Man-machine collaborative high real-time function test cross-domain optimization method, device and medium

Abstract

The application discloses a man-machine collaborative high real-time function test cross-domain optimization method, equipment and medium, relates to the electronic information technical field, and includes: extending a discrete optimization model into a continuous optimization model, setting the probability of simultaneous failure of multiple intermediate events in a fault tree as the minimum value of the test proportion of all intermediate events corresponding to the intersection term, and obtaining a continuous optimization model considering the intersection term; obtaining a continuous optimization model considering real-time by theoretically analyzing a system inherent reliability calculation formula and truncating the system inherent reliability calculation formula combined with expert experience; obtaining a continuous optimization model based on additional constraints by additionally establishing the relationship between decision variables and system inherent reliability; obtaining an optimal test strategy by solving the continuous optimization model based on additional constraints; and the function test cross-domain optimization method, equipment and medium solve the problem of model solving real-time, and improve the overall optimization capability of the model cross-domain.

Description

Technical Field

[0001] The present invention relates to the field of electronic information technology, and in particular to a cross-domain optimization method, device and medium for high real-time functional testing of human-machine collaboration. Background Art

[0002] In the context of flexible manufacturing research and application, existing test item selection methods incorporate expert experience from multiple process domains to trace the root causes of product quality, improve the accuracy of system reliability calculations, and design motherboard functional test strategies with cross-domain holistic optimization capabilities. However, most existing methods build discrete optimization models, meaning they often only select test items without selectively testing samples. In practice, the test strategies derived from these methods either test all or none of the samples for each test item. These methods can effectively reduce testing costs when the attribute values ​​of test items exhibit polarized distributions. However, in reality, the values ​​of test item attributes are not polarized; many attributes range between high and low values. If discrete optimization methods are still used, the resulting test strategy, if all samples with these attributes are tested, will significantly increase testing time. If samples with these attributes are not tested at all, the missed detection rate is likely to be high. Therefore, discrete optimization models are difficult to further reduce testing costs.

[0003] To address the above issues, a proposal was made to expand the discrete optimization model into a continuous optimization model. The expanded model achieves high-precision calculation of system reliability and further improves the model's overall cross-domain optimization capabilities. However, due to the increased complexity of the reliability calculation formula, the model's solution time is significantly increased, making it impossible to guarantee the real-time solution of the optimization model while maintaining the overall cross-domain optimization capabilities. The flexible manufacturing production model of multiple varieties and small batches is characterized by frequent product model switching and batch changes, which places high standards and strict requirements on the rapid adaptation of test strategy design for test production lines. In this context, the existing expanded model is difficult to adapt to the demand for rapid test strategy design in the motherboard functional test process, forming a new technical bottleneck. Summary of the Invention

[0004] Based on the technical problems existing in the background technology, the present invention proposes a cross-domain overall optimization method for high real-time functional testing in human-machine collaboration. By incorporating expert experience into the reliability analysis process, the problem of real-time model solving is solved, thereby further improving the model's cross-domain overall optimization capability.

[0005] The present invention proposes a cross-domain optimization method for high real-time functional testing of human-machine collaboration, comprising:

[0006] The discrete optimization model is extended to a continuous optimization model, a fault tree is established, and a probability of simultaneous failure of multiple intermediate events in the fault tree is set as a minimum value of corresponding test proportions of all intermediate events included in an intersection term, to obtain a continuous optimization model considering the intersection term;

[0007] The continuous optimization model considering real-time is obtained by theoretically analyzing a system inherent reliability calculation formula and truncating the system inherent reliability calculation formula according to expert experience.

[0008] The continuous optimization model based on additional constraints is obtained by additionally establishing a relationship between a test proportion of each test term and the system inherent reliability.

[0009] The continuous optimization model based on additional constraints is solved, and a test strategy is screened based on a training set, to obtain an optimal test strategy.

[0010] Further, in the fault tree, a mainboard failure is regarded as a top event, a test term failure is regarded as an intermediate event, and a bottom reason causing the test term failure is regarded as a bottom event of the fault tree.

[0011] Further, in the extension of the discrete optimization model to the continuous optimization model, the following is specifically performed:

[0012] A relationship between a fault coverage rate and the system inherent reliability is determined by performing reliability analysis on a mainboard function test process.

[0013] An influence of concurrent failure of intermediate events including the same bottom event on the system inherent reliability is analyzed based on the fault tree analysis.

[0014] The system reliability of the mainboard function test process is derived, so as to construct the continuous optimization model.

[0015] Further, the continuous optimization model considering real-time is obtained by theoretically analyzing the system inherent reliability calculation formula and truncating the system inherent reliability calculation formula according to expert experience, and the following is specifically performed:

[0016] Since the fault tree analysis method analyzes the system reliability from the perspective of system unreliability, and the sum of the system inherent reliability and the system inherent unreliability is equal to 1, the analysis of the system inherent unreliability calculation formula is equivalent to the analysis of the system inherent reliability calculation formula. Both are essentially equivalent.

