MAS-based microwave signal analyzer test task intelligent distribution method

Abstract

The invention belongs to the technical field of distributed testing, and particularly relates to an MAS-based microwave signal analyzer test task intelligent allocation method, which adopts a multi-agent system protocol of an extended contract network, introduces a trust mechanism, and analogy task allocation into an engineering bid invitation process. The whole process is further decomposed into four stages of bid invitation, bidding, bid winning and contract signing, namely four stages of test task issuing, signal analyzer autonomous response, intelligent selection of a target signal analyzer and test task starting, the test task is used as input, and the indexes of the signal analyzer, the test task completion time and quality and the like are used as evaluation standards; according to the method, intelligent distribution of the test tasks is carried out, trust factors of corresponding signal analyzers are modified according to results of completing the tasks, iterative optimization is carried out continuously, the accuracy and efficiency of intelligent test task distribution are improved, and therefore the overall test efficiency is improved.

Description

Technical Field

[0001] This application belongs to the field of distributed testing technology, specifically relating to an intelligent allocation method for test tasks of a microwave signal analyzer based on MAS. Background Technology

[0002] Performance testing of large-scale distributed electronic device systems relies on distributed testing systems. Specifically, based on the unique performance parameters of different electronic devices within the system, signal analyzers with corresponding specifications and functions are matched to perform the performance tests on those devices. In a distributed testing system, the electronic device under test (DUT) and the signal analyzers are distributed across the same network. The advantage is that each signal analyzer can conveniently test the electronic device. However, it also has significant drawbacks: when the testing capabilities of the signal analyzer do not match the performance of the DUT, the testing task cannot be completed, wasting time and the signal analyzer's testing resources. In practical applications, a distributed signal analysis testing system often handles multiple test tasks simultaneously, which undoubtedly increases the probability of mismatch between the signal analyzer and the DUT.

[0003] Existing distributed testing systems mostly employ manual task allocation, program-controlled task allocation, or "automated" task allocation based on traditional contract net models. Manual allocation offers high flexibility, but it is inefficient and prone to errors when dealing with numerous devices and parameters under test. Furthermore, the error rate of manual task allocation increases rapidly with the number of devices and parameters, leading to a sharp decline in overall testing efficiency. Program-controlled task allocation uses pre-coded methods to assign test tasks. However, this approach is only suitable for fixed systems and tasks, lacking flexibility. If the systems and tasks change, manual reprogramming and debugging of the program commands are required, which is time-consuming and labor-intensive. Task allocation based on traditional contract net models is better suited for solving the problem of allocating a single test task among multiple signal analyzers, i.e., suitable for single-task, single-winner, single-round task allocation scenarios. Summary of the Invention

[0004] To address the aforementioned problems, this invention proposes an intelligent task allocation method for microwave signal analyzers based on MAS (Microwave Signal Analyzer), the technical solution of which is as follows:

[0005] A method for intelligent allocation of microwave signal analyzer test tasks based on MAS includes the following steps: Utilizing the multi-agent system protocol of Extended Contract Network (MAS), a trust mechanism is introduced, and task allocation is decomposed into four stages: bidding, tendering, awarding, and contract signing. These stages correspond to four phases: test task release, autonomous response by the signal analyzer, intelligent selection of the target signal analyzer, and commencement of the test task. The test task is used as input, and the signal analyzer's performance, test task completion time, and quality are used as evaluation criteria for intelligent task allocation. The trust factor of the corresponding signal analyzer is modified based on the task completion results, and continuous iteration and optimization are performed to improve the accuracy and efficiency of intelligent test task allocation.

