A Python-based radio frequency test network construction and evaluation method and system

Through the construction and evaluation method of radio frequency test network based on Python, using technologies such as clustering, hashing algorithms and Bayesian networks, the problems of slow data processing speed and low accuracy of RF test networks are solved, and the efficiency of RF test networks is achieved Evaluation and optimization improve the accuracy and efficiency of test results.

CN118509333BActive Publication Date: 2024-11-08XIAN QIANJING DEFENSE TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410932150.X
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-07-12
Publication Date
2024-11-08
Estimated Expiration
2044-07-12

AI Technical Summary

Technical Problem

In the prior art, the data processing speed of the RF test network is slow, the data processing capability is poor and the algorithm accuracy is low, making it difficult to effectively apply to the evaluation and optimization of the RF test network.

Method used

Using the Python-based RF test network construction and evaluation method, design topology by determining the test objectives and requirements, write Python scripts to build a test network, collect performance parameters in real time, and use clustering, hashing algorithm, entropy weight method and Bayesian network Perform optimization and diagnosis.

Benefits of technology

The precise construction and evaluation of the RF test network is realized, which improves the accuracy and reliability of the test results, reduces the test time and improves efficiency.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118509333B_ABST
    Figure CN118509333B_ABST
Patent Text Reader

Abstract

The present invention relates to the field of testing technology, and in particular to a method and system for constructing and evaluating a radio frequency test network based on Python. The real-time test environment of the radio frequency test network is obtained, and various standard performance parameters of the radio frequency test network when working under the conditions of the real-time test environment are determined according to the real-time test environment; the actual performance parameters of the radio frequency test network in the pre-test process are collected; the actual performance parameters collected are clustered; the radio frequency test network is evaluated according to various actual performance parameters and standard performance parameters to obtain an evaluation result; if it is the first evaluation result, the radio frequency test network is controlled based on a Python script to perform actual testing on the test target; if it is the second evaluation result, indicating that the performance of the radio frequency test network is unqualified, the radio frequency test network is optimized, and the accurate construction and evaluation of the test network are realized, which can ensure the accuracy of the test results, thereby improving the performance and reliability of the radio frequency network.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of testing technology, and in particular to a Python-based radio frequency testing network construction and evaluation method and system. Background Art

[0002] The RF test network is a system used to evaluate and measure the performance of RF equipment. It usually includes components such as signal sources, power amplifiers, filters, RF switches, spectrum analyzers, and network analyzers. These components are precisely configured and connected to form a test path to simulate the signal transmission and processing process in the actual communication environment. The RF test network can be used to verify the design, performance, and quality of RF equipment, including the measurement of parameters such as signal frequency, amplitude, phase, bandwidth, distortion, and intermodulation characteristics. It is of great significance for the research and development, production, and maintenance of communication systems. Automated evaluation and optimization of RF test networks can greatly improve the efficiency of RF testing, reduce the time and errors of manual operations, and improve the accuracy and repeatability of tests. Machine learning algorithms have great potential in the field of automated RF testing. However, the current algorithms used to analyze and process data have limitations, such as slow processing speed, poor data processing capabilities, and low algorithm accuracy. Therefore, how to effectively apply these technologies to the evaluation and optimization of RF test networks remains a challenge. Summary of the invention

[0003] The present invention overcomes the deficiencies of the prior art and provides a Python-based radio frequency test network construction and evaluation method and system.

[0004] To achieve the above-mentioned purpose, the technical solution adopted by the present invention is:

[0005] The first aspect of the present invention discloses a method for constructing and evaluating a radio frequency test network based on Python, comprising the following steps:

[0006] Determine test objectives, test requirements and test hardware, and design test network topology and signal paths based on the test objectives, test requirements and test hardware;

[0007] Building a radio frequency test network according to the test network topology and signal path, and writing a Python script according to the radio frequency test network;

[0008] Acquire a real-time test environment of the radio frequency test network, and determine various standard performance parameters of the radio frequency test network when it works under the real-time test environment conditions according to the real-time test environment;

[0009] Based on the Python script, the RF test network is controlled to perform pre-testing work, and actual performance parameters of the RF test network during the pre-testing work are collected; the actual performance parameters collected are clustered to obtain various actual performance parameters;

[0010] The RF test network is evaluated according to various actual performance parameters and standard performance parameters to obtain an evaluation result; if the evaluation result is a first evaluation result, indicating that the performance of the RF test network is qualified, then the RF test network is controlled based on the Python script to perform actual testing on the test target;

[0011] If the evaluation result is the second evaluation result, indicating that the performance of the RF test network is unqualified, the RF test network is optimized; after the optimization is completed, the RF test network is controlled based on the Python script to perform actual testing on the test target.

[0012] Furthermore, in a preferred embodiment of the present invention, a real-time test environment of a radio frequency test network is obtained, and various standard performance parameters of the radio frequency test network when working under the real-time test environment conditions are determined according to the real-time test environment, specifically:

[0013] Obtain various standard performance parameters of the radio frequency test network under various preset test environment conditions based on the big data network retrieval, build a knowledge base, and import various standard performance parameters of the radio frequency test network under various preset test environment conditions into the knowledge base;

[0014] Acquire a real-time test environment of a radio frequency test network, and extract various preset test environments from the knowledge base; calculate hash values ​​between the real-time test environment and various preset test environments based on a hash algorithm to obtain a plurality of hash values;

[0015] Determine the similarity between the real-time test environment and various preset test environments according to each hash value to obtain a plurality of similarities; and sort the plurality of similarities in ascending order to sort out the maximum similarity degree;

[0016] A preset test environment corresponding to the maximum similarity is obtained, and a pairing tag is generated according to the preset test environment corresponding to the maximum similarity, and various standard performance parameters of the radio frequency test network when working under the real-time test environment conditions are paired in the knowledge base based on the pairing tag.

