Test data processing method and device and terminal equipment

Abstract

The invention provides a test data processing method and device and terminal equipment, and is suitable for the technical field of data processing.The method comprises the steps that feature extraction processing is conducted on to-be-processed test data according to a preset test data feature extraction matrix, and initial test data feature information is obtained; performing hierarchical filtering and reconstruction processing according to the initial test data feature information to obtain target test data feature information; and based on a preset test data processing model, according to the target test data feature information and the product test index threshold information, generating test data processing information. According to the method and the device, the fragmentation dilemma of multi-category product test data processing is broken, the efficiency and the accuracy of one-stop test service are improved, and a standardized and traceable test data processing result is provided for customers.

Description

Technical Field

[0001] This application belongs to the field of data processing technology, and in particular relates to test data processing methods, apparatus and terminal equipment. Background Technology

[0002] Current testing technologies are showing a trend towards intelligent and rapid development, with the penetration rate of technologies such as AI visual recognition, IoT real-time monitoring, and blockchain traceability gradually increasing.

[0003] Existing technologies conduct targeted tests based on the characteristics of different consumer products, such as electrical safety and energy efficiency testing for household appliances, heavy metal leaching and mechanical safety testing for infant and toddler toys, and migration detection for food contact products. At the same time, they rely on a laboratory management system (LIMS) to realize basic recording and report generation of test data, and provide standard alignment and certification application services for multiple countries / regions.

[0004] However, in existing technologies, there is a lack of a unified and standardized connection mechanism for testing processes of different categories, which leads to low efficiency when testing multiple product categories together and makes it difficult to meet the diversified needs of cross-border trade. Summary of the Invention

[0005] In view of this, embodiments of this application provide a test data processing method, apparatus, and terminal equipment, aiming to solve the problems of insufficient standardization and low data comparison efficiency in the existing technology for one-stop testing of multiple categories of consumer products.

[0006] The first aspect of this application provides a test data processing method, including:

[0007] Acquire multiple test data sets to be processed and multiple product test indicator threshold information;

[0008] Based on multiple preset test data feature extraction matrices, feature extraction processing is performed on the multiple test data to be processed to obtain multiple initial test data feature information;

[0009] Based on the feature information of the multiple initial test data, hierarchical filtering and reconstruction processing are performed to obtain the feature information of multiple target test data.

[0010] Based on a preset test data processing model, test data processing information is generated according to the multiple target test data feature information and multiple product test indicator threshold information.

[0011] A second aspect of this application provides a test data processing apparatus, comprising:

[0012] The module for acquiring test data to be processed and product test indicator threshold information is used to acquire multiple test data to be processed and multiple product test indicator threshold information.

[0013] The initial test data feature information generation module is used to perform feature extraction processing on the multiple test data to be processed according to multiple preset test data feature extraction matrices to obtain multiple initial test data feature information.

[0014] The target test data feature information generation module is used to perform hierarchical filtering and reconstruction processing based on the multiple initial test data feature information to obtain multiple target test data feature information.

[0015] The test data processing information generation module is used to generate test data processing information based on a preset test data processing model, according to the multiple target test data feature information and multiple product test indicator threshold information.

[0016] A third aspect of this application provides a terminal device, the terminal device including a memory and a processor, the memory storing a computer program executable on the processor, and the processor executing the computer program to implement the steps of the test data processing method described in the first aspect above.

[0017] A fourth aspect of this application provides a computer-readable storage medium, comprising: storing a computer program, wherein when executed by a processor, the computer program implements the steps of the test data processing method described in the first aspect above.

[0018] The beneficial effects of this application embodiment compared with the prior art are as follows: This application avoids the limitations of a single data dimension in product compliance judgment. It uses multiple test data feature extraction matrices to perform deep feature extraction on the test data to be processed. By performing hierarchical filtering and reconstruction processing on the extracted test data features, core features and redundant features are separated, which facilitates deep correlation analysis with multiple product test indicator threshold information. This enables efficient processing and accurate judgment of test data for multiple product categories, thereby improving the accuracy and effectiveness of compliance judgment and risk warning for multiple product categories. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a schematic diagram of the implementation flow of the test data processing method provided in Embodiment 1 of this application;

[0021] Figure 2 This is a schematic diagram of the implementation flow of the test data processing method provided in Embodiment 2 of this application;

[0022] Figure 3 This is a schematic diagram illustrating the implementation flow of the test data processing method provided in Embodiment 3 of this application;

[0023] Figure 4 This is a schematic diagram illustrating the implementation flow of the test data processing method provided in Embodiment 4 of this application;

[0024] Figure 5 This is a schematic diagram illustrating the implementation flow of the test data processing method provided in Embodiment 5 of this application;

[0025] Figure 6 This is a schematic diagram illustrating the implementation flow of the test data processing method provided in Embodiment Six of this application;

[0026] Figure 7 This is a schematic diagram of the implementation flow of the test data processing method provided in Embodiment 7 of this application;

[0027] Figure 8 This is a schematic diagram of the structure of the test data processing device provided in the embodiments of this application;

[0028] Figure 9 This is a schematic diagram of the terminal device provided in the embodiments of this application. Detailed Implementation

[0029] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0030] To illustrate the technical solution described in this application, specific embodiments are provided below.

[0031] Figure 1 A flowchart illustrating the implementation of the test data processing method provided in Embodiment 1 of this application is shown, and is described in detail below:

[0032] Step S101: Obtain multiple test data to be processed and multiple product test indicator threshold information.

