Optimization test method and device for rapid fault diagnosis, equipment and medium

Abstract

The invention discloses an optimization test method, device and equipment for rapid fault diagnosis and a medium, and belongs to the technical field of correlation matrix optimization and block type test optimization technologies for rapid fault diagnosis tests and testability analysis. After redundant tests and invalid tests in the correlation matrix are removed, the correlation matrix is optimized by considering the particularity of the whole row and the whole column, the optimized correlation matrix is subjected to block processing based on the correlation between the tests, and a large matrix is divided into a plurality of small correlation matrixes; judging whether the test group in each separated correlation matrix is an optimal test group or not; and directly adding the optimal test group into the final optimal test group, and performing test optimization on the non-optimal test group by using the improved discrete cuckoo algorithm. According to the method, optimization and blocking of the large-scale correlation matrix are realized, optimization of tests in the correlation matrix is carried out, and under the condition that performance indexes are ensured to be unchanged, the test cost is reduced, and the fault diagnosis speed is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of testability analysis, and in particular to an optimized test method, device, equipment and medium for rapid fault diagnosis. Background Art

[0002] Testability design is an important cutting-edge technology to improve system reliability, maintainability, safety, and security, and directly affects the fault detection rate, isolation rate, and system maintenance (testing) cost. The correlation matrix is ​​a 0-1 model used in testability design to describe the relationship between faults and tests. By judging whether each test in the correlation matrix is ​​normal, a test 0-1 sequence is obtained, which is compared with the row vector in the correlation matrix to achieve fault diagnosis.

[0003] In the prior art, for complex systems, the correlation matrix derived from test modeling is often large and complex, and contains many useless and redundant tests, which easily leads to excessively high time costs for subsequent system analysis and fault diagnosis. Faced with these technical problems, technicians in this field only remove duplicate columns, all-0 columns, or even all-1 rows and all-1 columns when processing the correlation matrix. Of course, it is also possible to further use intelligent algorithms for test optimization. These methods reduce the scale of the correlation matrix to a certain extent, but at the same time there are also problems such as insufficient removal of redundant tests, long test optimization time, and decreased indicators after optimization. Summary of the invention

[0004] In view of this, the purpose of the present invention is to provide an optimized test method, device, equipment and medium for rapid fault diagnosis, by analyzing unnecessary rows and columns in the correlation matrix and optimizing them, in the preliminary optimization process of the correlation matrix, the particularity of all-1 rows and all-1 columns is fully considered, after the preliminary optimization, the correlation matrix is ​​divided into blocks, and divided into relatively simple correlation matrices, on this basis, the improved discrete cuckoo algorithm is used for test optimization, and the time for test optimization is reduced while removing redundant tests as fully as possible, ensuring that the optimized correlation matrix index does not decrease. The specific scheme is as follows:

[0005] In a first aspect, the present application discloses an optimized test method for rapid fault diagnosis, comprising:

[0006] Redundant and invalid removal: remove redundant tests and invalid tests;

[0007] Fault model and test optimization, optimizing the failure modes that can be detected by all tests in the dependency matrix and the tests that can detect all failure modes;

[0008] Divide the test set into several related test groups based on the correlation between the tests;

[0009] The tests in the optimal test group are directly added to the final optimal test set, and the non-optimal test groups are optimized and added to the final optimal test set after optimization;

[0010] The tests of the optimal test set and the tests selected from the non-optimal test set are integrated to obtain the final optimal test set.

[0011] Optionally, the optimization of the fault modes that can be detected by all tests in the correlation matrix includes removing the all-1 row in the correlation matrix. If there is no all-0 column in the correlation matrix at this time, the optimization of the all-1 row is terminated; if an all-0 column appears after removing the all-1 row in the correlation matrix, the test corresponding to the all-0 column is added to the final preferred test set, and then the all-0 column is removed, and it is looped to determine whether there is an all-1 row. If so, another loop optimization is performed until there is no all-1 row, and the optimization of the all-1 row is terminated.

[0012] Optionally, the optimizing the tests that can detect all failure modes in the correlation matrix includes,

[0013] Remove the all-1 column in the correlation matrix. If there is no all-0 row in the correlation matrix at this time, the optimization of the all-1 column ends. If all-0 rows appear after removing the all-1 column in the correlation matrix, add the test corresponding to the all-1 column to the final preferred test set, and then remove the all-0 row, and loop to determine whether there is an all-1 column. If so, perform another loop optimization until there is no all-1 column, and then the optimization of the all-1 column ends.

[0014] Optionally, dividing the test set into a number of related test groups according to the correlation between the tests includes:

[0015] Selecting a test in a dependency matrix and setting a fault group that has an impact on the test;

[0016] The fault in the fault group also affects the second test other than the test, the second test constitutes a second test group, the second test is a directly related test of the test, the second test group is a directly related test group, and the test and the directly related test group constitute a related test group;

[0017] Selecting directly related tests of each test in the related test group that are not in the related test group and adding them to the related test group, until there are no tests outside the related test group in the correlation matrix that are directly related to the tests in the related test group;

[0018] Repeat the above operation for the remaining tests in the correlation matrix until all tests have corresponding related test groups;

[0019] The new correlation matrix formed by the relevant test groups is analyzed, and the relevant test groups are divided into optimal test groups and non-optimal test groups; wherein, the optimal test group refers to the test group whose fault detection rate or fault isolation rate will decrease when any test in the optimal test group is removed; the non-optimal test group is the test group other than the optimal test group.