[0017] The system inherent unreliability calculation formula is divided into sub-terms, the number of the sub-terms is consistent with the number of test terms, the first sub-term includes small terms, the first sub-term and the first The number of sub-items contained in the two sub-items is the same, so the two sub-items are divided into one group;

[0018] By analyzing the expressions of the group of sub-items after the division, it is found that the small items contained in some sub-items in the calculation formula of the inherent unreliability of the system cancel each other out.

[0019] The calculation formula of the system inherent reliability is truncated by combining expert experience, and a continuous optimization model based on real-time considerations is obtained, thereby reducing the complexity of the continuous optimization model while retaining the relevant information between measurement items.

[0020] Furthermore, by establishing the relationship between the test ratio of each test item and the inherent reliability of the system, a continuous optimization model based on additional constraints is obtained, specifically:

[0021] Based on the maintenance data of historical test data, the probability of occurrence of the bottom event is obtained as the unreliability of the bottom event. The mean and deviation of the probability of occurrence of the bottom event are obtained in combination with expert experience, and then the reliability range of the bottom event is determined;

[0022] Based on the reliability range of the bottom event, the fault tree is used to quantitatively calculate the system inherent reliability range of each test item corresponding to the intermediate event, and the test ratio of each test item is calculated accordingly. The optimization interval is set as a function based on the inherent reliability of the system.

[0023] Furthermore, the test ratio The optimization interval is specifically:

[0024] ;

[0025] in, and Respectively The lower and upper limits of the system inherent reliability of each test item, and The proportionality coefficients for the test ratio and the lower and upper limits of the system's inherent unreliability are set based on expert experience.

[0026] Furthermore, the continuous optimization model based on additional constraints is specifically:

[0027] ;

[0028] ;

[0029] ;

[0030] in, For the test ratio, is the system inherent reliability obtained based on the calculation formula of the system inherent reliability after truncation, Indicates the The average test time for each test item, Indicates the test time threshold for each motherboard, is the total number of test items, For the The test ratio of each test item, and are the lower and upper limits of the system’s inherent reliability, and The proportional coefficients of the test ratio and the lower and upper limits of the inherent unreliability of the system are set according to expert experience. Indicates the length is An array of all 1s.

[0031] Furthermore, three evaluation indicators are set to verify the effectiveness of the cross-domain overall optimization method for motherboard functional testing. The evaluation indicators include the average motherboard test time, the false negative rate and the average total cost of motherboard testing.

[0032] Furthermore, the false negative rate The calculation formula is as follows:

[0033] ;

[0034] in, Indicates the test ratio The randomly generated test strategy and the test status of each test item of each motherboard are Group correspondence, For test ratio The corresponding The false negative rate of group correspondences on the training set or test set.

[0035] Furthermore, the average test time of the motherboard The calculation formula is as follows:

[0036] ;

[0037] in, Indicates the test ratio Randomly generated test strategy The test status of each test item of each motherboard Group correspondence, For test ratio The corresponding The average test time of group correspondences on the training set or test set.

[0038] A computer device comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the above-mentioned method when executing the computer program.

[0039] A computer-readable storage medium stores a plurality of classification programs, wherein the plurality of classification programs are used to be called by a processor and execute the method described above.

[0040] The advantages of the human-machine collaborative high real-time functional test cross-domain optimization method, equipment and medium provided by the present invention are: expanding the discrete optimization model into a continuous optimization model, selecting reasonable probability values ​​for simultaneous failure of related modules, and incorporating expert experience into the reliability analysis process to reasonably truncate the system's inherent reliability calculation formula, while ensuring the accuracy of reliability calculation and significantly improving the real-time solution of the optimization model, and making the test strategy design more reasonable; providing a feasible path for the rapid design of a motherboard functional test strategy with cross-domain overall optimization capabilities, which helps to reduce costs and increase efficiency of flexible intelligent manufacturing processes. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 It is a schematic diagram of the process of the present invention;

[0042] Figure 2 This is a fault tree diagram. DETAILED DESCRIPTION

[0043] The technical solutions of the present invention are described in detail below through specific embodiments. Numerous specific details are set forth in the following description to facilitate a full understanding of the present invention. However, the present invention can be implemented in many other ways than those described herein, and those skilled in the art may make similar modifications without departing from the scope of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0044] like Figure 1 and 2 As shown, the present invention proposes a cross-domain overall optimization method for high real-time motherboard function testing in human-machine collaboration, including steps 1 to 4:

[0045] Step 1: Expand the discrete optimization model into a continuous optimization model, establish a fault tree, and set the probability of simultaneous failure of multiple intermediate events in the fault tree to the minimum value of the corresponding test ratio of all intermediate events included in the intersection term, thereby obtaining a continuous optimization model based on the intersection term;

[0046] Step 2: Through theoretical analysis of the system inherent reliability calculation formula and combining it with expert experience to truncate the system inherent reliability calculation formula, a continuous optimization model based on real-time considerations is obtained;

[0047] Step 3: By additionally establishing the relationship between the test ratio of each test item and the inherent reliability of the system, a continuous optimization model based on additional constraints is obtained;

[0048] Step 4: Solve the continuous optimization model based on additional constraints, screen the test strategy based on the training set, and obtain the optimal test strategy.

