Radar intelligent interference suppression software test evaluation method

Abstract

The invention discloses a radar intelligent interference suppression software test evaluation method, and relates to the technical field of radar test, and the method comprises the steps: S1, constructing a radar intelligent interference suppression software evaluation index system, and storing multi-dimensional indexes in the form of a knowledge graph; s2, designing a radar intelligent interference suppression software comprehensive evaluation algorithm, and quantitatively representing randomness and fuzziness in an evaluation process through effective data analysis, screening and integration of scoring opinions of multiple experts; s3, performing radar intelligent interference suppression software test evaluation; according to the method, a radar intelligent interference suppression software evaluation index system in a real application environment is constructed, the defects of a traditional index in the aspect of evaluating the operation capability and intelligence of radar software are overcome, and a comprehensive evaluation algorithm is provided; according to the method, the problem of evaluation deviation caused by large cognitive difference of people on the new concept of radar software intelligence can be solved, so that the evaluation result is more in line with subjective cognition.

Description

Technical Field

[0001] This invention relates to the field of radar testing technology, and specifically to a test and evaluation method for radar intelligent interference suppression software. Background Technology

[0002] Radar is widely used in meteorological monitoring, autonomous driving, aerospace and other fields, providing important security and technical support for modern society. With the popularization of wireless communication equipment, the electromagnetic environment is becoming increasingly complex. The interference caused by a large amount of noise and superposition of signals in the same frequency band has a significant impact on radar systems, making radar interference suppression software increasingly important.

[0003] Meanwhile, the development of artificial intelligence technology has also driven further improvements in the intelligence level of radar jamming suppression software. Conventional software testing in the laboratory stage is insufficient to comprehensively measure the performance of intelligent radar jamming suppression software, mainly due to the following problems:

[0004] (1) Traditional test evaluation indicators are mainly aimed at the design verification and acceptance testing phases of software, and rarely consider actual application scenarios;

[0005] (2) The evaluation techniques for intelligent software are not yet perfect, and the evaluation indicators, testing methods and evaluation processes for intelligent features have not yet been systematized.

[0006] To address the challenges of intelligent development in radar software, it is necessary to explore a set of software evaluation theories and methods that cover the actual application stage and reflect the level of intelligence, so as to support the intelligent upgrade of radar interference suppression functions. Summary of the Invention

[0007] To address the aforementioned problems, the purpose of this invention is to provide a testing and evaluation method for radar intelligent interference suppression software. By analyzing typical application scenarios, an evaluation index system suitable for real-world application environments and reflecting the performance of radar intelligent interference suppression software is constructed. A comprehensive evaluation algorithm that can reasonably quantify the randomness of evaluation is designed, and a scientific and effective testing and evaluation process is proposed. This solves the problem of evaluation result deviation caused by cognitive differences, thereby facilitating subsequent iterative optimization of interference suppression technology using evaluation results and improving its intelligence level.

[0008] To achieve the above objectives, the technical solution adopted by this invention is: a radar intelligent interference suppression software testing and evaluation method, comprising the following steps:

[0009] 1. Construct an evaluation index system for radar intelligent interference suppression software.

[0010] This section primarily addresses the challenge of conventional software testing metrics being insufficient to measure the performance of radar intelligent jamming suppression software. It constructs a relatively complete evaluation metric system from three aspects: basic software performance, task completion capability, and intelligence, and stores these multi-dimensional metrics in the form of a knowledge graph. The process includes the following steps:

[0011] X1-1: Analyze the working characteristics of radar intelligent interference suppression software in intelligent interference perception and intelligent decision-making, and determine the three dimensions of constructing evaluation indicators: software basic performance, software task completion capability, and intelligence.

[0012] X1-2: Based on real-world application scenarios, performance issues of radar intelligent interference suppression software are more easily exposed under conditions of dense target numbers and complex electromagnetic environments. Therefore, extreme application scenarios are designed for typical application scenarios such as target detection and target tracking by increasing the number of targets and the intensity of interference.