[0033] In this embodiment, the test data to be processed can refer to the full-dimensional raw test data collected by professional testing equipment for multiple categories of products, such as kitchen utensils, home appliances, baby toys, gifts and sundries, and household goods. This data can be collected under standard testing environments using equipment such as electrical performance testers, chemical substance analyzers, mechanical strength testers, and noise detectors. It can include electrical safety test data, chemical safety test data, mechanical performance test data, energy efficiency performance test data, and electromagnetic compatibility test data, facilitating subsequent standardized analysis, compliance assessment, and risk warning for test data of multiple product categories. The threshold information for multiple product test indicators can refer to compliance limit data for various test indicators set according to product access standards of different countries or regions, covering multiple regional standard limits such as EU REACH regulations, US FDA standards, and Chinese GB national standards. Specifically, the threshold information for multiple product test indicators can include electrical safety threshold data, such as insulation resistance limits, withstand voltage limits, and leakage current limits; and can include chemical safety threshold data, such as heavy metal leaching limits, phthalate content limits, and formaldehyde release limits, in text or numerical form. Threshold information for multiple product testing metrics can be obtained by reading the global product compliance standards database.

[0034] Step S102: Based on multiple preset test data feature extraction matrices, perform feature extraction processing on the multiple test data to be processed to obtain multiple initial test data feature information.

[0035] In this embodiment, optionally, the preset test data feature extraction matrix can be manually preset, and can be a matrix with multiple identical dimensions or multiple matrices with different dimensions, used for feature extraction from the test data to be processed. The test data to be processed can first be converted into matrix form using professional data processing software to generate test data matrix information. Then, multiple preset test data feature extraction matrices are convolved with the test data matrix information, and the results of the convolution calculations are weighted and summed based on preset weights (which can be manually preset). The weighted sum is then used as the initial test data feature information. In this embodiment, optionally, the test data to be processed can first be standardized. Standardization may include data normalization, preliminary screening of outliers, and data smoothing. Then, based on a pre-set test data classification mask, different categories of test data in the test data to be processed are identified, and electrical safety test data, chemical safety test data, mechanical performance test data, energy efficiency performance test data, and electromagnetic compatibility test data are extracted respectively. Then, multiple pre-set test data feature extraction matrices are used to perform convolution operations on the electrical safety test data, chemical safety test data, mechanical performance test data, energy efficiency performance test data, and electromagnetic compatibility test data respectively. Then, the results of the convolution of each test data feature extraction matrix are output separately, and the output results are spliced ​​together. The spliced ​​result is used as the initial test data feature information. In this embodiment, optionally, the multiple preset test data feature extraction matrices can be manually preset. The test data feature extraction matrices may include a preset test data compliance deviation feature matrix, a preset test data stability feature matrix, and a preset test data risk level feature matrix. The test data compliance deviation feature matrix is ​​used to perform convolution operations on the differences between various test data and standard thresholds to extract deviation amplitude and trend information and generate a deviation response map. The test data stability feature matrix is ​​used to calculate the data fluctuation amplitude and dispersion of repeated test data to quantify the consistency characteristics of the test data. The test data risk level feature matrix is ​​used to perform multi-level decomposition on the critical qualified test data to separate the fluctuation threshold features and risk correlation features. Then, the feature maps output by each test data feature extraction matrix are standardized and then stitched together, and the stitched feature map is used as the initial test data feature information.

[0036] Step S103: Perform hierarchical filtering and reconstruction processing based on the multiple initial test data feature information to obtain multiple target test data feature information.

[0037] In this embodiment, it is understood that the initial test data feature information contains a large number of basic features, and therefore has much repetitive, secondary, or noisy information. Therefore, it is necessary to separate the key features from redundant and noisy features in the initial test data feature information to filter out redundant and noisy information. This can be achieved by pre-setting a set of basic test data feature units, which can be manually preset. It is understood that a set of basic test data feature units includes multiple basic test data feature sub-units. Multiple sets of basic test data features can be generated by decomposing the initial test data feature information into linear combinations representing these basic test data feature units. The weight information of the sub-units is used to perform a weighted summation on multiple basic test data feature sub-units. This determines the basic test data feature sub-units and weight information combinations whose weighted summation result is closest to the initial test data feature information. Then, the weight values ​​in the weight information combination are arranged from largest to smallest. The weight values ​​of the first half are increased, the weight values ​​of the first half of the second half are decreased, and the weight values ​​of the last half of the second half are adjusted to 0. This is to highlight the effective information in the test data feature information and suppress and filter out redundant and noise information. Finally, the basic test data feature sub-units are weighted and summed using the adjusted weight values, and the result of the weighted summation is used as the target test data feature information.

[0038] Step S104: Based on the preset test data processing model, test data processing information is generated according to the multiple target test data feature information and multiple product test indicator threshold information.

[0039] In this embodiment, the preset test data processing model can be manually preset. It can involve first aligning the target test data feature information and multiple product test indicator threshold information; performing one-hot encoding on the regional standard classification information within the multiple product test indicator threshold information; and standardizing the text and numerical data within the multiple product test indicator threshold information. Then, through a preset mapping function (which can also be manually preset), the one-hot encoded regional standard classification information and the standardized text and numerical data are mapped to a high-dimensional space. This high-dimensional space allows for the mapping of test indicators and products... Using product category and regional standards as nodes, and "test indicator-product category association" and "indicator threshold-regional standard association" as edges, the target test data feature information and multiple product test indicator threshold information are respectively used as node attributes. By calculating the similarity between node attributes and nodes and edges, and artificially pre-setting a similarity threshold, nodes and edges with similarity greater than the similarity threshold are extracted. Then, the calculated similarity is used as a weight value, and the extracted nodes and edges are weighted and summed. The weighted sum is then used as the product compliance judgment result and risk warning level corresponding to the test data to be processed. It can be understood that the generated test data processing information may include compliance judgment results of multiple product categories, matching degree data of different regional standards, and risk level assessment information of test data.

[0040] The test data processing method provided in this application avoids the limitations of a single data dimension in product compliance determination. It utilizes multiple test data feature extraction matrices to perform deep feature extraction on the test data to be processed. By performing hierarchical filtering and reconstruction processing on the extracted test data features, core features and redundant features are separated. This facilitates in-depth correlation analysis with multiple product test indicator threshold information, thereby achieving efficient processing and accurate determination of test data for multiple product categories. This improves the accuracy and effectiveness of compliance determination and risk warning for multiple product categories.

