Automatic test system based on vector network analyzer
By establishing an automated testing system for vector network analyzers, dynamically adjusting the frequency scanning density and signal power level, and combining it with intelligent diagnostic algorithms, the problem of unstable measurement data in vector network analyzers was solved, achieving efficient and accurate fault identification and diagnosis.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- KUN YU GAO PIN (SU ZHOU) JI SHU YOU XIAN GONG SI
- Filing Date
- 2026-02-02
- Publication Date
- 2026-05-26
AI Technical Summary
In the existing technology, the scattering parameter measurement of vector network analyzers relies on fixed frequency scanning density and signal power level, lacking a dynamic adjustment mechanism, which leads to unstable measurement data and difficulty in accurately identifying fault types.
An automated testing system based on a vector network analyzer is adopted. The system defines product attributes through a knowledge base construction module, establishes relationships using a graph database, and generates a test specification knowledge base. The dynamic measurement data acquisition module adjusts scanning parameters in real time, and combines with the intelligent diagnosis module to match fault types and generate detailed diagnostic reports.
It improves testing accuracy and efficiency, enabling the acquisition of high-quality scattering parameter datasets under different conditions, accurately identifying fault types and generating reliable diagnostic results, thus supporting the quality control of filter products.
Smart Images
Figure CN122085016A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of radio frequency microwave technology, and in particular to an automated test system based on a vector network analyzer. Background Technology
[0002] A vector network analyzer is a device used to measure circuits, components, and systems in the radio frequency and microwave bands. It is mainly used to measure the scattering parameters of a network. Scattering parameters describe the transmission, reflection, and propagation characteristics of signals in a network and are widely used to test the performance of radio frequency components such as antennas, filters, amplifiers, and transmission lines.
[0003] With the advancement of radio frequency and microwave technologies, more and more equipment and systems need to be automated during the production process to improve testing efficiency and accuracy. Traditional fault diagnosis systems rely on human experience or rough error analysis, making it difficult to accurately and quickly identify fault types.
[0004] Furthermore, in existing technologies, the measurement of scattering parameter data relies on fixed frequency scan density and signal power level, lacking a dynamic adjustment mechanism. This results in measurement data that may be unstable under different test conditions, making it difficult to accurately capture changes in key frequency domain regions. Summary of the Invention
[0005] To address the issue that the measurement of scattering parameter data relies on fixed frequency scan density and signal power level, lacking a dynamic adjustment mechanism, this application provides an automated testing system based on a vector network analyzer. This automated testing system includes: The knowledge base construction module defines a set of attributes for filter products, uses a graph database to establish the association between the attribute set and test configuration parameters, and forms a product functional test specification knowledge base. The test item matching module extracts product identifiers and test requirement keywords from user input, matches them with product specifications in the product function test specification knowledge base, and outputs a list of test items containing specific parameter constraints. The command sequence generation module queries the corresponding parameter constraints from the product functional test specification knowledge base based on the list of items to be tested, and determines the execution order of the items through the test process rule engine to generate a sequence of measurement commands that can be executed by the vector network analyzer. The dynamic measurement data acquisition module executes a sequence of measurement commands and dynamically adjusts the scanning parameters based on the real-time scattering parameter characteristics to obtain a scattering parameter dataset that meets the preset requirements. The intelligent diagnosis and management module compares the output scattering parameter dataset with the standard limit curves stored in the product function test specification knowledge base point by point, executes the fault type matching algorithm, outputs the test item judgment result, and then associates the complete test data record with the corresponding product model in the knowledge base.
[0006] Optionally, the knowledge base construction module includes: The key attribute set definition module defines the set of key attributes for filter products, including type, topology, rated power, and impedance. The test configuration parameter determination module determines the test configuration parameters corresponding to the set of key attributes in the filter product. The relational network construction module uses a graph database to associate the defined set of key attributes with test configuration parameters to build a relational network. The product functional test specification knowledge base generation module verifies the constructed network of relationships and, once verified, solidifies it into the product functional test specification knowledge base.
[0007] Optionally, the command sequence generation module includes: The test item element structure extraction module parses the list of test items, extracts the test items and parameter constraints, and obtains structured test item elements. The test configuration parameter matching and completion module queries the product function test specification knowledge base based on the structured test item elements to obtain the corresponding complete test configuration parameters. The test execution order optimization and sorting module uses the test process rule engine to sort the test project elements with complete test configuration parameters to determine the optimal execution order; The instrument command sequence generation module converts the test item elements arranged in the optimal execution order into a sequence of measurement commands that can be executed by the vector network analyzer.