[0013] X1-3: Evaluation metrics in X1-1 that require special attention in analyzing radar intelligent jamming suppression software under extreme scenarios. Basic software performance mainly includes the general quality characteristics of the radar intelligent jamming suppression software and its resource consumption during operation; software task completion capability mainly refers to the ability of the radar intelligent jamming suppression software to complete tasks under extreme scenarios; intelligence mainly refers to the ability of the radar intelligent jamming suppression software to effectively schedule strategies in unknown environments, independently complete tasks, and adaptively adjust jamming suppression strategies using historical data and external sensing information to achieve self-learning and improvement.

[0014] X1-4: Based on the subordinate relationships between different indicators, represent all indicators as structured data;

[0015] X1-5: Extract triplet information from the structured data in X1-4, use indicators as graph nodes and the subordinate relationships between indicators as graph edges, and store the multi-level indicator knowledge graph in a graph database.

[0016] X1-6: Use the built-in query statements in the graph database to retrieve or manage the addition and deletion of indicators.

[0017] 2. Design a comprehensive evaluation algorithm for radar intelligent interference suppression software.

[0018] This section primarily addresses the evaluation bias caused by significant differences in public understanding of the novel concept of radar software intelligence. It proposes a comprehensive evaluation algorithm based on an improved fuzzy cloud model. This algorithm effectively filters and integrates scores from multiple experts through data analysis, quantifying the randomness and fuzziness in the evaluation process. The algorithm includes the following steps:

[0019] X2-1: Initialize the intelligence level The intelligence level was obtained based on expert experience and research data. of Grouped interval fractions;

[0020] X2-2: Calculation The sample mean, first-order sample absolute central moment, and sample variance of the scores in each interval;

[0021] X2-3: Calculate the hyperentropy of the multi-interval data cloud model using the sample statistics in X2-2. ,express The magnitude of cognitive ambiguity in the expert scoring data is used to calculate the expected value of the cloud model using a comprehensive cloud generation algorithm. Entropy This represents the sample point value that best represents the intelligence level assessment score and the fluctuation range of the intelligence level score;

[0022] X2-4: Set the threshold for hyperentropy. ,like This indicates significant ambiguity in the experts' scores for this intelligence level. The data should be re-examined and filtered, and steps X2-2 to X2-4 should be repeated; otherwise, this result should be used as the intelligence level. The comment cloud model, and make Repeat steps X2-2 to X2-4 to calculate the comment cloud model for the next intelligence level until the calculation of the last intelligence level is completed, and then proceed to step X2-5.

[0023] X2-5: Based on the idea of ​​game theory, the subjective and objective weights of the indicators obtained by the conventional weighting method are coordinated and balanced, and the objective function is optimized by minimizing the deviation between the combined weights and the subjective and objective weights.

[0024] X2-6: Based on expert experience, the rationality of each combination weight is scored, which serves as the membership information for each indicator weight. An indicator weight cloud model is constructed using a reverse cloud generator with known membership information. The weight matrix is ​​obtained by sequentially combining multiple indicator weights in the cloud model.

[0025] X2-7: Based on the test data for each indicator, a comprehensive evaluation cloud model is constructed using a reverse cloud generator with unknown membership information. The comprehensive evaluation matrix is ​​obtained by sequentially combining multiple indicators in the cloud model.

[0026] X2-8: Combine the weight matrix in X2-6 and the comprehensive evaluation matrix in X2-7 to obtain the intelligence scoring cloud model of the radar intelligent interference suppression software to be evaluated. The traditional single precise numerical evaluation result is replaced by three numerical characteristic parameters: expectation, entropy, and hyperentropy. The calculation formula is as follows:

[0027]

[0028]

[0029]

[0030] In the formula Indicates the number of indicators. , , They represent the first The expected value, entropy, and hyperentropy of the weighted cloud model for each indicator. , , They represent the first These indicators comprehensively evaluate the expectation, entropy, and hyperentropy of the cloud model;

[0031] X2-9: Calculate the similarity between the intelligence rating cloud model of the radar intelligent interference suppression software to be evaluated and the comment cloud model of each intelligence level, and select the intelligence level corresponding to the highest similarity between the two as the level evaluation result.

[0032] 3. Conduct testing and evaluation of radar intelligent interference suppression software.