[0049] This embodiment addresses the problem that existing methods are difficult to ensure the real-time performance of the optimization model solution while maintaining the cross-domain overall optimization capability, resulting in difficulty in meeting the requirements of the motherboard functional test process for rapid strategy design. This embodiment proposes a cross-domain overall optimization method for motherboard functional testing. By incorporating expert experience into the reliability analysis process and reasonably truncating the system's inherent reliability calculation formula, the real-time performance of the optimization model solution is significantly improved while ensuring the accuracy of the reliability calculation, and the test strategy design is made more reasonable. This embodiment provides a feasible path for the rapid design of a motherboard functional test strategy with cross-domain overall optimization capabilities, which helps to reduce costs and increase efficiency in flexible intelligent manufacturing processes.

[0050] In one embodiment, step 1 is to expand the discrete optimization model into a continuous optimization model, establish a fault tree, and set the intersection result of two intermediate events in the fault tree as the minimum value of the test ratio corresponding to all intermediate events included in the intersection term, thereby obtaining a continuous optimization model based on the intersection term, specifically:

[0051] Indicators like yield often only reflect the reliability of motherboard components over a short period of time. While yield can fluctuate frequently, the reliability of motherboard components, determined by fixed manufacturing processes, has a fixed value. Given that the reliability of the motherboard functional testing process depends on the reliability of the components being tested, if metrics like fault coverage are calculated based on the reliability of the motherboard functional testing process and an optimization model is established based on this, the resulting optimal solution is more likely to have stronger generalization capabilities than existing methods.

[0052] Functional testing is divided into two phases: motherboard functional testing and finished product functional testing. During the motherboard functional testing phase, the test strategy determines the test status of each test item and categorizes test items into tested and untested items based on the test strategy. A motherboard is considered a qualified product only if all tested items pass. If any item fails, it is considered a defective product and sent to a repair center for subsequent repair.

[0053] It is worth noting that if the motherboard has untested fault items (that is, potential defects not covered by the test strategy), the motherboard will be mistakenly judged as a good product. This situation is called a false negative motherboard. Such motherboards will be assembled into finished products in the subsequent production process and detected as defective products during the finished product functional testing stage. Because the finished product needs to be disassembled and repaired, and the repair cost of false negative motherboards is significantly higher than that of regular defective products, an excessively high false negative rate will directly lead to an increase in overall testing costs. During the finished product functional testing stage, all assembled finished products will undergo full functional testing to ensure that there are no missed detections at this stage and that all functional defects can be effectively detected.

[0054] During the motherboard functional testing phase, each test item is tested serially and independently. Therefore, the entire test system can be modeled as a reliability model consisting of two serially connected subsystems: one subsystem contains all tested items, and the other contains all untested items. The overall test reliability of the motherboard depends on the combined effect of these two subsystems.

[0055] Analysis shows that if a motherboard's fault happens to occur in a tested item, the fault will definitely be detected and the motherboard will be accurately judged as defective. Conversely, the generation of false negative motherboards is entirely due to the failure of the untested item. Therefore, for the functional test of the finished product, the tested item subsystem in the motherboard functional test can be considered completely reliable (reliability is 100%), and the unreliability of the entire test process is determined only by the untested item subsystem. Therefore, the inherent reliability of the system is calculated as follows

[0056] ; (1)

[0057] in, To test the strategy The total number of items included in is the test item index, To test the strategy Middle Test items, Indicates the Test items The test start signal is not active.

[0058] In step one, a reliability analysis of the functional test process was conducted, proposing that the use of system inherent reliability as a more accurate indicator of fault coverage is more accurate. Based on the fundamentals of reliability theory, system inherent reliability can be expressed as the product of the reliabilities of its component units. However, existing analysis methods for calculating system inherent reliability have a limitation: they fail to fully account for the impact of concurrent failures of modules containing the same underlying event on the overall system reliability.

[0059] Specifically, in fault tree modeling, we define the motherboard failure as the top event, the test item failure as the intermediate event, and the underlying factors that cause the test item failure as the bottom event. When using the fault tree method for quantitative calculations, the intermediate events are considered independent, ignoring the impact of repeated events in the minimum cut set on the calculation results. This simplification can lead to significant deviations in the calculation of the system's inherent reliability and ultimately affect the optimization of the test strategy.

[0060] by Figure 2 Taking the fault tree shown in the figure as an example, the calculation method of the probability of occurrence of the top event considering repeated events is explained in detail. In this fault tree, each event is connected by an OR gate, and the top event The fault may be caused by the fault of event M1, M2 or M3. By analyzing the fault tree structure, it can be found that events M1 and M2 contain the same event M5. Similarly, events M1, M2 and M3 also contain event M5.

[0061] To solve the problem of calculating the probability of concurrent failures of modules containing the same underlying event, we leveraged the module association information displayed in the fault tree. Specifically, we used the probability of event M5 to represent the probability of M1 and M2 occurring simultaneously. This approach effectively addressed the calculation bias issues found in existing methods.