[0020] Optionally, for the non-optimal test group, an improved discrete cuckoo algorithm is used to perform test optimization, wherein the improved discrete cuckoo algorithm is specifically:

[0021] Initialize the parameters of the cuckoo algorithm, where the cuckoo position is a binary vector and the nest position represents the solution to the test point optimization selection problem;

[0022] Calculate the fitness function of each current bird's nest position and construct a solution model; the solution model is related to the fault detection rate and the fault isolation rate;

[0023] Find the location of the optimal bird's nest and the optimal solution according to the fitness function, and determine whether it is the same as the optimal solution of the previous iteration. If the number of consecutive identical solutions reaches the preset threshold, it is considered to be converged, and then the fault detection rate and fault isolation rate are optimized;

[0024] For the one-fifth of the total number of bird nests with the worst location, they are removed with a preset probability, and if removed, new bird nests are generated to replace them;

[0025] For each nest location, use Levi's flight to get the new nest location;

[0026] The obtained new bird's nest position is a real number vector, the new bird's nest position is binarized, the new bird's nest position is compared with the original bird's nest position, and the better position is used as the current bird's nest position;

[0027] Determine whether the output conditions are met. If so, output the result. Otherwise, switch to the fitness function calculation and continue the cycle.

[0028] Optionally, in the improved discrete cuckoo algorithm, the quality of the cuckoo's choice of nest location is proportional to the value of the fitness function of the test point optimization selection problem. When multi-objective optimization is used to solve the objective function, the higher the system's fault detection rate, the higher the fault isolation rate, the higher the test cost of not being selected into the preferred test set, and the larger the fitness function value.

[0029] Optionally, when the algorithm converges, the fault detection rate and the fault isolation rate are optimized, specifically including the optimization of the fault detection rate and the optimization of the fault isolation rate.

[0030] Optimize the fault detection rate by adding the lowest-cost test that can detect the faults that were previously detectable but are currently undetectable to the current test solution set;

[0031] The fault isolation rate is optimized by adding all tests that can detect the faults that could have been isolated before but cannot be isolated now to the current test solution set.

[0032] In a second aspect, the present application discloses an optimized test device for rapid fault diagnosis, comprising:

[0033] Redundant and invalid removal module, used to remove redundant tests and invalid tests;

[0034] Optimization module for fault model and test optimization, optimizing the fault modes that can be detected by all tests in the correlation matrix and the tests that can detect all fault modes;

[0035] The block module is used to divide the test set into several related test groups according to the correlation between the tests;

[0036] The optimization module is used to directly add the tests in the optimal test group to the final optimal test set, and optimize the tests in the non-optimal test group, and then add them to the final optimal test set after optimization;

[0037] The output module is used to integrate the tests of the optimal test set and the tests selected from the non-optimal test set to obtain the final optimal test set.

[0038] In a third aspect, the present application discloses an electronic device, comprising:

[0039] Memory, used to store computer programs;

[0040] A processor is used to execute the computer program to implement the steps of the aforementioned disclosed optimized test method for rapid fault diagnosis.

[0041] In a fourth aspect, the present application discloses a computer-readable storage medium for storing a computer program; wherein, when the computer program is executed by a processor, the steps of the aforementioned disclosed optimized test method for rapid fault diagnosis are implemented.

[0042] As can be seen from the above, the present invention discloses an optimized test method, device, equipment and medium for rapid fault diagnosis, a correlation matrix optimization and block test optimization technology for rapid fault diagnosis testing, and the field of testability analysis technology. After removing redundant tests and invalid tests in the correlation matrix, the correlation matrix is ​​optimized considering the particularity of a full row and a full column, and the optimized correlation matrix is ​​processed in blocks based on the correlation between the tests, and the large matrix is ​​divided into multiple small correlation matrices; each test group in the separated correlation matrix is ​​judged once to see whether it is the optimal test group; the optimal test group is directly added to the final preferred test group, and the improved discrete cuckoo algorithm is used for test optimization of non-optimal test groups. The present invention realizes the optimization and blocking of large correlation matrices, as well as the optimization of tests in the correlation matrix, which reduces the test cost and improves the fault diagnosis speed while ensuring that the performance indicators remain unchanged. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0044] Figure 1 A flow chart of the optimized test method for rapid fault diagnosis disclosed in this application;

[0045] Figure 2 A full 1-row flow chart for optimizing the correlation matrix disclosed in this application;

[0046] Figure 3 A full 1-column flow chart for optimizing the correlation matrix disclosed in this application;

[0047] Figure 4 A flowchart of the correlation matrix partitioning method disclosed in this application;

[0048] Figure 5 The flowchart of the improved discrete cuckoo algorithm disclosed in this application;

[0049] Figure 6 This is a schematic diagram of the structure of an optimized test device for rapid fault diagnosis disclosed in this application;

[0050] Figure 7 This is a schematic diagram of the structure of the optimized test equipment for rapid fault diagnosis disclosed in this application. DETAILED DESCRIPTION

