A method and device for evaluating component reliability

Abstract

The present invention belongs to the technical field of electronic component reliability assessment, and discloses a component reliability assessment method and device, the method comprising: obtaining historical inspection data and production process parameters, reliability expectation values ​​and production process parameters of a batch to be assessed; statistically analyzing the historical inspection data to establish a product reliability baseline and failure rate statistical model parameters; using the baseline and expected values ​​to perform statistical estimation to obtain a production sample plan and a sampling plan; using the baseline after the associated parameters, the production process parameters of the batch to be assessed, the model parameters after the associated indicators and expected values ​​to perform statistical estimation and comparison, and determining a screening method and an inspection method; using a screening plan composed of a production sample plan and a screening method to assess screening reliability, obtain screening reliability data, and determine an inspection plan after comparing it with the product reliability baseline. The method uses historical data for component reliability assessment, improves data utilization, and reduces inspection time and expenses.

Description

Technical Field

[0001] The present invention belongs to the technical field of reliability assessment of electronic components for space and aerospace applications, and in particular relates to a component reliability assessment method and device. Background Art

[0002] Due to their unrepairability in orbit, electronic components must undergo reliability evaluation before they can be used in space missions to meet the reliability requirements of space missions.

[0003] The existing quality grade components have quality consistency inspection (QoS Class Identifier, QCI) specified in their specifications and are carried out batch by batch. Certified production lines are also subject to periodic inspection, including more than 160 hours of aging test and more than 1000 hours of life assessment test. These sampling inspection schemes based on binomial or hypergeometric distribution are actually parameter estimates of the production line quality level, i.e., the Acceptance Quality Limit (AQL) or the failure rate. The failure rate of components produced by modern production lines has reached the level of parts per million (PPM) or parts per billion (PPB). The test reliability assessment scheme has resulted in high time and economic costs and is actually impossible to carry out, although the sampling standards MIL-STD-105 and GJB179 stipulate a reduced sampling inspection scheme for components in a single batch.

[0004] Traditional aerospace component reliability assessment methods and specifications all use traditional reliability testing methods, which results in high component reliability assessment time and financial costs, making it difficult to meet the requirements for rapid response to aerospace component reliability assessment, especially for the rapid and low-cost needs of commercial aerospace that uses a large number of Commercial Off-The-Shelf (COTS) components. Summary of the invention

[0005] The purpose of the present invention is to overcome the defects of the prior art and to provide a component reliability evaluation method and device.

[0006] In one aspect, a component reliability assessment method is provided, wherein the reliability assessment includes screening reliability assessment and quality consistency inspection, including:

[0007] Step 1: Obtain historical component inspection data, historical production process parameters, reliability expectations of the batch of components to be evaluated, and production process parameter data of the batch of components to be evaluated, where the reliability expectations include: screening reliability expectations and quality consistency inspection expectations;

[0008] Step 2: Statistically analyze the historical inspection data of components, establish a product reliability baseline for a stable production line, obtain the statistical model parameters for component aging failure rate, associate the model parameters with batch aging indicators, and associate the product reliability baseline with historical production process parameters;

[0009] Step 3: Use the established product reliability baseline and the expected reliability value of the batch of components to be evaluated to make statistical estimates, and obtain the production sample plan of the batch of components to be evaluated and the sampling plan for quality consistency inspection;

[0010] Step 4: Use the product reliability baseline after correlation parameters, the production process parameters of the batch of components to be evaluated, the model parameters after correlation indicators, and the expected reliability values ​​to perform statistical estimation and comparison to determine the screening method for the batch of components to be evaluated and the inspection method in the quality consistency inspection;

[0011] Step 5: Using the screening scheme consisting of the production sample scheme and the screening method, the screening reliability of the batch of components to be evaluated is evaluated to obtain screening reliability data, wherein the screening reliability includes: the allowable defect rate of the aging batch and / or the percentage of the defect rate of the aging batch;

[0012] Step 6: Compare the screening reliability data with the product reliability baseline in step 2. Based on the comparison results, determine whether the batch of components to be evaluated requires quality consistency inspection and what inspection plan to use when necessary. The inspection plans include: the first plan: a quality consistency inspection plan consisting of the sampling plan and inspection method in steps 3 and 4; the second plan: a standard quality consistency inspection plan specified in the product specification.