[0062] This embodiment deeply analyzes the fault tree structure characteristics, quantitatively analyzes the impact of concurrent module failures containing the same basic event on system reliability, and derives the following calculation formula for the system inherent unreliability of the functional test process:

[0063] ; (1)

[0064] ; (2)

[0065] in, Top event The probability of occurrence, that is, the inherent unreliability of the system, is the total number of test items, is the test item index, For the Test items, For the The test start signal of the test item is not activated. Respectively The test start signal for the test item is not activated.

[0066] On this basis, the following optimization model is established:

[0067] ; (3)

[0068] ; (4)

[0069] ; (5)

[0070] in, To test the strategy, Indicates the The average test time for each test item, Indicates the test time threshold for each motherboard, For the The test start signal for a test item has been activated.

[0071] Since the test status in the test item selection method is single, each test item has only two states: tested and not tested. This two-category test status is effective for test items that are very necessary to test and those that are not necessary to test. However, for test items whose test necessity is between the two, whether they are set to tested or not tested, it is easy to make it difficult to further reduce the overall test cost. Therefore, this embodiment proposes a cross-domain overall optimization method for motherboard functional testing, which improves discrete testing into continuous testing, and each test item tests part of the motherboard.

[0072] Specifically:

[0073] remember is the test ratio of each test item, then The proportion of items not tested is , Respectively and The test ratio of the test items is Unreliability of the untested part of the test item It can be expressed as:

[0074] ; (6)

[0075] in, Indicates the The inherent reliability of the system for each test item is determined by the reliability of the motherboard components corresponding to the test item, which is expressed as The probability that a test item passes the test.

[0076] In the continuous planning problem, the motherboard functional test process is divided into two subsystems according to the test ratio: the subsystem composed of the untested part of all test items and the subsystem composed of the tested part of all test items. For the cost function test link, the reliability of the subsystem composed of the tested part is 100%. Therefore, the unreliability of the motherboard functional test process comes from the subsystem composed of the untested part of the test items.

[0077] Set the defective motherboard in the untested part of the motherboard function test process as the top event , No. The bottom event is that the untested part of the test item has a fault , the probability of occurrence of the bottom event is recorded as , then the top event The probability of occurrence Expressed as:

[0078] ; (7)

[0079]

[0080]

[0081] After the evolution of formula (7) to (9), the system inherent unreliability calculation formula (9) is finally obtained. All are item indexes. Refers to the first The failure of the untested part of each test item is regarded as the corresponding bottom event. It refers to the and The corresponding test ratio when the test items occur simultaneously, middle Representing an event The test ratio when failure occurs at the same time, Indicates test items The probability of simultaneous occurrence, Bottom event The probability of simultaneous occurrence, Indicates the The probability of failure of each test item is used to express the Unreliability of intermediate events The estimated value of The relationship with the reliability of the measurement items is as follows:

[0082] ; (10)

[0083] in, It represents the test ratio corresponding to the simultaneous occurrence of multiple events, which is closely related to the probability of occurrence of all events containing the same bottom event in the test item (i.e., the intermediate event). Take this as an example to illustrate. Indicates test items The corresponding test ratio when simultaneous events occur. 、 The test proportions of the corresponding test items are different, which makes it impossible to accurately determine the test proportion corresponding to the intersection result of two intermediate events. Therefore, it is set to the minimum value of the test proportions corresponding to all intermediate events contained in the intersection item, which can be expressed as:

[0084] ; (11)

[0085] Respectively and The proportion of tests for each test item (intermediate event).

[0086] Based on the above analysis, a continuous optimization model based on intersection considerations is established:

[0087] ; (12)

[0088] ; (13)

[0089] ; (14)

[0090] in, For the test ratio, is the inherent reliability of the system, that is, formula (12) is the reliability calculation formula, Indicates the The average test time for each test item, Indicates the test time threshold for each motherboard, is the total number of test items, For the The test ratio of each test item, The value range of the decision variables of the model.

[0091] The continuous optimization model based on intersection considerations achieves high-precision calculation of system reliability by expanding discrete test states into continuous test states and selecting reasonable probability values ​​for simultaneous failures of related modules, further optimizing the test strategy and reducing test costs.

[0092] In one embodiment, step 2 is to theoretically analyze the system inherent reliability calculation formula and truncate the system inherent reliability calculation formula based on expert experience to obtain a continuous optimization model based on real-time considerations, specifically:

[0093] In order to achieve high-precision calculation of reliability, the discrete optimization model is expanded to a continuous optimization model in step one. This increases the computational complexity of the system's inherent reliability calculation formula, and the calculation scale will increase with the increase in test items, resulting in a long time spent solving the continuous optimization model, which seriously affects the real-time acquisition of the test strategy and makes it difficult to meet the motherboard functional test process's demand for rapid strategy design.

[0094] Since the fault tree analysis method analyzes the system reliability from the perspective of system unreliability, and the sum of the system inherent reliability and the system inherent unreliability is always equal to 1, the calculation formula for analyzing the system inherent unreliability is essentially equivalent to the calculation formula for analyzing the system inherent reliability. Therefore, in order to solve the contradictory balance problem between the calculation accuracy and calculation complexity of the system reliability, this paper conducts a theoretical analysis of the formula (9) of the system inherent unreliability calculation formula based on expert guidance. The analysis process is as follows:

[0095] First, the structure of formula (9) is divided into sub-items, the number of sub-items is consistent with the number of test items, sub-items contain Small item, sub-items and The number of sub-items contained in the two sub-items is the same, so the two sub-items are divided into one group. Item, Item and Item, Item is taken as an example for analysis, .

