Product production line data collection and desensitization method and system based on multi-dimensional features

Abstract

The present invention discloses a product production line data collection and desensitization method and system based on multi-dimensional features, and relates to the field of electronic data processing technology. The method comprises the following steps: analyzing the completeness of data collection according to product production monitoring parameter data; analyzing the accuracy of data collection according to product production monitoring parameter data; desensitizing product production data to obtain product production desensitized data; analyzing the accuracy of desensitization processing according to product production desensitized data. The present invention analyzes the accuracy of desensitization processing according to product production desensitized data, repeatedly desensitizes product production data to obtain product production desensitized data, and analyzes the accuracy of desensitization processing according to product production desensitized data, until the accuracy of desensitization processing indicates that the product production data has been accurately desensitized, thereby achieving the effect of accurately desensitizing product production data, and solving the problem of overly inefficient data desensitization in the prior art.

Description

Technical Field

[0001] The present invention relates to the technical field of electronic data processing, and in particular to a method and system for collecting and desensitizing product production line data based on multi-dimensional features. Background Art

[0002] With the development of science and technology and the demand of the market, enterprises pay more and more attention to improving production efficiency and product quality. In order to achieve this goal, enterprises have begun to use advanced sensors, controllers and other equipment to realize the automation, digitization and intelligence of production lines. Multi-dimensional feature data acquisition refers to the collection of data from all links of the production line, including sensor data, equipment status data, product quality data, etc. The relevant data of the production line are obtained through data acquisition technology. These data contain rich information and are of great significance to the decision-making of the enterprise. For example, they are of great significance to the production optimization, fault diagnosis and quality management of the enterprise. However, these data often contain sensitive information, such as corporate business secrets, personal privacy, etc. If these sensitive information is leaked, it will cause serious losses to enterprises and individuals, so it is necessary to desensitize the data. Data desensitization technology refers to replacing, hiding and other processing of sensitive information in the original data, so that the processed data no longer contains sensitive information, thereby ensuring data security.

[0003] The existing data collection and desensitization system obtains data through the data collection module, and desensitizes the data through the desensitization module to realize the data collection and desensitization functions.

[0004] For example, the invention patent with announcement number: CN111460512B announces an automatic desensitizing data collection and distribution system and method, including: a data collection module, a desensitizing module, a user communication module, and a storage module, wherein: the data collection module is used to obtain target data; the desensitizing module is used to perform forced desensitization on the target data based on forced desensitization items; the storage module is used to store the target data after forced desensitization; the user communication module is used to receive user-side data requests; the storage module is also used to store the permissions corresponding to the user side; the desensitizing module is also used to perform permission desensitization on the target data after forced desensitization based on the permissions of the user side; the user communication module is also used to send the target data after permission desensitization to the user side.

[0005] For example, the invention patent announcement with announcement number: CN107301353B discloses a method for desensitizing streaming intensive data and a method for desensitizing data equipment, including: the equipment includes: field programmable gate array FPGA, ternary content addressable memory TCAM, double rate synchronous dynamic random access memory DDR SDRAM, multi-core network processor, and expandable connection backplane; the method includes: taking online collection, real-time filtering, and offline association of metadata streams in network communication applications as the goal, targeted classification, aggregation, association, and based on standardized data sharing level rules and policy configuration and customized whitelists, creating bounded data generalization application models and technologies as well as innovative streaming intensive data desensitization models and equipment.

[0006] However, in the process of implementing the technical solution of the invention in the embodiments of the present application, the present application found that the above technology has at least the following technical problems:

[0007] In the prior art, in the actual production line, there may be some hidden or difficult-to-measure features, which have a great impact on the quality and production efficiency of the product, but are often ignored, and these multi-dimensional features are not fully considered, resulting in insufficient comprehensiveness of the data. The accuracy of data collection directly affects the subsequent data analysis and processing results. In the actual production line, due to various reasons, such as sensor accuracy, environmental interference, etc., the collected data may contain errors, which will have an adverse effect on the subsequent data analysis and processing. The purpose of data desensitization is to protect the company's business secrets and user privacy. However, the existing data desensitization methods may have some loopholes, such as some desensitization algorithms are easily cracked, resulting in the leakage of sensitive information, or when performing data desensitization processing, some important features of the data may be lost, affecting the availability of the data, resulting in improper desensitization processing, and there is a problem of data desensitization being too inefficient. Summary of the invention

[0008] The embodiments of the present application solve the problem of inefficient data desensitization in the prior art by providing a product production line data collection and desensitization method and system based on multi-dimensional features, and achieve accurate and effective desensitization of product production data.

[0009] An embodiment of the present application provides a product production line data collection and desensitization method based on multi-dimensional features, including the following steps: obtaining product production data by monitoring the product production line and collecting data; obtaining product production monitoring parameter data from the product production line and product production data; analyzing the completeness of data collection based on the product production monitoring parameter data; when the completeness of data collection is represented by the complete collection of product production data, analyzing the accuracy of data collection based on the product production monitoring parameter data; when the accuracy of data collection is represented by the accuracy of the collected product production data, desensitizing the product production data to obtain product production desensitized data; and analyzing the accuracy of desensitization processing based on the product production desensitized data.