[0051] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0052] In this field, as the requirements for equipment fault diagnosis capabilities continue to increase, the optimization of test points for observing system states in equipment testability design also puts forward higher requirements. The correlation matrix is ​​a 0-1 model used in testability design to describe the relationship between faults and tests. For the optimization problem of the correlation matrix, technicians in this field first directly delete all 0 rows, all 0 columns, all 1 rows, and all 1 columns in the matrix, and then perform subsequent optimization on the entire matrix after a round of optimization. The subsequent optimization of the correlation matrix is ​​mainly to remove some of the tests from the test group, and ensure the simplicity of the test under the condition of ensuring the testability index, thereby reducing the test cost. Test optimization is mainly divided into two parts. The first is to remove redundant tests, and the test with and without the test can detect the same set of faults. The performance in the correlation matrix is ​​to remove the columns with the same columns. After the redundant tests are removed, the test optimization is specifically manifested in selecting the optimal test subset from a large number of test sets based on the fault-test correlation matrix of the system, so that the test set can meet the required testability index and the test cost is as small as possible, which has become a technical problem that needs to be solved urgently.

[0053] Currently, the test optimization problem is a non-deterministic problem of polynomial complexity, that is, an NP-complete problem, and all its solutions have 2 n where n is the number of tests. However, to compare each solution is a huge workload for projects with a large number of tests. In this field, metaheuristic search algorithms are intelligent optimization algorithms based on multi-agent interaction, and have great application prospects for finding global optimal solutions to NP-hard problems, multi-objective optimization, and nonlinear optimization problems. Metaheuristic search algorithms such as genetic algorithms (GA), particle swarm algorithms (PSO), quantum evolutionary algorithms (QEA), and hybrid algorithms such as binary particle swarm-genetic algorithms have been used to solve the test point optimization selection problem, and the solution efficiency and solution quality have been improved to a certain extent.

[0054] During the initial optimization of the correlation matrix, there are currently two ways to handle all-1 rows and all-1 columns, namely, directly deleting them completely or completely retaining them. In most cases, all-1 rows and all-1 columns are meaningless for fault diagnosis, so they are basically not completely retained. If all-1 columns are directly removed, all-0 rows may appear, and removing all-0 rows may cause the originally detectable faults to be undetectable. The same is true for directly removing all-1 rows. Therefore, if there are all-1 rows and all-1 columns in the correlation matrix, they need to be optimized.

[0055] In the test optimization problem, the metaheuristic algorithm used in this field is directly applied to a large correlation matrix, which often results in a high running time cost and a local optimum, resulting in the optimal test subset being obtained after a long run and still not meeting the required testability indicators. However, if the correlation matrix can be split, the algorithm's running efficiency can be improved.

[0056] At the same time, when selecting tests, the overall testability index (such as fault detection rate and fault isolation rate) is more important than the cost of one or two more tests. Ensuring that the testability index of the correlation matrix composed of the preferred test subset is optimal will help the design and detection of the system. However, when most heuristic algorithms process large correlation matrices, the final result may lead to a decrease in testability index.

[0057] Based on this, the present invention discloses an optimization test for rapid fault diagnosis. The present invention optimizes by analyzing unnecessary rows and columns in a correlation matrix, and then divides the matrix according to the correlation between tests, and uses an improved discrete cuckoo algorithm to perform test optimization.

[0058] See also Figure 1 As shown, the embodiment of the present application discloses an optimized test method for rapid fault diagnosis, the method comprising:

[0059] Step 100, redundant and invalid removal, removing redundant tests and invalid tests;

[0060] Redundant tests with the same effect are usually represented by the same column in the correlation matrix. For these tests, the test with the lowest cost is retained and the rest are removed. Invalid tests are represented by all-0 columns in the correlation matrix. The invalid tests cannot detect any faults, so they are also removed in this embodiment.

[0061] This step removes redundant tests and invalid tests with the same effect in the correlation matrix. Select a group of tests with the same detection characteristics that can detect a group of the same faults, that is, compare the columns in the correlation matrix D, Ti = Tj (i ≠ j). Only the test with the lowest cost is retained in each group. If there is more than one lowest test, the more than one lowest test is formed into a set, which is a replaceable test set, that is, one of the tests in the optional set can be added to the final optimal test set.

[0062] Step 200, fault model and test optimization, optimizing the fault modes that can be detected by all tests in the correlation matrix and the tests that can detect all fault modes;

[0063] In this embodiment, the failure mode that can be detected by all tests in the correlation matrix is ​​optimized. The failure mode is represented by a full 1 row in the correlation matrix, as shown in the attached figure. Figure 2 As shown, the optimization method is:

[0064] Remove all 1 rows in the correlation matrix. If there are no all 0 columns in the correlation matrix at this time, the all 1 row optimization ends;

[0065] If an all-0 column appears after removing the all-1 row in the correlation matrix, it is necessary to add the test t corresponding to the all-0 column to the final preferred test set, and then remove the all-0 column. After a round of optimization, determine again whether there is an all-1 row. If so, perform another round of optimization until there is no all-1 row, and then the all-1 row optimization is completed.

[0066] Further, in this embodiment, the test that can detect all failure modes in the correlation matrix is ​​optimized, and the test is represented by a full 1 column in the correlation matrix, as shown in the attached figure. Figure 3 As shown, the optimization method is:

[0067] Remove all 1 columns from the correlation matrix. If there are no all 0 rows in the correlation matrix, the all 1 column optimization ends.