[0096] No. The expression of the sub-item is:

[0097] ; (15)

[0098] Formula (15) is expanded to obtain formula (16):

[0099] ; (16)

[0100] No. The expression of the term is:

[0101] ; (17)

[0102] Formula (17) is expanded to obtain formula (18):

[0103] ; (18)

[0104] No. The expression of the term is:

[0105] ; (19)

[0106] Formula (19) is expanded to obtain formula (20):

[0107] ; (20)

[0108] No. The expression of the term is:

[0109] ;(twenty one)

[0110] Formula (21) is expanded to obtain formula (22):

[0111] ;(twenty two)

[0112] The relationship between the above four items in formula (9) is:

[0113] ;(twenty three)

[0114] Formula (23) is combined to obtain formula (24):

[0115] ;(twenty four)

[0116] For ease of understanding, formula (16) is the expanded form of formula (15), is the index of the sub-item in the sub-item, ,Right now , which corresponds to the first term in formula (16), namely . No. sub-items and The number of sub-items contained in the items is the same. According to formula (16) and formula (22), the result corresponding to the intersection item contained in formula (22) is a subset of the result corresponding to the intersection item contained in formula (16). That is to say, the probability of occurrence of the intersection item in formula (22) is less than or equal to the probability of occurrence of the intersection item contained in formula (16). At the same time, the test ratio of the intersection item in formula (22) is less than or equal to the test ratio of the intersection item contained in formula (16). Therefore, the overall result of the first part of formula (24) is greater than or equal to 0. Similarly, the result of the second half of formula (24) is less than or equal to 0, so the overall calculation result is close to 0. The first part of formula (24) is: , the second half of formula (24) is: .

[0117] Based on the above analysis, we know that in the actual calculation process, some of the sub-items in formula (9) contain small items that cancel each other out. Therefore, combined with the above analysis and expert experience, the system inherent unreliability calculation formula is reasonably truncated to retain the relevant information between the measurement items while reducing the complexity of the model. The results after truncation are as follows:

[0118] ; (25)

[0119] ; (26)

[0120] ; (27)

[0121] Based on the above analysis, the formulas (12)-(14) of the continuous optimization model based on intersection term consideration are improved to a continuous optimization model based on real-time consideration, specifically:

[0122] ; (28)

[0123] ; (29)

[0124] ; (30)

[0125] in, is the system inherent reliability obtained based on the calculation formula of the system inherent reliability after truncation.

[0126] In the continuous optimization model based on real-time considerations, by analyzing the system inherent reliability calculation formula, it is known that the impact of multiple events failing at the same time on the system inherent reliability is very small. Therefore, the system inherent reliability calculation formula is truncated, which not only ensures the calculation accuracy of the system reliability, but also can greatly reduce the model solution time. When facing changes in demand, the test strategy can be quickly adjusted.

[0127] In one embodiment, step three, by additionally establishing a relationship between the decision variables and the inherent reliability of the system, a continuous optimization model based on additional constraints is obtained, specifically:

[0128] The inherent reliability of an intermediate event (test item) essentially reflects the corresponding product pass rate of the test item. Test items with high inherent reliability have a relatively high product pass rate and a relatively low probability of a faulty motherboard. Conversely, test items with low inherent reliability have a relatively low product pass rate and a relatively high probability of a faulty motherboard. Therefore, during testing, to detect as many faulty motherboards as possible, the test strategy should be designed with a relatively low test ratio for test items with high inherent reliability and a relatively high test ratio for test items with low inherent reliability. This approach will optimize test efficiency and quality.

[0129] The decision variable value range of the continuous optimization model is For test items with relatively low inherent reliability and high average test time, a very low test ratio is often set. When the yield fluctuates, it is easy to cause the optimized test ratio to be too low. The generalization ability on the test set is poor, and the effect of reducing test costs is weakened.

[0130] Due to the test ratio It is closely related to the inherent reliability of the test items. Test items with high inherent reliability should be tested at a lower ratio, while test items with low inherent reliability should be tested at a higher ratio. Therefore, the test ratio is adjusted based on the process mechanism. The optimization interval of , the specific calculation process is as follows:

[0131] Based on the maintenance data of historical test data, the probability of occurrence of the bottom event is obtained as the unreliability of the bottom event. In order to take into account the calculation error, the mean and deviation of the probability of occurrence of the bottom event are obtained through mathematical statistics or combined with expert experience to determine the reliability range of the bottom event;

[0132] Based on the reliability range of the bottom event, the reliability range of each test item corresponding to the intermediate event is quantitatively calculated using the fault tree, and then the system inherent reliability of each test item is obtained, and the test ratio of each test item is calculated accordingly. The optimization interval of is set as a function based on the inherent reliability of the system:

[0133] ; (31)

[0134] in For the The test ratio of each test item, and Respectively The lower and upper limits of the system inherent reliability of each test item, and The test proportion coefficients are the lower and upper limits of the test proportion and the system inherent unreliability, respectively, and are set according to expert experience.