[0017] Furthermore, in a preferred embodiment of the present invention, the collected actual performance parameters are clustered to obtain various actual performance parameters, specifically:

[0018] Introduce the fuzzy clustering algorithm, initialize several initial fuzzy clustering centers, and regard each actual performance parameter as a data point;

[0019] Calculate the Chebyshev distance between each data point and each initial fuzzy cluster center, construct a sorting table, and import the Chebyshev distance between each data point and each initial fuzzy cluster center into the sorting table for sorting, so as to obtain a sorting result based on the numerical value;

[0020] Selecting the initial fuzzy cluster center corresponding to the minimum Chebyshev distance in the sorting results as the membership cluster center of the corresponding data point, and obtaining the membership cluster center to which each data point belongs;

[0021] Assign each data point to its affiliated cluster center, calculate the membership degree between each data point and its affiliated cluster center, and compare the membership degree between each data point and its affiliated cluster center with a preset membership degree threshold;

[0022] If the degree of membership between all data points and their affiliated cluster centers is greater than the preset membership threshold, the iteration is stopped, and the data points in each affiliated cluster center are obtained, and the clustering results are output according to the data points in each affiliated cluster center;

[0023] If there is a situation where the degree of membership between one or more data points and their affiliated cluster centers is not greater than a preset membership threshold, the initial fuzzy cluster centers are reinitialized, and re-iteration is performed based on the reinitialized initial fuzzy cluster centers until the degrees of membership between all data points and their affiliated cluster centers are greater than the preset membership threshold, then the iteration is stopped, and the data points in each affiliated cluster center are obtained, and the clustering results are output based on the data points in each affiliated cluster center;

[0024] Various actual performance parameters are obtained according to the clustering results.

[0025] Furthermore, in a preferred embodiment of the present invention, the RF test network is evaluated according to various actual performance parameters and standard performance parameters to obtain evaluation results, specifically:

[0026] Calculate the difference between each actual performance parameter and the corresponding standard performance parameter respectively; and preset the deviation threshold of each performance parameter in advance, and compare the difference between each actual performance parameter and the corresponding standard performance parameter with the deviation threshold;

[0027] If the difference between a certain actual performance parameter and the corresponding standard performance parameter is greater than the deviation threshold, the actual performance parameter is marked as an abnormal performance parameter;

[0028] If the difference between a certain actual performance parameter and the corresponding standard performance parameter is not greater than the deviation threshold, the actual performance parameter is marked as a normal performance parameter;

[0029] Analyze the status of each actual performance parameter; if all actual performance parameters are normal performance parameters, generate a first evaluation result; if there is a situation where at least one actual performance parameter is an abnormal performance parameter, generate a second evaluation result.

[0030] Further, in a preferred embodiment of the present invention, if the evaluation result is the second evaluation result, indicating that the performance of the radio frequency test network is unqualified, the radio frequency test network is optimized, specifically:

[0031] Obtain the functional elements of all test hardware in the RF test network, and obtain abnormal performance parameters of the RF test network;

[0032] Calculate the weight value between the functional elements of each test hardware and the abnormal performance parameters of the radio frequency test network based on the entropy weight method, and determine the correlation between the functional elements of each test hardware and the abnormal performance parameters of the radio frequency test network according to the weight value;

[0033] Acquire test hardware corresponding to a correlation degree greater than a preset correlation degree, and define it as suspicious test hardware; and acquire real-time operation data of the suspicious test hardware;

[0034] Constructing a Bayesian network, and importing the real-time operation data of the suspicious test hardware into the Bayesian network for diagnosis, to obtain the diagnosis result of the suspicious test hardware;

[0035] Obtain the abnormal probability of the suspicious test hardware in the diagnosis result; obtain the suspicious test hardware whose abnormal probability is greater than the preset abnormal probability, and define it as abnormal test hardware; and obtain the abnormal type of the abnormal test hardware in the diagnosis result.

[0036] Furthermore, in a preferred embodiment of the present invention, a Bayesian network is constructed, and the real-time operation data of the suspicious test hardware is imported into the Bayesian network for diagnosis, and the diagnosis result of the suspicious test hardware is obtained, specifically:

[0037] Based on big data network retrieval, possible abnormal types of each test hardware are obtained, and characteristic variables that affect each abnormal type are defined;

[0038] Acquire characteristic data corresponding to various abnormal types, and determine the causal relationship between characteristic variables according to the characteristic data; construct a Bayesian network according to the causal relationship between characteristic variables, where the nodes of the Bayesian network represent characteristic variables and the edges represent the causal relationship between characteristic variables;

[0039] Assign a prior probability to each feature variable, that is, the probability of the variable state without any information; create a conditional probability table for each node in the Bayesian network, which represents the conditional probability under the state of the parent node;

[0040] Acquire real-time operation data of the suspicious test hardware, import the real-time operation data of the suspicious test hardware into the Bayesian network, and use the Bayesian theorem to update the posterior probability of each node in the network;

[0041] Compare the posterior probability of each node with the preset probability threshold respectively; obtain the node corresponding to the posterior probability greater than the preset probability threshold, and define it as the associated node;

[0042] Obtain the posterior probability of the associated node, determine the abnormal probability of the suspicious test hardware based on the posterior probability of the associated node, and obtain the node position of the associated node, determine the abnormal type of the suspicious test hardware based on the node position of the associated node; generate a diagnostic result based on the abnormal probability and abnormal type of the suspicious test hardware.

[0043] Furthermore, in a preferred embodiment of the present invention, the following steps are also included:

[0044] Formulate optimization plans in advance for various abnormalities of various test hardware in the RF test network;

[0045] Construct a database, and import the optimization solutions corresponding to various abnormal types of each test hardware in the radio frequency test network that have been formulated in advance into the database to obtain a characteristic database;

[0046] Obtain the abnormal type of the abnormal test hardware, and obtain a corresponding optimization solution according to the abnormal type of the abnormal test hardware in the characteristic database;

[0047] The optimization scheme is sent to the control terminal, and the radio frequency test network is optimized and adjusted based on the optimization scheme.