[0068] If all-0 rows appear after removing the all-1 column in the correlation matrix, it is necessary to add the test t corresponding to the all-1 column to the final preferred test set, and then remove the all-0 row. After a round of optimization, determine again whether there is an all-1 column. If so, perform another round of optimization until there is no all-1 column, and then the all-1 column optimization is completed.

[0069] Step 300, dividing the test set into a number of related test groups according to the correlation between the tests;

[0070] In this embodiment, as shown in the attached Figure 4 As shown, the test set is divided into several related test groups, including:

[0071] Step 301, select a test t in the correlation matrix, and set the fault group that affects the test t to be S = [s0, s1, ..., sm];

[0072] Step 302, the fault in the fault group S also affects other tests except the test t. Suppose these test groups are T1=[t0, t1,…, tn], then these tests are directly related tests of t. It can be understood that t0~tn represent all tests corresponding to 1 in the row where the fault of set S is located, that is, the tests that will be set to 1 when the fault occurs. t and the directly related test group T1 constitute the related test group T.

[0073] Step 303, further selecting directly related tests of each test in the related test group T that is not in T and adding them to T, until there is no test other than T in the correlation matrix that is directly related to the test in T;

[0074] It should be noted that the purpose of this step is to ensure that all directly related tests of all tests included in the related test group T are in T, and the tests outside T are not directly related to any test in T. At this time, each test in T is a related test, and T is a related test group.

[0075] Step 304: Repeat the above operation for the remaining tests in the correlation matrix until all tests have corresponding related test groups.

[0076] Further, the remaining tests are all the tests except the tests that have been selected into the relevant test group, that is, all the tests - the tests that have been selected into the relevant test group = the remaining tests.

[0077] In this embodiment, all original tests are divided into multiple groups, and the tests in each group are related, while any two tests in different test groups, such as group 1 and group 2, are not related. It should be emphasized that the above-mentioned related test group is recorded as T, and a specific related test group is used as an example for explanation. In fact, the related test group T is a related test group among multiple groups, in order to specifically illustrate that the tests in the related test group are obtained according to the correlation.

[0078] It can be understood that the matrix expression in this embodiment is to change the correlation matrix into a block diagonal matrix by exchanging rows or columns of the correlation matrix, such as: Among them, A and B are matrices that can be separated and tested separately for optimization, as shown in the attached Figure 4 shown.

[0079] Step 305, analyzing the new correlation matrix formed by the related test groups, and dividing the related test groups into optimal test groups and non-optimal test groups;

[0080] The optimal test group refers to the test group in which the fault detection rate or fault isolation rate of the correlation matrix formed by the test group will decrease when any test in the test group is removed. Such an optimal test group may exist in the separated test groups, and the optimal test group can be directly added to the final solution set. The non-optimal test group is the test group other than the optimal test group.

[0081] Step 400, directly add the tests in the optimal test group to the final optimal test set, perform test optimization on the non-optimal test group, and then add them to the final optimal test set after optimization.

[0082] For non-optimal test groups, the improved discrete cuckoo algorithm is used to test and optimize them, and then they are added to the final optimized test set.

[0083] The specific steps of the improved discrete cuckoo algorithm are as follows:

[0084] Step 401, initializing the parameters of the cuckoo algorithm, the cuckoo position is a binary vector, and the nest position represents the solution to the test point optimization selection problem;

[0085] Initialize the parameters of the cuckoo algorithm. The position of the cuckoo is represented by a binary vector. Each cuckoo's choice of nest position represents a solution to the test point optimization selection problem. Let the cuckoo choose the nest position x i =[ts 1 ,ts 2 ,…,ts n ] is a binary vector, i.e. ts 1 ~ts n The values ​​are 0 or 1, 0 means not to select the test corresponding to the value as the optimal test for the final selection, 1 means to select it, and n is the number of tests in the candidate test set T. k Selected (ie t k ∈T,t k Indicates test) time ts k =1, k∈[1,n], otherwise ts k =0.

[0086] The number of solutions in each iteration is m, where m = number of tests + 20. m initial solution vectors M = (x 1 ,x 2 …,x m ) as the initial location of each cuckoo's nest, FDR r and FIR r are the original fault detection rate and the original fault isolation rate;

[0087] Step 402, calculating the fitness function of each current bird's nest position and constructing a solution model; the solution model is related to the fault detection rate and the fault isolation rate;

[0088] The fitness function of each current bird's nest position is calculated. Considering the importance of the two indicators of fault detection rate and fault isolation rate in industrial practical applications, in this embodiment, the testability index of the correlation matrix after test optimization is consistent with that before test optimization, while ensuring the minimum test cost. The solution model for the test point optimization selection problem can be obtained as shown in formula (1):

[0089]

[0090] Among them, FDR r and FIR r To test the fault detection rate and fault isolation rate of the correlation matrix before optimization, c i is the cost of the ith test.

[0091] It should be noted that since the quality of the cuckoo's nest location selection is proportional to the value of the fitness function of the test point optimization selection problem, it is only necessary to solve the fitness function. Thus, the problem of formula (1) is transformed into a maximum optimization problem. When multi-objective optimization is used to solve the objective function, the higher the system's fault detection rate, the higher the fault isolation rate, and the higher the test cost not selected in Ts (Ts represents the test set selected from T), the greater the fitness function value. The fitness function of the test point optimization selection problem is shown in formula (2):

[0092]

[0093] Among them, Ts represents the test set selected from T, C represents the test cost set, and is the fault detection rate and fault isolation rate of the correlation matrix of the optimal test, rFD and rFI are the fault detection rate and fault isolation rate of the original correlation matrix, q is a constant, which can usually be taken as 1.2.