[0010] Furthermore, the specific process of obtaining product production data is: monitoring the product production line through a torque sensor to obtain the screw tightening processing torque; monitoring the product production line through a timer to obtain the screw tightening processing time; the product production data includes the screw tightening processing torque and the screw tightening processing time.

[0011] Furthermore, the specific analysis process of analyzing the data collection completeness according to the product production monitoring parameter data is as follows: obtaining data collection completeness parameter data according to the product production monitoring parameter data; preprocessing the data collection completeness parameter data; analyzing the preprocessed data collection completeness parameter data to obtain a data collection completeness assessment coefficient, wherein the data collection completeness assessment coefficient is used to represent the completeness of collecting product production data; and analyzing the data collection completeness according to the data collection completeness assessment coefficient.

[0012] Furthermore, the specific method for obtaining the data collection completeness evaluation coefficient is as follows: obtaining the number of data collection times and numbering the data collection times; obtaining the number of data acquisition times and numbering the data acquisition times; obtaining data collection completeness parameter data; constructing a data collection completeness evaluation coefficient calculation formula based on the data collection completeness parameter data; the specific data collection completeness evaluation coefficient calculation formula is:

[0013]

[0014]

[0015] In the formula, It is the data collection completeness evaluation coefficient of the product production data at the j0th data collection, j0=1,2,...,j, j0 is the number of data collection times, j is the total number of data collection times, It represents the type of data collected when the product production data is collected at the h0th data collection at the j0th data collection, h0=1,2,...,h, h0 represents the number of data acquisition times, h represents the total number of data acquisition times, It represents the memory space occupied by the data at the h0th data acquisition of the product production data at the j0th data acquisition. It is represented as the reference value of the collected data type at the j0th data collection. It is represented as the reference value of the memory space occupied by the data at the j0th data collection, D1 is represented as the weight ratio of the collected data type in the data collection completeness evaluation coefficient, and D2 is represented as the weight ratio of the memory space occupied by the data in the data collection completeness evaluation coefficient.

[0016] Furthermore, the specific analysis process for analyzing the data collection completeness according to the data collection completeness assessment coefficient is: obtain the data collection completeness assessment coefficient, obtain the first data collection completeness threshold and the second data collection completeness threshold, compare the data collection completeness assessment coefficient with the first data collection completeness threshold and the second data collection completeness threshold, when the data collection completeness assessment coefficient is greater than the first data collection completeness threshold and less than the second data collection completeness threshold, it indicates that the product production data is completely collected; when in other situations, it indicates that the product production data is not completely collected, and repeat the process of monitoring the product production line and collecting data to obtain product production data to analyze the data collection completeness according to the data collection completeness assessment coefficient, until the data collection completeness indicates that the product production data is completely collected.

[0017] Furthermore, the specific analysis process of analyzing the data collection accuracy according to the product production monitoring parameter data is as follows: obtaining data collection accuracy parameter data according to the product production monitoring parameter data; normalizing the data collection accuracy parameter data; obtaining a data collection accuracy assessment coefficient according to the normalized data collection accuracy parameter data, wherein the data collection accuracy assessment coefficient is used to reflect the accuracy of the collected product production data; analyzing the data collection accuracy according to the data collection accuracy assessment coefficient; the specific analysis process of analyzing the data collection accuracy according to the data collection accuracy assessment coefficient is as follows: obtaining the data collection accuracy assessment coefficient, obtaining a first data collection accuracy threshold and a second data collection accuracy threshold, when the data collection accuracy assessment coefficient is greater than the first data collection accuracy threshold and less than the second data collection accuracy threshold, it indicates that the content of the collected product production data is accurate; when in other situations, it indicates that the content of the collected product production data is inaccurate, and repeatedly obtaining product production data by monitoring the product production line and collecting data until the data collection accuracy is analyzed according to the product production monitoring parameter data, until the data collection accuracy indicates that the content of the collected product production data is accurate.

[0018] Furthermore, the specific analysis process of desensitizing the product production data to obtain the desensitized product production data is: the product production desensitized data is obtained by encrypting and masking the product production data; the data masking includes productive masking and random masking.

[0019] Furthermore, the specific analysis process of analyzing the accuracy of desensitizing processing based on the desensitized data of product production is: obtaining desensitizing processing accuracy parameter data based on the desensitized data of product production; obtaining a desensitizing processing accuracy evaluation index based on the desensitizing processing accuracy parameter data, and the desensitizing processing accuracy evaluation index is used to reflect the desensitization situation of the product production data; analyzing the accuracy of the desensitizing processing based on the desensitizing processing accuracy evaluation index.