[0033] The steps include the following:

[0034] X3-1: Select the radar intelligent interference suppression software to be evaluated, and select indicators that meet the principles of measurability and comparability from the evaluation indicator knowledge graph in step 1 according to the evaluation requirements.

[0035] X3-2: The test equipment generates intelligent test scenarios of varying difficulty in batches. The number of test scenarios for each type of scenario is increased by modifying the interference ratio or replacing the composite interference type, and then outputs the results to the signal simulation equipment.

[0036] X3-3: The signal simulation equipment simulates target and interference signals in real time according to preset scenario requirements and outputs them to the radar to be evaluated.

[0037] X3-4: The test equipment synchronously collects test data such as the output signal of the interference suppression software in real time, analyzes and calculates the collected data, and obtains evaluation index data;

[0038] X3-5: The assessment of intelligence should be based on the basic operational effectiveness meeting the standards. If the indicators of the two dimensions of software basic performance and software task completion capability meet the task requirements, proceed to step X3-6. If the requirements are not met, it indicates that the basic operational capability of the interference suppression software to be assessed does not meet the standards, and it is considered that it does not have true intelligence, thus ending the intelligence assessment of the software.

[0039] X3-6: Using the comprehensive evaluation algorithm designed in section 2, the intelligence scoring cloud model and level evaluation results of the radar intelligent interference suppression software to be evaluated are obtained.

[0040] Compared with the prior art, the technical solution adopted in this invention has the following beneficial effects:

[0041] (1) An evaluation index system for radar intelligent interference suppression software in a real application environment was constructed, which made up for the shortcomings of traditional indexes in evaluating the operation capability and intelligence of radar software. Furthermore, the multi-level evaluation indexes were stored in the form of a knowledge graph, which is logically clear and more conducive to subsequent index analysis and retrieval.

[0042] (2) A comprehensive evaluation algorithm for radar intelligent interference suppression software is proposed. It can handle the evaluation bias caused by the large differences in people's understanding of the new concept of radar software intelligence. At the same time, it can judge and improve the confidence of expert scoring data through data analysis, screen and integrate the opinions of multiple experts, and replace the traditional evaluation results of a single precise value with three digital feature parameters, which is more in line with human subjective cognition.

[0043] (3) A test and evaluation process for radar intelligent interference suppression software is proposed, which can reasonably evaluate the intelligence level of radar intelligent interference suppression software. Attached Figure Description

[0044] Figure 1 This document outlines the knowledge graph construction and retrieval process for radar intelligent interference suppression software evaluation indicators in this embodiment.

[0045] Figure 2 This is the flowchart of the radar intelligent interference suppression software comprehensive evaluation algorithm in this embodiment.

[0046] Figure 3 This is a test and evaluation process for radar intelligent interference suppression software in this embodiment. Detailed Implementation

[0047] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0048] The knowledge graph construction and retrieval process for radar intelligent interference suppression software evaluation indicators according to embodiments of the present invention is as follows: Figure 1 As shown, the implementation steps are as follows:

[0049] X1-1: Analyze the working characteristics of radar intelligent interference suppression software in intelligent interference perception and intelligent decision-making, and determine the three dimensions of constructing evaluation indicators: software basic performance, software task completion capability, and intelligence.

[0050] X1-2: Based on real-world application environments, performance issues of radar intelligent interference suppression software are more easily exposed under conditions of dense target numbers and complex electromagnetic environments. Therefore, extreme application scenarios are designed for typical application scenarios such as target detection and target tracking by increasing the number of targets and the intensity of interference.

[0051] X1-3: Evaluation indicators of the three dimensions mentioned in X1-1 that need special attention in the analysis of radar intelligent interference suppression software under extreme scenarios. In an example of the present invention, the evaluation indicator system is shown in Table 1.

[0052] Table 1: Evaluation Index System for Radar Intelligent Interference Suppression Software (Example)

[0053]

[0054] X1-4: Based on the hierarchical relationship between different indicators, represent all indicators as structured data in JSON format, such as {"Primary Indicator":"General Quality Characteristics"}.

[0055] X1-5: Extract triplet information from the structured data in X1-4, use indicators as graph nodes and the hierarchical relationships between indicators as graph edges, and use Python to connect with the Neo4j graph database to store the multi-level indicator knowledge graph in the graph database.