[0048] The second aspect of the present invention discloses a Python-based radio frequency test network construction and evaluation system, the radio frequency test network construction and evaluation system includes a memory and a processor, the memory stores a radio frequency test network construction and evaluation method program, when the radio frequency test network construction and evaluation method program is executed by the processor, any one of the steps of the radio frequency test network construction and evaluation method is implemented.

[0049] The present invention solves the technical defects existing in the background technology, and has the following beneficial effects: the present invention automatically analyzes the radio frequency test data in the pre-test process, evaluates the problems existing in the test network, considers the influence of environmental factors on the performance of the test network, optimizes the structure and parameters of the test network, realizes the accurate construction and evaluation of the test network, can ensure the accuracy of the test results, thereby improving the performance and reliability of the radio frequency network, and adopts a simple and easy-to-implement algorithm, which can effectively reduce the test time and improve the test efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, drawings of other embodiments can be obtained based on these drawings without paying creative work.

[0051] Figure 1 It is a flowchart of the overall method of building and evaluating RF test networks based on Python;

[0052] Figure 2 This is a partial method flow chart of a Python-based RF test network construction and evaluation method;

[0053] Figure 3 This is a system block diagram of a Python-based RF test network construction and evaluation system. DETAILED DESCRIPTION

[0054] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0055] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0056] like Figure 1 As shown, the first aspect of the present invention discloses a method for constructing and evaluating a radio frequency test network based on Python, comprising the following steps:

[0057] S102: Determine test objectives, test requirements and test hardware, and design a test network topology and signal path according to the test objectives, test requirements and test hardware;

[0058] Among them, the test requirements include the test parameter range and test accuracy requirements; the test hardware includes RF signal generator, RF power meter, spectrum analyzer, network analyzer, power meter, test cable and connector; the test target is the object to be tested; the test target, test requirements and test hardware are all determined by relevant technical personnel;

[0059] S104: Building a radio frequency test network according to the test network topology and signal path, and writing a Python script according to the radio frequency test network;

[0060] The topology includes signal splitting, combining, filtering and amplification. Signal paths include linear, cascade, ring and star paths. Python scripts can be written using software such as PyCharm.

[0061] S106: Acquire a real-time test environment of the radio frequency test network, and determine various standard performance parameters of the radio frequency test network when it works under the real-time test environment conditions according to the real-time test environment;

[0062] Among them, performance parameters include signal-to-noise ratio, signal distortion, frequency response, frequency coverage, bandwidth, phase noise, reflection loss, insertion loss, measurement speed and accuracy, etc.; test environment includes temperature, humidity, magnetic field strength, etc.;

[0063] S108: Controlling the radio frequency test network to perform pre-test work based on the Python script, and collecting actual performance parameters of the radio frequency test network during the pre-test work; performing clustering processing on the collected actual performance parameters to obtain various actual performance parameters;

[0064] S110: Evaluate the radio frequency test network according to various actual performance parameters and standard performance parameters to obtain an evaluation result; if the evaluation result is a first evaluation result, indicating that the performance of the radio frequency test network is qualified, then control the radio frequency test network to perform actual testing on the test target based on the Python script;

[0065] S112: If the evaluation result is the second evaluation result, indicating that the performance of the RF test network is unqualified, the RF test network is optimized; after the optimization is completed, the RF test network is controlled based on the Python script to perform actual testing on the test target.

[0066] It should be noted that, first of all, it is necessary to clarify the goals and requirements of the test network, including the RF parameters, frequency range, signal type, test speed, etc. to be tested; according to the goals and requirements of the test network, design the topology of the test network, including signal sources, test equipment (such as spectrum analyzers, network analyzers, signal generators, etc.), RF components (such as amplifiers, filters, antennas, etc.) and RF cables and connectors connecting them; physically connect all selected devices and components, ensure that all cables and connectors are properly connected, and follow proper RF wiring and shielding practices; then evaluate, calibrate and optimize the test network to ensure the accuracy of the measurement.

[0067] like Figure 2 As shown, further, in a preferred embodiment of the present invention, the real-time test environment of the radio frequency test network is obtained, and various standard performance parameters of the radio frequency test network when working under the real-time test environment conditions are determined according to the real-time test environment, specifically:

[0068] S202: Obtain various standard performance parameters of the radio frequency test network under various preset test environment conditions based on the big data network retrieval, construct a knowledge base, and import various standard performance parameters of the radio frequency test network under various preset test environment conditions into the knowledge base;

[0069] S204: Acquire a real-time test environment of the radio frequency test network, and extract various preset test environments from the knowledge base; calculate hash values ​​between the real-time test environment and various preset test environments based on a hash algorithm to obtain a plurality of hash values;

[0070] Among them, the principle of hashing algorithm to calculate the similarity between two sets of environmental parameters mainly involves mapping the environmental parameter sets into numerical space and then comparing the differences between these numerical representations. Specifically, the hashing algorithm converts each set of environmental parameters into a fixed-length hash value, which can be regarded as points in the numerical space. By comparing the distance or similarity measure (such as cosine similarity, Euclidean distance, etc.) between these hash values, the similarity between two sets of environmental parameters can be evaluated. This method is widely used in many applications because it is both efficient and able to capture the key differences between environmental parameter sets;

[0071] S206: Determine the similarity between the real-time test environment and various preset test environments according to each hash value to obtain a plurality of similarities; and sort the plurality of similarities in ascending order to sort out the maximum similarity degree;

[0072] S208: Acquire a preset test environment corresponding to the maximum similarity, generate a pairing tag according to the preset test environment corresponding to the maximum similarity, and pair various standard performance parameters of the radio frequency test network when working under the real-time test environment conditions in the knowledge base based on the pairing tag.