[0013] In an improved component reliability evaluation method, after step 2, the method further includes:

[0014] Compare the historical screening reliability data of components with the product reliability baseline to obtain the batch fluctuation cycle;

[0015] The shortest period in the batch fluctuation cycle that is below the lower limit of the product reliability baseline is significantly compared with the attenuation inspection scheme in the inspection standard to obtain the quality consistency inspection cycle of the stable production line.

[0016] In an improved component reliability assessment method, the step 2 performs statistical analysis on the component historical inspection data to establish a product reliability baseline for a stable production line, specifically including:

[0017] Perform binomial, normal, Weibull, exponential, or chi-square statistical analysis on historical component inspection data to establish a product reliability baseline for a stable production line.

[0018] In an improved component reliability assessment method, the step 2 performs statistical analysis on the historical inspection data of components to obtain the statistical model parameters of component aging failure rate, and associates the model parameters with the batch aging index, specifically including:

[0019] The Weibull distribution statistical analysis is performed on the historical inspection data of components to obtain the shape parameters, scale parameters and position parameters of the component failure rate statistical model. The shape parameters are associated with the defect percentage, and the scale parameters are associated with the aging time.

[0020] In an improved component reliability assessment method, step 3 specifically includes:

[0021] The established product reliability baseline and the expected reliability value of the batch of components to be evaluated are used to perform a posteriori probability statistical estimation, and the production sample plan of the batch of components to be evaluated and the sampling plan in the quality consistency inspection are obtained.

[0022] In an improved component reliability evaluation method, the product reliability baseline includes: an upper limit, a center, and a lower limit, and step 6 specifically includes:

[0023] Compare the screening reliability data with the product reliability baseline in step 2. If the screening reliability data is higher than the center, no quality consistency test is performed; if the screening reliability data is higher than the lower limit and lower than or equal to the center, the first solution is used to perform quality consistency test on the components in the batch to be evaluated; if the statistical analysis result of the screening reliability data is lower than or equal to the lower limit in step 2, the second solution is used to perform quality consistency test on the components in the batch to be evaluated.

[0024] In an improved component reliability evaluation method, the following is also included:

[0025] Step 7: Return the obtained screening reliability data and the quality consistency inspection data obtained after the inspection to the component historical inspection data.

[0026] In an improved component reliability evaluation method, the component reliability evaluation method is applicable to components produced by a stable production line using statistical process control.

[0027] On the other hand, a device based on the above-mentioned component reliability evaluation method is provided, comprising:

[0028] A data acquisition module is used to acquire historical inspection data of components, historical production process parameters, reliability expectations of the batch of components to be evaluated, and production process parameters of the batch of components to be evaluated, wherein the reliability expectations include: screening reliability expectations and quality consistency inspection expectations;

[0029] Product reliability baseline establishment unit, used to perform statistical analysis on historical component inspection data, establish product reliability baselines for stable production lines, and associate product reliability baselines with historical production process parameters;

[0030] A aging failure rate statistical model building unit is used to perform statistical analysis on historical component inspection data, obtain component aging failure rate statistical model parameters, and associate the model parameters with batch aging indicators;

[0031] The production sample and sampling plan determination unit is used to make statistical estimates using the established product reliability baseline and the expected reliability value of the batch of components to be evaluated, and obtain the production sample plan of the batch of components to be evaluated and the sampling plan in the quality consistency inspection;

[0032] The screening and inspection method determination unit uses the product reliability baseline after the associated parameters, the production process parameters of the batch of components to be evaluated, the model parameters after the associated indicators, and the reliability expectation value to perform statistical estimation and comparison to determine the screening method of the batch of components to be evaluated and the inspection method in the quality consistency inspection;

[0033] A screening reliability evaluation unit is used to evaluate the screening reliability of the batch of components to be evaluated by using a screening scheme consisting of a production sample scheme and a screening method, and obtain screening reliability data, wherein the screening reliability includes: the allowable defect rate of a single aging batch or the percentage of the defective rate of the aging batch;

[0034] A comparison unit is used to compare the screening reliability data with the product reliability baseline, and based on the comparison results, determine whether a quality consistency inspection is required for the batch of components to be evaluated and what inspection scheme to use when necessary, wherein the inspection schemes include: a first scheme: a quality consistency inspection scheme consisting of a sampling scheme and an inspection method; a second scheme: a standard quality consistency inspection scheme specified in the product specification.