[0135] Based on the above analysis, the formulas (28)-(30) of the continuous optimization model based on real-time considerations are improved to a continuous optimization model based on additional constraints, specifically:

[0136] ; (32)

[0137] ; (33)

[0138] ; (34)

[0139] in, For the test ratio, is the system inherent reliability obtained based on the calculation formula of the system inherent reliability after truncation, Indicates the length is An array of all 1s.

[0140] It should be noted that since the total number of measurement items set is , and The lower and upper bounds of the system inherent reliability of the test item are arrays of length m. Since the system inherent unreliability of the test item needs to be calculated, a full 1 array of length is needed.

[0141] The continuous optimization model (32) - (34) based on additional constraints ensures that each test item can be allocated to a reasonable proportion of test resources by introducing additional constraints for decision variables. This improvement can effectively improve the fault detection rate and reduce the test cost even for test items with low system inherent reliability and long average test time.

[0142] In one embodiment, three evaluation indexes are used to verify the effectiveness of the method: the average test time of the motherboard, the false negative rate (FNR), and the average total test cost of the motherboard. The average test time of the motherboard and the defect level can evaluate the test efficiency and test quality of the test strategy, respectively. The calculation process of the indexes is as follows:

[0143] First, the number of actually qualified motherboards is calculated as:

[0144] ; (35)

[0145] wherein, represents the motherboard index, is the number of batch motherboards tested, represents the test result of the th test item of the th product, represents the passing test of the th test item of the th motherboard, is true for logical AND operation.

[0146] The actual number of defective motherboards is:

[0147] ; (36)

[0148] The number of detected defective motherboards when using the test strategy is:

[0149] ; (37)

[0150] wherein, represents the test failure of the th test item of the th product.

[0151] The number of detected defective motherboards The number of undetected defective motherboards for:

[0152] . (38)

[0153] The false negative rate indicates the ratio of the number of undetected defective motherboards to the total number of defective motherboards, and evaluates the detection capability of the test strategy for defective motherboards. The formula for calculating the false negative rate FNR is as follows:

[0154] ; (39)

[0155] in, For testing strategies The false negative rate.

[0156] Average test time refers to the average test time of testing a motherboard, which is used to evaluate the test efficiency of the test strategy. The average test time formula is as follows:

[0157] ; (40)

[0158] in, Indicates the The first product The test time for each test item.

[0159] Average motherboard testing cost It is the sum of the test time and repair time of a motherboard, which is used to evaluate the overall test effect of the test strategy. The calculation formula is as follows:

[0160] ; (41)

[0161] in, Indicates the average repair time of the motherboard. Indicates the average repair time of finished notebooks, To test strategies during the motherboard functional testing phase The number of faulty motherboards detected. The repair time for motherboards and finished products is determined based on statistical analysis of the actual test process and combined with the experience and evaluation of field experts.

[0162] The randomness of high-precision test strategies in the actual test process is specifically manifested in the process of determining the test status according to the test strategy. The test status of each test item of each motherboard can be expressed as:

[0163] ; (42)

[0164] in, Indicates the The first motherboard The test status of each test item, ⌉ indicates distance The nearest integer value is the number of motherboard batches tested.

[0165] For the same test ratio , test status May be different, resulting in different calculated evaluation index values. In order to determine the overall test effect of a test ratio, for each test ratio , randomly generated The test status of each time is calculated, and the evaluation index is calculated. The average value of the evaluation index is used to represent the test ratio The overall test effect.

[0166] The average test cost is used to evaluate the comprehensive test effect of the test strategy. The average test cost under continuous planning Expressed as:

[0167] ; (43)

[0168] in, Indicates the test ratio The first random number generated by formula (42) Group testing strategy, Indicates the test status The comprehensive evaluation index value on the training set or test set is calculated using formula (41).

[0169] Average test time is used to evaluate test efficiency, average test time under continuous planning Expressed as:

[0170] ; (44)

[0171] in, In test state The average test time on the training set or test set is calculated using formula (40).

[0172] FNR is used to evaluate the test quality of the test strategy. The calculation formula under continuous planning is:

[0173]

[0174] in, In test state The false negative rate on the training set or test set, that is, the fault missed detection rate, is calculated using formula (39).

[0175] This embodiment expands the discrete optimization model into a continuous optimization model to select reasonable probabilities of simultaneous failure of related modules. By incorporating expert experience into the reliability analysis process, the system inherent reliability calculation formula is reasonably truncated, thereby significantly improving the real-time performance of the optimization model while ensuring the accuracy of the system inherent reliability calculation. By additionally establishing the relationship between decision variables and system reliability, necessary constraints are added to the optimization model, further enhancing the generalization ability of the model.

[0176] In order to verify the effectiveness of this embodiment, the following comparison method is set up:

[0177] (1) A continuous optimization method without truncation. This method differs from the present embodiment in that the system inherent reliability calculation formula is not truncate.

[0178] (2) A continuous optimization method without setting additional constraints. The difference between this method and the present embodiment is that the constraints on the test ratio optimization range are canceled, and the test ratio optimization range of each test item is set to .