[0073] It should be noted that the various standard performance parameters of the RF test network under various preset test environment conditions can be retrieved in the big data network by setting keywords. For example, first clarify the specific parameters you want to retrieve, such as signal strength, signal-to-noise ratio, frequency response, etc., and then determine the data source for the search, such as academic papers, technical documents, industry standards, professional forums, etc. In the selected data source, use keywords related to the target parameters to search, for example, search for "RF test", "signal-to-noise ratio", etc., so as to retrieve standard performance parameters such as frequency response, frequency coverage, bandwidth, phase noise, reflection loss, etc. when the RF test network is working under a certain temperature and humidity range.

[0074] Furthermore, in a preferred embodiment of the present invention, the collected actual performance parameters are clustered to obtain various actual performance parameters, specifically:

[0075] Introduce the fuzzy clustering algorithm, initialize several initial fuzzy clustering centers, and regard each actual performance parameter as a data point;

[0076] Calculate the Chebyshev distance between each data point and each initial fuzzy cluster center, construct a sorting table, and import the Chebyshev distance between each data point and each initial fuzzy cluster center into the sorting table for sorting, so as to obtain a sorting result based on the numerical value;

[0077] Selecting the initial fuzzy cluster center corresponding to the minimum Chebyshev distance in the sorting results as the membership cluster center of the corresponding data point, and obtaining the membership cluster center to which each data point belongs;

[0078] Assign each data point to its affiliated cluster center, calculate the membership degree between each data point and its affiliated cluster center, and compare the membership degree between each data point and its affiliated cluster center with a preset membership degree threshold;

[0079] Among them, the preset membership threshold is set to 98%;

[0080] If the degree of membership between all data points and their affiliated cluster centers is greater than the preset membership threshold, the iteration is stopped, and the data points in each affiliated cluster center are obtained, and the clustering results are output according to the data points in each affiliated cluster center;

[0081] If there is a situation where the degree of membership between one or more data points and their affiliated cluster centers is not greater than a preset membership threshold, the initial fuzzy cluster centers are reinitialized, and re-iteration is performed based on the reinitialized initial fuzzy cluster centers until the degrees of membership between all data points and their affiliated cluster centers are greater than the preset membership threshold, then the iteration is stopped, and the data points in each affiliated cluster center are obtained, and the clustering results are output based on the data points in each affiliated cluster center;

[0082] Various actual performance parameters are obtained according to the clustering results.

[0083] It should be noted that fuzzy C-means clustering is a clustering algorithm based on fuzzy logic, which is used to group the actual performance parameters collected into different categories. Through fuzzy C-means clustering, the performance parameters can be effectively divided into several categories with similar characteristics, thereby providing a basis for further performance analysis and optimization. The fuzzy C-means clustering algorithm has a fast convergence speed and can effectively improve the efficiency of data processing. However, fuzzy C-means clustering also has certain limitations, such as being sensitive to the initial cluster center and being highly dependent on parameter selection; such as clustering some invalid or noise data into a valid membership cluster center, which affects the quality of cluster data and thus affects the evaluation results. Therefore, if there is a situation where the membership between one or more data points and their membership cluster center is not greater than the preset membership threshold, it is necessary to reinitialize the initial fuzzy cluster center and re-iterate according to the reinitialized initial fuzzy cluster center, that is, to re-iterate by re-determining the initial fuzzy cluster center to obtain a better clustering effect and improve the quality of cluster data.

[0084] Furthermore, in a preferred embodiment of the present invention, the RF test network is evaluated according to various actual performance parameters and standard performance parameters to obtain evaluation results, specifically:

[0085] Calculate the difference between each actual performance parameter and the corresponding standard performance parameter respectively; and preset the deviation threshold of each performance parameter in advance, and compare the difference between each actual performance parameter and the corresponding standard performance parameter with the deviation threshold;

[0086] The deviation threshold is set to 2 to 10, preferably 5;

[0087] If the difference between a certain actual performance parameter and the corresponding standard performance parameter is greater than the deviation threshold, the actual performance parameter is marked as an abnormal performance parameter;

[0088] If the difference between a certain actual performance parameter and the corresponding standard performance parameter is not greater than the deviation threshold, the actual performance parameter is marked as a normal performance parameter;

[0089] Analyze the status of each actual performance parameter; if all actual performance parameters are normal performance parameters, generate a first evaluation result; if there is a situation where at least one actual performance parameter is an abnormal performance parameter, generate a second evaluation result.

[0090] It should be noted that if the difference between a certain actual performance parameter and the corresponding standard performance parameter is greater than the deviation threshold, it means that the deviation between the actual performance parameter and the standard state is too large, and the actual performance parameter is marked as an abnormal performance parameter; if the difference between a certain actual performance parameter and the corresponding standard performance parameter is not greater than the deviation threshold, it means that the deviation between the actual performance parameter and the standard state is not large, and the actual performance parameter is marked as a normal performance parameter; if all actual performance parameters are normal performance parameters, it means that the performance of the RF test network is qualified, and at this time, there is no need to optimize the built RF test network, and the RF test network can be controlled based on the Python script to perform actual testing on the test target. If there is a situation where at least one actual performance parameter is an abnormal performance parameter, it means that the performance of the RF test network is unqualified, and the RF test network needs to be optimized and adjusted.

[0091] Further, in a preferred embodiment of the present invention, if the evaluation result is the second evaluation result, indicating that the performance of the radio frequency test network is unqualified, the radio frequency test network is optimized, specifically:

[0092] Obtain the functional elements of all test hardware in the RF test network, and obtain abnormal performance parameters of the RF test network;

[0093] Among them, the functional elements of the test hardware refer to the functional role corresponding to the test hardware. For example, the RF signal generator can generate RF signals, so the RF signal range in the RF test network is related to the RF signal generator;

[0094] Calculate the weight value between the functional elements of each test hardware and the abnormal performance parameters of the radio frequency test network based on the entropy weight method, and determine the correlation between the functional elements of each test hardware and the abnormal performance parameters of the radio frequency test network according to the weight value;