[0056] X1-6: Use the built-in query "cypher" in the Neo4j graph database to retrieve or manage indicators.

[0057] In this embodiment, the comprehensive evaluation algorithm flow for radar intelligent interference suppression software is as follows: Figure 2 As shown, the implementation steps are as follows:

[0058] In step X2-1, the intelligence level is initialized. The intelligence level was obtained based on expert experience and research data. of Group interval fractions , Proceed to step X2-2.

[0059] In step X2-2, calculate Sample mean of scores in each interval First-order sample absolute central moments and sample variance Proceed to step X2-3.

[0060] In step X2-3, the hyperentropy of the multi-interval data cloud model is calculated using the sample statistics from X2-2. ,express The magnitude of cognitive ambiguity in the expert scoring data is used to calculate the expected value of the cloud model using a comprehensive cloud generation algorithm. Entropy This represents the sample point value that best represents the intelligence level assessment score and the fluctuation range of the intelligence level score.

[0061] In step X2-4, the threshold for hyperentropy is set. ,like This indicates significant ambiguity in the experts' scoring data for this intelligence level. The data needs to be re-examined and filtered; after removing excessively biased interval-type evaluation scores, steps X2-2 to X2-4 should be repeated. Otherwise, this result will be used as the intelligence level. The comment cloud model, and make Repeat steps X2-2 to X2-4 to calculate the comment cloud model for the next intelligence level, until the calculation for the last intelligence level is completed. In one embodiment of the invention, referring to the equipment intelligence level standard, the intelligence level is divided into 10 levels, resulting in 10 comment cloud models in the form of... The comment cloud model for intelligence level 1 is (0.6018, 0.2006, 0.0014), proceed to step X2-5.

[0062] In steps X2-5, the G1 sequence method and the CRITIC weighting method are selected to calculate the subjective and objective weights respectively. Based on the idea of ​​game theory, the obtained subjective and objective weights of the indicators are coordinated and balanced. The optimized objective function is:

[0063]

[0064] in, Represents the set of subjective and objective weight vectors. , Indicates the number of indicators; express A linear combination coefficient, This represents the combined weight vector set. Proceed to step X2-6.

[0065] In steps X2-6, the scores given by eight experts on the rationality of each weight combination are used as the membership information of each indicator weight. An inverse cloud generator with known membership information is then used to construct an indicator weight cloud model. In one embodiment of the invention, the weight cloud model for the indicator "suppression-type interference cancellation ratio" is as follows: Proceed to step X2-7.

[0066] In steps X2-7, the 50 sets of index test data obtained in steps X3-4 are standardized to eliminate the influence of dimensional differences. Then, a comprehensive evaluation cloud model is constructed using a reverse cloud generator with unknown membership information. In one embodiment of the present invention, the comprehensive evaluation cloud model for the index "suppression-type interference cancellation ratio" is as follows: Proceed to step X2-8.

[0067] In step X2-8, the weight matrix in X2-6 and the comprehensive evaluation matrix in X2-7 are combined to obtain the intelligence scoring cloud model of the radar intelligent interference suppression software to be evaluated. In one embodiment of the present invention, the intelligence scoring cloud model of the radar intelligent interference suppression software to be evaluated is as follows: Proceed to step X2-9.

[0068] In step X2-9, a cloud model similarity algorithm based on normalized Euclidean space is used to calculate the similarity between the intelligence rating cloud model in X2-8 and the comment cloud models for each intelligence level. The calculation method is as follows:

[0069]

[0070] A test and evaluation process for radar intelligent interference suppression software according to an embodiment of the present invention is as follows: Figure 3 As shown, the implementation steps are as follows:

[0071] X3-1: Select the radar intelligent interference suppression software to be evaluated, and based on the evaluation requirements, screen indicators that meet the principles of measurability and comparability from the evaluation indicator knowledge graph in X1-5. In one example of this invention, 16 evaluation indicators were selected.

[0072] X3-2: The testing equipment generates intelligent test scenarios of varying difficulty, including suppression, deception, and composite interference. The number of test scenarios for each type is increased by modifying the interference-to-noise ratio or replacing the composite interference type, and the results are output to the signal simulation equipment. Six different levels of interference are superimposed on a conventional uniform motion target detection scenario. The intelligent test scenarios are shown in Table 2.