[0094] Step 403, find the location of the optimal bird's nest and the optimal solution according to the fitness function, and determine whether it is the same as the optimal solution of the previous iteration. If the number of consecutive identical solutions reaches a preset threshold, it is considered to be converged, and the fault detection rate optimization and fault isolation rate optimization are performed once;

[0095] Find the optimal nest location based on the fitness function The fitness function fmax of the optimal solution is used to determine whether it is the same as the optimal solution of the previous iteration. If the number of consecutive identical solutions reaches 10, it is considered to have converged to a certain extent, and the fault detection rate optimization and fault isolation rate optimization are performed once.

[0096] In this embodiment, it is impossible to detect or isolate the faults that were originally undetectable or impossible to isolate without adding new tests. Therefore, the optimal values ​​of the fault detection rate and isolation rate of the correlation matrix of the test composition after the optimization test are consistent with the indicators before the optimization. When the algorithm converges to a certain degree, it is proposed to optimize the fault detection rate and fault isolation rate. The specific method is:

[0097] 1) Optimizing the fault detection rate: Add the lowest-cost test to the current test solution set that can detect faults that were previously detectable but are currently undetectable.

[0098] 2) Optimize the fault isolation rate: Add all tests that can detect faults that could be isolated originally but cannot be isolated currently to the current test solution set.

[0099] Step 404: For the bird nests with the worst position among the total number of nests, one fifth are removed with a probability of p. Usually, p can be p=0.8. If removed, a new bird nest is randomly generated to replace it.

[0100] Among them, the total number of groups is the total number of solutions, and the total group with the worst position is determined according to the fitness function.

[0101] Step 405: using Levy flight to obtain a new bird's nest position for each bird's nest position;

[0102] Wherein, the Levy flight formula is shown in formula (3) and formula (4):

[0103]

[0104]

[0105] Where x is the solution, i.e. the location of the bird's nest, and α>0 is the step-size proportional factor, which can usually be 1. λ=1, Among them, u~N(0,σ2), v~N(0,1), Here, β=1.5 is taken.

[0106] Step 406, the new bird's nest position is obtained as a real vector, the new bird's nest position is binarized, the new bird's nest position is compared with the original bird's nest position, and the better position is used as the current bird's nest position;

[0107] Get the new nest location It will become a real vector, so the new position needs to be binarized. The Sigmoid function can realize the mapping from the real number domain to the 0-1 interval, as shown in equations (5) and (6) respectively:

[0108]

[0109] Formula (5) is translated and scaled, and the improved formula (6) has a better binarization processing effect.

[0110]

[0111] The binary calculation formula for the new nest location that the cuckoo is looking for is shown in formula (7):

[0112]

[0113] In the formula, rand is a random number in the range of 0-1.

[0114] The obtained new position is compared with the original position, and the better position is retained, that is, the position corresponding to the higher fitness function.

[0115] Step 407, determine whether the output condition is met, if so, output the optimal solution xb, otherwise go to step 402;

[0116] Determine whether the output condition is met, that is, whether the number of times the optimal solution remains unchanged is greater than the unchanged threshold. Preferably, the unchanged threshold can be set to 20, 30, 40 times, etc. If yes, output the optimal solution x b , if not satisfied, go to step 402. The flowchart is as shown in the attached Figure 5 shown.

[0117] Step 500, integrating the tests of the optimal test set and the tests selected from the non-optimal test set to obtain a final optimal test set.

[0118] In order to better illustrate the technical solution of the present invention, a specific example is used to describe the present invention in detail. Table 1 below is a correlation matrix of this embodiment.

[0119] Table 1

[0120] D t1 t2 t3 t4 t5 t6 t7 t8 t9 t10 t11 t12 t13 t14 t15 t16 t17 t18 t19 t20 t21 t22 t23 t24 t25 t26 s1 1 1 1 1 1 0 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 s2 0 0 1 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 s3 0 1 1 1 1 0 1 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 s4 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 s5 0 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 s6 0 1 1 1 1 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 s7 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 s8 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 s9 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 1 s10 0 0 0 0 0 0 0 0 0 1 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 1 s11 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 1 s12 0 0 0 0 0 0 0 0 0 1 0 1 0 1 0 0 1 0 0 0 0 0 1 0 0 1 s13 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 1 0 0 1 0 0 0 0 1 s14 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 1 s15 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0 0 0 1 0 1 0 0 0 0 1 s16 0 0 0 0 0 0 0 0 0 1 0 1 0 1 0 0 1 0 0 1 0 1 1 0 0 1 s17 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 1 0 0 0 0 0 0 0 0 1 s18 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 0 0 0 0 0 0 0 0 1 s19 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 0 1 s20 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

[0121] As shown in Table 1, the system of this embodiment has a total of 20 faults and 26 alternative tests. The elements in the table indicate whether the corresponding test can test the corresponding fault. If it can be tested, it is 1, otherwise it is 0. The initial fault detection rate and fault isolation rate of the correlation matrix are both 100%, that is, all faults can be detected and isolated. The test cost is 1.

[0122] Specifically, the technical solution of the present invention is:

[0123] First, three redundant test groups (t4, t5), (t11, t19), and (t20, t22) are obtained, and the test with the minimum cost is retained in each redundant test group.