[0041] Figure 2 The flowchart illustrating the implementation of the test data processing method provided in Embodiment 2 of this application is shown. The difference between this method and Embodiment 1 is that step S102 specifically includes:

[0042] Step S201: Based on the preset test data block dimension information, the multiple test data to be processed are divided into blocks to obtain multiple test data block information.

[0043] In this embodiment, the preset test data block dimension information can be manually preset and may include test data block category dimension information and test data block numerical dimension information. The test data block category dimension information corresponds to product test categories such as electrical safety and chemical safety, while the test data block numerical dimension information corresponds to the numerical range of different test items within the same category. The test data to be processed can be divided into blocks based on the preset test data block dimension information, grouping test data within the same category dimension and the same numerical dimension range into the same block, thereby generating multiple test data block information to ensure that different types of test data can be classified and analyzed.

[0044] Step S202: Extract the numerical information of the multiple test data block information to obtain the numerical information of the multiple test data blocks.

[0045] In this embodiment, it can be understood that each test data block contains a set of test data to be processed belonging to the same category and numerical dimension. The specific numerical content in each test data block is extracted one by one to obtain information such as the single test value, the range of numerical fluctuation and the average value in each block. These extracted contents are used as the test data block numerical information.

[0046] Step S203: According to the preset test data block value selection order, the multiple test data block value information is selected and processed to obtain multiple selected test data block value information and multiple unselected test data block value information.

[0047] In this embodiment, the preset order of selecting test data block values ​​can be preset manually. It can be selected according to the priority of the test data block category dimension information from high to low. Electrical safety test data block values ​​are selected first, followed by chemical safety test data block values, and then mechanical performance test data block values. Within the same category dimension, values ​​are selected in descending order. The first test data block value is selected as the selected test data block value, and the remaining values ​​in the block that were not selected for the first time are considered as unselected test data block values.

[0048] Step S204: Based on the numerical information of multiple selected test data blocks, the numerical information of multiple unselected test data blocks, and multiple preset test data feature extraction matrices, multiple initial test data feature information are obtained.

[0049] In this embodiment, the preset test data feature extraction matrix can be manually preset and may include a test data compliance deviation feature matrix, a test data stability feature matrix, and a test data risk level feature matrix. The process involves first converting the selected test data block numerical information into a one-dimensional vector, then performing convolution operations with multiple preset test data feature extraction matrices to generate multiple test data feature response vectors. For unselected test data block numerical information, the numerical correlation degree between it and the selected test data block numerical information is calculated. The correlation degree is determined by quantifying the proportion of the difference between the two. Then, the correlation degree information is multiplied element-wise with the test data feature response vector to obtain a weighted feature vector of the test data. All weighted feature vectors of the test data are then concatenated, and the concatenated result is used as the initial test data feature information. By distinguishing between selected and unselected numerical information and assigning different weights, the characteristics of core test data are effectively highlighted, and interference from secondary data is suppressed. For example, in electrical safety test data, the characteristics of selected withstand voltage values ​​are enhanced, while the characteristics of unselected auxiliary monitoring values ​​are weakened, improving the targeting of feature extraction.

[0050] The test data processing method provided in this application dynamically adjusts the segmentation dimensions for core test items of different product categories to ensure that key test data such as electrical safety and chemical safety are not missed. By distinguishing between selected and unselected test data segmentation numerical information and combining it with feature extraction matrix processing, the method effectively strengthens core data features and suppresses redundant data interference, thereby improving the accuracy of test data feature extraction. This provides high-quality feature data for subsequent hierarchical filtering and compliance judgment, ensuring the effectiveness and reliability of test data processing for multiple product categories.

[0051] Figure 3 The flowchart illustrating the implementation of the test data processing method provided in Embodiment 3 of this application is shown. The difference between this method and Embodiment 2 is that step S204 specifically includes:

[0052] Step S301: Calculate the number of selected test data blocks and the number of unselected test data blocks to obtain the number of test data blocks.

[0053] In this embodiment, the number of selected test data block values ​​and the number of unselected test data block values ​​within a single test data block information can be counted separately. Then, the counts of the two types of values ​​are added together to obtain the total number of values ​​corresponding to the test data block information. Finally, the total number of values ​​of all test data block information is summarized to generate test data block quantity information.

[0054] Step S302: Based on the numerical information of the multiple selected test data blocks and the numerical information of the multiple unselected test data blocks, calculate the numerical deviation information of the multiple test data blocks.

[0055] In this embodiment, the difference between the numerical information of each selected test data block and the numerical information of all unselected test data blocks in the corresponding block can be calculated one by one to obtain the absolute difference between each unselected value and the selected value. Then, these absolute differences can be used as the test data block numerical deviation value information, and the degree of difference between different values ​​can be quantified by the deviation value.

[0056] Step S303: Calculate the local deviation information of the test data blocks based on the numerical deviation information of the multiple test data blocks.

[0057] In this embodiment, the numerical deviation values ​​of multiple test data blocks within a single test data block can be aggregated. First, median filtering is performed to remove abnormal deviation values. Then, a Gaussian kernel function is used for smoothing. Subsequently, a weighted average is calculated on the smoothed deviation values ​​based on manually set weight information. The result of the weighted average is used as the local deviation value information of the test data block, which is used to characterize the overall severity of the numerical differences within the block, such as the concentration of withstand voltage deviation in the electrical safety block.

[0058] Step S304: Determine whether the local deviation value information of the test data block is greater than or equal to the preset local deviation threshold information of the test data block; if yes, generate test data value range information based on the multiple selected test data block values ​​and the multiple unselected test data block values; if no, skip the multiple selected test data block values ​​and the multiple unselected test data block values.

[0059] In this embodiment, the preset threshold information for the local deviation of test data block values ​​can be manually preset. When the local deviation value of the test data block values ​​is greater than or equal to the threshold information, it is considered that the numerical differences within the test data block are significant and may contain key features that affect compliance determination. Therefore, the numerical information of the selected test data block and the numerical information of the unselected test data block within the current block are combined to generate test data value range information covering the numerical range and deviation distribution of the block, which is used for subsequent deep feature extraction. When the local deviation value of the test data block values ​​is less than the threshold information, it is considered that the numerical distribution within the block is uniform and there are no key difference features. Therefore, the two types of numerical information of the block are skipped, and there is no need to generate test data value range information to improve processing efficiency.