[0008] Optionally, the test process rule engine is used to sort the test project elements with complete test configuration parameters to determine the optimal execution order, including: Receive test project elements with complete test configuration parameters and analyze the logical dependencies between the test project elements; Based on the logical dependencies obtained from the analysis, sorting rules in the test process rule engine aimed at improving test efficiency are applied. Prioritize test item elements with complete test configuration parameters based on sorting rules, and output a sequence of test item elements arranged in the optimal execution order.
[0009] Optionally, the dynamic measurement data acquisition module includes: The measurement control module controls the vector network analyzer to execute a sequence of measurement commands and acquire real-time scattering parameter datasets. The data analysis and decision-making module uses a parameter decision-making algorithm to analyze the trends of the acquired scattering parameter dataset and dynamically adjust the frequency scanning density and signal power level. The adjustment measurement module performs measurements based on the dynamically adjusted frequency scan density and signal power level, and outputs a scattering parameter dataset that meets the preset requirements.
[0010] Optionally, the trend of the acquired scattering parameter dataset is analyzed using a parameter decision algorithm, and the frequency scanning density and signal power level are dynamically adjusted, including: Frequency domain gradient analysis was performed on the acquired scattering parameter dataset to calculate the amplitude change rate at each frequency point; Based on the amplitude change rate of each frequency point obtained from frequency domain gradient analysis, key frequency domain regions are identified, and the frequency band boundaries where the change amplitude exceeds the preset interval are marked to obtain the amplitude change characteristics of scattering parameters. Based on the marked key frequency domain regions, calculate the signal-to-noise ratio under the current measurement configuration and evaluate the stability of data quality. Based on the data quality stability assessment results obtained from the identified key frequency domain region distribution and signal-to-noise ratio indicators, the frequency scan density adjustment coefficient is determined. Based on the amplitude variation characteristics of scattering parameters in key frequency domain regions and the frequency scan density adjustment coefficient, the signal power level compensation value is calculated.
[0011] Optionally, based on the amplitude change rate at each frequency point obtained from frequency domain gradient analysis, key frequency domain regions are identified, and the boundaries of frequency bands where the change amplitude exceeds a preset interval are marked, thus obtaining the amplitude change characteristics of the scattering parameters, including: Calculate the absolute value of the rate of change of amplitude at each frequency point to obtain the gradient amplitude; The gradient magnitude is compared with a preset threshold range to filter all candidate frequency points that exceed the threshold range; Cluster analysis is performed on the selected candidate frequency points to merge adjacent frequency points on the frequency axis and form a continuous key frequency domain region. Calculate the statistical characteristics of each key frequency domain region, and output the final list of key frequency domain regions and their corresponding amplitude variation characteristics.
[0012] Optionally, the intelligent diagnostics and management module includes: The scattering parameter dataset preprocessing module preprocesses the measured scattering parameter dataset, matches the processed dataset with the standard limit curves stored in the product functional test specification knowledge base at frequency points, and establishes a standardized data mapping relationship. The multi-dimensional error feature extraction module calculates the amplitude and phase deviation at each frequency point based on the established standardized data mapping relationship, and completes multi-dimensional error assessment by combining the preset tolerance threshold and extracting error distribution features. The fault type matching and analysis module uses a fault type matching algorithm to analyze the extracted error distribution features, identify specific fault types, calculate the corresponding fault confidence, and generate fault type matching results. The comprehensive judgment report generation module takes into account the error distribution characteristics data and the fault type matching results to generate a test report containing detailed judgment conclusions, and associates the metadata, process data and test report generated during the test with the corresponding product model.
[0013] Optionally, the extracted error distribution features are analyzed using a fault type matching algorithm to identify specific fault types and calculate the corresponding fault confidence scores, generating fault type matching results including: Read the error distribution feature data output by the multi-dimensional error feature extraction module, extract the amplitude deviation statistics, phase anomaly distribution, and continuity of the out-of-standard frequency band from it, and construct a set of feature indicators; The extracted feature index set is matched with the fault type templates stored in the product function test specification knowledge base in multiple dimensions, and the most suitable fault type is identified based on the matching results. Based on the matching degree between the feature index set and the fault type, the statistical characteristics of historical fault data, and the current test environment parameters, the confidence score of each candidate fault type is calculated using the Bayesian inference algorithm. By integrating the identified fault types, calculated confidence scores, and feature index sets, a structured fault type matching result is generated.