[0006] Preferably, when the system generates a test task:

[0007] S11. The task processor first decomposes the task, obtains the task's performance requirements, and forms a negotiation agreement; S12. Based on the system knowledge base, the trust value database of each signal analyzer, and environmental requirements, a response requirement is formed, i.e., a bidding contract is created, the scope of publication is determined, and the task bidding notice is issued to the signal analyzers within the specific scope through the communication processor; the trust value of each signal analyzer is calculated as follows:

[0008]

[0009] Among them, T i,0 Let T be the initial confidence value of the i-th signal analyzer. i,τ-1 Let be the trust value of the i-th signal analyzer after its last task, k be the weighting coefficient, η be the adjustment coefficient, η>0 indicates positive nonlinearity, η<0 indicates negative nonlinearity, and τ be the number of times the signal analyzer performs the task.

[0010] Preferably, the signal analyzer responds autonomously: After receiving the task notification, the signal analyzer that meets the system requirements decomposes the task content by its task processor, obtains the environmental requirements and performance indicators of the test task, and then determines whether it can complete the test task based on its own knowledge base; if the signal analyzer's own performance meets the relevant requirements, it further determines whether to bid based on its own working status; if it decides to bid, it sends a bid application to the system through its own communication processor.

[0011] Preferably, the intelligent selection of target signal analyzers: After receiving the bidding application for the signal analyzer, the distributed test system takes the performance index requirements of the test task, the current trust value of the bid signal analyzer, and the environmental requirements as input data, inputs them into the target intelligent selection network model based on deep learning, outputs the weights of the bid signal analyzers, and determines the instrument to carry the test task according to the weight of each signal analyzer.

[0012] Preferably, the weight calculation formula is as follows:

[0013]

[0014] Among them, P i To define the performance requirements for the i-th signal analyzer, T i Let E be the current trust value of the i-th signal analyzer. i To determine the environmental requirements for the i-th signal analyzer, W i,t R is the recommended weight of the i-th signal analyzer at time t. i,t W represents the actual weight of the i-th signal analyzer at time t (the weight change caused by manual intervention due to a sudden malfunction). i,0 Let f(·) be the initial weight of the i-th signal analyzer, and f(·) be the mapping function, ∈ i,t For the feedback correction term, δ is the competition coefficient, which represents the similarity between two signal analyzers competing for the same task. α, β, and γ are the dimension contribution coefficients, which represent the basic contribution ratio of "performance, trust, and environment" of the intelligent model learning to the weights. Initially, α + β + γ = 1, and subsequently optimized with data iteration. λ is the correction coefficient (0 < λ < 0.1), which controls the correction amplitude and avoids oscillation.

[0015] Preferably, the test task is started: after selecting the target signal analyzer, the test task is sent to the designated signal analyzer. The signal analyzer then starts to execute the test task and reports the execution results to the distributed test system. The system evaluates the performance of the signal analyzer's test task execution results based on the task execution time, the accuracy and comprehensiveness of the result data, etc., and modifies the trust value of the corresponding signal analyzer based on the evaluation results.

[0016] Compared with the prior art, the beneficial effects of this application are as follows:

[0017] (1) This invention proposes a smart task allocation method for microwave signal analyzer test based on MAS. Compared with existing methods, it has higher communication efficiency and task allocation accuracy when solving the task allocation problem of multiple tasks, multiple targets and multiple rounds, which can effectively improve test efficiency.

[0018] (2) The intelligent allocation method for test tasks of microwave signal analyzer based on MAS designed in this invention is implemented in pure software, with high program portability and low cost.

[0019] (3) The present invention introduces a trust mechanism to intelligently allocate test tasks and modify the trust factor of the corresponding signal analyzer according to the results of task completion, continuously iterating and optimizing to further improve the accuracy and efficiency of intelligent test task allocation. Attached Figure Description

[0020] Figure 1 Flowchart of intelligent task allocation scheme for testing;

[0021] Figure 2 Intelligent task allocation process for testing. Detailed Implementation

[0022] The technical solution of this application will be described in detail below with reference to specific embodiments and accompanying drawings. It should be understood that the embodiments and specific features in the embodiments are detailed descriptions of the technical solution of this application, rather than limitations thereof. Specific technical features can be combined with each other.

