provides a design method for an uncertain acceptance sampling scheme based on performance degradation, which specifically comprises the following steps:

[0006] Step 1: Establish a product performance degradation process model;

[0007] Step 2: Establish an uncertain acceptance sampling model based on the product performance degradation process model in step 1 to obtain the production risk and user risk;

[0008] Among them, the uncertain acceptance sampling model includes acceptance index module, acceptance reliability module and two-party risk module;

[0009] Acceptance index module, used to obtain the reliable life of the product;

[0010] Receiving credibility module, including the receiving credibility model of the product;

[0011] Based on the acceptance index module and the receiving credibility module, a risk model for both parties is constructed to obtain the risks of the producer and the user;

[0012] Step 3: Based on the producer risk and user risk obtained from the two-party risk model in step 2, determine the uncertain acceptance sampling plan.

[0013] Optionally, a performance degradation process model of the product is established based on uncertainty theory.

[0014] Optionally, the performance degradation process model of the product is expressed as:

[0015] (1)

[0016] in, represents the normalized stress level s Next time t The performance degradation process of the product; is the normalized stress level s Next time t The amount of performance degradation of the product; Indicates the performance degradation of the product at the initial moment; is the performance degradation rate of the product; is the diffusion coefficient; Indicates time t The mean is 0 and the standard deviation is Normal uncertainty distribution of ; For the moment t The standard deviation function of represents the indicative function; Indicates time t The mean is 0 and the standard deviation is Normal uncertain distribution.

[0017] Optionally, the expression for the performance degradation rate of the product is:

[0018] (2)

[0019] in, , a and b are all model parameters.

[0020] Optionally, when the stress type is temperature stress, the normalized stress level is calculated using the Arrhenius model s ; When the stress type is not temperature stress, the normalized stress level is calculated using the power law model or exponential model s , the expression is:

[0021] (3)

[0022] in, and They respectively represent the lower limit and upper limit of the product's working stress level.

[0023] Optionally, the indicative function The expression is:

[0024] . (4)

[0025] Optionally, substitute formulas (2) to (4) into formula (1) to express the product performance degradation process model as follows:

[0026] . (5)

[0027] Optionally, the timed truncation uncertain acceptance sampling scheme is a multi-objective optimization model including a threshold value of the sample size invested in the acceptance test, the acceptance test time, and the sum of degradation amounts of all samples invested in the acceptance test.

[0028] Optionally, based on the multi-objective optimization model, a multi-objective optimization algorithm is used to obtain a Pareto solution set of multiple groups of timed truncation uncertain acceptance sampling schemes;

[0029] Each set of timed truncation uncertain acceptance sampling schemes in the Pareto solution set is taken as a decision execution unit, and a relative efficiency model for obtaining decision execution units is established.

[0030] Based on the relative efficiency model of decision execution units, according to the linear programming equation, the input and output weights of each decision execution unit are optimized to obtain the optimal uncertain acceptance sampling scheme.

[0031] Compared with the prior art, the present invention has at least the following beneficial effects: it can provide an acceptance sampling plan under small sample conditions, thereby achieving the purpose of scientific acceptance of batch products with small sample characteristics. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] The drawings are only for the purpose of illustrating particular embodiments and are not to be construed as limiting the invention.

[0033] Figure 1 The flowchart of the design method of the uncertain acceptance sampling scheme based on performance degradation of the present invention. DETAILED DESCRIPTION

[0034] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments can be combined with each other. In addition, the present invention can also be implemented in other ways different from those described herein, and therefore, the protection scope of the present invention is not limited by the specific embodiments disclosed below.

[0035] A specific embodiment of the present invention, as Figure 1 , discloses a design method for an uncertain acceptance sampling scheme based on performance degradation, which specifically includes the following steps:

[0036] Step 1: Establish a performance degradation process model of the product;

[0037] Specifically, the performance degradation process model of the product is established based on uncertainty theory, and the expression is:

[0038] (1)

[0039] in, Y ( s , t ) represents the normalized stress level s Next time t The performance degradation process of the product; is the normalized stress level s Next time t The amount of performance degradation of the product; Y 0 Indicates the performance degradation of the product at the initial moment; e s is the performance degradation rate of the product; s is the diffusion coefficient; Indicates time t The mean is 0 and the standard deviation is Normal uncertainty distribution of ; For the moment t The standard deviation function of represents the indicative function; Indicates time t The mean is 0 and the standard deviation is Normal uncertain distribution.