[0014] In summary, this application includes at least one of the following beneficial technical effects: 1. This application improves testing accuracy by establishing a relational network, optimizing command sequences, dynamically adjusting measurement parameters, and combining multi-dimensional error analysis and fault type matching algorithms. It identifies potential fault types and generates detailed diagnostic reports, greatly enhancing the efficiency and accuracy of functional testing.
[0015] 2. This application uses frequency domain gradient analysis and parameter decision algorithms to intelligently analyze and adjust the measurement process of scattering parameter data, accurately identify key frequency domain regions and dynamically optimize frequency scanning density and signal power level, thereby improving data quality and test accuracy, and ensuring that high-quality scattering parameter datasets can be obtained under different conditions.
[0016] 3. This application can accurately identify fault types and calculate their confidence levels by preprocessing the scattering parameter dataset, extracting error features, and performing fault type matching analysis. It can also generate reliable fault diagnosis results by combining multi-dimensional data and historical fault information, providing strong support for the quality control of filter products. Attached Figure Description
[0017] Figure 1 This is a schematic diagram of the system in this application. Detailed Implementation
[0018] The embodiments of this application are described in detail below, and examples of the embodiments are shown in the accompanying drawings.
[0019] In the description of this specification, the references to "certain embodiments," "one embodiment," "some embodiments," "illustrative embodiment," "example," "specific example," or "some examples" refer to specific features, structures, materials, or characteristics described in connection with the described embodiment or example, which are included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.
[0020] This application discloses an automatic testing system based on a vector network analyzer, referring to... Figure 1 The automated testing system based on a vector network analyzer includes: The knowledge base construction module defines a set of attributes for filter products, uses a graph database to establish the association between the attribute set and test configuration parameters, and forms a product functional test specification knowledge base.
[0021] Preferably, the knowledge base construction module includes: The key attribute set definition module defines the set of key attributes for filter products, including type, topology, rated power, and impedance. The test configuration parameter determination module determines the test configuration parameters corresponding to the set of key attributes in the filter product. The relational network construction module uses a graph database to associate the defined set of key attributes with test configuration parameters to build a relational network. The product functional test specification knowledge base generation module verifies the constructed network of relationships and, once verified, solidifies it into the product functional test specification knowledge base.
[0022] It needs to be explained that the core attributes of filter products need to be clearly defined, which typically include, but are not limited to: Type: The type of filter (e.g., low-pass, high-pass, band-pass, band-stop, etc.); Topology: The design architecture of the filter; Rated power: The maximum operating power of the filter; Impedance: The operating impedance of the filter, which usually needs to be specified during the design process (such as 50Ω, 75Ω, etc.). Based on the technical requirements and standards of the filter product, determine which attributes are critical and define them in detail; For each key attribute (such as type, topology, rated power, impedance, etc.), determine the relevant functional test configuration parameters, for example: For different types, different frequency response tests may be required; For the topology, it is necessary to test its frequency response curve and circuit characteristics; For rated power, the power test range needs to be set; For impedance, impedance matching and transmission loss testing may be required; Graph databases (such as Neo4j) are used to build relational models. Graph databases can efficiently store and query complex relational data. In a graph database, each node represents a key attribute or test configuration parameter, and the edges between nodes represent the relationships between them.
[0023] Test and verify the relationships to ensure that each test configuration parameter matches the actual performance requirements of the product; for example, check whether the test type matches the corresponding test content and whether the power test range is reasonable; once the verification is passed, fix the relationships and configuration parameters to form a formal product functional test specification knowledge base.
[0024] The test item matching module extracts product identifiers and test requirement keywords from user input, matches them with product specifications in the product functional test specification knowledge base, and outputs a list of test items containing specific parameter constraints.
[0025] It should be explained that the input includes product identifiers (such as product model, serial number, etc.) and test requirement keywords (such as test items, parameter descriptions, etc.). The system searches for entries related to the input product identifier in the knowledge base. It can find the test specifications corresponding to the product model through direct matching or fuzzy search, parse the test requirement keywords provided by the user, and identify the test items (such as frequency response, power test, impedance, etc.). Generate a list of test items containing specific parameter constraints; each item in the list will include: 1. Test item name: such as frequency response test, power test, etc.; 2. Specific parameter constraints: such as frequency range (e.g., 10MHz to 1000MHz), power range (e.g., 1W to 50W), maximum permissible deviation (e.g., ±5%), etc.