[0023] This invention primarily addresses the challenge of task allocation in distributed collaborative testing systems. It designs an intelligent task allocation method for microwave signal analyzers based on MAS (Multi-Agent System Protocol). Employing the extended contract network (EMB) multi-agent system protocol and introducing a trust mechanism, the method uses test tasks as input and signal analyzer specifications, task completion time, and quality as evaluation criteria for intelligent task allocation. The trust factor of the corresponding signal analyzer is modified based on the task completion results, continuously iterating and optimizing to improve the accuracy and efficiency of intelligent test task allocation, thereby enhancing testing efficiency. The test task allocation scheme of this invention is as follows: Figure 1 As shown, the intelligent assignment process for test tasks is as follows: Figure 2 As shown, it will be explained in detail below.

[0024] Step 1: Test Task Issuance. When the system generates a test task, the task processor first decomposes the task, obtains the task's performance requirements, and forms a negotiation agreement. Then, based on the system knowledge base, the trust value database of each signal analyzer, and environmental requirements, response requirements are formed, i.e., a tender contract is created, the scope of issuance is determined, and the task tender notice is issued to the signal analyzers within the specific scope through the communication processor. The trust value of each signal analyzer is calculated as follows:

[0025]

[0026] Among them, T i,0 Let T be the initial confidence value of the i-th signal analyzer. i,τ-1 Let be the trust value of the i-th signal analyzer after its last task, k be the weighting coefficient, η be the adjustment coefficient, η>0 indicates positive nonlinearity, η<0 indicates negative nonlinearity, and τ be the number of times the signal analyzer performs the task.

[0027] Step 2: Signal Analyzer Autonomous Response. Upon receiving the task notification, the signal analyzer that meets the system requirements decomposes the task content using its task processor, obtaining the environmental and performance requirements of the test task. Then, based on its own knowledge base, it determines whether it can complete the test task. If the signal analyzer's performance meets the relevant requirements, it further determines whether to submit a bid based on its own operating status (whether it is within the calibration validity period, etc.). If it decides to bid, it sends a bid application to the system through its own communication processor.

[0028] Step 3: Intelligent Selection of Target Signal Analyzer. After receiving the bidding applications for signal analyzers, the distributed testing system uses the performance requirements of the test task, the current trust values ​​of the bidding signal analyzers, and the environmental requirements as input data. This data is fed into a deep learning-based intelligent target selection network model, which outputs the weights of the bidding signal analyzers. Based on the weights of each signal analyzer, the system determines the instrument to undertake the test task. The weight calculation formula is shown below:

[0029]

[0030] Among them, P i To define the performance requirements for the i-th signal analyzer, T i Let E be the current trust value of the i-th signal analyzer. i To determine the environmental requirements for the i-th signal analyzer, W i,t R is the recommended weight of the i-th signal analyzer at time t. i,t W represents the actual weight of the i-th signal analyzer at time t (the weight change caused by manual intervention due to a sudden malfunction). i,0 Let f(·) be the initial weight of the i-th signal analyzer, and f(·) be the mapping function, ∈ i,t For the feedback correction term, δ is the competition coefficient, which represents the similarity between two signal analyzers competing for the same task. α, β, and γ are the dimension contribution coefficients, which represent the basic contribution ratio of "performance, trust, and environment" of the intelligent model learning to the weights. Initially, α + β + γ = 1, and subsequently optimized with data iteration. λ is the correction coefficient (0 < λ < 0.1), which controls the correction amplitude and avoids oscillation.

[0031] Step 4: Start the test task. After selecting the target signal analyzer, the distributed testing system distributes the test task to the designated signal analyzer. The signal analyzer then begins executing the test task and reports the results to the distributed testing system. The system evaluates the performance of the signal analyzer's test task execution based on factors such as task execution time, accuracy, and completeness of the result data, and adjusts the trust value of the corresponding signal analyzer accordingly.

[0032] Step 5: When a new test task needs to be assigned, repeat steps 1 to 4.