[0073] Table 2: Test Scenarios for Interference Suppression Intelligence at Different Difficulty Levels

[0074]

[0075] X3-3: The signal simulation equipment simulates target, interference and other signals in real time according to the preset scenario requirements and outputs them to the radar to be evaluated.

[0076] X3-4: The test equipment synchronously collects test data such as the output signal of the interference suppression software in real time, and analyzes and calculates the collected data. In one example of this invention, 50 sets of valid data for 16 indicators are obtained, with the interference suppression effective probability... For example, the calculation formula is as follows:

[0077]

[0078] in This represents the total number of tests conducted under this type of interference scenario. This indicates the number of times the radar software can effectively complete target detection tasks without being affected by interference.

[0079] X3-5: If the indicators of both basic software performance and software task completion capability meet the task requirements, these requirements can be set according to actual task needs. For example, in one embodiment of this invention, CPU utilization is required in interference scenarios. Target measurement accuracy The software meets the requirements; proceed to step X3-6.

[0080] X3-6: Using the comprehensive evaluation algorithm designed in section 2, the intelligence scoring cloud model and level evaluation results of the radar intelligent interference suppression software to be evaluated are obtained.

[0081] In one embodiment of the present invention, the similarity between the intelligence rating cloud model of the radar intelligent interference suppression software to be evaluated and the 10 comment cloud models are 0.4680, 0.4030, 0.0714, 0.0269, 0.0171, 0.0039, 0.0009, 0.0006, 0.0004, and 0.0002, respectively. The 10 comment cloud models represent intelligence levels 1 to 10. The intelligence level corresponding to the highest similarity (0.4680) is selected as the level evaluation result, and the intelligence level of the radar intelligent interference suppression software to be evaluated is 1.

[0082] Although the present invention has been disclosed above with reference to preferred embodiments, the embodiments and accompanying drawings are not intended to limit the invention. Any person skilled in the art can make various changes or modifications without departing from the spirit and scope of the invention, and these changes will also be within the protection scope of the invention. Therefore, the protection scope of the present invention should be defined by the scope of the claims of this application.

Claims

1. A method for testing and evaluating radar intelligent interference suppression software, characterized in that, Includes the following steps: S1: Construct an evaluation index system for radar intelligent interference suppression software and store multi-dimensional indicators in the form of a knowledge graph; S2: Design a comprehensive evaluation algorithm for radar intelligent interference suppression software. Through effective data analysis and screening, and by integrating the scoring opinions of multiple experts, the randomness and ambiguity in the evaluation process are quantitatively represented. S3: Conduct testing and evaluation of radar intelligent interference suppression software.

2. The radar intelligent interference suppression software testing and evaluation method as described in claim 1, characterized in that, S1 specifically includes the following steps: S1-1: Analyze the working characteristics of radar intelligent interference suppression software in intelligent interference perception and intelligent decision-making, and determine the three dimensions of constructing evaluation indicators: software basic performance, software task completion capability and intelligence. S1-2: Design extreme application scenarios for target detection or target tracking applications; S1-3: Analyze the evaluation indicators in three dimensions under extreme application scenarios; S1-4: Based on the subordinate relationships between different indicators, represent all indicators as structured data; S1-5: Extract triplet information from structured data, use indicators as graph nodes and the subordinate relationships between indicators as graph edges, and store the multi-level indicator knowledge graph in a graph database. S1-6: Use the built-in query statements in the graph database to retrieve or manage the addition and deletion of indicators.

3. The radar intelligent interference suppression software testing and evaluation method as described in claim 2, characterized in that, In S1-1: The basic performance of the software includes the general quality characteristics of the radar intelligent interference suppression software and the resource consumption during operation; Software task completion capability refers to the ability of radar intelligent interference suppression software to complete tasks under extreme scenarios. Intelligence refers to the ability of radar intelligent interference suppression software to perform strategy scheduling and complete tasks independently in unknown environments, and to adaptively adjust interference suppression strategies using historical data and external sensing information, thereby achieving self-learning and improvement capabilities.