[0026] The command sequence generation module queries the corresponding parameter constraints from the product functional test specification knowledge base based on the list of items to be tested, and determines the execution order of the items through the test process rule engine to generate a sequence of measurement commands that can be executed by the vector network analyzer.
[0027] Preferably, the command sequence generation module includes: The test item element structure extraction module parses the list of test items, extracts the test items and parameter constraints, and obtains structured test item elements.
[0028] The test configuration parameter matching and completion module queries the product functional test specification knowledge base based on structured test item elements to obtain the corresponding complete test configuration parameters.
[0029] The test execution order optimization and sorting module uses the test process rule engine to sort the test project elements with complete test configuration parameters to determine the optimal execution order.
[0030] Preferably, the specific steps for using a test process rule engine to sort test item elements with complete test configuration parameters and determine the optimal execution order are as follows: Receive test project elements with complete test configuration parameters and analyze the logical dependencies between the test project elements; Based on the logical dependencies obtained from the analysis, sorting rules in the test process rule engine aimed at improving test efficiency are applied. Prioritize test item elements with complete test configuration parameters based on sorting rules, and output a sequence of test item elements arranged in the optimal execution order.
[0031] It should be explained that test item elements typically include specific configuration parameters for each test item, such as frequency range, power level, error tolerance, test conditions, etc. Each test item element represents an independent test item and comes with its required configuration parameters.
[0032] For example, a test item might include a frequency response test configuration with parameters such as frequency range (10MHz to 1GHz), amplitude error (±2dB), etc. Each test item may depend on the execution order or results of other items, for example: Prerequisites: Some test items may depend on the results of other test items to be performed. For example, "impedance test" may need to be performed after "frequency response test" because its measurement process depends on previous frequency response data. Parallel dependency: Some tests can be executed in parallel without needing to be in a specific order. Test Process Rule Engine: The rule engine is a decision-making platform used to optimize the test process based on a set of rules, including: 1. Minimize testing time: Schedule shorter tests earlier in the test; 2. Minimize resource conflicts: Avoid performing multiple tests on the same device or resources to prevent resource conflicts; Prioritize important or essential tests: For example, certain tests may be the basis for pass / fail determination, so they should be prioritized.
[0033] The rules engine assigns priorities to each test item; for example: 1. If certain tests are crucial to subsequent tests, they will have higher priority; for example, frequency response testing may be crucial to subsequent power testing, so it should be performed first. 2. Sort the test items according to their complexity, time required, and resources required to improve the efficiency of the testing process.
[0034] The instrument command sequence generation module converts the test item elements arranged in the optimal execution order into a sequence of measurement commands that can be executed by the vector network analyzer.
[0035] It should be explained that, based on the optimized test sequence, the specific parameters of each test item (such as frequency range, power level, etc.) are converted into commands that the vector network analyzer can understand and execute; the commands will specify the instrument's operation, such as setting the frequency range, measuring scattering parameters, etc.
[0036] The dynamic measurement data acquisition module executes a sequence of measurement commands and dynamically adjusts the scanning parameters based on the real-time scattering parameter characteristics to obtain a scattering parameter dataset that meets the preset requirements.
[0037] Preferably, the dynamic measurement data acquisition module includes: The measurement control module controls the vector network analyzer to execute a sequence of measurement commands and acquire real-time scattering parameter datasets.
[0038] The data analysis and decision-making module uses a parameter decision-making algorithm to analyze the trends of the acquired scattering parameter dataset and dynamically adjust the frequency scanning density and signal power level.
[0039] Preferably, the specific steps for dynamically adjusting the frequency scanning density and signal power level by analyzing the trend of the acquired scattering parameter dataset using a parameter decision algorithm are as follows: Frequency domain gradient analysis was performed on the acquired scattering parameter dataset to calculate the amplitude change rate at each frequency point.
[0040] Based on the amplitude change rate of each frequency point obtained from frequency domain gradient analysis, key frequency domain regions are identified, and the frequency band boundaries where the change amplitude exceeds the preset interval are marked to obtain the amplitude change characteristics of scattering parameters.