[0035] Compared with the prior art, the advantages of the present invention are:

[0036] Advantage 1: Using historical inspection data and production process data for component reliability assessment improves data utilization;

[0037] Advantage 2: There is a quantitative basis for determining the QCI cycle time. Since the product reliability baseline is established and batch fluctuation cycle data is obtained, data support can be provided for determining the QCI cycle time, which is lower than the traditional prescribed time.

[0038] Advantage 3: Time and cost reduction. Due to the use of past data, according to the characteristics of Bayesian statistics and other results, the number of test samples and the test time will be reduced. Therefore, under the premise of keeping the confidence level unchanged, the test time and cost are reduced.

[0039] Advantage 4: Suitable for rapid response to mass-produced space missions and commercial space missions. Due to the reduced inspection costs, COTS components can be used for rapid response to space missions and commercial space missions. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 is one of the flow charts of the component reliability evaluation method according to an embodiment of the present invention;

[0041] Figure 2 This is the second flowchart of the component reliability evaluation method according to the embodiment of the present invention;

[0042] Figure 3 It is a schematic structural diagram of a component reliability evaluation device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0043] Component reliability assessment includes screening reliability assessment and quality consistency inspection. With the large investment brought by the advancement of component technology, fast and efficient reliability assessment is also required. In the 1990s, the built-in reliability (BIR) assessment method was proposed. Its main method is to control the production process, such as the most widely used statistical process control (SPC) to control and record production process data. The purpose of BIR is to reduce or even eliminate the dependence on test reliability, conduct reliability assessment quickly and efficiently, and reduce subsequent testing costs and shorten time to market.

[0044] At present, for component production lines that use SPC stabilization, there are already a large amount of production process data and component inspection data. However, these data have not been used by the existing aerospace component test reliability assessment methods, that is, the current reliability assessment is only related to the batch to be assessed, which also results in a waste of test data.

[0045] The solution of the embodiment of the present invention integrates BIR and component historical inspection data to evaluate the reliability of components, that is, integrates production process data and component historical inspection data to evaluate the production process control level and component reliability level, such as process satisfaction index (C pk ), first-time finished product rate (yield), acceptance quality level (AQL) and parts per million (PPM), etc. The solution of the embodiment of the present invention is applicable to a production line with a stable production process control level.

[0046] The technical solution of the present invention is described in detail below with reference to the accompanying drawings and embodiments.

[0047] like Figure 1As shown, it is a flow chart of a component reliability evaluation method according to an embodiment of the present invention, comprising the following steps:

[0048] Step 101: Obtain historical component inspection data, historical production process parameters, reliability expectations of a batch of components to be evaluated, and production process parameter data of the batch of components to be evaluated.

[0049] The historical inspection data of components include: historical screening reliability data and historical quality consistency inspection data; the expected reliability values ​​include: screening reliability expected values ​​and quality consistency inspection expected values. The process parameter data of the batch of components to be evaluated include: substrate tape-out production process data and packaging process data.

[0050] Step 102: Perform statistical analysis on the historical inspection data of components, establish a product reliability baseline for a stable production line, and associate the product reliability baseline with historical production process parameters.

[0051] Specifically, binomial, normal, Weibull, exponential or chi-square statistical analysis methods can be used to establish a product reliability baseline for a stable production line. The product reliability baseline includes a lower limit, a center and an upper limit.