[0020] Furthermore, the specific analysis process of analyzing the accuracy of desensitizing processing according to the desensitizing processing accuracy evaluation index is as follows: obtaining the desensitizing processing accuracy evaluation index, obtaining the first desensitizing processing accuracy threshold and the second desensitizing processing accuracy threshold, when the desensitizing processing accuracy evaluation index is greater than the first desensitizing processing accuracy threshold and less than the second desensitizing processing accuracy threshold, it indicates that the product production data has been accurately desensitized; when in other situations, it indicates that the desensitizing processing of the product production data is inaccurate, and the product production data is repeatedly desensitized to obtain the product production desensitized data until the desensitization processing accuracy is analyzed according to the product production desensitized data, until the desensitization processing accuracy indicates that the product production data has been accurately desensitized.

[0021] An embodiment of the present application provides a product production line data collection and desensitization system based on multi-dimensional features, including: a monitoring module, a parameter acquisition module, a first analysis module, a second analysis module, a desensitization module and a third analysis module; the monitoring module: is used to acquire product production data by monitoring the product production line and performing data collection; the parameter acquisition module: is used to acquire data on the product production line and product production data to obtain product production monitoring parameter data; the first analysis module: is used to analyze the completeness of data collection according to the product production monitoring parameter data; the second analysis module: is used to analyze the accuracy of data collection according to the product production monitoring parameter data when the completeness of data collection is represented by the complete collection of product production data; the desensitization module: is used to desensitize the product production data to obtain product production desensitized data when the accuracy of data collection is represented by the accuracy of the content of the collected product production data; the third analysis module: is used to analyze the accuracy of desensitization processing according to the product production desensitized data.

[0022] One or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages:

[0023] 1. Obtain product production monitoring parameter data through data acquisition of product production lines and product production data, analyze data collection completeness according to product production monitoring parameter data, when data collection completeness indicates complete collection of product production data, analyze data collection accuracy according to product production monitoring parameter data, when data collection accuracy indicates that the content of collected product production data is accurate, desensitize product production data to obtain product production desensitized data, thereby analyzing desensitization processing accuracy according to product production desensitized data, repeatedly desensitize product production data to obtain product production desensitized data, and analyze desensitization processing accuracy according to product production desensitized data, until desensitization processing accuracy indicates that product production data has been accurately desensitized, thereby realizing accurate desensitization of product production data, and effectively solving the problem of inefficient data desensitization in the prior art.

[0024] 2. Obtain data collection completeness parameter data based on product production monitoring parameter data, pre-process the data collection completeness parameter data, analyze the pre-processed data collection completeness parameter data to obtain the data collection completeness assessment coefficient, and then analyze the data collection completeness according to the data collection completeness assessment coefficient, repeat the process of monitoring the product production line and collecting data to obtain product production data, and analyzing the data collection completeness according to the data collection completeness assessment coefficient, until the data collection completeness indicates that the product production data is completely collected, thereby achieving complete collection of product production data.

[0025] 3. Obtain data collection accuracy parameter data based on product production monitoring parameter data, normalize the data collection accuracy parameter data, and obtain the data collection accuracy assessment coefficient based on the normalized data collection accuracy parameter data. Then, analyze the data collection accuracy based on the data collection accuracy assessment coefficient, and repeat the process of monitoring the product production line and collecting data to obtain product production data, and analyzing the data collection accuracy based on the product production monitoring parameter data, until the data collection accuracy indicates that the content of the collected product production data is accurate, thereby ensuring the accuracy of the content of the collected product production data. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 A flow chart of a product production line data collection and desensitization method based on multi-dimensional features provided in an embodiment of the present application;

[0027] Figure 2 A curve diagram of the data collection completeness evaluation coefficient provided in an embodiment of the present application;

[0028] Figure 3 A schematic diagram of the structure of a product production line data collection and desensitization system based on multi-dimensional features provided in an embodiment of the present application. DETAILED DESCRIPTION

[0029] The embodiments of the present application solve the problem of inefficient data desensitization in the prior art by providing a product production line data collection and desensitization method and system based on multi-dimensional features. The method acquires data from the product production line and product production data to obtain product production monitoring parameter data, analyzes the completeness of data collection according to the product production monitoring parameter data, and when the completeness of data collection indicates that the product production data is completely collected, analyzes the accuracy of data collection according to the product production monitoring parameter data. When the accuracy of data collection indicates that the content of the collected product production data is accurate, the product production data is desensitized to obtain product production desensitized data, and then analyzes the accuracy of desensitization processing according to the product production desensitized data. The product production data is desensitized repeatedly to obtain product production desensitized data until the accuracy of desensitization processing is analyzed according to the product production desensitized data, and the accurate desensitization processing indicates that the product production data has been accurately desensitized, thereby achieving accurate and effective desensitization of product production data.