[0179] (3) A continuous optimization method without integrating correlation information. This method differs from the present embodiment in that, when analyzing the reliability of the functional test process, the test items are regarded as independent components in the system, and the connections between the test items are ignored.

[0180] (4) A binary optimization method considering reliability instability. The difference between this method and the present embodiment is that the test status of the test item in this method has only two settings: 0 and 1.

[0181] (5) A binary optimization method based on budget maximum coverage. The difference between this method and the present embodiment is that this method takes maximizing the fault coverage in the training set as the optimization model objective function, and the test status of the test item has only two settings: 0 and 1.

[0182] Among them, maximizing the fault coverage in the training set as the objective function of the optimization model is an existing method. For details, see the literature PAN R, ZHANG Z, LI X, et al. Black-Box Test-Coverage Analysis and Test-Cost Reduction Based on a Bayesian Network Model[C] / / 2019 IEEE 37th VLSI TestSymposium(VTS). 2019: 1-6.

[0183] The evaluation indicators of the method of this embodiment and its comparative method on the test set are shown in Table 1, and the model solution time of each method is shown in Table 2;

[0184] Table 1 Evaluation indicators of the method of the embodiment and its comparison method on the test set

[0185]

[0186] Table 2 Model solution time of the method of this embodiment and its comparative method

[0187]

[0188] Comparing the proposed method in Table 1 with the continuous optimization method without truncation, it can be seen that the truncation of the system inherent reliability calculation formula results in a slightly higher average total test cost of the obtained test strategy by 1% due to the loss of some relevant information to a certain extent. However, as can be seen from Table 2, the model solution time of the proposed method is only about 4% of that of the continuous optimization method without truncation. This treatment significantly reduces the model solution time, which allows for faster adjustment of the test strategy when facing the demand for rapid strategy design, reducing unnecessary testing costs.

[0189] The data in Tables 1 and 2 show that the method of this embodiment can effectively reduce the missed detection rate of defective motherboards, thereby reducing the total motherboard testing cost by up to 7.58%. By adding test states, this method enhances the flexibility of test strategy adjustment, thereby improving the detection rate of faulty motherboards and further reducing testing costs by 2.53%. Furthermore, by establishing the relationship between decision variables and system reliability under expert guidance and adding necessary constraints to the optimization model, the model optimization results are more reasonable and the model's generalization capability is further enhanced.

[0190] As an embodiment;

[0191] Taking a typical notebook motherboard as the research object, by converting the binary state of the test item into a continuous test state, the discrete optimization model was expanded into a continuous optimization model, achieving high-precision calculation of system reliability. On this basis, a human-computer collaborative method was used to reasonably truncate the system's inherent reliability calculation formula, significantly improving the real-time performance of the calculation while ensuring calculation accuracy. Furthermore, based on the experience of domain experts, a mapping relationship between decision variables and system reliability was constructed. By introducing key constraints, the optimization model architecture was improved. This not only enhanced the model's generalization ability in different production scenarios, but also designed a test strategy that meets the needs of flexible manufacturing. The specific implementation steps are as follows:

[0192] A1. Based on historical maintenance data, calculate the probability of occurrence of the bottom event and the reliability range of the bottom event;

[0193] A2. Quantitatively calculate the unreliability range of the test item corresponding to the intermediate event based on the established fault tree;

[0194] A3. Based on expert experience, set the lower limit test ratio coefficient in formula (31) 100, upper limit test ratio coefficient is 2500, and then the test ratio is obtained The optimization range of , that is, the optimization range of decision variables, the results are shown in Table 3;

[0195] Table 3 Test ratio The optimization interval

[0196]

[0197] A4. Test ratio based on A3 The optimization range is , and the following continuous optimization model is established:

[0198] ; (46)

[0199] ; (47)

[0200] ; (48)

[0201] in, For test ratio The lower limit of For test ratio upper limit.

[0202] A5. Based on the progress analysis of the motherboard manufacturing process and expert experience evaluation, the test time threshold for each motherboard is set at 75 seconds;

[0203] A6. Combined with the statistical analysis of historical maintenance data and comprehensive evaluation of expert experience, the motherboard repair cost Set to 0.5h, the cost of finished product rework Set to 4h;

[0204] A7. Solve the continuous optimization model established in A4 to obtain the optimal test strategy. Calculate the evaluation indicators corresponding to the test strategies solved by the proposed method and the comparative method. The detailed results are shown in Table 1. By analyzing the evaluation indicators in Table 1, the effectiveness of the method proposed in this embodiment is verified.

[0205] In one embodiment, a computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the method described above when executing the computer program. The processor is responsible for performing computation and control functions; the memory includes a non-volatile storage medium and a memory, wherein the non-volatile storage medium stores an operating system, a computer program, and a database, and the memory provides a support environment for the operation of the operating system and the program; and a network interface enables network communication between the device and an external terminal. The database is specifically used to store data required for the relevant method, and when the processor executes the computer program, the method is implemented.