[0095] Among them, the entropy weight method is a weight determination method based on information entropy theory. It measures the discreteness of indicators by calculating the entropy value of each indicator, thereby determining the weight of each indicator in the overall evaluation system. In the RF test network, the principle of determining the correlation between each test hardware functional element and abnormal performance parameters by the entropy weight method can be summarized as follows: According to the abnormal performance parameters of the RF test network, the entropy value of each functional element is calculated using the entropy weight method to reflect its discreteness under different working conditions; then the weight of each functional element is determined according to the entropy value. The higher the weight, the stronger the correlation between the functional element and the abnormal performance parameters of the RF test network; finally, by analyzing the functional elements with a higher degree of correlation, the abnormal situation in the RF test network can be more accurately identified and located, thereby providing a basis for hardware diagnosis and performance optimization;

[0096] Obtain test hardware corresponding to a correlation greater than a preset correlation, which is defined as suspicious test hardware; and obtain real-time operation data of the suspicious test hardware; wherein the preset correlation is set to 98%; the real-time operation data of the suspicious test hardware includes real-time voltage, real-time current, real-time temperature, real-time resistance, etc.;

[0097] Constructing a Bayesian network, and importing the real-time operation data of the suspicious test hardware into the Bayesian network for diagnosis, to obtain the diagnosis result of the suspicious test hardware;

[0098] Obtain the abnormal probability of the suspicious test hardware in the diagnosis result; obtain the suspicious test hardware whose abnormal probability is greater than the preset abnormal probability, and define it as abnormal test hardware; and obtain the abnormal type of the abnormal test hardware in the diagnosis result. The preset abnormal probability is set to 90%.

[0099] It should be noted that if the correlation of a certain test hardware is greater than the preset correlation, it means that the abnormal performance parameters produced in the RF test network may be caused by the abnormal reasons of the test hardware, and it is marked as suspicious test hardware. Then, by obtaining the real-time operation data of the suspicious test hardware and diagnosing the suspicious test hardware in combination with the Bayesian network model, the diagnosis result of the suspicious test hardware is obtained, so as to further determine the abnormal test hardware. This method can quantify the correlation between each hardware functional element and the abnormal performance parameter, so as to more accurately identify the hardware components that may cause the test network abnormality and improve the accuracy and reliability of the diagnosis.

[0100] Furthermore, in a preferred embodiment of the present invention, a Bayesian network is constructed, and the real-time operation data of the suspicious test hardware is imported into the Bayesian network for diagnosis, and the diagnosis result of the suspicious test hardware is obtained, specifically:

[0101] Based on big data network retrieval, possible abnormal types of each test hardware are obtained, and characteristic variables that affect each abnormal type are defined;

[0102] Acquire characteristic data corresponding to various abnormal types, and determine the causal relationship between characteristic variables according to the characteristic data; construct a Bayesian network according to the causal relationship between characteristic variables, where the nodes of the Bayesian network represent characteristic variables and the edges represent the causal relationship between characteristic variables;

[0103] Assign a prior probability to each feature variable, that is, the probability of the variable state without any information; create a conditional probability table for each node in the Bayesian network, which represents the conditional probability under the state of the parent node;

[0104] Acquire real-time operation data of the suspicious test hardware, import the real-time operation data of the suspicious test hardware into the Bayesian network, and use the Bayesian theorem to update the posterior probability of each node in the network;

[0105] Compare the posterior probability of each node with the preset probability threshold respectively; obtain the nodes corresponding to the posterior probability greater than the preset probability threshold, and define them as associated nodes; wherein the preset probability threshold is set to 90%;

[0106] Obtain the posterior probability of the associated node, determine the abnormal probability of the suspicious test hardware based on the posterior probability of the associated node, and obtain the node position of the associated node, determine the abnormal type of the suspicious test hardware based on the node position of the associated node; generate a diagnostic result based on the abnormal probability and abnormal type of the suspicious test hardware.

[0107] It should be noted that the possible abnormal types of each test hardware are retrieved based on the big data network, such as overheating, abnormal voltage, abnormal current, etc., and the characteristic variables that affect each abnormal type are defined, such as temperature, voltage, current, usage time, etc. Expert knowledge is used in combination with the characteristic data to determine the relationship between variables, for example, temperature may affect voltage, voltage may affect current, etc. A directed acyclic graph is constructed, in which nodes represent variables and edges represent causal relationships between variables. Conditional probabilities are assigned to each variable in the network. These probabilities represent the probability of a variable taking a specific value given the value of its parent variable. These probabilities are obtained through expert knowledge or experimental data. Then, real-time operating data of the suspected test hardware is collected, which will be used to update the probabilities in the Bayesian network. The Bayesian theorem is used to update the probabilities of the variables in the network based on the collected data, which usually involves the propagation of conditional probabilities, which can be achieved through various algorithms such as variable elimination and belief propagation. Then, by finding the abnormal types with the highest posterior probabilities, determine which abnormal types have the highest probability. The Bayesian network has good robustness and can perform effective reasoning and diagnosis even when faced with incomplete or uncertain data. This step can diagnose the abnormal probability and abnormal type of each suspicious test hardware, which helps to analyze and locate the abnormal situation, thereby optimizing the RF test network and improving the reliability of the test results.

[0108] Furthermore, in a preferred embodiment of the present invention, the following steps are also included:

[0109] Formulate optimization plans in advance for various abnormalities of various test hardware in the RF test network;

[0110] Construct a database, and import the optimization solutions corresponding to various abnormal types of each test hardware in the radio frequency test network that have been formulated in advance into the database to obtain a characteristic database;

[0111] Obtain the abnormal type of the abnormal test hardware, and obtain a corresponding optimization solution according to the abnormal type of the abnormal test hardware in the characteristic database;

[0112] The optimization scheme is sent to the control terminal, and the radio frequency test network is optimized and adjusted based on the optimization scheme.