[0030] The technical solution in the embodiment of the present application is to solve the above-mentioned problem that data desensitization is too inefficient. The overall idea is as follows:

[0031] Acquire product production data by monitoring product production lines and collecting data; acquire product production monitoring parameter data by acquiring data from product production lines and product production data; obtain data collection completeness parameter data based on product production monitoring parameter data; preprocess data collection completeness parameter data; analyze preprocessed data collection completeness parameter data to obtain data collection completeness evaluation coefficient; analyze data collection completeness based on data collection completeness evaluation coefficient; when data collection completeness indicates complete collection of product production data, obtain data collection accuracy parameter data based on product production monitoring parameter data; and evaluate data collection accuracy. The parameter data is normalized; the data collection accuracy evaluation coefficient is obtained based on the normalized data collection accuracy parameter data analysis; the data collection accuracy is analyzed based on the data collection accuracy evaluation coefficient; when the data collection accuracy is represented by the accuracy of the collected product production data, the product production data is desensitized to obtain product production desensitized data; desensitized processing accuracy parameter data is obtained based on the product production desensitized data; desensitized processing accuracy evaluation index is obtained based on the desensitized processing accuracy parameter data; the desensitized processing accuracy is analyzed based on the desensitized processing accuracy evaluation index, thereby achieving the effect of accurately and effectively desensitizing the product production data.

[0032] In order to better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0033] like Figure 1 As shown, it is a flow chart of the product production line data collection and desensitization method based on multi-dimensional features provided in an embodiment of the present application, and the method includes the following steps: obtaining product production data by monitoring the product production line and collecting data; obtaining product production monitoring parameter data by acquiring data from the product production line and product production data; analyzing the completeness of data collection according to the product production monitoring parameter data; when the completeness of data collection is represented by the complete collection of product production data, analyzing the accuracy of data collection according to the product production monitoring parameter data; when the accuracy of data collection is represented by the accuracy of the content of the collected product production data, desensitizing the product production data to obtain product production desensitized data; analyzing the accuracy of desensitization processing according to the product production desensitized data.

[0034] In this embodiment, the product production line is a product production line that requires screw tightening processing, such as a mechanical product production line.

[0035] Furthermore, the specific process of obtaining product production data is: monitoring the product production line through a torque sensor to obtain the screw tightening processing torque; monitoring the product production line through a timer to obtain the screw tightening processing time; the product production data includes the screw tightening processing torque and the screw tightening processing time.

[0036] In this embodiment, the product production data includes not only the screw tightening processing torque and the screw tightening processing time, but also production-related data such as temperature and humidity. The temperature and humidity data can be obtained through temperature sensors and humidity sensors.

[0037] Furthermore, the specific analysis process of analyzing the data collection completeness based on the product production monitoring parameter data is as follows: obtain data collection completeness parameter data based on the product production monitoring parameter data; preprocess the data collection completeness parameter data; analyze the preprocessed data collection completeness parameter data to obtain a data collection completeness assessment coefficient, which is used to represent the completeness of collecting product production data; and analyze the data collection completeness based on the data collection completeness assessment coefficient.

[0038] In this embodiment, the data collection completeness parameter data is preprocessed, including data cleaning of the data collection completeness parameter data, identifying and processing missing values ​​in the data collection completeness parameter data, and filling missing values ​​(for example, using the mean, median or nearest neighbor value), or deleting records containing missing values; scaling the data collection completeness parameter data to a suitable range. By preprocessing the data collection completeness parameter data, a more accurate data collection completeness evaluation coefficient can be obtained based on the data collection completeness parameter data, which is conducive to a more detailed analysis of the data collection completeness.

[0039] Furthermore, the data collection completeness assessment coefficient can be obtained by the following methods, including but not limited to: conducting a detailed review of the product production data by using data review tools or programming languages ​​(such as Python's Pandas library) to determine the number and location of missing values; counting the total number of all missing values ​​in the product production data. If some columns (or rows) in the product production data have missing values, it is also necessary to calculate the number of missing values ​​in each column (or row) separately; determining the total data volume of the product production data, including the total number of all columns (or rows) and the total number of locations with data in each column (or row); using the following formula to calculate the proportion of missing values ​​in each column (or row): missing value proportion = (total number of missing values / total data volume) * 100%, this proportion can intuitively show the degree of missing data in each column (or row); summarizing the missing value proportions of each column (or row), and evaluating the degree of missing values ​​of the entire data set by calculating the average value of the missing value proportions of all columns (or rows), that is, obtaining the data collection completeness assessment coefficient. The data collection completeness evaluation coefficient can also be obtained by calculation. The specific method for obtaining the data collection completeness evaluation coefficient is: obtain the number of data collection times and number the data collection times; obtain the number of data acquisition times and number the data acquisition times; data collection completeness parameter data; construct a data collection completeness evaluation coefficient calculation formula based on the data collection completeness parameter data; the specific data collection completeness evaluation coefficient calculation formula is:

[0040]

[0041]

[0042] In the formula, It is the data collection completeness evaluation coefficient of the product production data at the j0th data collection, j0=1,2,...,j, j0 is the number of data collection times, j is the total number of data collection times, It represents the type of data collected when the product production data is collected at the h0th data collection at the j0th data collection, h0=1,2,...,h, h0 represents the number of data acquisition times, h represents the total number of data acquisition times, It represents the memory space occupied by the data at the h0th data acquisition of the product production data at the j0th data acquisition. It is represented as the reference value of the collected data type at the j0th data collection. It is represented as the reference value of the memory space occupied by the data at the j0th data collection, D1 is represented as the weight ratio of the collected data type in the data collection completeness evaluation coefficient, and D2 is represented as the weight ratio of the memory space occupied by the data in the data collection completeness evaluation coefficient.