[0041] Preferably, the specific implementation steps for identifying key frequency domain regions and marking the frequency band boundaries where the change amplitude exceeds a preset interval based on the amplitude change rate at each frequency point obtained from frequency domain gradient analysis are as follows: Calculate the absolute value of the rate of change of amplitude at each frequency point to obtain the gradient amplitude; The gradient magnitude is compared with a preset threshold range to filter all candidate frequency points that exceed the threshold range; Cluster analysis is performed on the selected candidate frequency points to merge adjacent frequency points on the frequency axis and form a continuous key frequency domain region. Calculate the statistical characteristics of each key frequency domain region, and output the final list of key frequency domain regions and their corresponding amplitude variation characteristics.
[0042] Based on the marked key frequency domain regions, the signal-to-noise ratio (SNR) under the current measurement configuration is calculated to evaluate the stability of data quality.
[0043] Based on the data quality stability assessment results obtained from the identified key frequency domain region distribution and signal-to-noise ratio indicators, the frequency scan density adjustment coefficient is determined.
[0044] Based on the amplitude variation characteristics of scattering parameters in key frequency domain regions and the frequency scan density adjustment coefficient, the signal power level compensation value is calculated.
[0045] It should be explained that, for each frequency point, the rate of change of amplitude, i.e., the frequency domain gradient, is calculated by setting two adjacent frequency points. and and through and Obtain the frequency point amplitude change rate ; To eliminate the influence of positive and negative values, the absolute value of the amplitude change rate at each frequency point is calculated; The calculated frequency point amplitude change rate is compared with a preset threshold, and frequency points with gradient amplitude exceeding the preset threshold are selected. Cluster analysis is performed on the selected candidate frequency points to merge adjacent frequency points into a continuous key frequency domain region. In this way, a set of key frequency domain regions is obtained, and each region represents the region with a large change amplitude in the frequency band. The signal quality of each frequency band is evaluated based on the power ratio of the signal and noise. The higher the signal-to-noise ratio, the better the signal quality. Frequency scanning density is related to the signal-to-noise ratio (SNR). For frequency bands with high SNR, the scanning density should be appropriately reduced to decrease unnecessary measurement points. For frequency bands with low SNR or large amplitude variations, the scanning density can be increased. Based on the amplitude variation characteristics of each frequency band and the frequency scanning density adjustment coefficient, the signal power level compensation value for each frequency band is calculated. The specific expression is as follows: In the formula, Frequency point The final signal power compensation value; This indicates the current signal power level (serving as the reference power for calculating compensation). Indicates the first i The first transmitting unit and the first j Between receiving units, frequency points The signal frequency domain amplitude response at that location; Indicates the first i The first transmitting unit and the first i Between receiving units, frequency points The signal frequency domain amplitude response at that location; Indicates the target frequency point for calculating the amplitude gradient; This represents the reference frequency point for calculating the amplitude gradient; Frequency point ; signal power; Frequency point Noise power.
[0046] The adjustment measurement module performs measurements based on the dynamically adjusted frequency scan density and signal power level, and outputs a scattering parameter dataset that meets the preset requirements.
[0047] Specific examples are as follows: Suppose we are testing a microwave filter whose performance is measured in the frequency range of 1 GHz to 10 GHz using a vector network analyzer to obtain a dataset of scattering parameters. The goal is to optimize the measurement process by dynamically adjusting the scan density and signal power to improve test efficiency. Data from the following two adjacent frequency points has been collected from the vector network analyzer: At frequency =2GHz, measured -15dB, signal power =10dBm, noise power =2dBm; At frequency =2.1GHz, measured -10dB, signal power =9dBm, noise power =2.1dBm; Calculate the gradient of amplitude change between frequency points to determine which frequency bands have larger change amplitudes. Based on the above data, the amplitude change rate between frequency points is calculated to be 50dB / GHz. The threshold was set at 30dB / GHz. Since the amplitude change rate of 50dB / GHz is greater than this threshold, the frequency point was determined to be a key frequency domain region. Data from more frequency bands (e.g., 2.2GHz, 2.3GHz, etc.) were collected, and similar amplitude change rates were obtained. Adjacent frequency points were merged through cluster analysis, and a key frequency domain region was finally obtained, such as 2.0GHz-2.5GHz. Based on the signal amplitude change rate and frequency scan density adjustment coefficient, the signal power level compensation value is calculated. By substituting these values into the formula for calculating the signal power compensation value, the following is obtained: =25.15; Finally, the adjusted settings of frequency scan density and signal power are applied to the vector network analyzer to perform new measurements and collect a scattering parameter dataset that meets the preset requirements, ensuring more accurate and stable data in the key frequency band.