[0033] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for intelligent allocation of test tasks for a microwave signal analyzer based on MAS, characterized in that, Includes the following steps: By adopting the multi-agent system protocol of the Extended Contract Network and introducing a trust mechanism, task allocation is decomposed into four stages: bidding, tendering, winning the bid, and signing the contract. These stages are: test task release, autonomous response by the signal analyzer, intelligent selection of the target signal analyzer, and commencement of the test task. The test task is used as input, and the signal analyzer indicators, test task completion time, and quality are used as evaluation criteria to intelligently allocate test tasks. The trust factor of the corresponding signal analyzer is modified according to the results of task completion, and continuous iteration and optimization are carried out to improve the accuracy and efficiency of intelligent test task allocation.

2. The intelligent task allocation method for microwave signal analyzer based on MAS according to claim 1, characterized in that, When the system generates a test task: S11. The task processor first decomposes the task, obtains the task's performance requirements, and forms a negotiation agreement. S12. Based on the system knowledge base and the trust value database of each signal analyzer, and combined with environmental requirements, a response requirement is formed, i.e., a bidding contract is created, the scope of distribution is determined, and the task bidding notice is issued to the signal analyzers within the specific scope through the communication processor; the trust value of each signal analyzer is calculated as follows: Among them, T i,0 Let T be the initial confidence value of the i-th signal analyzer. i,τ-1 Let be the trust value of the i-th signal analyzer after its last task, k be the weighting coefficient, η be the adjustment coefficient, η>0 indicates positive nonlinearity, η<0 indicates negative nonlinearity, and τ be the number of times the signal analyzer performs the task.

3. The intelligent task allocation method for microwave signal analyzer test based on MAS according to claim 1, characterized in that, Autonomous response of the signal analyzer: After receiving the task notification, the signal analyzer that meets the system requirements decomposes the task content by its task processor, obtains the environmental requirements and performance indicators of the test task, and then determines whether it can complete the test task based on its own knowledge base; if the signal analyzer's own performance meets the relevant requirements, it further determines whether to bid based on its own working status; if it decides to bid, it sends a bid application to the system through its own communication processor.

4. The intelligent task allocation method for microwave signal analyzer test based on MAS according to claim 1, characterized in that, Intelligent selection of target signal analyzer: After receiving the bidding application for the signal analyzer, the distributed test system takes the performance index requirements of the test task, the current trust value of the bid signal analyzer, and the environmental requirements as input data, inputs them into the target intelligent selection network model based on deep learning, outputs the weights of the bid signal analyzers, and determines the instrument to carry the test task according to the weight of each signal analyzer.

5. The intelligent task allocation method for microwave signal analyzer test based on MAS according to claim 4, characterized in that, The weight calculation formula is as follows: Among them, P i To define the performance requirements for the i-th signal analyzer, T i Let E be the current trust value of the i-th signal analyzer. i To determine the environmental requirements for the i-th signal analyzer, W i,t R is the recommended weight of the i-th signal analyzer at time t. i,t W represents the actual weight of the i-th signal analyzer at time t. i,0 Let f(·) be the initial weight of the i-th signal analyzer, and f(·) be the mapping function, ∈ i,t For the feedback correction term, δ is the competition coefficient, representing the similarity between two signal analyzers competing for the same task; α, β, and γ are the dimension contribution coefficients, representing the basic contribution ratio of "performance, trust, and environment" of the intelligent model learning to the weights; and λ is the correction coefficient.

6. The intelligent task allocation method for microwave signal analyzer test based on MAS according to claim 1, characterized in that, Start the test task: After selecting the target signal analyzer, the test task is sent to the designated signal analyzer. The signal analyzer then starts to execute the test task and reports the execution results to the distributed test system. The system evaluates the performance of the signal analyzer's test task execution results based on the task execution time, the accuracy and comprehensiveness of the result data, etc., and modifies the trust value of the corresponding signal analyzer based on the evaluation results.

Images