[0043] In this embodiment, the opinions and suggestions of experts in the field are collected. Through questionnaire surveys, expert interviews, etc., the experts are asked to score the importance of indicators that affect the data collection completeness evaluation coefficient. The weight ratio of each indicator is obtained based on the experts' scores, that is, the specific values ​​of D1 and D2 are obtained. The value ranges of D1 and D2 are both (0, 1), and the sum of D1 and D2 is equal to 1.

[0044] The types of collected data refer to all types of data included in the product production data, the reference value of the types of collected data refers to all types of data included in the product production data when it is accurately collected, the memory space occupied by the data refers to the size of the memory space occupied by the product production data when it is stored, and the reference value of the memory space occupied by the data refers to the size of the memory space occupied by the product production data when it is accurately collected. Access the product production data through API interfaces and database queries, and obtain CSV, JSON or readme files of the product production data. These files usually contain basic information such as the type, structure, and size of the data set, thereby obtaining the types of collected data. Use the tools provided by the operating system or programming language to count the size of the product production data, that is, to obtain the memory space occupied by the data. Under the condition of accurate collection, the accurate product production data is analyzed by the above method to obtain the reference value of the types of collected data and the reference value of the memory space occupied by the data.

[0045] Under normal circumstances, the more types of collected data, the larger the memory space occupied by the corresponding data, and the smaller the types of collected data, the smaller the memory space occupied by the corresponding data. In the algorithm of this embodiment, by using the hyperbolic cosine function, the dependent variable gradually increases when the independent variable increases, reducing the influence of extreme values ​​on the data collection completeness evaluation coefficient, and by applying the hyperbolic cosine function to the data collection completeness parameter data, a certain smoothing effect is played in the smoothing process.

[0046] like Figure 2 As shown, it is a curve diagram of the data collection completeness evaluation coefficient provided by the embodiment of the present application, with the type of collected data as the independent variable and the data collection completeness evaluation coefficient as the dependent variable. Through the increase and decrease analysis of curves 1, 2 and 3, it can be seen that when the type of collected data is larger, the data collection completeness evaluation coefficient is larger; when the type of collected data is smaller, the data collection completeness evaluation coefficient is smaller. The statistical table of data collection completeness evaluation coefficient is shown in Table 1:

[0047] Table 1 Statistics of data collection completeness evaluation coefficient

[0048]

[0049]

[0050] From Table 1, we can see that in Curve 1, when other parameters are the same, the larger the type of collected data, that is, when the types of collected data are 16, 18 and 20, the data collection completeness evaluation coefficients are 1.04, 1.09 and 1.13 respectively, and the data collection completeness evaluation coefficient gradually increases; similarly, when the types of collected data are smaller, the data collection completeness evaluation coefficient is smaller. The above changes are also true for Curve 2 and Curve 3.

[0051] For curve 1, curve 2 and curve 3, when the types of collected data of curve 1, curve 2 and curve 3 are 181, when the memory space occupied by the data of curve 1, curve 2 and curve 3 are 14MB, 15MB and 16MB respectively, and other parameters are consistent, the data collection completeness evaluation coefficients of curve 1, curve 2 and curve 3 are 1.09, 1.11 and 1.13 respectively. It can be seen that when the types of collected data and other parameters are consistent, the larger the memory space occupied by the data, the larger the data collection completeness evaluation coefficient.

[0052] Furthermore, the specific analysis process for analyzing the data collection completeness according to the data collection completeness assessment coefficient is: obtain the data collection completeness assessment coefficient, obtain the first data collection completeness threshold and the second data collection completeness threshold, compare the data collection completeness assessment coefficient with the first data collection completeness threshold and the second data collection completeness threshold, when the data collection completeness assessment coefficient is greater than the first data collection completeness threshold and less than the second data collection completeness threshold, it means that the product production data is completely collected; when it is in other situations, it means that the product production data is not completely collected, and repeat the process of monitoring the product production line and collecting data to obtain product production data until the data collection completeness is analyzed according to the data collection completeness assessment coefficient, until the data collection completeness indicates that the product production data is completely collected.

[0053] In this embodiment, the first threshold value for complete data collection and the second threshold value for complete data collection are obtained based on the experience and knowledge of experts in related fields. The first threshold value for complete data collection is The value range is [1.08, 1.09], and the second threshold for complete data collection is The value range of is [1.17, 1.18]. The data collection completeness evaluation coefficient is compared with the first data collection completeness threshold and the second data collection completeness threshold. When the data collection completeness evaluation coefficient is greater than the first data collection completeness threshold and less than the second data collection completeness threshold, that is, when When it is, it means to collect product production data completely; when it is in other situations, that is, when or , it means that the product production data has not been completely collected. Repeat the process of monitoring the product production line and collecting data to obtain product production data until the data collection completeness situation is analyzed according to the data collection completeness assessment coefficient, until the data collection completeness situation indicates that the product production data has been completely collected.