[0048] The intelligent diagnosis and management module compares the output scattering parameter dataset with the standard limit curves stored in the product function test specification knowledge base point by point, executes the fault type matching algorithm, outputs the test item judgment result, and then associates the complete test data record with the corresponding product model in the knowledge base.
[0049] Preferably, the intelligent diagnosis and management module includes: The scattering parameter dataset preprocessing module preprocesses the measured scattering parameter dataset, matches the processed dataset with the standard limit curves stored in the product functional test specification knowledge base at frequency points, and establishes a standardized data mapping relationship.
[0050] The multi-dimensional error feature extraction module calculates the amplitude and phase deviation at each frequency point based on the established standardized data mapping relationship, and completes multi-dimensional error assessment by combining the preset tolerance threshold and extracting error distribution features.
[0051] The fault type matching analysis module uses a fault type matching algorithm to analyze the extracted error distribution features, identify specific fault types, calculate the corresponding fault confidence, and generate fault type matching results.
[0052] Preferably, the specific implementation steps for analyzing the extracted error distribution features using a fault type matching algorithm, identifying specific fault types, calculating the corresponding fault confidence scores, and generating fault type matching results are as follows: Read the error distribution feature data output by the multi-dimensional error feature extraction module, extract the amplitude deviation statistics, phase anomaly distribution, and continuity of the out-of-standard frequency band from it, and construct a set of feature indicators; The extracted feature index set is matched with the fault type templates stored in the product function test specification knowledge base in multiple dimensions, and the most suitable fault type is identified based on the matching results. Based on the matching degree between the feature index set and the fault type, the statistical characteristics of historical fault data, and the current test environment parameters, the confidence score of each candidate fault type is calculated using the Bayesian inference algorithm. By integrating the identified fault types, calculated confidence scores, and feature index sets, a structured fault type matching result is generated.
[0053] It should be explained that the expression that best matches the fault type is: In the formula, Indicates the set of candidate fault types W In the process, based on posterior probability and feature matching degree evaluation, the one that best matches the measurement data is... E Fault type; Indicates the first n One fault type; Represents the set of all candidate fault types; Represents a given measurement feature E Fault type The posterior probability; Indicates measurement characteristics E The first in k One eigenvalue; Indicates the fault type Next, the k The mean of each feature; Indicates the dimension of the feature vector; This represents the k-th feature in the feature vector.
[0054] Bayesian inference algorithms update confidence in the occurrence of an event by combining existing prior knowledge with newly collected measurement data. In the context of fault diagnosis, Bayesian inference is used to calculate confidence scores for different fault types (or fault modes) given measurement data (a set of feature indicators). After selecting the fault mode that best matches the measurement data, a structured matching result needs to be generated. This structured result will include: 1. Fault mode type: The type of fault ultimately identified (e.g., overheating fault, short circuit, etc.). 2. Confidence score: The confidence score for each failure mode indicates the degree of match between the failure mode and the measurement data; 3. Feature index set: Feature data that supports the fault diagnosis, such as amplitude deviation and phase change at certain frequency points, to help explain why this fault mode was selected.
[0055] The comprehensive judgment report generation module takes into account the error distribution characteristics data and the fault type matching results to generate a test report containing detailed judgment conclusions, and associates the metadata, process data and test report generated during the test with the corresponding product model.
[0056] Specific examples are as follows: In a factory, during the diagnosis of a radio frequency device (such as a signal transmission device), the device's scattering parameters ( S 11 , S 21 Abnormal fluctuations occurred during the test, and the measured values were... S 11 and S 21 The data at different frequencies are as follows: In the scattering parameter dataset preprocessing module, the measured data are matched with the standard limit curve, a standard limit curve is set, and the specified limits are defined. S 11 The amplitude should not exceed -10dB, and S 21 The amplitude should be greater than -4dB; then all data should be standardized.