[0040] Furthermore, when Y ( s , t ) and time t When there is a linear relationship, f ( t )= t ;when Y ( s , t ) has a nonlinear relationship with time, f ( t )= t β , t β Indicates time t of β Power, β is a constant.

[0041] Furthermore, the expression of the performance degradation rate of the product is:

[0042] (2)

[0043] in, , a and b are all model parameters; s represents the normalized stress level.

[0044] Furthermore, for the normalized stress level s , when the stress type is temperature, the normalized stress level value is calculated by the Arrhenius model; when the stress type is non-temperature stress, the normalized stress level value is calculated by the power law model or exponential model, and the expression is:

[0045] (3)

[0046] in, and They respectively represent the lower limit and upper limit of the product's working stress level.

[0047] Furthermore, the indicative function The expression is:

[0048] (4)

[0049] The present invention adopts an indicative function to associate performance degradation with the performance increasing or decreasing over time, making the performance degradation model more accurate.

[0050] Furthermore, in order to better illustrate the present invention, according to formulas (1) to (4), the present invention adopts a specific performance degradation process model shown in formula (5), which is expressed as follows:

[0051] (5)

[0052] It should be noted that the present invention selects f ( t )= t To illustrate the relationship between performance degradation process and time, other f ( t ) is also applicable to the method of the present invention.

[0053] Step 2: Establish an uncertain acceptance sampling model based on the product performance degradation process model in step 1 to obtain the production risk and user risk;

[0054] Specifically, the uncertain acceptance sampling model includes an acceptance index module, a receiving reliability module and a two-party risk module.

[0055] (1) Acceptance index module, used to obtain the reliable life of the product.

[0056] According to the relationship between the product performance margin and reliability, that is, reliability is the confidence that the performance margin is greater than 0, combined with the product performance degradation process model in step 1, that is, formula (5), the product reliability model is constructed, and the expression is:

[0057] (6)

[0058] in, represents the normalized stress level s Next time t Reliability of the product; Indicates the failure threshold of the product.

[0059] make α , based on the reliability of the performance margin, the reliable life of the product is obtained, and the expression is:

[0060] (7)

[0061] in, Indicates the reliable life of the product.

[0062] (2) Receiving reliability module, including the receiving reliability model of the product.

[0063] Furthermore, the acceptance requirement constraint is established: the sum of the degradation of all samples put into the acceptance test The threshold value that does not exceed the sum of the degradation values ​​of all samples put into the acceptance test x .

[0064] Establish the acceptance reliability model of batch products, the expression is:

[0065] (8)

[0066] in, Indicates the acceptance confidence of a batch of products; It represents the sum of degradation of all samples put into acceptance test; The threshold value representing the sum of degradation amounts of all samples put into the acceptance test is a constant; n The sample size devoted to the acceptance test; Indicates i Samples at time t The amount of performance degradation.

[0067] It can be understood that the acceptance confidence of a product is the possibility that the product will be accepted when the results of the acceptance test meet the pre-given requirements.

[0068] Further, i Samples at time t The performance degradation The performance degradation model obeys equation (5), i Samples at time t The performance degradation is an uncertain variable. i Samples at time t The performance degradation Normal uncertainty distribution The expression is:

[0069] (9)

[0070] in, Indicates i Samples at time t The performance degradation Less than or equal to a constant y reliability;

[0071] No. i Samples at time t The performance degradation The inverse of the normal uncertain distribution The expression is:

[0072] (10)

[0073] in, Indicates unreliability. , .

[0074] Furthermore, the sum of the degradation values ​​of all samples put into the acceptance test is The inverse of the uncertain distribution The expression is:

[0075] (11)

[0076] in, Indicates n Samples at time t The performance degradation amount follows the inverse of the normal uncertainty distribution.

[0077] Therefore, according to the uncertainty theory, the sum of the degradation of all samples put into the acceptance test is The uncertainty distribution of is its inverse uncertainty distribution Then find the inverse, that is, the sum of the degradation of all samples put into the acceptance test The uncertain distribution The expression is:

[0078] (12)

[0079] in, The sum of the degradation of all samples put into the acceptance test Less than or equal to a constant z reliability;

[0080] Furthermore, the product acceptance reliability model is:

[0081] . (13)

[0082] (3) A bilateral risk model is constructed based on the acceptance index module and the receiving credibility module to obtain the production risk and the user risk.

[0083] Specifically, establish a product acceptance reliability model L Reliable life of the product t sα The functional relationship is expressed as:

[0084] . (14)

[0085] Through the production acceptance indicators (ie: t sα0 ) and user acceptance indicators (i.e.: t sα1 ) to obtain the producer risk and the user risk, the expression is:

[0086] (15)

[0087] in, α p Risks for the producer; β c Risks of the user; L ( t sα0 )and L ( t sα1 ) are the production acceptance indicators t sα0 And the user acceptance index is t sα1 The acceptance reliability of the product.