[0060] Step S305: Count the number of times the test data block value information is selected to obtain the test data block value selection count information.

[0061] In this embodiment, each time a selection operation is completed on the numerical information of the test data block within a single test data block, the selection count is incremented by 1, thereby recording the number of selections completed within the current block and generating test data block numerical selection count information, ensuring that the selection process of all numerical information within the block is traceable and quantifiable.

[0062] Step S306: Determine whether the number of times the test data block value is selected is less than the number of test data blocks; if yes, return to step S203; if no, proceed to step S307.

[0063] In this embodiment, the number of times test data blocks are selected is compared with the number of test data blocks. When the number of selections is less than the total number, it means that there are still unselected test data block values ​​within the test data block, so the selection process continues. When the number of selections is equal to the total number, it means that all values ​​within the block have been selected, and there is no need to continue iterating, ensuring that the values ​​within the block are fully parsed and avoiding the omission of key data.

[0064] Step S307: Based on multiple preset test data feature extraction matrices, feature extraction processing is performed on the test data value range information to obtain multiple initial test data feature information.

[0065] In this embodiment, the multiple preset test data feature extraction matrices can be manually preset and may include a test data compliance deviation feature matrix and a test data stability feature matrix. The test data value range information can be first converted into a two-dimensional tensor, and then convolved with the two feature extraction matrices respectively. The test data compliance deviation feature matrix extracts the deviation trend features between the values ​​in the value range and the standard threshold, while the test data stability feature matrix extracts the fluctuation pattern features of the values ​​in the value range. The results of the two convolution calculations are then concatenated and used as the independent variable of the ReLU function. After ReLU function calculation, the output function value serves as multiple initial test data feature information, enhancing the representation capability of key difference features.

[0066] The test data processing method provided in this application accurately identifies regions with significant numerical differences, filters out worthless data blocks by combining local deviation threshold judgment, improves the targeting and efficiency of data processing, ensures that the numerical information in each test data block is fully analyzed, avoids the omission of key features, and deeply extracts test data value range information based on feature extraction matrix to strengthen compliance deviation and stability features, thereby improving the accuracy of initial test data feature information, providing high-quality support for subsequent target feature extraction and compliance judgment, and ensuring the accuracy of test data processing for multiple product categories.

[0067] Figure 4 The flowchart illustrating the implementation of the test data processing method provided in Embodiment 4 of this application is shown. Its difference from Embodiment 3 described above lies in:

[0068] The test data value range information includes multiple core numerical information of the test data value range and multiple edge numerical information of the test data value range;

[0069] Multiple preset test data feature extraction matrices include preset test data linear feature extraction weight matrices and preset test data nonlinear feature extraction weight matrices;

[0070] Step S307 specifically includes:

[0071] Step S401: Based on the preset linear feature extraction weight matrix of the test data, the edge numerical information of the multiple test data value ranges is weighted to generate multiple linear weighted information of the edge numerical information of the test data value ranges.

[0072] In this embodiment, the preset linear feature extraction weight matrix for the test data can be manually preset, or it can be a manually set two-dimensional matrix. The weight values ​​are linearly distributed along the numerical change trend, used to highlight the linear variation characteristics of the numerical information at the edge of the test data value range. This can be achieved by performing element-wise multiplication of the linear feature extraction weight matrix with the numerical information at the edge of the test data value range, summing the results, and then using the summed result as the linear weighted information of the numerical information at the edge of the test data value range. This is used to enhance linearly correlated edge features, for example, highlighting the linear change trend of withstand voltage value with test time in electrical safety test data, or strengthening the linear fluctuation characteristics of heavy metal leaching with temperature in chemical safety test data.

[0073] Step S402: Based on the preset nonlinear feature extraction weight matrix of the test data, the numerical information of the edge values ​​of the multiple test data ranges is weighted to generate multiple nonlinear weighted information of the edge values ​​of the test data ranges.

[0074] In this embodiment, the preset weight matrix for extracting nonlinear features of the test data can be manually preset, or it can be a manually set two-dimensional matrix. The weight values ​​adopt an exponential or logarithmic distribution to highlight the nonlinear abrupt changes in the numerical information at the edge of the test data value range. This can be achieved by performing element-wise multiplication of the weight matrix with the numerical information at the edge of the test data value range, summing the results, and using this summed result as the nonlinear weighted information at the edge of the test data value range. This enhances the nonlinearly related edge features, for example, highlighting the stress abrupt changes during material fracture in mechanical performance test data, or amplifying the nonlinear fluctuation peaks of power consumption in energy efficiency test data.

[0075] Step S403: Based on the linear weighted information of the edge values ​​of the multiple test data ranges, the non-linear weighted information of the edge values ​​of the multiple test data ranges, and the core value information of the multiple test data ranges, multiple initial test data feature information are obtained.

[0076] In this embodiment, the core numerical information of the test data range can be first converted into a vector of the same dimension as the linearly weighted information of the edge values ​​of the test data range. Then, the converted core numerical information of the test data range is concatenated with the linearly weighted information and the non-linearly weighted information of the edge values ​​of the test data range to generate a three-dimensional feature vector containing a core numerical benchmark, linear change characteristics, and non-linear abrupt change characteristics. This three-dimensional feature vector is then used as the initial test data feature information. Alternatively, in this embodiment, the core numerical information of the test data range can represent the benchmark level of the test data, the linearly weighted information can reflect the stable change trend of the values, and the non-linearly weighted information can reflect the abrupt change pattern of the values. The values ​​of the core numerical benchmark, the linear change characteristics, and the non-linear abrupt change characteristics can be divided by the maximum value to eliminate dimensional differences. Finally, the three types of information after eliminating dimensional differences are concatenated to generate the initial test data feature information.