[0113] It should be noted that the optimization schemes corresponding to various abnormal types of each test hardware in the RF test network are formulated in advance by relevant technical personnel, such as the optimization scheme corresponding to the temperature abnormality of the RF signal generator, and the optimization schemes include the control procedures for various hardware. Build a database, and import the optimization schemes corresponding to various abnormal types of each test hardware in the RF test network formulated in advance into the database to obtain a characteristic database, and then, after diagnosing the abnormal type of the abnormal test hardware through the Bayesian network, match the abnormal type of the abnormal test hardware with the various abnormal types in the characteristic database, so as to quickly match the corresponding optimization scheme in the characteristic database; and send the optimization scheme to the control terminal, optimize and adjust the RF test network based on the optimization scheme, and after the optimization is completed, control the RF test network based on the Python script to perform actual testing on the test target.

[0114] In summary, the present invention automatically analyzes the RF test data in the pre-test process, evaluates the problems existing in the test network, takes into account the impact of environmental factors on the performance of the test network, optimizes the structure and parameters of the test network, and realizes the accurate construction and evaluation of the test network. It can ensure the accuracy of the test results, thereby improving the performance and reliability of the RF network, and adopts a simple and easy-to-implement algorithm, which can effectively reduce the test time and improve the test efficiency.

[0115] In addition, the method further comprises the following steps:

[0116] Draw a standard electromagnetic distribution diagram of the radio frequency test network when it works under real-time test environment conditions according to the test network topology and signal path;

[0117] In the process of pre-testing the radio frequency test network, actual electromagnetic intensity values ​​of several preset position nodes are obtained, and an actual electromagnetic distribution map of the radio frequency test network is drawn according to the actual electromagnetic intensity values ​​of several preset position nodes;

[0118] Calculating the cosine similarity between the actual electromagnetic distribution map and the standard electromagnetic distribution map based on a cosine similarity algorithm; and comparing the cosine similarity with a preset cosine similarity threshold; wherein the preset cosine similarity threshold is set to 0.8;

[0119] If the cosine similarity is greater than a preset cosine similarity threshold, it means that the electromagnetic intensity in the test environment is normal;

[0120] If the cosine similarity is not greater than the preset cosine similarity threshold, it means that the electromagnetic intensity in the test environment is abnormal, indicating that there is a signal interference source in the test environment; the actual electromagnetic distribution diagram is subjected to matrix conversion processing to obtain a first gray level co-occurrence matrix, and the standard electromagnetic distribution diagram is subjected to matrix conversion processing to obtain a second gray level co-occurrence matrix;

[0121] Compare the elements at the same matrix position in the first gray level co-occurrence matrix and the second gray level co-occurrence matrix; if there is a difference between the elements at the same matrix position, mark the position corresponding to the position as a singular point;

[0122] A closed area connected by all the singular points is obtained, and the closed area connected by all the singular points is marked as the location of the signal interference source.

[0123] It should be noted that the principle of converting the electromagnetic distribution map into a grayscale co-occurrence matrix involves converting the electromagnetic parameters of the image (such as intensity, phase, and polarization) into grayscale values ​​and calculating the co-occurrence relationship between these grayscale values. Specifically, this process includes the following steps: First, the electromagnetic parameters of each pixel in the electromagnetic distribution map are converted into grayscale values. This is usually achieved by mapping the parameters to integers between 0 and 255 to represent different grayscale levels. For each pixel, an angle (θ) and a distance (d) are selected, which define the range of pixel pairs considered in the co-occurrence matrix. The angle is usually between 0° and 180°, while the distance can be selected according to the characteristics of the image and the required resolution. For each pixel, find its corresponding co-occurrence pixel according to the selected angle and distance. Then, calculate the co-occurrence relationship between the grayscale values ​​of the original pixel and the co-occurrence pixel, and store these relationships in a matrix. This matrix is ​​the grayscale co-occurrence matrix. By converting the electromagnetic distribution map into a gray-level co-occurrence matrix, the texture characteristics of the image can be quantitatively analyzed, thereby better understanding and explaining the distribution characteristics of the electromagnetic field, so that the signal interference source can be quickly analyzed and located later.

[0124] It should be noted that during the operation of the RF test network, if there is a source of signal interference (i.e., an electromagnetic interference source), due to the multipath effect or interference from adjacent frequency bands, it will cause problems with the signal quality of the RF test network, thereby affecting the test process and test results. Therefore, in the present invention, it is possible to effectively evaluate whether there is a signal interference source in the RF test network. If there is a signal interference source, and the position of the signal interference source can be located, by locating the interference source and taking measures to reduce its impact, the performance of the entire RF test network can be optimized, the data transmission rate can be improved, and the bit error rate can be reduced. By identifying and eliminating the interference source, the signal quality in the RF test network can be improved, thereby improving the reliability and performance of the communication system.

[0125] In addition, the method further comprises the following steps:

[0126] Obtain abnormal test hardware in the radio frequency test network and the abnormal type of the abnormal test hardware;

[0127] A fault tree analysis algorithm is introduced, and abnormal test hardware in the radio frequency test network is defined as a top event, and the remaining test hardware in the radio frequency test network is defined as a sub-event, and a fault tree is constructed according to the top event and the sub-event;

[0128] According to the abnormal type of abnormal test hardware, Boolean logic is used to analyze the association probability between the top event and each sub-event step by step from the top event in the fault tree;

[0129] A sub-event with a correlation probability greater than a preset correlation probability is defined as correlation test hardware, and all correlation test hardware is marked as suspicious test hardware; wherein the preset correlation probability is set to 95%.

[0130] It should be noted that in the RF test network, various test hardware are usually interconnected and dependent on each other. When a certain test hardware fails, it may cause a chain reaction to some test hardware in the test network, causing these test hardware to fail as well. Therefore, if abnormal test hardware is detected in the test network, it is necessary to further analyze whether the remaining test hardware associated with the abnormal test hardware also has the possibility of failure, and then mark the test hardware with the possibility of failure as suspicious, and then use the Bayesian network to diagnose it, so as to achieve the function of comprehensive evaluation and inventory of the RF test network.