4. The radar intelligent interference suppression software testing and evaluation method as described in claim 1, characterized in that, S2 specifically includes the following steps: S2-1: Initialize the intelligence level The intelligence level was obtained based on expert experience and research data. of Grouped interval fractions; S2-2: Calculation The sample mean, first-order sample absolute central moment, and sample variance of the scores in each interval; S2-3: Calculate the hyperentropy of the multi-interval data cloud model using the sample statistical data from S2-2. , express The magnitude of cognitive ambiguity in the expert scoring data is used to calculate the expected value of the cloud model using a comprehensive cloud generation algorithm. Entropy , This represents the sample point values ​​that can represent the intelligence level G assessment score. This indicates the range of fluctuation that represents the G-score value, which represents the intelligence level. S2-4: Set the threshold for hyperentropy ,like This indicates that there is ambiguity in the expert scoring data for intelligence level G. The data should be re-examined and filtered, and steps S2-2 to S2-4 should be repeated; otherwise, the result should be used as the intelligence level. The comment cloud model, and make Repeat steps S2-2 to S2-4 to calculate the comment cloud model for the next intelligence level until the calculation of the last intelligence level is completed, and then proceed to step S2-5. S2-5: Based on the idea of ​​game theory, the subjective and objective weights of the indicators obtained by the conventional weighting method are coordinated and balanced, and the objective function is optimized by minimizing the deviation between the combined weights and the subjective and objective weights. S2-6: Based on expert experience, score the rationality of each combination weight as the membership information of each indicator weight. Use the inverse cloud generator with known membership information to construct the indicator weight cloud model. The weight matrix is ​​obtained by sequentially combining multiple indicator weights in the cloud model. S2-7: Based on the test data for each indicator, construct a comprehensive evaluation cloud model using a reverse cloud generator with unknown membership information. The comprehensive evaluation matrix is ​​obtained by sequentially combining multiple indicators in the cloud model. S2-8: Combine the weight matrix in S2-6 and the comprehensive evaluation matrix in S2-7 to obtain the intelligence scoring cloud model of the radar intelligent interference suppression software to be evaluated. The traditional single precise numerical evaluation result is replaced by three numerical characteristic parameters: expectation, entropy, and hyperentropy. The calculation formula is as follows: ； ； ； in, Indicates the number of indicators. , , They represent the first The expected value, entropy, and hyperentropy of the weighted cloud model for each indicator. , , They represent the first These indicators comprehensively evaluate the expectation, entropy, and hyperentropy of the cloud model; S2-9: Calculate the similarity between the intelligence rating cloud model of the radar intelligent interference suppression software to be evaluated and the comment cloud model of each intelligence level, and select the intelligence level corresponding to the high similarity between the two as the level evaluation result.

5. The radar intelligent interference suppression software testing and evaluation method as described in claim 1, characterized in that, S3 specifically includes the following steps: S3-1: Select the radar intelligent interference suppression software to be evaluated, and select indicators that meet the principles of measurability and comparability from the knowledge graph of intelligent evaluation indicators according to the evaluation requirements. S3-2: The test equipment generates intelligent test scenarios of varying difficulty in batches. The number of test scenarios for each type of scenario is increased by modifying the interference ratio or replacing the composite interference type, and then outputs the results to the signal simulation equipment. S3-3: The signal simulation equipment simulates the target and interference signals in real time according to the preset scenario requirements and outputs the results to the radar to be evaluated. S3-4: The test equipment synchronously collects the test data of the output signal of the interference suppression software in real time, analyzes and calculates the collected data, and obtains the evaluation index data. S3-5: The assessment of intelligence is based on the basic performance meeting the standards. If the indicators of the two dimensions of software basic performance and software task completion ability meet the task requirements, proceed to step S3-6. If the requirements are not met, it means that the basic ability of the interference suppression software to be assessed does not meet the standards, and it is considered that it does not have true intelligence, and the intelligence assessment of the software ends. S3-6: Using the designed radar intelligent interference suppression software comprehensive evaluation algorithm, the intelligence scoring cloud model and level evaluation results of the interference suppression software to be evaluated are obtained.

Images