[0052] The product reliability baseline specifically includes ① the production process control baseline, such as the process satisfaction index (C pk ), first-time finished product rate (yield), acceptance quality level (AQL) and parts per million (PPM), etc.; ② The reliability baseline of screening and quality consistency inspection specified in the detailed specifications of components, such as screening pass rate and quality consistency inspection pass rate; ③ The pass rate baseline of test items in screening and quality consistency inspection specified in the general and detailed specifications of components, such as high-stability storage, temperature cycling, constant acceleration, PIND, sealing, aging, X-ray, bonding tension and chip shear force, etc.

[0053] For the product reliability baseline, taking yield as an example, the component inspection data that has existed in a component production line so far is (t1, y1), (t2, y2), ┈, (t n ,y n ), where t i and i Represent the i-th batch time and its first-time yield respectively.

[0054] With y i Taking the normal statistical distribution model N(E,σ) as an example, we can calculate the two parameters of the normal statistical distribution model, namely the mean E and variance σ of y. With E as the center of y, E+mσ and E-mσ as the upper and lower limits, usually m (positive integer) = 1, 2, 3..., the baseline of the first yield y is established.

[0055] Simply put, we can use the defects generated during the production process as a uniform distribution model, then the average rejection rate of components is R = 1-(y / m) α , α represents the ratio of the defect density of the production line test to the reliability defect density, and m represents the maximum value of components that can be contained in a substrate (such as a wafer). R or AQL = 1-R represents the reliability level of the components in continuous production. The average rejection rate R (rejection) and AQL of the components can also be calculated according to the concentrated distribution model of the defects generated during the production process.

[0056] You can also draw y i The curve fluctuates around E over time, and then we get y i The average time interval below the lower limit MTBF y , which will serve as the basis for determining the quality consistency inspection cycle in the future.

[0057] For a fixed production line, the first-time yield, parts per million (PPM) and process satisfaction index (C pk ), that is, they can be converted into each other, so each production line can use the product reliability baseline expression method according to the actual situation.

[0058] The above method is further used to obtain the reliability baseline of screening and quality consistency inspection and the baseline of the pass rate of its test items, including high-stability storage, temperature cycling, constant acceleration, PIND, sealing, aging, X-ray, bonding tension and chip shear force, etc., and correlate them with the production process parameters. For example, the bonding tension is correlated with the production process parameters such as bonding pressure, bonding power and bonding layer thickness.

[0059] Step 103: Statistically analyze the historical inspection data of components to obtain the statistical model parameters of component aging failure rate, and associate the obtained statistical model parameters of aging failure rate with the batch screening index.

[0060] Specifically, the Weibull distribution statistical analysis can be performed on the historical inspection data of components to obtain the shape parameters, scale parameters and position parameters of the component failure rate statistical model, and the obtained shape parameters can be associated with the defect percentage, and the scale parameters can be associated with the aging time.

[0061] A large amount of data analysis shows that the aging failure of components follows the Weibull distribution W (t0, b, c). Among them, t0 is the position parameter, which represents the time without failure or the shortest life, usually t0 = 0; b is the shape parameter, and the aging failure rate of components produced on a stable production line decreases with time, so b is less than 1; c is the scale parameter representing the characteristic life, that is, after the time exceeds c, the failure rate no longer changes significantly and approaches a constant.

[0062] For example, the aging data of all batches of components in a component production line so far are (t1, m1), (t2, m2), ┈, (t n , m n ), where t i and m i Represent the i-th time and its failure number respectively.

[0063] After fitting the component aging data according to the Weibull distribution, the estimated values ​​of the three parameters t0, b, and c can be obtained, and the fitting degree R 2 The requirement is higher than 0.9. For burn-in with a fixed end or a fixed number of ends, the revised estimation method is used to obtain the estimated values ​​of the three parameters t0, b, and c.

[0064] With these three parameters, the aging failure rate λ of components produced by this stable production line can be calculated according to the Weibull distribution failure rate formula, that is:

[0065] λ(t)=1-exp((t-t0) / c)b

[0066] By matching the aging batch allowable defect rate (percentage defect allowance, PDA) or the aging batch defect rate percentage (Lot Tolerance Percent Defective, LTPD) with the failure rate λ, the aging time corresponding to the inspection requirement PDA or LTPD is obtained.