[0131] like Figure 3 As shown, the second aspect of the present invention discloses a Python-based RF test network construction and evaluation system, the RF test network construction and evaluation system includes a memory 20 and a processor 80, the memory 20 stores a RF test network construction and evaluation method program, when the RF test network construction and evaluation method program is executed by the processor 80, any one of the RF test network construction and evaluation method steps is implemented.

[0132] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

[0133] The units described above as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units; they may be located in one place or distributed on multiple network units; some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0134] In addition, all functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may be separately used as a unit, or two or more units may be integrated into one unit; the above-mentioned integrated units may be implemented in the form of hardware or in the form of hardware plus software functional units.

[0135] A person of ordinary skill in the art can understand that: all or part of the steps of implementing the above method embodiment can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above method embodiment; and the aforementioned storage medium includes: a mobile storage device, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and other media that can store program codes.

[0136] Alternatively, if the above-mentioned integrated unit of the present invention is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiment of the present invention can be essentially or partly reflected in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the methods of each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as mobile storage devices, ROM, RAM, magnetic disks or optical disks.

[0137] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art who is familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A Python-based RF test network construction and evaluation method, It is characterized in that The following steps are involved: Determine test objectives, test requirements and test hardware, and design test network topology and signal paths based on the test objectives, test requirements and test hardware; Building a radio frequency test network according to the test network topology and signal path, and writing a Python script according to the radio frequency test network; Acquire a real-time test environment of the radio frequency test network, and determine various standard performance parameters of the radio frequency test network when it works under the real-time test environment conditions according to the real-time test environment; Based on the Python script, the RF test network is controlled to perform pre-test work, and actual performance parameters of the RF test network during the pre-test work are collected; Perform clustering processing on the collected actual performance parameters to obtain various actual performance parameters; Evaluate the RF test network according to various actual performance parameters and standard performance parameters to obtain evaluation results; If the evaluation result is the first evaluation result, indicating that the performance of the radio frequency test network is qualified, then the radio frequency test network is controlled based on the Python script to perform actual testing on the test target; If the evaluation result is the second evaluation result, indicating that the performance of the radio frequency test network is unqualified, the radio frequency test network is optimized; After the optimization is completed, the RF test network is controlled based on the Python script to perform actual testing on the test target; The collected actual performance parameters are clustered to obtain various actual performance parameters, which are as follows: Introduce the fuzzy clustering algorithm, initialize several initial fuzzy clustering centers, and regard each actual performance parameter as a data point; Calculate the Chebyshev distance between each data point and each initial fuzzy cluster center, construct a sorting table, and import the Chebyshev distance between each data point and each initial fuzzy cluster center into the sorting table for size sorting to obtain a sorting result based on numerical value size; Selecting the initial fuzzy cluster center corresponding to the minimum Chebyshev distance in the sorting results as the membership cluster center of the corresponding data point, and obtaining the membership cluster center to which each data point belongs; Assign each data point to its affiliated cluster center, calculate the degree of membership between each data point and its affiliated cluster center, and compare the degree of membership between each data point and its affiliated cluster center with a preset membership threshold; If the degree of membership between all data points and their affiliated cluster centers is greater than the preset membership threshold, the iteration is stopped, and the data points in each affiliated cluster center are obtained, and the clustering results are output according to the data points in each affiliated cluster center; If there is a situation where the degree of membership between one or more data points and their affiliated cluster centers is not greater than a preset membership threshold, the initial fuzzy cluster centers are reinitialized, and re-iteration is performed based on the reinitialized initial fuzzy cluster centers until the degrees of membership between all data points and their affiliated cluster centers are greater than the preset membership threshold, then the iteration is stopped, and the data points in each affiliated cluster center are obtained, and the clustering results are output based on the data points in each affiliated cluster center; Obtaining various actual performance parameters according to the clustering results; If the evaluation result is the second evaluation result, indicating that the performance of the radio frequency test network is unqualified, the radio frequency test network is optimized, specifically: Obtain the functional elements of all test hardware in the RF test network, and obtain abnormal performance parameters of the RF test network; Calculate the weight value between the functional elements of each test hardware and the abnormal performance parameters of the radio frequency test network based on the entropy weight method, and determine the correlation between the functional elements of each test hardware and the abnormal performance parameters of the radio frequency test network according to the weight value; Acquire test hardware corresponding to a correlation degree greater than a preset correlation degree, and define it as suspicious test hardware; and acquire real-time operation data of the suspicious test hardware; Constructing a Bayesian network, and importing the real-time operation data of the suspicious test hardware into the Bayesian network for diagnosis, to obtain the diagnosis result of the suspicious test hardware; Obtaining the abnormal probability of the suspicious test hardware in the diagnosis result; obtaining the suspicious test hardware whose abnormal probability is greater than the preset abnormal probability, defining it as abnormal test hardware; and obtaining the abnormal type of the abnormal test hardware in the diagnosis result; Among them, a Bayesian network is constructed, and the real-time operation data of the suspicious test hardware is imported into the Bayesian network for diagnosis to obtain the diagnosis result of the suspicious test hardware, which is specifically: Based on big data network retrieval, possible abnormal types of each test hardware are obtained, and characteristic variables that affect each abnormal type are defined; Acquire characteristic data corresponding to various abnormal types, and determine the causal relationship between characteristic variables according to the characteristic data; construct a Bayesian network according to the causal relationship between characteristic variables, where the nodes of the Bayesian network represent characteristic variables and the edges represent the causal relationship between characteristic variables; Assign a prior probability to each feature variable, that is, the probability of the variable state without any information; create a conditional probability table for each node in the Bayesian network, which represents the conditional probability under the state of the parent node; Acquire real-time operation data of the suspicious test hardware, import the real-time operation data of the suspicious test hardware into the Bayesian network, and use the Bayesian theorem to update the posterior probability of each node in the network; Compare the posterior probability of each node with the preset probability threshold respectively; obtain the node corresponding to the posterior probability greater than the preset probability threshold, and define it as the associated node; Obtaining the posterior probability of the associated node, determining the abnormal probability of the suspicious test hardware according to the posterior probability of the associated node, and obtaining the node position of the associated node, determining the abnormal type of the suspicious test hardware according to the node position of the associated node; generating a diagnosis result according to the abnormal probability and abnormal type of the suspicious test hardware; Among them, the RF test network is evaluated according to various actual performance parameters and standard performance parameters, and the evaluation results are obtained, specifically: Calculate the difference between each actual performance parameter and the corresponding standard performance parameter respectively; and preset the deviation threshold of each performance parameter in advance, and compare the difference between each actual performance parameter and the corresponding standard performance parameter with the deviation threshold; If the difference between a certain actual performance parameter and the corresponding standard performance parameter is greater than the deviation threshold, the actual performance parameter is marked as an abnormal performance parameter; If the difference between a certain actual performance parameter and the corresponding standard performance parameter is not greater than the deviation threshold, the actual performance parameter is marked as a normal performance parameter; Analyze the status of each actual performance parameter; if all the actual performance parameters are normal performance parameters, generate a first evaluation result; if at least one actual performance parameter is an abnormal performance parameter, generate a second evaluation result; The following steps are also included: Formulate optimization plans in advance for various abnormalities of various test hardware in the RF test network; Construct a database, and import the optimization solutions corresponding to various abnormal types of each test hardware in the radio frequency test network that have been formulated in advance into the database to obtain a characteristic database; Obtain the abnormal type of the abnormal test hardware, and obtain a corresponding optimization solution according to the abnormal type of the abnormal test hardware in the characteristic database; Sending the optimization scheme to the control terminal, and optimizing and adjusting the radio frequency test network based on the optimization scheme; The following steps are also included: Draw a standard electromagnetic distribution diagram of the radio frequency test network when it works under real-time test environment conditions according to the test network topology and signal path; In the process of pre-testing the radio frequency test network, actual electromagnetic intensity values ​​of several preset position nodes are obtained, and an actual electromagnetic distribution map of the radio frequency test network is drawn according to the actual electromagnetic intensity values ​​of several preset position nodes; Calculating the cosine similarity between the actual electromagnetic distribution map and the standard electromagnetic distribution map based on a cosine similarity algorithm; and comparing the cosine similarity with a preset cosine similarity threshold; If the cosine similarity is greater than a preset cosine similarity threshold, it means that the electromagnetic intensity in the test environment is normal; If the cosine similarity is not greater than the preset cosine similarity threshold, it means that the electromagnetic intensity in the test environment is abnormal, indicating that there is a signal interference source in the test environment; the actual electromagnetic distribution diagram is subjected to matrix conversion processing to obtain a first gray level co-occurrence matrix, and the standard electromagnetic distribution diagram is subjected to matrix conversion processing to obtain a second gray level co-occurrence matrix; Compare the elements at the same matrix position in the first gray level co-occurrence matrix and the second gray level co-occurrence matrix; if there is a difference between the elements at the same matrix position, mark the position corresponding to the position as a singular point; Obtain a closed area connected by all singular points, and mark the closed area connected by all singular points as the location of the signal interference source; The steps of converting the electromagnetic distribution map into a gray-level co-occurrence matrix are as follows: The electromagnetic parameters of each pixel in the electromagnetic distribution map are converted into gray values ​​between 0 and 255 to represent different gray levels; An angle θ and a distance d are selected for each pixel, and the angle θ and the distance d define the range of pixel pairs in the co-occurrence matrix; the angle θ is between 0° and 180°, and the distance d is selected according to the characteristics of the image and the required resolution; For each pixel, find its corresponding co-occurrence pixel according to the selected angle θ and distance d, then calculate the co-occurrence relationship between the grayscale values ​​of the original pixel and the co-occurrence pixel, and generate a grayscale co-occurrence matrix according to the co-occurrence relationship between the grayscale values ​​of the original pixel and the co-occurrence pixel; The following steps are also included: Obtain abnormal test hardware in the radio frequency test network and the abnormal type of the abnormal test hardware; A fault tree analysis algorithm is introduced, and abnormal test hardware in the radio frequency test network is defined as a top event, and the remaining test hardware in the radio frequency test network is defined as a sub-event, and a fault tree is constructed according to the top event and the sub-event; According to the abnormal type of abnormal test hardware, Boolean logic is used to analyze the association probability between the top event and each sub-event step by step from the top event in the fault tree; A sub-event with a correlation probability greater than a preset correlation probability is defined as correlation test hardware, and all correlation test hardware is marked as suspicious test hardware.