[0124] Then, the optimization is performed to optimize all 1 rows and all 1 columns, where t25 is removed and added to the necessary test group. After optimization, a new correlation matrix is ​​obtained, as shown in Table 2.

[0125] Table 2

[0126] D t1 t2 t3 t4 t6 t7 t8 t9 t10 t11 t12 t13 t14 t15 t16 t17 t18 t20 t21 t23 t24 s1 1 1 1 1 0 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 s2 0 0 1 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 s3 0 1 1 1 0 1 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 s4 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 s5 0 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 s6 0 1 1 1 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 s7 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 s8 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 s9 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 1 0 s10 0 0 0 0 0 0 0 0 1 0 0 1 0 0 0 0 0 0 0 0 0 s11 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 1 0 0 s12 0 0 0 0 0 0 0 0 1 0 1 0 1 0 0 1 0 0 0 1 0 s13 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 1 0 1 0 0 s14 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 s15 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 1 0 0 s16 0 0 0 0 0 0 0 0 1 0 1 0 1 0 0 1 0 1 0 1 0 s17 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 1 0 0 0 0 0 s18 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 0 0 0 0 0 s19 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0 0

[0127] Third, the new correlation matrix shown in Table 2 is segmented to obtain four correlation matrices, as shown in Tables 3 to 6 respectively.

[0128] Table 3

[0129] D t1 t2 t3 t4 t7 t8 t9 s1 1 1 1 1 1 1 1 s2 0 0 1 0 1 0 0 s3 0 1 1 1 1 0 1 s5 0 1 1 1 0 0 0 s6 0 1 1 1 0 1 0 s7 1 1 0 0 0 0 0 s8 0 1 0 0 0 0 0

[0130] Table 4

[0131] D t6 s4 1

[0132] Table 5

[0133] D t10 t11 t12 t13 t14 t15 t16 t17 t18 t20 t21 t23 s9 1 0 0 0 0 0 0 0 0 0 0 1 s10 1 0 0 1 0 0 0 0 0 0 0 0 s11 1 0 0 0 0 0 0 0 0 0 1 0 s12 1 0 1 0 1 0 0 1 0 0 0 1 s13 0 0 0 1 1 0 0 0 1 0 1 0 s15 1 1 0 0 0 0 0 0 0 0 1 0 s16 1 0 1 0 1 0 0 1 0 1 0 1 s17 0 0 0 0 1 1 0 1 0 0 0 0 s18 0 0 0 0 1 1 1 1 0 0 0 0 s19 0 0 0 0 0 1 1 0 0 0 0 0

[0134] Table 6

[0135] Dt24

[0136] s141

[0137] Among them, the correlation matrices corresponding to Table 2 and Table 6 are optimal, and the tests of the correlation matrices in the two tables are all necessary tests. The correlation matrices corresponding to Table 3 and Table 5 are not the optimal correlation matrices. In this embodiment, the improved discrete cuckoo algorithm is used to test and select them, and one of the optimal solutions is taken, which are [1, 1, 1, 0, 1, 1, 0] and [1, 1, 0, 1, 0, 0, 1, 1, 0, 1, 0, 1], respectively, and then merged with the previous necessary tests to obtain the final selected test: [t1, t2, t4, t6, t7, t8, t10, t11, t16, t17, t20, t21, t23, t24, t25]. The constructed correlation matrix is ​​shown in Table 7. The test cost is reduced from the original 26 to 15. At the same time, its fault detection rate and fault isolation rate are both 100%, ensuring optimality.

[0138] Table 7

[0139] D t1 t2 t4 t6 t7 t8 t10 t11 t16 t17 t20 t21 t23 t24 t25 s1 1 1 1 0 1 1 0 0 0 0 0 0 0 0 0 s2 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 s3 0 1 1 0 1 0 0 0 0 0 0 0 0 0 0 s4 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 s5 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 s6 0 1 1 0 0 1 0 0 0 0 0 0 0 0 0 s7 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 s8 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 s9 0 0 0 0 0 0 1 0 0 0 0 0 1 0 0 s10 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 s11 0 0 0 0 0 0 1 0 0 0 0 1 0 0 0 s12 0 0 0 0 0 0 1 0 0 1 0 0 1 0 0 s13 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 s14 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 s15 0 0 0 0 0 0 1 1 0 0 0 1 0 0 0 s16 0 0 0 0 0 0 1 0 0 1 1 0 1 0 0 s17 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 s18 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0 s19 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 s20 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