[0067] Here, step 102 and step 103 may be performed in parallel or in series. Step 102 may be performed first and then step 103. Step 103 may also be performed first and then step 102.

[0068] Step 104: Statistical estimation is performed using the established product reliability baseline and the expected reliability value of the batch of components to be evaluated to obtain a production sample plan for the batch of components to be evaluated and a sampling plan for quality consistency inspection.

[0069] Specifically, a posteriori probability statistical estimation can be performed on the established product reliability baseline and the expected reliability value of the batch of components to be evaluated, and the production sample plan of the batch of components to be evaluated and the sampling plan for the quality consistency inspection of the batch of components to be evaluated can be obtained.

[0070] The production sample plan refers to the number of components produced in a batch for screening, which is a form of sampling.

[0071] Perform posterior probability statistical estimation on the established product reliability baseline and the reliability expectation value of the components in the batch to be evaluated, and obtain the production sample plan for the components in the batch to be evaluated and the sampling plan in the quality consistency inspection of the components in the batch to be evaluated, specifically as follows:

[0072] The sampling plan is an inspection that takes n samples and allows m (<n) failures, that is, the n(m) sampling plan, and its purpose is to evaluate a certain PDA or LTPD required by the inspection party. Without actual data or prior data on the reliability level of components, such as the Acceptable Quality Level (AQL), the higher the LTPD or PDA requirement, the larger the number of samples n, and the higher the inspection cost.

[0073] Take the AQL data obtained in step 102 as the prior probability. After giving PDA or LTPD (i.e., the posterior probability) and a confidence level = 0.9, the sampling plan n(m) for sampling inspection can be determined.

[0074] The basis of the method for evaluating the posterior probability through the prior probability is the total probability formula. Here, taking the Bayesian estimation of PDA as an example, let the prior probability P0 = (n(m)|AQL), and the posterior probability be P = (PDA|n(m)), then there is:

[0075] P = (PDA|n(m)) = P(n(m)|PDA)·P0 / P(n(m))

[0076] When the confidence level is set to 0.9, that is, the posterior probability P = (PDA|n(m)) = 0.9, then there is:

[0077] P(n(m)|PDA)·P0 / P(n(m)) = 0.9

[0078] And P(n(m)|PDA) can be calculated using the binomial distribution or the hypergeometric distribution. The total probability P(n(m)) is a fixed value. Therefore, after giving PDA, the value of n(m) can be obtained, that is, the A group and the screening sampling plan for the batch to be evaluated are obtained. The n(m) algorithm for a given LTPD is the same.

[0079] The method for determining the sampling plan in quality consistency is the same as the A group and the screening sampling plan for the batch to be evaluated above. Here, the number of production samples ≥ the number of samples in the quality consistency inspection.

[0080] Step 105: Use the product reliability baseline after associating parameters, the production process parameters of the components in the batch to be evaluated, the model parameters after associating indicators, and the reliability expectation value to perform statistical estimation and comparison, and determine the screening method for the components in the batch to be evaluated and the inspection method in the quality consistency inspection, that is, whether to conduct a certain test.

[0081] Specifically, if the product reliability baseline determined by the production process parameters of the batch to be evaluated, such as high-stability storage, temperature cycling, constant acceleration, PIND, sealing, X-ray, bond tension and chip shear force, is not lower than the obtained product reliability baseline, the center will no longer conduct the test; if it is lower than the center is required to conduct the test.

[0082] For a given PDA or LTPD requirement of a batch to be evaluated, the aging time is determined according to the Weibull distribution given in step 103 .

[0083] The number of samples in the above screening is carried out according to the production sample plan determined in step 104, and is carried out 100%.

[0084] Step 106: Using the screening scheme consisting of the production sample scheme and the screening method, the screening reliability of the batch of components to be evaluated is evaluated to obtain screening reliability data.

[0085] Among them, screening reliability includes: the allowable defect rate of a single aging batch or the percentage of defective rate of the aging batch.