[0077] The test data processing method provided in this application enhances the linear and nonlinear abrupt change characteristics of numerical information at the edge of the test data value range, thereby achieving refined capture of the test data change patterns. This allows for in-depth analysis of the compliance risks and performance fluctuations behind the test data, and is used to determine the stable compliance status of values ​​in electrical safety tests, identify the abrupt change risks of component content in chemical testing, and monitor the stress change patterns of materials in mechanical performance tests. This fully explores the characteristic value of the test data and effectively improves the accuracy and reliability of test data processing.

[0078] Figure 5The flowchart illustrating the implementation of the test data processing method provided in Embodiment 5 of this application is shown. The difference between this method and Embodiment 1 is that step S103 specifically includes:

[0079] Step S501: Based on the preset test data feature low-pass filter matrix and the preset test data feature high-pass filter matrix, filter the multiple initial test data feature information to obtain multiple initial test data feature steady-state information and multiple initial test data feature fluctuation information.

[0080] In this embodiment, both the preset low-pass filter matrix and the preset high-pass filter matrix for test data features can be manually preset. The preset low-pass filter matrix can be a mean filter kernel, used to extract the steady-state components from the initial test data feature information, i.e., the baseline features of the test data, such as the average value in multiple repeated tests and the stable fluctuation level within the compliance threshold range. The preset high-pass filter matrix can be a Laplace kernel, used to extract the fluctuation components, i.e., the abnormal change features of the test data, such as peak values ​​exceeding the normal fluctuation range and sudden changes deviating from the baseline level. This can be achieved by convolving the test data feature low-pass filter matrix and the test data feature high-pass filter matrix with the initial test data feature information, and then using the results of the convolution operation as the steady-state information and fluctuation information of the initial test data features, respectively. For example, in energy efficiency test data, the steady-state information of the initial test data features can reflect the normal energy consumption level of the equipment, while the fluctuation information of the initial test data features can capture instantaneous peaks or abnormal decreases in energy consumption.

[0081] Step S502: Based on the preset test data feature low-pass filter matrix and the preset test data feature high-pass filter matrix, filter the steady-state information of multiple initial test data features and the fluctuation information of multiple initial test data features to obtain multiple target test data feature information.

[0082] In this embodiment, a second low-pass filter is performed on the initial test data feature steady-state information by convolving a preset test data feature low-pass filter matrix with the initial test data feature steady-state information. This further compresses subtle fluctuations in the steady-state features and generates a simpler baseline feature representation, such as simplifying the average energy consumption value from multiple tests into a core baseline value. A second high-pass filter is performed on the initial test data feature fluctuation information by convolving a preset test data feature high-pass filter matrix with the initial test data feature fluctuation information. This enhances the sharpness of the fluctuation features, such as highlighting the difference between the peak energy consumption spike and the baseline level. The steady-state information and fluctuation information after the second filtering are then concatenated to generate the target test data feature information.

[0083] The test data processing method provided in this application avoids mutual interference between steady-state features and fluctuation features, improves the distinguishability of test data features, performs secondary filtering on the two types of features after decomposition, filters out noise interference and redundant components in the initial test data feature information, so as to effectively meet the accurate processing needs of test data for multiple product categories and improve the accuracy and reliability of test data compliance judgment and risk warning.

[0084] Figure 6 The flowchart illustrating the implementation of the test data processing method provided in Embodiment Six of this application is shown. Its difference from Embodiment Five described above lies in:

[0085] The initial test data feature steady-state information includes initial test data security dimension steady-state information, initial test data performance dimension steady-state information, and test data compliance dimension steady-state information;

[0086] The initial test data fluctuation information includes initial test data security dimension fluctuation information, initial test data performance dimension fluctuation information, and test data compliance dimension fluctuation information;

[0087] Step S502 specifically includes:

[0088] Step S601: Based on the preset test data feature low-pass filter matrix, filter the initial test data security dimension steady-state information to obtain the target test data security dimension steady-state information.

[0089] In this embodiment, the preset test data feature low-pass filter matrix can be manually preset or can use a large-scale mean kernel. It can be achieved by convolving the test data feature low-pass filter matrix with the initial test data's safety dimension steady-state information, using the convolution result as the target test data's safety dimension steady-state information. This achieves secondary filtering of the initial test data's safety dimension steady-state information, weakening subtle fluctuations in the safety dimension's steady-state features and strengthening the baseline characteristics of core safety indicators such as electrical safety and chemical safety.

[0090] Step S602: Based on the preset high-pass filtering matrix of test data features, the security dimension fluctuation information of the initial test data is filtered to obtain the security dimension fluctuation information of the target test data.

[0091] In this embodiment, the preset test data feature high-pass filter matrix can be manually preset, and can be a sharpening kernel or a convolution kernel with high center weights and low neighborhood weights. It can be calculated by convolving the test data feature high-pass filter matrix with the initial test data safety dimension fluctuation information, and using the result of the convolution calculation as the target test data safety dimension fluctuation information to enhance the initial test data safety dimension fluctuation information, highlighting fluctuation details in the safety dimension that exceed the normal range, such as the instantaneous peak fluctuation of leakage current.

[0092] Step S603: Based on the preset test data feature low-pass filter matrix, filter the steady-state information of the initial test data performance dimension to obtain the steady-state information of the target test data performance dimension.

[0093] In this embodiment, the preset test data feature low-pass filter matrix can be manually preset. It can be achieved by convolving the preset test data feature low-pass filter matrix with the steady-state information of the initial test data performance dimension, and then using the result of the convolution calculation as the steady-state information of the target test data performance dimension to extract the macroscopic benchmark features of the performance dimension and weaken minor fluctuations in performance indicators, such as the stable level of the equipment's normal operating power.

[0094] Step S604: Based on the preset high-pass filter matrix for test data features, filter the performance dimension fluctuation information of the initial test data to obtain the performance dimension fluctuation information of the target test data.