[0140] The present invention proposes an optimization test method for rapid fault diagnosis. First, redundant tests and invalid tests in a correlation matrix are removed. The correlation matrix is ​​optimized based on the particularity of all-1 rows and all-1 columns. The optimized correlation matrix is ​​processed in blocks, and the large matrix is ​​divided into several smaller correlation matrices. A test group in each separated correlation matrix is ​​judged once to see whether it is the optimal test group. The optimal test group is directly added to the final preferred test group, and the non-optimal test group is tested by using an improved discrete cuckoo algorithm. The present invention realizes the optimization and block division of a large correlation matrix, and the optimization of the tests in the correlation matrix. Under the condition of ensuring that the performance index remains unchanged, the test cost is reduced and the fault diagnosis speed is improved. The beneficial technical effects of the present invention include: first, when pre-processing the correlation matrix, the all-1 rows and all-1 columns are optimized to avoid the omission of the test and the decrease in the fault detection rate. Second, in the present invention, a separation correlation test method is proposed to divide the correlation matrix into blocks to solve the problem that the correlation matrix is ​​too large, resulting in too high calculation time cost and poor results, and its feasibility is verified. Running the correlation matrix in blocks significantly reduces the time cost and obtains better test optimization results. It can also intuitively show the correlation between tests. Third, the discrete cuckoo algorithm is optimized for optimization, and the number of times the optimal solution remains unchanged in each iteration is used as the judgment of the degree of convergence and the output condition, which reduces the probability of falling into the local optimum to a certain extent. Fourth, in order to achieve the optimal final testability index, the optimization of fault detection rate and fault isolation rate is proposed and applied in the improved discrete cuckoo algorithm, ensuring that the testability index of the correlation matrix constructed by the optimized test is consistent with the original one, and the index is optimal without considering the test cost.

[0141] The present invention also proposes an optimized test device for rapid fault diagnosis, as shown in the attached Figure 6 The device 10 includes a redundant invalid removal module, an optimization module, a block division module, a selection module and an output module.

[0142] Redundant and invalid removal module, used to remove redundant tests and invalid tests;

[0143] Optimization module for fault model and test optimization, optimizing the fault modes that can be detected by all tests in the correlation matrix and the tests that can detect all fault modes;

[0144] The block module is used to divide the test set into several related test groups according to the correlation between the tests;

[0145] The optimization module is used to directly add the tests in the optimal test group to the final optimal test set, and optimize the tests in the non-optimal test group, and then add them to the final optimal test set after optimization;

[0146] The output module is used to integrate the tests of the optimal test set and the tests selected from the non-optimal test set to obtain the final optimal test set.

[0147] Figure 7 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. Specifically, it may include: at least one processor 21, at least one memory 22, a power supply 23, a communication interface 24, an input / output interface 25, and a communication bus 26. The memory 22 is used to store a computer program, which is loaded and executed by the processor 21 to implement the relevant steps in the optimization test method for rapid fault diagnosis performed by the electronic device disclosed in any of the aforementioned embodiments.

[0148] In this embodiment, the power supply 23 is used to provide working voltage for each hardware device on the electronic device 20; the communication interface 24 can create a data transmission channel between the electronic device 20 and the external device, and the communication protocol it follows is any communication protocol that can be applied to the technical solution of the present application, and is not specifically limited here; the input and output interface 25 is used to obtain external input data or output data to the outside world, and its specific interface type can be selected according to specific application needs and is not specifically limited here.

[0149] Among them, the processor 21 may include one or more processing cores, such as a 4-core processor, an 8-core processor, etc. The processor 21 can be implemented in at least one hardware form of DSP (Digital Signal Processing), FPGA (Field-Programmable Gate Array), and PLA (Programmable Logic Array). The processor 21 may also include a main processor and a coprocessor. The main processor is a processor for processing data in the awake state, also known as a CPU (Central Processing Unit); the coprocessor is a low-power processor for processing data in the standby state. In some embodiments, the processor 21 may be integrated with a GPU (Graphics Processing Unit), which is responsible for rendering and drawing the content to be displayed on the display screen. In some embodiments, the processor 21 may also include an AI (Artificial Intelligence) processor, which is used to process computing operations related to machine learning.

[0150] In addition, the memory 22, as a carrier for storing resources, can be a read-only memory, a random access memory, a disk or an optical disk, etc. The resources stored thereon include an operating system 221, a computer program 222 and data 223, etc. The storage method can be temporary storage or permanent storage.

[0151] Among them, the operating system 221 is used to manage and control the hardware devices and computer programs 222 on the electronic device 20, so as to realize the operation and processing of the massive data 223 in the memory 22 by the processor 21, which can be Windows, Unix, Linux, etc. In addition to including a computer program that can be used to complete the optimized test method for rapid fault diagnosis performed by the electronic device 20 disclosed in any of the aforementioned embodiments, the computer program 222 can also further include a computer program that can be used to complete other specific tasks. In addition to data transmitted from an external device received by the electronic device, the data 223 can also include data collected by its own input and output interface 25, etc.

[0152] Furthermore, an embodiment of the present application also discloses a computer-readable storage medium, in which a computer program is stored. When the computer program is loaded and executed by a processor, the steps of the optimized test method for rapid fault diagnosis disclosed in any of the aforementioned embodiments are implemented.

[0153] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0154] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in the above description according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0155] The steps of the method or algorithm described in conjunction with the embodiments disclosed herein may be implemented directly using hardware, a software module executed by a processor, or a combination of the two. The software module may be placed in a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

[0156] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0157] The above is a detailed introduction to an optimized test method, device, equipment and storage medium for rapid fault diagnosis provided by the present invention. Specific examples are used in this article to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea; at the same time, for general technical personnel in this field, according to the idea of ​​the present invention, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. An optimized test method for rapid fault diagnosis, characterized in that: include: Redundant and invalid removal: remove redundant tests and invalid tests; Fault model and test optimization, optimizing the failure modes that can be detected by all tests in the dependency matrix and the tests that can detect all failure modes; Divide the test set into several related test groups based on the correlation between the tests; The tests in the optimal test group are directly added to the final optimal test set, and the non-optimal test groups are optimized and added to the final optimal test set after optimization; The tests of the optimal test set and the tests selected from the non-optimal test set are integrated to obtain the final optimal test set.