[0054] In actual production lines, there may be some hidden or difficult-to-measure data that have a great impact on product quality and production efficiency, but are often overlooked. Therefore, it is necessary to fully consider these multi-dimensional characteristics and avoid incomplete data collection by ensuring the comprehensiveness of the data. Here, the comprehensiveness of data collection is analyzed through the data collection completeness evaluation coefficient to avoid more mistakes caused by subsequent analysis of incomplete data.

[0055] Furthermore, the specific analysis process of analyzing the data collection accuracy according to the product production monitoring parameter data is as follows: obtaining data collection accuracy parameter data according to the product production monitoring parameter data; normalizing the data collection accuracy parameter data; obtaining a data collection accuracy assessment coefficient according to the normalized data collection accuracy parameter data, the data collection accuracy assessment coefficient is used to reflect the accuracy of the collected product production data; analyzing the data collection accuracy according to the data collection accuracy assessment coefficient; the specific analysis process of analyzing the data collection accuracy according to the data collection accuracy assessment coefficient is as follows: obtaining the data collection accuracy assessment coefficient, obtaining a first data collection accuracy threshold and a second data collection accuracy threshold, when the data collection accuracy assessment coefficient is greater than the first data collection accuracy threshold and less than the second data collection accuracy threshold, it indicates that the content of the collected product production data is accurate; when in other situations, it indicates that the content of the collected product production data is inaccurate, and repeatedly obtaining product production data by monitoring the product production line and collecting data until the data collection accuracy is analyzed according to the product production monitoring parameter data, until the data collection accuracy indicates that the content of the collected product production data is accurate.

[0056] In this embodiment, the data collection accuracy parameter data is normalized, including scaling the range of the data collection accuracy parameter data to a fixed interval, such as [0, 1] or [-1, 1], through maximum-minimum normalization. Normalization can prevent certain data with large values ​​from having an excessive impact on the data collection accuracy evaluation coefficient. Normalization scales the data to the same scale so that the data can be directly compared without considering their physical units or dimensions. And the accuracy of the obtained data collection accuracy evaluation coefficient is higher, which is conducive to a more detailed and accurate analysis of the data collection accuracy.

[0057] The data collection accuracy assessment coefficient can be obtained through a variety of methods, including: calculating the mean square error (MSE) of each data in the product production data. The smaller the MSE value, the closer each data in the product production data is to the actual value corresponding to the data, and the more accurate the data; the mean absolute error (MAE). MAE gives the average size of the error between each data in the product production data and the actual value corresponding to the data. The smaller the value, the more accurate the data; combining the above indicators for comprehensive evaluation to obtain the data collection accuracy assessment coefficient. The data collection accuracy assessment coefficient can also be obtained through the following calculation. The specific calculation formula for the data collection accuracy assessment coefficient is:

[0058]

[0059]

[0060] In the formula, It is expressed as the data collection accuracy evaluation coefficient of the product production data at the j0th data collection, It is represented by the screw tightening processing time when the h0th data of the product production data is acquired during the j0th data collection. It is represented by the screw tightening torque when the h0th data of the product production data is acquired during the j0th data collection. It is expressed as the reference value of the screw tightening processing time at the j0th data collection, It is represented as the reference value of the screw tightening processing torque at the j0th data collection, C1 is represented as the weight ratio of the collected data type in the data collection accuracy evaluation coefficient, and C2 is represented as the weight ratio of the memory space occupied by the data in the data collection accuracy evaluation coefficient.

[0061] The principal component analysis (PCA) method can be used to obtain the weight ratio of the types of collected data and the memory space occupied by the data. The specific process is: collect indicator data that affect the evaluation coefficient of the accuracy of data collection, including the types of collected data and the memory space occupied by the data, standardize or normalize the collected data to ensure that the data of each indicator are in the same dimension, use the Pearson correlation matrix to obtain the correlation between the indicators, perform eigendecomposition on the Pearson correlation matrix to obtain eigenvalues ​​and corresponding eigenvectors, the eigenvalues ​​represent the variance explained by each eigenvector, and according to the size of the eigenvalues, select the first few eigenvectors that can best explain the data variance. These eigenvectors are called principal components, and the coefficients of the principal components are the weights of each indicator. The weights reflect the importance of the indicator in the overall analysis. The principal components are the types of collected data and the memory space occupied by the data, and then the weights of the types of collected data and the memory space occupied by the data are obtained, that is, to obtain the specific values ​​of C1 and C2. The value ranges of C1 and C2 are both (0, 1), and the sum of C1 and C2 is equal to 1.

[0062] The screw tightening processing time refers to the time consumed to complete the screw tightening processing task in the product production line represented by the product production data. The screw tightening processing time reference value refers to the time consumed to complete the screw tightening processing task in the product production line represented by the product production data when the product production data is accurately collected. The screw tightening processing torque refers to the processing torque used to complete the screw tightening processing task in the product production line represented by the product production data. The screw tightening processing torque reference value refers to the processing torque used to complete the screw tightening processing task in the product production line represented by the product production data when the product production data is accurately collected. The screw tightening processing time and screw tightening processing torque are obtained by reading and analyzing the product production data through database query. When the product production data is accurately collected, the product production data is read and analyzed to obtain the screw tightening processing time reference value and screw tightening processing torque reference value.