[0095] In this embodiment, the preset test data feature high-pass filter matrix can be manually preset. It can be achieved by convolving the preset test data feature high-pass filter matrix with the initial test data performance dimension fluctuation information, and then using the result of the convolution calculation as the target test data performance dimension fluctuation information to enhance the contrast of performance dimension fluctuation details, such as the abnormal fluctuation amplitude of equipment operating noise.

[0096] Step S605: The target test data feature information is obtained by fusing the following information: the initial test data security dimension steady-state information, the initial test data performance dimension steady-state information, the target test data security dimension steady-state information, the test data compliance dimension steady-state information, the initial test data security dimension fluctuation information, the initial test data performance dimension fluctuation information, the target test data performance dimension fluctuation information, and the test data compliance dimension fluctuation information.

[0097] In this embodiment, the steady-state information of the initial test data's security dimension, performance dimension, and compliance dimension can be concatenated to generate steady-state structural feature information of the test data. Then, the fluctuation information of the initial test data's security dimension, performance dimension, and compliance dimension can be concatenated to generate fluctuation structural feature information of the test data. Finally, the steady-state information of the target test data's security dimension and the fluctuation information of the initial test data's security dimension are multiplied, and the result of the multiplication is used as the test data's security dimension fluctuation suppression information. This process utilizes the steady-state features of the security dimension. Suppressing interference from normal fluctuations, such as slight leakage current fluctuations during normal equipment operation, and enhancing risk-related fluctuation signals, such as abnormal fluctuations in leakage current approaching the threshold; performing a dot product calculation on the performance dimension fluctuation information of the target test data and the steady-state performance dimension information of the initial test data, and using the result of the dot product calculation as the performance dimension fluctuation suppression information of the test data; further fusing the steady-state structural feature information of the test data, the fluctuation structural feature information of the test data, the safety dimension fluctuation suppression information of the test data, and the performance dimension fluctuation suppression information of the test data to generate the target test data feature information, thereby reducing the risk of misjudgment caused by test environment interference or equipment errors.

[0098] The test data processing method provided in this application performs dimensional secondary filtering on the initial test data feature information to accurately extract the steady-state benchmark features and fluctuation anomaly features of each dimension. Through in-depth analysis and fusion processing of the three dimensions of safety, performance and compliance features, it fully captures the potential risk features in the test data, thereby improving the accuracy and effectiveness of test data processing for multiple product categories and providing reliable feature support for product compliance determination.

[0099] Figure 7 The flowchart illustrating the implementation of the test data processing method provided in Embodiment Seven of this application is shown. The difference between this method and Embodiment One is that step S104 specifically includes:

[0100] Step S701: Extract semantic information from the threshold information of the multiple product test indicators to obtain product safety threshold information, product performance threshold information, product category adaptation threshold information, and product market access threshold information.

[0101] In this embodiment, natural language processing technology can be used to segment, identify named entities, and classify the text and numerical association information in the threshold information of multiple product test indicators, thereby achieving semantic information extraction. For example, from "Under the EU REACH regulation, the threshold for heavy metal leaching in infant toys is ≤0.01mg / kg, the threshold for electrical toy withstand voltage is ≥250V, and it is suitable for children under 3 years old", the product safety threshold information is extracted as heavy metal leaching ≤0.01mg / kg and withstand voltage ≥250V, the product performance threshold information is the electrical withstand voltage compliance requirement, the product category suitability threshold information is the suitability standard for infant toys for children under 3 years old, and the product market access threshold information is the threshold corresponding to the EU REACH regulation. Then, the extracted information is organized by category to generate standardized threshold information for each dimension.

[0102] Step S702: Extract semantic association information from the threshold information of the multiple product test indicators to obtain product safety performance association threshold information, product category threshold adaptation association information, and product market access threshold matching information.

[0103] In this embodiment, association rule mining algorithms can be used to analyze product testing threshold data from multiple regions and categories globally, constructing probabilistic association networks for "safety-performance," "category-threshold," and "market-threshold." For example, the statistical results show that the association strength between "household appliances - insulation resistance threshold and power performance threshold" is 80%, the matching association probability of "infant and toddler toys - heavy metal leaching threshold" is 90%, and the matching accuracy of "EU market - phthalate content threshold" is 95%. This quantifies the association relationships between thresholds in each dimension, generating product safety performance association threshold information, product category threshold matching association information, and product market access threshold matching information.

[0104] Step S703: Generate multiple product test threshold feature information based on the product safety threshold information, product performance threshold information, product category adaptation threshold information, product market access threshold information, product safety performance related threshold information, product category threshold adaptation related information, and product market access threshold matching information.

[0105] In this embodiment, the product category adaptation threshold information is first one-hot encoded, and the numerical thresholds of the product safety threshold information and product performance threshold information are normalized, i.e., mapped to the [0,1] interval. The product market access threshold information is then converted into a Boolean vector. This Boolean vector is then concatenated with the correlation strength values ​​of the product safety performance correlation threshold information, the product category threshold adaptation correlation information, and the product market access threshold matching information to form multi-dimensional product test threshold feature information containing basic threshold features and correlation features. For example, for infant toys in the EU market, a vector containing features such as "Category - Infant Toys", "Market - EU", "Safety Threshold - Heavy Metals ≤ 0.01 mg / kg", "Correlation - Infant Toys - Heavy Metal Threshold Adaptation (0.9)", and "Correlation - EU Market - Heavy Metal Threshold Matching Rate (0.95)" is generated.

[0106] Step S704: Based on the preset test data processing model, generate multiple test data processing information according to the multiple target test data feature information and multiple product test threshold feature information.

[0107] In this embodiment, the preset test data processing model can be manually preset. It can involve first calculating the cosine similarity between the target test data feature information and the product test threshold feature information, then using this cosine similarity as the product compliance matching degree. This is combined with the correlation strength of product safety performance-related threshold information and product market access threshold matching information to weighted adjust the compliance matching degree, obtaining compliance judgment results for each dimension. Then, the weighted adjusted results are combined with the core features in the target test data feature information to output test data processing information including product compliance status, matching degrees for each threshold, risk warning level, and corresponding market access recommendations. This information is used to accurately characterize the compliance status and market adaptability of the product test data.