2. According to the method for constructing and evaluating a radio frequency test network based on Python in claim 1, It is characterized in that Acquire a real-time test environment of the radio frequency test network, and determine various standard performance parameters of the radio frequency test network when working under the real-time test environment conditions according to the real-time test environment, specifically: Obtain various standard performance parameters of the radio frequency test network under various preset test environment conditions based on the big data network retrieval, build a knowledge base, and import various standard performance parameters of the radio frequency test network under various preset test environment conditions into the knowledge base; Acquire a real-time test environment of a radio frequency test network, and extract various preset test environments from the knowledge base; Calculating hash values ​​between the real-time test environment and various preset test environments based on a hash algorithm to obtain a plurality of hash values; Determine the similarity between the real-time test environment and various preset test environments according to each hash value to obtain a plurality of similarities; and sort the plurality of similarities in ascending order to sort out the maximum similarity degree; A preset test environment corresponding to the maximum similarity is obtained, and a pairing tag is generated according to the preset test environment corresponding to the maximum similarity, and various standard performance parameters of the radio frequency test network when working under the real-time test environment conditions are paired in the knowledge base based on the pairing tag.

3. A Python-based RF test network construction and evaluation system, It is characterized in that The RF test network construction and evaluation system includes a memory and a processor, and the memory stores a RF test network construction and evaluation method program. When the RF test network construction and evaluation method program is executed by the processor, the RF test network construction and evaluation method steps as described in any one of claims 1-2 are implemented.