[0088] It can be understood that the production side risk is the reliability that acceptable products are mistakenly rejected; the user side risk is the reliability that products that should be rejected are mistakenly accepted.

[0089] Further, t sα0 > t sα1 .

[0090] Reliability of product acceptance L Acceptance criteria t sα The risk of both parties is obtained by the functional relationship between the production side risk being the confidence level that acceptable products are mistakenly rejected, and the user side risk being the confidence level that products that should be rejected are mistakenly accepted.

[0091] Step 3: Based on the producer risk and user risk obtained from the two-party risk model in step 2, determine the uncertain acceptance sampling plan.

[0092] Specifically, the timed truncation uncertain acceptance sampling scheme includes the sample size of the acceptance test, the acceptance test time, and the threshold of the sum of the degradation of all samples put into the acceptance test. x ,Right now( n , T , x ).

[0093] Based on the producer risk and user risk obtained by the two-party risk model in step 2, the present invention aims to minimize producer risk, user risk, and test cost, and constructs a solution to the timed truncation uncertain acceptance sampling scheme ( n , T , x ), that is, the design variable is ( n , T , x ) is a multi-objective optimization model, and its expression is:

[0094] (16)

[0095] in, TC is the total cost of acceptance sampling; is the unit sample cost; is the unit test time cost; α 0 and β 0 are the upper limits of the risks for the producer and the user, respectively; n max and T max They represent the maximum sample size and maximum test time that can be invested in the test respectively; α p and β c The calculation formula of is shown in formula (15).

[0096] Preferably, n Maximum threshold for the sum of product degradation x No more than 40 n , n Is a positive integer.

[0097] Furthermore, based on the multi-objective optimization model, the multi-objective optimization algorithm NSGA-II is used to obtain the Pareto solution set of multiple groups of timed truncation uncertain acceptance sampling schemes;

[0098] Each set of fixed-time truncation uncertain acceptance sampling schemes in the Pareto solution set ( n , T , x ) is used as a decision execution unit DMU, ​​and a relative efficiency model of the decision execution unit DMU is established, which is expressed as:

[0099] (17)

[0100] in, Represents the relative efficiency of the decision execution unit DMU; and Respectively represent w The decision execution unit DMU i The input and j Outputs; and Respectively represent j output and i The weight of the input, , >0; Number Indicates the total number of decision execution units DMU; X , Y Represent the number of inputs and outputs respectively.

[0101] Further, let y w1 = n , y w2 = T , Y =2, x w1 = , x w2 = , x w3 = , x w4 = TC , X=4, based on the relative efficiency model of the decision execution unit DMU, ​​according to the linear programming equation, the input and output weights of each decision execution unit DMU are optimized to obtain the optimal uncertain acceptance sampling scheme ( n , T , x ), the expression is:

[0102] (18)

[0103] in, Indicates w 0 The relative efficiency of the decision making units (DMUs); and Respectively represent w 0 The decision execution unit DMU j output and i The weight of the input; and Respectively represent w 0 The decision execution unit DMU j output and i Inputs; Indicates w 0 A decision execution unit DMU.

[0104] make is the output variable of each decision execution unit DMU, ​​the larger the better; let is the input variable, the smaller the better.

[0105] The present invention provides y w1 = n , y w2 = T , Y =2, x w1 = , x w2 = , x w3 = , x w4 = TC , X =4 The more information about the batch of products to be accepted is obtained by the timed truncation uncertain acceptance sampling, the more accurate the result of the acceptance sampling; the smaller the total cost of the acceptance sampling, the smaller the price paid for the acceptance sampling; the smaller the risk to the producer and the user.

[0106] In order to illustrate the effectiveness of the method proposed in the present invention, the above technical solution of the present invention is described in detail below through an electrical connector embodiment, and the specific implementation steps are as follows:

[0107] The failure threshold of the electrical connector performance degradation y th =100, the confidence reliability corresponding to the quantile life α = 0.9, diffusion coefficient s =0.6, manufacturer acceptance index t sα0 =500, user acceptance index t sα1 =50, unit sample cost of electrical connector a =100, the unit test time cost of the electrical connector b =20, number of test samples 1≤ n ≤10, test time 1≤ T ≤500, upper limit of risk for the producer α 0 =20%, user risk β 0 =20%. At the same time, the genetic generation of NSGA-II is set to 20000 and the population size is set to 60.