[0108] The test data processing method provided in this application avoids the one-sidedness of judging product compliance based on a single threshold dimension, thereby effectively reducing the probability of misjudgment of compliance in cross-border trade of multiple product categories and improving the accuracy, robustness and market adaptability of test data processing.

[0109] Corresponding to the method in the above embodiments, Figure 8 A structural block diagram of the test data processing apparatus provided in the embodiments of this application is shown. For ease of explanation, only the parts related to the embodiments of this application are shown. Figure 8 The test data processing device in the example can be the execution subject of the test data processing method provided in the aforementioned embodiment 1.

[0110] Reference Figure 8 The test data processing device includes:

[0111] The module 810 for acquiring test data to be processed and product test indicator threshold information is used to acquire multiple test data to be processed and multiple product test indicator threshold information.

[0112] The initial test data feature information generation module 820 is used to perform feature extraction processing on the multiple test data to be processed according to multiple preset test data feature extraction matrices to obtain multiple initial test data feature information.

[0113] The target test data feature information generation module 830 is used to perform hierarchical filtering and reconstruction processing based on the multiple initial test data feature information to obtain multiple target test data feature information.

[0114] The test data processing information generation module 840 is used to generate test data processing information based on a preset test data processing model, according to the multiple target test data feature information and multiple product test indicator threshold information.

[0115] The process by which each module in the test data processing device provided in this application implements its respective function can be specifically referred to the foregoing. Figure 1 The description of Embodiment 1 shown will not be repeated here.

[0116] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0117] It should be understood that, when used in this application specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or a collection thereof.

[0118] It should also be understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0119] As used in this application specification and the appended claims, the term "if" may be interpreted, depending on the context, as "when," "once," "in response to determination," or "in response to detection." Similarly, the phrase "if determined" or "if detected [the described condition or event]" may be interpreted, depending on the context, as meaning "once determined," "in response to determination," "once detected [the described condition or event]," or "in response to detection [the described condition or event]."

[0120] References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0121] The test data processing method provided in this application can be applied to terminal devices such as mobile phones, tablets, wearable devices, in-vehicle devices, augmented reality / virtual reality devices, laptops, super mobile personal computers, netbooks, and personal digital assistants. This application does not impose any restrictions on the specific type of terminal device.

[0122] Figure 9 This is a schematic diagram of the structure of a terminal device provided in an embodiment of this application. For example... Figure 9 As shown, the terminal device 9 of this embodiment includes: at least one processor 90 ( Figure 9 Only one is shown in the diagram), and a memory 91 stores a computer program 92 that can run on the processor 90. When the processor 90 executes the computer program 92, it implements the steps in the various test data processing method embodiments described above, for example... Figure 1 Steps S101 to S104 are shown. Alternatively, when the processor 90 executes the computer program 92, it implements the functions of each module / unit in the above-described device embodiments, for example... Figure 8 The functions of modules 810 to 840 are shown.

[0123] The terminal device 9 can be a desktop computer, laptop, handheld computer, cloud server, or other computing device. The terminal device may include, but is not limited to, a processor 90 and a memory 91. Those skilled in the art will understand that... Figure 9 This is merely an example of terminal device 9 and does not constitute a limitation on terminal device 9. It may include more or fewer components than shown, or combine certain components, or different components. For example, the terminal device may also include input transmission devices, network access devices, buses, etc.

[0124] The processor 90 may be a central processing unit, or it may be other general-purpose processors, digital signal processors, application-specific integrated circuits, off-the-shelf programmable gate arrays or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc.

[0125] In some embodiments, the memory 91 may be an internal storage unit of the terminal device 9, such as a hard disk or memory of the terminal device 9. The memory 91 may also be an external storage device of the terminal device 9, such as a plug-in hard disk, smart memory card, secure digital card, flash memory card, etc., equipped on the terminal device 9. Furthermore, the memory 91 may include both internal and external storage units of the terminal device 9. The memory 91 is used to store operating systems, applications, bootloaders, data, and other programs, such as the program code of the computer program. The memory 91 can also be used to temporarily store data that has been sent or will be sent.

[0126] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0127] This application also provides a terminal device, which includes at least one memory, at least one processor, and a computer program stored in the at least one memory and executable on the at least one processor. When the processor executes the computer program, it causes the terminal device to implement the steps in any of the above method embodiments.

[0128] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps described in the various method embodiments above.

[0129] This application provides a computer program product that, when run on a terminal device, enables the terminal device to implement the steps described in the various method embodiments above.

[0130] If the integrated module / unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory, a random access memory, an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc.

[0131] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0132] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0133] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0134] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A test data processing method, characterized in that, include: Acquire multiple test data sets to be processed and multiple product test indicator threshold information; Based on multiple preset test data feature extraction matrices, feature extraction processing is performed on the multiple test data to be processed to obtain multiple initial test data feature information; Based on the feature information of the multiple initial test data, hierarchical filtering and reconstruction processing are performed to obtain the feature information of multiple target test data. Based on a preset test data processing model, test data processing information is generated according to the multiple target test data feature information and multiple product test indicator threshold information.

2. The test data processing method as described in claim 1, characterized in that, The step of performing feature extraction processing on the multiple test data to be processed based on multiple preset test data feature extraction matrices to obtain multiple initial test data feature information specifically includes: Based on the preset test data block dimension information, the multiple test data to be processed are divided into blocks to obtain multiple test data block information. Extract the numerical information of the multiple test data blocks to obtain the numerical information of the multiple test data blocks; According to the preset test data block value selection order, the multiple test data block value information is selected and processed to obtain multiple selected test data block value information and multiple unselected test data block value information. Based on the numerical information of multiple selected test data blocks, the numerical information of multiple unselected test data blocks, and multiple preset test data feature extraction matrices, multiple initial test data feature information are obtained.