2. The optimized test method for rapid fault diagnosis according to claim 1, characterized in that: The optimization of the failure modes detectable by all tests in the dependency matrix includes: Remove all 1 rows in the correlation matrix. If there are no all 0 columns in the correlation matrix, the optimization of all 1 rows ends. If an all-0 column appears after removing the all-1 row in the correlation matrix, the test corresponding to the all-0 column is added to the final preferred test set, and then the all-0 column is removed, and a loop is performed to determine whether there is an all-1 row. If so, another loop optimization is performed until there is no all-1 row, and the all-1 row optimization is completed.

3. The optimized test method for rapid fault diagnosis according to claim 1, characterized in that: The optimization of the tests in the dependency matrix that can detect all failure modes includes, Remove all 1 columns from the correlation matrix. If there are no all 0 rows in the correlation matrix, the all 1 column optimization ends. If all-0 rows appear after removing the all-1 columns in the correlation matrix, add the test corresponding to the all-1 columns to the final preferred test set, then remove the all-0 rows, and loop to determine whether there is an all-1 column. If so, perform another loop optimization until there is no all-1 column, then the all-1 column optimization ends.

4. The optimized test method for rapid fault diagnosis according to claim 1, characterized in that: The method of dividing the test set into several related test groups according to the correlation between the tests includes: Selecting a test in a dependency matrix and setting a fault group that has an impact on the test; The fault in the fault group also affects the second test other than the test, the second test constitutes a second test group, the second test is a directly related test of the test, the second test group is a directly related test group, and the test and the directly related test group constitute a related test group; Selecting directly related tests of each test in the related test group that are not in the related test group and adding them to the related test group, until there are no tests outside the related test group in the correlation matrix that are directly related to the tests in the related test group; Repeat the above operation for the remaining tests in the correlation matrix until all tests have corresponding related test groups; The new correlation matrix formed by the relevant test groups is analyzed, and the relevant test groups are divided into optimal test groups and non-optimal test groups; wherein, the optimal test group refers to the test group whose fault detection rate or fault isolation rate will decrease when any test in the optimal test group is removed; the non-optimal test group is the test group other than the optimal test group.

5. The optimized test method for rapid fault diagnosis according to claim 1, characterized in that: For the non-optimal test group, an improved discrete cuckoo algorithm is used to test and optimize it. The improved discrete cuckoo algorithm is specifically: Initialize the parameters of the cuckoo algorithm, where the cuckoo position is a binary vector and the nest position represents the solution to the test point optimization selection problem; Calculate the fitness function of each current bird's nest position and construct a solution model; the solution model is related to the fault detection rate and the fault isolation rate; Find the location of the optimal bird's nest and the optimal solution according to the fitness function, and determine whether it is the same as the optimal solution of the previous iteration. If the number of consecutive identical solutions reaches the preset threshold, it is considered to be converged, and then the fault detection rate and fault isolation rate are optimized; For the one-fifth of the total number of bird nests with the worst location, they are removed with a preset probability, and if removed, new bird nests are generated to replace them; For each nest location, use Levi's flight to get the new nest location; The obtained new bird's nest position is a real number vector, the new bird's nest position is binarized, the new bird's nest position is compared with the original bird's nest position, and the better position is used as the current bird's nest position; Determine whether the output conditions are met. If so, output the result. Otherwise, switch to the fitness function calculation and continue the cycle.

6. The optimized test method for rapid fault diagnosis according to claim 5, characterized in that: In the improved discrete cuckoo algorithm, the quality of the cuckoo's choice of nest location is proportional to the value of the fitness function of the test point optimization selection problem. When multi-objective optimization is used to solve the objective function, the higher the system's fault detection rate, the higher the fault isolation rate, the higher the test cost that is not selected into the preferred test set, and the larger the fitness function value.

7. The optimized test method for rapid fault diagnosis according to claim 5, characterized in that: When the algorithm converges, the fault detection rate and fault isolation rate are optimized, including the optimization of fault detection rate and fault isolation rate. Optimize the fault detection rate by adding the lowest-cost test that can detect the faults that were previously detectable but are currently undetectable to the current test solution set; The fault isolation rate is optimized by adding all tests that can detect the faults that could have been isolated before but cannot be isolated now to the current test solution set.

8. An optimized test device for rapid fault diagnosis, characterized in that: include: Redundant and invalid removal module, used to remove redundant tests and invalid tests; Optimization module for fault model and test optimization, optimizing the fault modes that can be detected by all tests in the correlation matrix and the tests that can detect all fault modes; The block module is used to divide the test set into several related test groups according to the correlation between the tests; The optimization module is used to directly add the tests in the optimal test group to the final optimal test set, and optimize the tests in the non-optimal test group, and then add them to the final optimal test set after optimization; The output module is used to integrate the tests of the optimal test set and the tests selected from the non-optimal test set to obtain the final optimal test set.

9. An electronic device, characterized in that: include: Memory, used to store computer programs; A processor, configured to execute the computer program to implement the steps of the optimized test method for rapid fault diagnosis as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: Used to store computer programs; wherein, when the computer program is executed by a processor, the steps of the optimized test method for rapid fault diagnosis as described in any one of claims 1 to 7 are implemented.

Images