[0086] Step 107: Compare the screening reliability data with the product reliability baseline. If the screening reliability data is higher than the center, execute step 108; if the screening reliability data is higher than the lower limit and lower than or equal to the center, execute step 109; if the statistical analysis result of the screening reliability data is lower than or equal to the lower limit, execute step 110;

[0087] Step 108: No quality consistency inspection is performed on the batch of components to be evaluated.

[0088] Step 109: Use the quality consistency inspection plan composed of the sampling plan and the inspection method to conduct a quality consistency inspection on the batch of components to be evaluated, obtain the evaluation quality consistency inspection data, and then execute step 111;

[0089] Step 110: Perform a quality consistency test on the batch of components to be evaluated using the quality consistency test plan specified in the product specification to obtain evaluation quality consistency test data, and then execute step 111;

[0090] Step 111: Return the obtained screening reliability data and evaluation quality consistency inspection data to the component historical inspection data.

[0091] The above steps compare the screening reliability data with the product reliability baseline in step 102, and based on the comparison results, determine whether the batch of components to be evaluated requires quality consistency inspection and what inspection scheme to use when necessary, wherein the inspection scheme includes: the first scheme: a quality consistency inspection scheme consisting of the sampling scheme and inspection method in steps 104 and 105; the second scheme: a standard quality consistency inspection scheme specified in the product specification.

[0092] This completes the reliability assessment of components.

[0093] In order to determine the production line quality consistency inspection cycle, preferably, after step 102, the reliability evaluation method of the present invention further includes:

[0094] Compare the historical screening reliability data of components with the product reliability baseline to obtain the batch fluctuation cycle;

[0095] The shortest period in the batch fluctuation cycle that is below the lower limit of the product reliability baseline is significantly compared with the weakening inspection scheme in the inspection standard to obtain the quality consistency inspection cycle of the production line.

[0096] Here, the historical screening reliability data of all components can be compared with the product reliability baseline to obtain the batch fluctuation cycle.

[0097] Figure 2 A flowchart of another component reliability evaluation method for implementing the above process is given.

[0098] The scheme of the embodiment of the present invention integrates a large amount of existing data, mathematical statistics concepts and built-in reliability mechanisms of component production lines that are stabilized by statistical process control, and quantitatively compares batch fluctuations with the weakening inspection scheme in MIL-STD-105 or GJB179, thereby determining the QCI cycle time; statistically estimates the product reliability baseline, the production process parameter data and task requirements (reliability expectations) of the batch of components to be evaluated, and obtains the QCI sample sampling scheme and the inspection sample sampling scheme of the batch to be evaluated; statistically estimates the statistical model parameters and the production process parameter data and task requirements (reliability expectations) of the batch of components to be evaluated, and obtains the inspection scheme of the batch to be evaluated, technically realizing the use of existing data, mathematical statistics concepts and built-in reliability for component reliability evaluation, and the time and financial costs used are less than 1 / 2 of those of traditional evaluation methods.

[0099] In addition, the present invention also provides a device based on the above-mentioned component reliability evaluation method, such as Figure 3 As shown, including:

[0100] The data acquisition module 31 is used to acquire historical inspection data of components, historical production process parameters, reliability expectations of the batch of components to be evaluated, and production process parameters of the batch of components to be evaluated, wherein the reliability expectations include: screening reliability expectations and quality consistency inspection expectations;

[0101] A product reliability baseline establishing unit 32, for statistically analyzing the historical inspection data of components, establishing a product reliability baseline for a stable production line, and associating the product reliability baseline with historical production process parameters;

[0102] A aging failure rate statistical model building unit 33 is used to perform statistical analysis on historical component inspection data, obtain component aging failure rate statistical model parameters, and associate the model parameters with batch aging indicators;

[0103] The production sample and sampling plan determination unit 34 is used to perform statistical estimation using the established product reliability baseline and the expected reliability value of the batch of components to be evaluated, and obtain the production sample plan of the batch of components to be evaluated and the sampling plan in the quality consistency inspection;

