Communication system optimal test strategy generation method based on nested heuristic function

Abstract

The invention discloses a communication system optimal test strategy generation method based on a nested heuristic function, and the method comprises the steps: constructing a fault test model H, carrying out the preprocessing of a fault dependence matrix D in the fault test model H, obtaining a test t0 with the minimum test cost according to the preprocessed fault dependence matrix D0, and finally, carrying out the optimal test strategy generation of a communication system according to the test t0. A binary tree is generated by using a nested heuristic function and serves as an optimal test strategy, so that the problems of difficulty in solving and high calculation cost of an optimal binary tree generation problem in a large-scale communication system are solved.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of complex equipment system test strategy generation, and more specifically relates to a communication system optimal test strategy generation method based on a nested heuristic function. BACKGROUND

[0002] With the rapid development of electronic information technology, the demand for efficient information interaction in various fields of the national economy has significantly increased, which promotes the communication system to be deeply integrated into the fields of public information network, industrial control, aerospace, emergency rescue and livelihood services, forming a core infrastructure supporting the improvement of social operation efficiency, industrial digital transformation and national security barrier construction. It not only bears the basic functions of high-speed data transmission, multi-dimensional signal interaction, cross-domain resource collaborative scheduling, but also shoulders the key missions of dynamic guarantee of service quality, multi-scene service adaptation and emergency response in extreme environments.

[0003] In the whole life cycle of the communication system, from deployment and operation to upgrading and iteration, fault diagnosis is always the core link to ensure high reliability and high availability of the system. Therefore, how to solve the optimal diagnosis strategy in the communication system fault diagnosis has become the core technology to improve fault positioning accuracy, shorten fault troubleshooting time and reduce the whole life cycle operation cost. The generation of the optimal test strategy aims to ensure that the diagnosis cost is minimized under the premise of the maximum fault isolation rate, and is one of the core research directions in the field of fault diagnosis. The generation of the optimal test strategy is usually realized by constructing a multi-signal flow graph model of the system, which describes the relationship between the fault state and the available test of the system, and converts the optimal test strategy generation problem into an optimal binary tree generation problem (OBP, Optimal Binary search tree generation Problem). Under this framework, the optimal binary tree is used to isolate faults step by step, and the appropriate test path is selected to realize efficient fault diagnosis.

[0004] However, with the increase of system size and complexity, the difficulty of solving the optimal binary tree is significantly improved. In dealing with large-scale complex systems, the existing algorithms often fail to converge to the global optimal solution of the optimal binary tree, resulting in high diagnosis cost and affecting the maintainability and reliability of the system. Therefore, how to realize efficient and accurate generation of the optimal test strategy in complex systems has become a difficult problem to be solved in the field of fault diagnosis.

[0005] In order to solve the above problems, the application provides a communication system optimal test strategy generation method based on a nested heuristic function.

[0006] The communication system optimal test strategy generation method based on the nested heuristic function not only solves the problems of slow convergence and high calculation cost of the traditional optimal binary tree generation method in a large-scale communication system, but also greatly improves the calculation efficiency and accuracy of the algorithm through the fine heuristic function design and layer-by-layer nesting, and is suitable for more complex fault diagnosis scenarios. SUMMARY

[0007] The application aims to efficiently generate an optimal fault diagnosis strategy by using the nested heuristic function to enhance the search capability of the solution space and gradually converging to the global optimal solution with the increase of the nesting number, thereby solving the difficulty of solving the optimal binary tree generation problem and the high calculation cost problem in a large-scale communication system.

[0008] To achieve the above application purposes, the application provides a communication system optimal test strategy generation method based on a nested heuristic function, which comprises the following steps:

[0009] (1) constructing a fault test model ;

[0010] (2) preprocessing a fault dependency matrix D

[0011] (3) obtaining a test with the minimum test cost according to the preprocessed fault dependency matrix ;

[0012] (4) generating a binary tree based on the nested heuristic function and taking the binary tree as an optimal test strategy.

[0013] The application purposes of the application are achieved in the following manner:

[0014] The application purposes are achieved in the following manner: The application purposes are achieved in the following manner: The application purposes are achieved in the following manner: Obtaining a test with minimum test cost , and finally according to the test , a binary tree is generated by using a nested heuristic function, and is used as an optimal test strategy, so that the difficulty in solving the optimal binary tree generation problem in a large-scale communication system and the high calculation cost problem are solved.