[0057] The tolerance thresholds are set as follows: right S 11 The amplitude deviation has a tolerance threshold of ±2dB. right S 21 The amplitude deviation has a tolerance threshold of ±1dB. S 11 Amplitude deviation: For a frequency of 1.0 GHz, the standard value is -10 dB, the actual measurement is -10 dB, and the deviation is 0 dB (within the tolerance range). For a frequency of 1.5GHz, the standard value is -10dB, but the actual measurement is -12dB, which is a deviation of 2dB (exceeding the standard). For a frequency of 2.0 GHz, the standard value is -10 dB, the actual measurement is -11 dB, and the deviation is 1 dB (within the tolerance range). S21 Amplitude deviation: For a frequency of 1.0 GHz, the standard value is -4 dB, the actual measurement is -3 dB, and the deviation is 1 dB (within the tolerance range). For a frequency of 1.5GHz, the standard value is -4dB, but the actual measurement is -2dB, which is a deviation of 2dB (exceeding the standard). For a frequency of 2.0 GHz, the standard value is -4 dB, the actual measurement is -4 dB, and the deviation is 0 dB (within the tolerance range). Based on the above, the error distribution characteristics are extracted; Fault Type 1: Overheating Fault – This fault mode is characterized by… S 11 The amplitude deviation is large, and the frequency bands are continuously exceeding the standard; Fault Type 2: Short Circuit – This fault mode is characterized by… S 21 The amplitude deviation is large, and the phase shows obvious abnormalities; Prior probability: Overheating fault: P ( F 1) = 0.6; Circuit short circuit: P( F 2) = 0.4; Likelihood: Overheating fault: P(E|) F 1) = 0.8; Circuit short circuit: P(E| F 2) = 0.7; Based on the expression that best matches the fault type, ; Therefore, the score for the overheating fault is 0.4105; The score for the short circuit was 0.2579. The corresponding fault type is overheating fault.
[0058] It should be noted that the calculation formulas and all parameters involved in the calculations in this application have been dimensionless beforehand. The process of dimensionless processing is well known in the industry and will not be described here.
[0059] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of this application.
Claims
1. An automated testing system based on a vector network analyzer, characterized in that, This automated testing system based on a vector network analyzer includes: The knowledge base construction module defines a set of attributes for filter products, uses a graph database to establish the association between the attribute set and test configuration parameters, and forms a product functional test specification knowledge base. The test item matching module extracts product identifiers and test requirement keywords from user input, matches them with product specifications in the product function test specification knowledge base, and outputs a list of test items containing specific parameter constraints. The command sequence generation module queries the corresponding parameter constraints from the product functional test specification knowledge base based on the list of items to be tested, and determines the execution order of the items through the test process rule engine to generate a sequence of measurement commands that can be executed by the vector network analyzer. The dynamic measurement data acquisition module executes a sequence of measurement commands and dynamically adjusts the scanning parameters based on the real-time scattering parameter characteristics to obtain a scattering parameter dataset that meets the preset requirements. The intelligent diagnosis and management module compares the output scattering parameter dataset with the standard limit curves stored in the product function test specification knowledge base point by point, executes the fault type matching algorithm, outputs the test item judgment result, and then associates the complete test data record with the corresponding product model in the knowledge base.
2. The automatic testing system based on a vector network analyzer according to claim 1, characterized in that, The knowledge base construction module includes: The key attribute set definition module defines the set of key attributes for filter products, including type, topology, rated power, and impedance. The test configuration parameter determination module determines the test configuration parameters corresponding to the set of key attributes in the filter product. The relational network construction module uses a graph database to associate the defined set of key attributes with test configuration parameters to build a relational network. The product functional test specification knowledge base generation module verifies the constructed network of relationships and, once verified, solidifies it into the product functional test specification knowledge base.
3. The automatic testing system based on a vector network analyzer according to claim 1, characterized in that, The command sequence generation module includes: The test item element structure extraction module parses the list of test items, extracts the test items and parameter constraints, and obtains structured test item elements. The test configuration parameter matching and completion module queries the product function test specification knowledge base based on the structured test item elements to obtain the corresponding complete test configuration parameters. The test execution order optimization and sorting module uses the test process rule engine to sort the test project elements with complete test configuration parameters to determine the optimal execution order; The instrument command sequence generation module converts the test item elements arranged in the optimal execution order into a sequence of measurement commands that can be executed by the vector network analyzer.
4. The automatic testing system based on a vector network analyzer according to claim 3, characterized in that, The process of using a test flow rule engine to sort test project elements with complete test configuration parameters and determine the optimal execution order includes: Receive test project elements with complete test configuration parameters and analyze the logical dependencies between the test project elements; Based on the logical dependencies obtained from the analysis, sorting rules in the test process rule engine aimed at improving test efficiency are applied. Prioritize test item elements with complete test configuration parameters based on sorting rules, and output a sequence of test item elements arranged in the optimal execution order.