[0104] The screening and inspection method determination unit 35 uses the product reliability baseline after the associated parameters, the production process parameters of the batch of components to be evaluated, the model parameters after the associated indicators, and the reliability expectation value to perform statistical estimation and comparison to determine the screening method of the batch of components to be evaluated and the inspection method in the quality consistency inspection;

[0105] The screening reliability evaluation unit 36 ​​is used to evaluate the screening reliability of the batch of components to be evaluated by using the screening scheme composed of the production sample scheme and the screening method, and obtain screening reliability data, wherein the screening reliability includes: the allowable defect rate of the burn-in batch or the percentage of the defect rate of the burn-in batch; and

[0106] The comparison unit 37 is used to compare the screening reliability data with the product reliability baseline, and based on the comparison result, determine whether the batch of components to be evaluated needs quality consistency inspection and what inspection scheme to use when necessary, wherein the inspection scheme includes: the first scheme: a quality consistency inspection scheme consisting of a sampling scheme and an inspection method; the second scheme: a standard quality consistency inspection scheme specified in the product specification.

[0107] In view of the high reliability evaluation time and cost caused by the traditional reliability evaluation method for aerospace components, and the fact that a large amount of component reliability test data in the past has not been used by the traditional reliability evaluation method, the embodiment of the present invention proposes a component reliability evaluation method with built-in reliability and big data for a component production line that uses SPC stability. This reliability evaluation method is based on the mechanism that reliability is designed, produced and managed, and uses a large amount of previous test data and mathematical statistics methods to reduce the number of sampling samples and test items in the traditional method. Under the condition of ensuring the confidence of the evaluation, it achieves the purpose of fast and low cost, and meets the needs of aerospace missions and commercial aerospace for low-cost and fast evaluation of aerospace components, especially COTS components.

[0108] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention is described in detail with reference to the embodiments, it should be understood by those skilled in the art that any modification or equivalent replacement of the technical solutions of the present invention does not depart from the spirit and scope of the technical solutions of the present invention and should be included in the scope of the claims of the present invention.

Claims

1. A component reliability assessment method, the reliability assessment comprising screening reliability assessment and quality consistency inspection, the method comprising: Step 1: Obtain historical component inspection data, historical production process parameters, reliability expectations of the batch of components to be evaluated, and production process parameter data of the batch of components to be evaluated, where the reliability expectations include: screening reliability expectations and quality consistency inspection expectations; Step 2: Statistically analyze the historical inspection data of components, establish a product reliability baseline for a stable production line, obtain the statistical model parameters for component aging failure rate, associate the model parameters with batch aging indicators, and associate the product reliability baseline with historical production process parameters; Step 3: Use the established product reliability baseline and the expected reliability value of the batch of components to be evaluated to make statistical estimates, and obtain the production sample plan of the batch of components to be evaluated and the sampling plan for quality consistency inspection; Step 4: Use the product reliability baseline after correlation parameters, the production process parameters of the batch of components to be evaluated, the model parameters after correlation indicators, and the expected reliability values ​​to perform statistical estimation and comparison to determine the screening method for the batch of components to be evaluated and the inspection method in the quality consistency inspection; Step 5: Using the screening scheme consisting of the production sample scheme and the screening method, the screening reliability of the batch of components to be evaluated is evaluated to obtain screening reliability data, wherein the screening reliability includes: the allowable defect rate of the aging batch and / or the percentage of the defect rate of the aging batch; Step 6: Compare the screening reliability data with the product reliability baseline in step 2. Based on the comparison results, determine whether the batch of components to be evaluated requires quality consistency inspection and what inspection plan to use when necessary. The inspection plans include: the first plan: a quality consistency inspection plan composed of the sampling plan and inspection method in steps 3 and 4; the second plan: a standard quality consistency inspection plan specified in the product specification. The reliability evaluation of the components is now completed.

2. The component reliability evaluation method according to claim 1, characterized in that: The historical inspection data includes: historical screening reliability data, and after step 2, further includes: Compare the historical screening reliability data of components with the product reliability baseline to obtain the batch fluctuation cycle; The shortest period in the batch fluctuation cycle that is below the lower limit of the product reliability baseline is significantly compared with the attenuation inspection scheme in the inspection standard to obtain the quality consistency inspection cycle of the stable production line.