[0015] Meanwhile, the optimal test strategy generation method for a communication system based on a nested heuristic function has the following beneficial effects:

[0016] (1) The search capability of the solution space is enhanced by using the nested heuristic function, and with the increase of the nesting number, the global optimal solution is gradually converged, so that the optimal fault diagnosis strategy is efficiently generated.

[0017] (2) The present application is suitable for various types of heuristic functions and has strong flexibility; the present application can effectively control the calculation complexity and error range while ensuring the quality of the diagnosis strategy by precisely controlling the nesting number, thereby improving the operability; in addition, the nested heuristic function can strongly explore the solution space and gradually approach the global optimal solution through multiple recursions; with the increase of the nesting number, the search accuracy of the algorithm is continuously improved, and the problems of difficult solution space processing and slow convergence speed of the traditional method when facing complex systems are effectively solved. BRIEF DESCRIPTION OF DRAWINGS

[0018] Figure 1 is a flowchart of the optimal test strategy generation method for a communication system based on a nested heuristic function; DETAILED DESCRIPTION

[0019] The specific embodiments of the present application will be described below with reference to the accompanying drawings, so that those skilled in the art can better understand the present application. It should be particularly noted that in the following description, when the detailed description of known functions and designs may obscure the main content of the present application, these descriptions will be omitted here.

[0020] EMBODIMENT

[0021] Figure 1 is a flowchart of the optimal test strategy generation method for a communication system based on a nested heuristic function.

[0022] In this embodiment, as shown in Figure 1 , the present application provides an optimal test strategy generation method for a communication system based on a nested heuristic function, comprising the following steps:

[0023] (1) Constructing a fault test model H:

[0024] ;

[0025] Wherein, S represents a set of potential faults of a system to be tested, , represents the ith fault state, represents the fault-free state, represents the fault state category; T represents all available test sets of the system to be tested, , represents the jth available test, and n represents the total number of available tests; D represents the fault dependency matrix, which is specifically represented as:

[0026] ;

[0027] wherein, represents the test information under , represents failure to be detected, represents ability to be detected, ;

[0028] P is a set of prior probabilities corresponding to S, , represents the prior probability of , represents the test cost of , represents whether is selected when isolating from other fault states, =0, represents that is not selected when isolating from other fault states, =1, represents whether is selected when isolating from other fault states;

[0029] (2), pre-processing is performed on ;

[0030] (2.1), setting the minimum number of isolation units ;

[0031] (2.2), according to , dividing into multiple subsets ; is the smallest fuzzy set that cannot be isolated, that is:

[0032] , ;​

[0033] (2.3), traverse each subset , count the number of faulty states in each , denoted as ;

[0034] If , the subset is deleted, if , the subset is kept, finally, all the kept subsets are merged to obtain the updated fault dependency matrix ;

[0035] In this embodiment, as shown in Table 1, an example of generating the fault dependency matrix is shown. When the number of minimum isolation cells is , is merged from two faults (the two have the same test vector in ).

[0036] Table 1 is the fault dependency matrix optimization diagram when ;

[0037]

[0038] (3), the test cost corresponding to each test in is sorted in ascending order, and the test with the smallest test cost is recorded as ;

[0039] (4), a binary tree is generated based on a nested heuristic function:

[0040] (4.1), define the node increment-decrement rule of the nested heuristic function: let be a set of tests that can effectively split the node , , ;

[0041] If the node in the set to be transferred is effectively tested into two child nodes , , and , then , otherwise does not change, Similarly, finally remove the node from ;

[0042] (4.2), initialization , construct the set of faults to be transferred ;

[0043] (4.3), initialization of the number of nesting ; take the node as the root node

[0044] (4.4), effective partition of the root node using the nesting heuristic function

[0045] (4.4.1), randomly select as the next test, generate the child nodes , using the node increment-decrement rule of the nesting heuristic function ;

[0046] (4.4.2), initialize the set ;

[0047] (4.4.3), select a node from , for the convenience of subsequent description, the node is denoted as ;

[0048] (4.4.4), randomly select as the next test, generate the child nodes , using the node increment-decrement rule of the nesting heuristic function ;

[0049] (4.4.5), calculate the initial heuristic value :

[0050] ;

[0051] wherein is the test cost of , and represents the probability sum of all faults contained by the node

[0052] (4.4.6), return to step (4.4.4) and continue to select the next , until all tests in are traversed

[0053] (4.4.7), take as the final test of , and then use to Split the node, and add the two child nodes obtained by the split to ;

[0054] (4.4.8), calculate the test cost spent on the node ; :

[0055] ;

[0056] wherein, P (F) represents the probability sum of all faults contained in ; C (F) represents the test cost of ;