5. The automatic testing system based on a vector network analyzer according to claim 1, characterized in that, The dynamic measurement data acquisition module includes: The measurement control module controls the vector network analyzer to execute a sequence of measurement commands and acquire real-time scattering parameter datasets. The data analysis and decision-making module uses a parameter decision-making algorithm to analyze the trends of the acquired scattering parameter dataset and dynamically adjust the frequency scanning density and signal power level. The adjustment measurement module performs measurements based on the dynamically adjusted frequency scan density and signal power level, and outputs a scattering parameter dataset that meets the preset requirements.
6. The automatic testing system based on a vector network analyzer according to claim 5, characterized in that, The step of using a parameter decision algorithm to analyze the trend of the acquired scattering parameter dataset and dynamically adjust the frequency scanning density and signal power level includes: Frequency domain gradient analysis was performed on the acquired scattering parameter dataset to calculate the amplitude change rate at each frequency point; Based on the amplitude change rate of each frequency point obtained from frequency domain gradient analysis, key frequency domain regions are identified, and the frequency band boundaries where the change amplitude exceeds the preset interval are marked to obtain the amplitude change characteristics of scattering parameters. Based on the marked key frequency domain regions, calculate the signal-to-noise ratio under the current measurement configuration and evaluate the stability of data quality. Based on the data quality stability assessment results obtained from the identified key frequency domain region distribution and signal-to-noise ratio indicators, the frequency scan density adjustment coefficient is determined. Based on the amplitude variation characteristics of scattering parameters in key frequency domain regions and the frequency scan density adjustment coefficient, the signal power level compensation value is calculated.
7. The automatic testing system based on a vector network analyzer according to claim 6, characterized in that, The amplitude change rate at each frequency point obtained based on frequency domain gradient analysis is used to identify key frequency domain regions, mark the frequency band boundaries where the change amplitude exceeds a preset interval, and obtain the amplitude change characteristics of the scattering parameters, including: Calculate the absolute value of the rate of change of amplitude at each frequency point to obtain the gradient amplitude; The gradient magnitude is compared with a preset threshold range to filter all candidate frequency points that exceed the threshold range; Cluster analysis is performed on the selected candidate frequency points to merge adjacent frequency points on the frequency axis and form a continuous key frequency domain region. Calculate the statistical characteristics of each key frequency domain region, and output the final list of key frequency domain regions and their corresponding amplitude variation characteristics.
8. The automatic testing system based on a vector network analyzer according to claim 1, characterized in that, The intelligent diagnosis and management module includes: The scattering parameter dataset preprocessing module preprocesses the measured scattering parameter dataset, matches the processed dataset with the standard limit curves stored in the product functional test specification knowledge base at frequency points, and establishes a standardized data mapping relationship. The multi-dimensional error feature extraction module calculates the amplitude and phase deviation at each frequency point based on the established standardized data mapping relationship, and completes multi-dimensional error assessment by combining the preset tolerance threshold and extracting error distribution features. The fault type matching and analysis module uses a fault type matching algorithm to analyze the extracted error distribution features, identify specific fault types, calculate the corresponding fault confidence, and generate fault type matching results. The comprehensive judgment report generation module takes into account the error distribution characteristics data and the fault type matching results to generate a test report containing detailed judgment conclusions, and associates the metadata, process data and test report generated during the test with the corresponding product model.
9. The automatic testing system based on a vector network analyzer according to claim 8, characterized in that, The step of using a fault type matching algorithm to analyze and extract error distribution features, identify specific fault types, calculate corresponding fault confidence levels, and generate fault type matching results includes: Read the error distribution feature data output by the multi-dimensional error feature extraction module, extract the amplitude deviation statistics, phase anomaly distribution, and continuity of the out-of-standard frequency band from it, and construct a set of feature indicators; The extracted feature index set is matched with the fault type templates stored in the product function test specification knowledge base in multiple dimensions, and the most suitable fault type is identified based on the matching results. Based on the matching degree between the feature index set and the fault type, the statistical characteristics of historical fault data, and the current test environment parameters, the confidence score of each candidate fault type is calculated using the Bayesian inference algorithm. By integrating the identified fault types, calculated confidence scores, and feature index sets, a structured fault type matching result is generated.