3. The component reliability evaluation method according to claim 1, characterized in that: In step 2, statistical analysis is performed on the historical inspection data of components to establish a product reliability baseline for a stable production line, specifically including: Perform binomial, normal, Weibull, exponential, or chi-square statistical analysis on historical component inspection data to establish a product reliability baseline for a stable production line.

4. The component reliability evaluation method according to claim 1, characterized in that: The step 2 performs statistical analysis on the historical inspection data of components to obtain the statistical model parameters of component aging failure rate, and associates the model parameters with the batch aging index, specifically including: The Weibull distribution statistical analysis is performed on the historical inspection data of components to obtain the shape parameters, scale parameters and position parameters of the component failure rate statistical model. The shape parameters are associated with the defect percentage, and the scale parameters are associated with the aging time.

5. The component reliability evaluation method according to claim 1, characterized in that: The step 3 specifically includes: The established product reliability baseline and the expected reliability value of the batch of components to be evaluated are used to perform a posteriori probability statistical estimation, and the production sample plan of the batch of components to be evaluated and the sampling plan in the quality consistency inspection are obtained.

6. The component reliability evaluation method according to claim 1, characterized in that: The product reliability baseline includes: an upper limit, a center, and a lower limit. Step 6 specifically includes: Compare the screening reliability data with the product reliability baseline in step 2. If the screening reliability data is higher than the center, no quality consistency test is performed; if the screening reliability data is higher than the lower limit and lower than or equal to the center, the first solution is used to perform quality consistency test on the components in the batch to be evaluated; if the statistical analysis result of the screening reliability data is lower than or equal to the lower limit in step 2, the second solution is used to perform quality consistency test on the components in the batch to be evaluated.

7. The component reliability evaluation method according to claim 1, characterized in that: Also includes: Step 7: Return the obtained screening reliability data and the quality consistency inspection data obtained after the inspection to the component historical inspection data.

8. The component reliability evaluation method according to any one of claims 1 to 7, characterized in that: The component reliability evaluation method is applicable to components produced by a stable production line using statistical process control.

9. A device based on the component reliability evaluation method according to claim 1, characterized in that: include: A data acquisition module is used to acquire historical inspection data of components, historical production process parameters, reliability expectations of the batch of components to be evaluated, and production process parameters of the batch of components to be evaluated, wherein the reliability expectations include: screening reliability expectations and quality consistency inspection expectations; Product reliability baseline establishment unit, used to perform statistical analysis on historical component inspection data, establish product reliability baselines for stable production lines, and associate product reliability baselines with historical production process parameters; A aging failure rate statistical model building unit is used to perform statistical analysis on historical component inspection data, obtain component aging failure rate statistical model parameters, and associate the model parameters with batch aging indicators; The production sample and sampling plan determination unit is used to make statistical estimates using the established product reliability baseline and the expected reliability value of the batch of components to be evaluated, and obtain the production sample plan of the batch of components to be evaluated and the sampling plan in the quality consistency inspection; The screening and inspection method determination unit uses the product reliability baseline after the associated parameters, the production process parameters of the batch of components to be evaluated, the model parameters after the associated indicators, and the reliability expectation value to perform statistical estimation and comparison to determine the screening method of the batch of components to be evaluated and the inspection method in the quality consistency inspection; A screening reliability evaluation unit is used to evaluate the screening reliability of the batch of components to be evaluated by using a screening scheme consisting of a production sample scheme and a screening method, and obtain screening reliability data, wherein the screening reliability includes: an allowable defect rate of a single burn-in batch or a percentage of a defect rate of a burn-in batch; and A comparison unit is used to compare the screening reliability data with the product reliability baseline, and based on the comparison results, determine whether the batch of components to be evaluated requires quality consistency inspection and what inspection scheme to use when necessary, wherein the inspection schemes include: the first scheme: a quality consistency inspection scheme consisting of a sampling scheme and an inspection method; the second scheme: a standard quality consistency inspection scheme specified in the product specification.

Images