[0057] (4.4.9), calculate the cost of the sub-tree of the root node :

[0058] ;

[0059] wherein, N (F) represents the child node of the sub-tree with as the root node;

[0060] (4.4.10), in the same way, process the nodes in according to steps (4.4.3)-(4.4.9), and after all the nodes in are processed, delete the node from ;

[0061] (4.4.11), calculate the heuristic value for the next nesting:

[0062] ;

[0063] wherein, , N (F) represents the child node of the sub-tree ;

[0064] (4.4.12), return to step (4.4.1), and continue to select the next , until all the tests in are iterated through;

[0065] (4.4.13) determine the maximum number of nestings ;

[0066] S1, initialize the iteration number of the inner loop , set the maximum relative error ;

[0067] S2, define the transition cost of any node :

[0068] ;

[0069] wherein, is the test cost;

[0070] S3, randomly extract different subsets of }, ; take each subset as a node of a binary tree;

[0071] S4, calculate the transition cost of each node:

[0072] ;

[0073] S5, when , calculate the initial heuristic value according to step (4.4.5):

[0074] ;

[0075] when , calculate the heuristic value according to step (4.4.11):

[0076] ;

[0077] S6, according to , calculate using the maximum likelihood estimation method; according to the error theory, further calculate the transition cost increment distribution probability ;

[0078] S7, calculate the relative error :

[0079] ;

[0080] wherein, is the test cost of , and is the number of all tests;

[0081] S8, compare the relative error with the maximum relative error , if , then the round corresponding to ​​​ ' as the maximum number of nesting , i.e. '; else, let , and return to step S3;

[0082] (4.5), compare the number of nesting with the maximum number of nesting , if , let , and go to step (4.6); else go to step (4.7);

[0083] (4.6), effectively split the leaf node using the nesting heuristic function;

[0084] (4.6.1), randomly select a node from , and mark it as ;

[0085] (4.6.2), randomly select as the next test, and generate the child nodes , using the node increment / decrement rule of the nesting heuristic function;

[0086] (4.6.3), process according to steps (4.4.2)-(4.4.4), (4.4.6)-(4.4.12), and return to step (4.5) after the processing is completed until ;

[0087] (4.7), judge whether is empty, if , go to step (4.8), else select a node from as the root node, and return to step (4.3) until ;

[0088] (4.8), output the binary tree as the optimal test strategy.

[0089] ​In the embodiment, the experiment is carried out on a typical MIMO communication module. The MIMO communication module is a core communication component based on the multiple-input multiple-output technology, which deploys multiple sets of antennas at the transmitting end and the receiving end, and realizes efficient data transmission through key technologies such as spatial multiplexing, spatial diversity, and beamforming. Spatial multiplexing splits data into sub-streams and sends them in parallel to improve the rate, spatial diversity uses multiple antenna signal copies to counteract channel fading, and beamforming focuses signal energy to enhance transmission efficiency. The module can significantly improve data transmission rate, enhance signal reliability, improve spectrum utilization, and also expand coverage, and is widely used in 5G, Wi-Fi, satellite communication and other fields, and is a key support for improving the performance of modern wireless communication systems. In the embodiment, the related data is obtained by injecting faults into the MIMO communication module, the D matrix is established, the algorithm is compared, and the related test results are obtained.

[0090] Table 2 is a comparison of the cost of the present application and other algorithms.

[0091]

[0092] In the embodiment, as shown in Table 2, it is a comparison chart of the test cost of the algorithm PSO_MLP1 (particle swarm optimization-multilayer perceptron) used in the present application and other algorithms QGA (quantum genetic algorithm), HPSO (hybrid particle swarm optimization algorithm), Growing (growing algorithm), Hybrid (hybrid algorithm). The test cost required by PSO_MLP1 is 3.480, which is lower than 5.521, 5.724, 5.444, and 3.537 of the comparison algorithms, and the present application has superiority.

[0093] Although the above describes the specific embodiments of the present application in order to facilitate the understanding of the present application by those skilled in the art, it should be clear that the present application is not limited to the scope of the specific embodiments, and for those skilled in the art, it is obvious that various changes are within the spirit and scope of the present application defined and determined by the appended claims, and all the inventions utilizing the concept of the present application are included in the protection.