[0108] According to the uncertain acceptance sampling scheme obtained in step 3, combined with the model parameters given in the electrical connector embodiment, the NSGA-II algorithm can be used to obtain a set of uncertain acceptance sampling schemes that meet the Pareto solution, and the top five uncertain acceptance sampling schemes ranked in efficiency in the Pareto solution set are given according to the DEA algorithm, as shown in Table 1.

[0109] Table 1 The top five uncertain acceptance sampling schemes with the highest efficiency in the Pareto solution set

[0110]

[0111] Therefore, it can be seen from Table 1 that the acceptance sampling plan with the highest relative efficiency is ( n , T , x ) = (1, 25,15), risks and acceptance costs of both parties ( α 0 , β 0 , TC )=(3.21%, 11.56%, 600).

[0112] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by any technician familiar with the technical field within the technical scope disclosed by the present invention should be covered within the protection scope of the present invention.

Claims

1. A design method for an uncertain acceptance sampling scheme based on performance degradation, characterized in that: Tool The process includes the following steps: Step 1: Establish the performance degradation process model of the product, the expression is: (1) in, represents the normalized stress level s Next time t The performance degradation process of the product; is the normalized stress level s Next time t The amount of performance degradation of the product; Indicates the performance degradation of the product at the initial moment; is the performance degradation rate of the product; is the diffusion coefficient; Indicates time t The mean is 0 and the standard deviation is Normal uncertainty distribution of ; For the moment t The standard deviation function of represents the indicative function; Indicates time t The mean is 0 and the standard deviation is Normal uncertainty distribution of ; Step 2: Establish an uncertain acceptance sampling model based on the product performance degradation process model in step 1 to obtain the production risk and user risk; Among them, the uncertain acceptance sampling model includes acceptance index module, acceptance reliability module and two-party risk module; Acceptance index module, used to obtain the reliable life of the product; According to the relationship between the product's performance margin and reliability, combined with the product's performance degradation process model in step 1, the product's reliability model is constructed, and the expression is: (6) in, represents the normalized stress level s Next time t Reliability of the product; Indicates the failure threshold of the product; and b are all model parameters; make α , based on the reliability of the performance margin, the reliable life of the product is obtained, and the expression is: (7) in, Indicates the reliable life of the product; Receiving credibility module, including the receiving credibility model of the product; Based on the acceptance index module and the receiving credibility module, a risk model for both parties is constructed to obtain the risks of the producer and the user; Step 3: Based on the producer risk and user risk obtained from the two-party risk model in step 2, determine the uncertain acceptance sampling plan.

2. The method for designing an uncertain acceptance sampling scheme according to claim 1, characterized in that: The performance degradation process model of the product is established based on uncertainty theory.

3. The method for designing an uncertain acceptance sampling scheme according to claim 1, characterized in that: The expression of product performance degradation rate is: (2) in, a is the model parameter.

4. The method for designing an uncertain acceptance sampling scheme according to claim 3, characterized in that: When the stress type is temperature stress, the normalized stress level is calculated using the Arrhenius model s ; When the stress type is not temperature stress, the normalized stress level is calculated using the power law model or exponential model. s , the expression is: (3) in, and They respectively represent the lower limit and upper limit of the product's working stress level.

5. The method for designing an uncertain acceptance sampling scheme according to claim 4, characterized in that: Indicative Function The expression is: 。(4) 6. The method for designing an uncertain acceptance sampling scheme according to claim 5, characterized in that: Substituting formulas (2) to (4) into formula (1), the product performance degradation process model is expressed as: 。(5) 7. The method for designing an uncertain acceptance sampling scheme according to any one of claims 1 to 6, characterized in that: The timed truncation uncertain acceptance sampling scheme is a multi-objective optimization model that includes the sample size of the acceptance test, the acceptance test time and the threshold of the sum of the degradation amounts of all samples put into the acceptance test.

8. The method for designing an uncertain acceptance sampling scheme according to claim 7, characterized in that: Based on the multi-objective optimization model, the multi-objective optimization algorithm is used to obtain the Pareto solution set of multiple groups of timed truncation uncertain acceptance sampling schemes. Each set of timed truncation uncertain acceptance sampling schemes in the Pareto solution set is taken as a decision execution unit, and a relative efficiency model for obtaining decision execution units is established. Based on the relative efficiency model of decision execution units, according to the linear programming equation, the input and output weights of each decision execution unit are optimized to obtain the optimal uncertain acceptance sampling scheme.