[0206] In one embodiment, the present invention also provides a computer-readable storage medium, on which several classification programs are stored, and the several classification programs are used to be called by the processor and execute the method as described above. Ordinary technicians in this field can understand that all or part of the steps of implementing the above method embodiment can be completed by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium, and when the program is executed, it executes the steps of the above method embodiment; and the aforementioned storage medium includes: ROM, RAM, disk or optical disk and other media that can store program codes.

[0207] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0208] The above description is only a preferred specific embodiment of the present invention, which is used to help understand the technical solution of the present invention, but the scope of protection of the present invention is not limited to this. Any technician familiar with this technical field, within the technical scope disclosed by the present invention, can make equivalent replacements or changes based on the technical solution and inventive concept of the present invention, which should be covered by the scope of protection of the present invention.

Claims

1. A cross-domain optimization method for high real-time functional testing of human-machine collaboration, characterized by: include: Step 1: Expand the discrete optimization model into a continuous optimization model, establish a fault tree, and set the probability of simultaneous failure of multiple intermediate events in the fault tree to the minimum value of the corresponding test ratio of all intermediate events included in the intersection term, thereby obtaining a continuous optimization model based on the intersection term; Step 2: Through theoretical analysis of the system inherent reliability calculation formula and combining it with expert experience to truncate the system inherent reliability calculation formula, a continuous optimization model based on real-time considerations is obtained; Step 3: By additionally establishing the relationship between the test ratio of each test item and the inherent reliability of the system, a continuous optimization model based on additional constraints is obtained; Step 4: Solve the continuous optimization model based on the additional constraints, and screen the test strategies based on the training set to obtain the optimal test strategy; Step 2 is as follows: The calculation formula of the system inherent unreliability is divided into sub-items, the number of sub-items is consistent with the number of test items, sub-items contain Small item, sub-items and The number of sub-items contained in each sub-item is the same, so the two sub-items are divided into one group; by analyzing the expression of the divided sub-items, it is found that the sub-items contained in some sub-items in the system inherent unreliability calculation formula cancel each other out; because the sum of the system inherent reliability and the system inherent unreliability is always equal to 1, the system inherent reliability calculation formula is truncated based on expert experience, and a continuous optimization model based on real-time considerations is obtained; Step three is as follows: Based on the maintenance data of historical test data, the probability of occurrence of the bottom event is obtained as the unreliability of the bottom event. The mean and deviation of the probability of occurrence of the bottom event are obtained in combination with expert experience, and then the reliability range of the bottom event is determined. Based on the reliability range of the bottom event, the fault tree is used to quantitatively calculate the system inherent reliability range of each test item corresponding to the intermediate event, and the test proportion of each test item is calculated accordingly. The optimization interval is set as a function based on the inherent reliability of the system.

2. The cross-domain optimization method for functional testing according to claim 1, characterized in that: In the fault tree, the mainboard failure is regarded as the top event, the test item failure is regarded as the intermediate event, and the underlying cause of the test item failure is regarded as the bottom event of the fault tree.

3. The cross-domain optimization method for functional testing according to claim 2, characterized in that: In expanding the discrete optimization model to a continuous optimization model, specifically: By conducting reliability analysis on the motherboard functional test process, the relationship between fault coverage and system inherent reliability is determined; Based on the fault tree analysis, the impact of concurrent failures of intermediate events containing the same bottom event on the inherent reliability of the system is analyzed; The system reliability of the motherboard functional test process is derived to build a continuous optimization model.

4. The cross-domain optimization method for functional testing according to claim 1, characterized in that: Test ratio The optimization interval is specifically: ； in, and Respectively The lower and upper limits of the system inherent reliability of each test item, and The proportionality coefficients for the test ratio and the lower and upper limits of the system's inherent unreliability are set based on expert experience.

5. The cross-domain optimization method for functional testing according to claim 1, characterized in that: The continuous optimization model based on additional constraints is specifically: ； ； ； in, For the test ratio, is the system inherent reliability obtained based on the calculation formula of the system inherent reliability after truncation, Indicates the The average test time for each test item, Indicates the test time threshold for each motherboard, is the total number of test items, For the The test ratio of each test item, and are the lower and upper limits of the system’s inherent reliability, and are the proportional coefficients of the test ratio to the lower and upper limits of the system inherent unreliability, set according to expert experience, Indicates the length is An array of all 1s.

6. The cross-domain optimization method for functional testing according to claim 1, characterized in that: Three evaluation indicators are set to verify the effectiveness of the cross-domain overall optimization method for motherboard functional testing. The evaluation indicators include the average motherboard test time, false negative rate, and the average total cost of motherboard testing. False negative rate The calculation formula is as follows: ； in, Indicates the test ratio The randomly generated test strategy and the test status of each test item of each motherboard are Group correspondence, For test ratio The corresponding The false negative rate of group correspondence on the training set or test set; Average motherboard test time The calculation formula is as follows: ； in, Indicates the test ratio Randomly generated test strategy The test status of each test item of each motherboard Group correspondence, For test ratio The corresponding The average test time of group correspondences on the training set or test set.

7. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the method according to any one of claims 1 to 6 is implemented.

8. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a plurality of classification programs, which are used to be called by a processor and execute the method according to any one of claims 1 to 6.

Images