3. The test data processing method as described in claim 2, characterized in that, The step of obtaining multiple initial test data feature information based on multiple selected test data block numerical information, multiple unselected test data block numerical information, and multiple preset test data feature extraction matrices specifically includes: The number of test data blocks is obtained by statistically analyzing the numerical information of the selected test data blocks and the numerical information of the unselected test data blocks. Based on the numerical information of the multiple selected test data blocks and the numerical information of the multiple unselected test data blocks, the numerical deviation information of the multiple test data blocks is calculated. Based on the numerical deviation information of the multiple test data blocks, the local deviation information of the test data blocks is calculated. When the local deviation value of the test data block is greater than or equal to the preset local deviation threshold value of the test data block, test data value range information is generated based on the multiple selected test data block values ​​and the multiple unselected test data block values. The number of times the numerical information of each test data block was selected was counted to obtain the information on the number of times the numerical information of each test data block was selected. Determine whether the number of times the test data block value is selected is less than the number of test data blocks. If so, return to the step of selecting the multiple test data block values ​​according to the preset test data block value selection order to obtain multiple selected test data block values ​​and multiple unselected test data block values. If not, then based on multiple preset test data feature extraction matrices, feature extraction processing is performed on the test data value range information to obtain multiple initial test data feature information.

4. The test data processing method as described in claim 3, characterized in that, The test data value range information includes multiple core numerical information of the test data value range and multiple edge numerical information of the test data value range; Multiple preset test data feature extraction matrices include preset test data linear feature extraction weight matrices and preset test data nonlinear feature extraction weight matrices; The step of performing feature extraction processing on the value range information of the test data based on multiple preset test data feature extraction matrices to obtain multiple initial test data feature information specifically includes: Based on a preset linear feature extraction weight matrix for test data, the edge numerical information of the multiple test data value ranges is weighted to generate multiple linear weighted edge numerical information of the test data value ranges. Based on a preset weight matrix for extracting nonlinear features from test data, the numerical information at the edge of the multiple test data value ranges is weighted to generate nonlinear weighted information at the edge of the multiple test data value ranges. Based on the linear weighted information of the edge values ​​of the multiple test data ranges, the non-linear weighted information of the edge values ​​of the multiple test data ranges, and the core value information of the multiple test data ranges, multiple initial test data feature information are obtained.

5. The test data processing method as described in claim 1, characterized in that, The step of performing hierarchical filtering and reconstruction processing based on the multiple initial test data feature information to obtain multiple target test data feature information specifically includes: Based on the preset test data feature low-pass filter matrix and the preset test data feature high-pass filter matrix, the multiple initial test data feature information is filtered to obtain multiple initial test data feature steady-state information and multiple initial test data feature fluctuation information. Based on the preset test data feature low-pass filter matrix and the preset test data feature high-pass filter matrix, the steady-state information of multiple initial test data features and the fluctuation information of multiple initial test data features are filtered to obtain multiple target test data feature information.

6. The test data processing method as described in claim 5, characterized in that, The initial test data feature steady-state information includes initial test data security dimension steady-state information, initial test data performance dimension steady-state information, and test data compliance dimension steady-state information; The initial test data fluctuation information includes initial test data security dimension fluctuation information, initial test data performance dimension fluctuation information, and test data compliance dimension fluctuation information; The step of filtering multiple initial test data feature steady-state information and multiple initial test data feature fluctuation information according to a preset test data feature low-pass filter matrix and a preset test data feature high-pass filter matrix to obtain multiple target test data feature information specifically includes: Based on the preset test data feature low-pass filter matrix, the steady-state information of the security dimension of the initial test data is filtered to obtain the steady-state information of the security dimension of the target test data. Based on the preset high-pass filtering matrix of test data features, the security dimension fluctuation information of the initial test data is filtered to obtain the security dimension fluctuation information of the target test data. Based on the preset test data feature low-pass filter matrix, the steady-state information of the performance dimension of the initial test data is filtered to obtain the steady-state information of the performance dimension of the target test data. Based on the preset high-pass filtering matrix of test data features, the performance dimension fluctuation information of the initial test data is filtered to obtain the performance dimension fluctuation information of the target test data. The target test data feature information is obtained by fusing the following information: the steady-state information of the initial test data security dimension, the steady-state information of the initial test data performance dimension, the steady-state information of the target test data security dimension, the steady-state information of the test data compliance dimension, the fluctuation information of the initial test data security dimension, the fluctuation information of the initial test data performance dimension, the fluctuation information of the target test data performance dimension, and the fluctuation information of the test data compliance dimension.

7. The test data processing method as described in claim 1, characterized in that, The step of generating test data processing information based on a preset test data processing model, according to the multiple target test data feature information and multiple product test indicator threshold information, specifically includes: Semantic information is extracted from the threshold information of the multiple product test indicators to obtain product safety threshold information, product performance threshold information, product category adaptation threshold information, and product market access threshold information. Semantic association information is extracted from the threshold information of the multiple product test indicators to obtain product safety performance related threshold information, product category threshold adaptation related information, and product market access threshold matching information. Based on the product safety threshold information, product performance threshold information, product category adaptation threshold information, product market access threshold information, product safety performance related threshold information, product category threshold adaptation related information, and product market access threshold matching information, multiple product test threshold feature information are generated. Based on a preset test data processing model, multiple test data processing information is generated according to the multiple target test data feature information and multiple product test threshold feature information.

8. A test data processing device, characterized in that, include: The module for acquiring test data to be processed and product test indicator threshold information is used to acquire multiple test data to be processed and multiple product test indicator threshold information. The initial test data feature information generation module is used to perform feature extraction processing on the multiple test data to be processed according to multiple preset test data feature extraction matrices to obtain multiple initial test data feature information. The target test data feature information generation module is used to perform hierarchical filtering and reconstruction processing based on the multiple initial test data feature information to obtain multiple target test data feature information. The test data processing information generation module is used to generate test data processing information based on a preset test data processing model, according to the multiple target test data feature information and multiple product test indicator threshold information.

9. A terminal device, characterized in that, The terminal device includes a memory and a processor. The memory stores a computer program that can run on the processor. When the processor executes the computer program, it implements the steps of the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1 to 7.

Images