Claims

1. A method for generating an optimal test strategy for a communication system based on a nested heuristic function, characterized by, The method comprises the following steps: (1) constructing a fault test model H: ； where S represents the potential fault set of the system under test, , represents the ith fault state, represents the no fault state, represents the fault state category; T represents all available test sets of the system under test, , represents the jth available test, n represents the total number of available tests; D represents the fault dependency matrix, which is specifically represented as: ； wherein represents in under test information, represents failed to pass detection, represents able to pass detection; P is the prior probability set corresponding to S. , express The prior probability of T; C is the set of test costs corresponding to T. , express The test cost; B is the test selection matrix. , Indicates isolation When in other fault states Whether it is selected or not =0 indicates isolation. When in other fault states Not selected =1 indicates isolation. When in other fault states Is it selected? (2) To Preprocessing is required; (2.1), setting minimum isolation unit number ; (2.2) according to , the subsets are divided into a plurality of sub-sets is the smallest fuzzy set that is not separable, i.e.:​ ， ； (2.3) Traverse each subset Statistical analysis of each Includes fault status The number of them is denoted as ; If , the subset is deleted, if , the subset is kept, and finally all the kept subsets are merged to obtain the updated fault dependency matrix ; (3) To Sort the test costs of each test in ascending order, and denote the test with the lowest test cost as . ; (4) Combined test A binary tree is generated based on the nested heuristic function and used as the optimal test strategy.

2. The method of claim 1, wherein the nested heuristic-based communication system optimal test policy generation method is characterized by, The step (4) specifically comprises: (4.1), define the node increment-decrement rule of the nested heuristic function: let in order to effectively split the node whose test constitutes a set, , ; if the set to be moved the middle node is effectively tested divided into , two child nodes, and then , otherwise does not change, for the same reason, finally remove the node from ; (4.2), initialization , constructing a set to be transferred ; (4.3), initializing the nesting number ; setting the node as the root node; (4.4) performing efficient partitioning of the root node using a nested heuristic function ​ (4.4.1), randomly selected As For the following tests, the node addition / removal rule of the nested heuristic function is used to generate the child nodes , ; (4.4.2), initializing the set ; (4.4.3), from selecting a node , for convenience of subsequent description, the node is denoted as ; (4.4.4), randomly selected As For the next test, the node addition / removal rule of the nested heuristic function is used to generate the child nodes , ; (4.4.5), computing the initialization heuristic value : ； wherein, is the test cost, denotes the probability sum of all faults contained by the node. (4.4.6), return to step (4.4.4) and continue selecting the next until all tests in are completed; (4.4.7), will be As The final test under The node splitting is performed on The two child nodes obtained by the node splitting are added to ; (4.4.8), computing tests at the node costs spent : ； wherein, denotes the probability sum of all faults contained, denotes the test cost of (4.4.9), compute for the root node subtree of cost: ； wherein, represents a child node of the sub-tree with root node . (4.4.10), and similarly, the nodes in are processed according to steps (4.4.3)-(4.4.9) , and after all the nodes in are processed, the nodes are deleted from ; (4.4.11) computing the heuristic value for the next nest : ； wherein , is a child node; (4.4.12), return to step (4.4.1) and continue selecting the next , until all tests in are completed; (4.4.13) determining the maximum number of nesting ; S1, initializing the number of iterations of the inner loop , setting the maximum relative error ; S2, define any node of the transition cost: ； wherein, To test Splitting The cost of; S3, Random Selection Different subsets { }, ; each subset As a node in a binary tree; S4, calculating the transition cost of each node: ； S5. When the initialization heuristic value is calculated according to step (4.4.5): ； When the heuristic value is calculated according to step (4.4.11): ； S6、According to , the maximum likelihood estimation method is used to calculate ; according to the error theory, the transition cost increment distribution probability is calculated again ; S7, calculating the relative error : ； wherein, is the test cost, is the number of all tests; S8, compare the relative error with the size of the maximum relative error If then set the corresponding round as the maximum nesting number i.e. '; otherwise, let then return to step S3; (4.5) compare the nesting number with the size of the maximum nesting number If then let and go to step (4.6); otherwise go to step (4.7); (4.6), performing efficient partitioning using nested heuristics on leaf nodes ​ (4.6.1), from a node is randomly selected from ; (4.6.2), randomly selected As For the following tests, the node addition / removal heuristic was used to generate the child nodes , ; (4.6.3), processing according to steps (4.4.2)-(4.4.4), (4.4.6)-(4.4.12), and returning to step (4.5) after the processing is completed until ; (4.7), judging whether empty, if then entering step (4.8), otherwise selecting a node from as a root node, and then returning to step (4.3) until ; (4.8) outputting the binary tree as the optimal test strategy.

Images