[0063] Under normal circumstances, there is a certain correlation between the time required for screw tightening and the torque used. When tightening screws of the same specifications and processing requirements, the greater the processing torque used, the easier it is to perform screw tightening, and the shorter the time required to complete the screw tightening. Therefore, within a certain range, the smaller the screw tightening torque, the longer the screw tightening time, and the larger the screw tightening torque, the shorter the screw tightening time. In the algorithm of this embodiment, by using the natural logarithm function to perform logarithmic transformation on abnormally large or small values, their influence on the overall mean or median is reduced, and the influence of extreme values ​​on the data collection accuracy evaluation coefficient is reduced, so that the accuracy of the data collection accuracy evaluation coefficient is higher.

[0064] Consult with experts in the field of data desensitization stations, set a preliminary threshold based on historical data, conduct preliminary tests to see if the results meet expectations, and continuously adjust the threshold until the optimal value is found, that is, to obtain the first threshold for accurate data collection and the second threshold for accurate data collection. The first threshold for accurate data collection is β1, and the value range of β1 is [0.685, 0.697]. The second threshold for accurate data collection is β2, and the value range of β2 is [0.705, 0.716]. When the data collection accuracy evaluation coefficient is greater than the first threshold for accurate data collection and less than the second threshold for accurate data collection, that is, when When the product production data collected is accurate, it means that the content is accurate; when it is in other situations, that is, when or , it indicates that the collected product production data is inaccurate. Repeat the process of monitoring the product production line and collecting data to obtain product production data until the data is collected accurately based on the product production monitoring parameter data analysis. The data collection is accurate, indicating that the collected product production data is accurate.

[0065] In actual production lines, due to various reasons, such as sensor accuracy, environmental interference, etc., the collected data may contain errors, which directly affects the subsequent data analysis and processing results. If subsequent analysis is performed, more erroneous analysis will be caused. Therefore, here we analyze the data collection accuracy evaluation coefficient to clarify the accuracy of data collection.

[0066] Furthermore, the specific analysis process of desensitizing the product production data to obtain the desensitized product production data is as follows: the product production data is encrypted and masked to obtain the desensitized product production data; the data masking includes productive masking and random masking.

[0067] In this embodiment, the specific process of encryption processing is: encrypting the sensitive data in the product production data, and only users with decryption keys can access the original data. The specific process of generative mask processing is: generating data similar to the original data but not containing sensitive information. The specific process of random mask processing is: replacing the sensitive data in the product production data with randomly generated data, for example, replacing the numerical data with different random numerical values.

[0068] In addition to the above-mentioned processing, other methods can be used for desensitization, including partial hiding, complete hiding and pseudo-anonymous processing. The specific process of partial hiding is: hiding part of the data, such as only showing the first few digits of the ID number. The specific process of complete hiding is: completely hiding the sensitive data and leaving only the non-sensitive part, such as hiding all information in front of the credit card number and only showing the last four digits. The specific process of pseudo-anonymous processing is: converting sensitive data into a non-sensitive alternative identifier (pseudo-anonymous), such as replacing the temperature with an anonymous identifier.

[0069] Furthermore, the specific analysis process of analyzing the accuracy of desensitizing processing based on the desensitized data of product production is: obtaining desensitizing processing accuracy parameter data based on the desensitized data of product production; obtaining a desensitizing processing accuracy evaluation index based on the desensitizing processing accuracy parameter data, and the desensitizing processing accuracy evaluation index is used to reflect the desensitization situation of product production data; analyzing the accuracy of desensitizing processing based on the desensitizing processing accuracy evaluation index.

[0070] In this embodiment, the desensitization processing accuracy evaluation index can be obtained by the following method, and the specific process is: before starting the desensitization process, determine the minimum and maximum values ​​of each type of data in the product production data, and these values ​​represent the total range of the data; use the selected desensitization algorithm to process the product production data to generate desensitized product production desensitized data; perform a similar process on the desensitized product production desensitized data to determine the minimum and maximum values ​​of each type of data in the product production desensitized data; compare the minimum and maximum values ​​of the product production desensitized data with the minimum and maximum values ​​of the product production data. If the minimum and maximum values ​​of the desensitized product production desensitized data are the same or very close to the minimum and maximum values ​​of the product production data, the desensitization process successfully maintains the overall range of the data. If the minimum and maximum values ​​of the desensitized product production desensitized data are significantly different from the product production data, the desensitization algorithm has not processed the data correctly; compare the minimum and maximum values ​​of each type of data after desensitization with the minimum and maximum values ​​before desensitization and perform a comprehensive analysis to obtain the desensitization processing accuracy evaluation index. In addition to the above method, the desensitization processing accuracy evaluation index can also be obtained by calculation. The specific calculation formula for the desensitization processing accuracy evaluation index is:

[0071]

[0072] In the formula, It is represented as the evaluation index of the accuracy of desensitization processing of product production data at the j0th data collection. It is represented by the maximum number of encryption times when the product production data is acquired at the h0th data collection time during the j0th data collection. It is expressed as the minimum number of encryption times when the product production data is acquired at the h0th data collection time during the j0th data collection. It is represented by the number of desensitized data at the time of the h0th data acquisition of the product production data at the j0th data collection. It is expressed as the maximum number of standard encryption times at the j0th data collection. It is expressed as the minimum number of standard encryption times at the j0th data collection. It is expressed as the standard value of the number of desensitized data at the j0th data collection, Y1 is expressed as the weight ratio of the maximum number of encryption times in the desensitization processing accuracy assessment index, Y2 is expressed as the weight ratio of the minimum number of encryption times in the desensitization processing accuracy assessment index, Y3 is expressed as the weight ratio of the number of desensitized data in the desensitization processing accuracy assessment index, and e is expressed as a natural constant.

[0073] Establish a hierarchical structure, including: target layer: accurately obtain the desensitization accuracy evaluation index; criterion layer: the maximum number of encryptions, the minimum number of encryptions and the number of desensitized data; scheme layer: desensitization accuracy parameter data. Perform pairwise comparisons: for each factor in each layer, compare them two by two and give relative importance. The comparison is usually based on a scale of 1 to 9, where 1 means that the two factors are equally important and 9 means that one factor is much more important than the other. Construct a pairwise comparison matrix: organize the results of the pairwise comparison into a matrix form. The elements on the diagonal of the matrix are always 1 (because the factors are compared with themselves), while the elements A_ab on the non-diagonal represent the importance of factor a relative to factor b. Calculate relative weights: for each factor in the criterion layer, by calculating the maximum eigenvalue and corresponding eigenvector of the pairwise comparison matrix, which represents the relative weight of the factor, its relative weight relative to the target layer is obtained. That is, to obtain the specific values ​​of Y1, Y2 and Y3, the value ranges of Y1, Y2 and Y3 are all (0, 1), and the sum of Y1, Y2 and Y3 is equal to 1.

[0074] The maximum number of encryption times refers to the maximum number of encryption times of the data after the product production data is desensitized, the maximum number of standard encryption times refers to the maximum number of encryption times of the data after the product production data is accurately desensitized, the minimum number of encryption times refers to the minimum number of encryption times of the data after the product production data is desensitized, the minimum number of standard encryption times refers to the minimum number of encryption times of the data after the product production data is accurately desensitized, the number of desensitized data refers to the number of data desensitized in the product production data, and the standard value of the number of desensitized data refers to the number of data desensitized after the product production data is accurately desensitized. The encryption times are obtained by reading and analyzing the data of the product production desensitized data, and the encryption times are sorted from large to small, and the encryption times ranked first and last are taken as the maximum number of encryption times and the minimum number of encryption times. The amount of data desensitized in the product production desensitized data is obtained by statistical analysis, that is, the number of desensitized data. The above analysis process is performed on the product production desensitized data obtained after the product production data is accurately desensitized to obtain the maximum value of the standard encryption times, the minimum value of the standard encryption times and the standard value of the number of desensitized data.

[0075] For data that needs to be desensitized, the higher the importance of the data, the greater the number of encryptions of the data, and the lower the importance of the data, the smaller the number of encryptions of the data. The maximum number of encryptions and the minimum number of encryptions together reflect the importance of the data. In the algorithm of this embodiment, by combining the exponential function and the inverse tangent function, the result is compressed to a smaller range to achieve the effect of smoothing the data, thereby improving the reliability of the desensitization processing accuracy evaluation index.

[0076] The purpose of data desensitization is to protect the business secrets of enterprises and the privacy of users. However, existing data desensitization methods may have some loopholes. For example, some desensitization algorithms are easy to be cracked, resulting in the leakage of sensitive information. When performing data desensitization, some important features of the data may be lost, affecting the availability of the data. For example, after desensitizing a phone number, it may no longer be used in scenarios such as telemarketing. Data desensitization requires customized processing for different data types and scenarios, which increases the complexity of desensitization. For some non-professionals, it is difficult to master these desensitization methods, resulting in improper desensitization. Due to various factors, the data desensitization effect is not good. If the data with poor desensitization effect is shared, etc., it may lead to the leakage of important information of the product production line. Therefore, it is necessary to analyze the effect of data desensitization. Here, the desensitization accuracy evaluation index is used to analyze the effect of data desensitization.

[0077] Furthermore, the specific analysis process of analyzing the accuracy of desensitizing processing according to the desensitizing processing accuracy evaluation index is as follows: obtaining the desensitizing processing accuracy evaluation index, obtaining the first desensitizing processing accuracy threshold and the second desensitizing processing accuracy threshold, when the desensitizing processing accuracy evaluation index is greater than the first desensitizing processing accuracy threshold and less than the second desensitizing processing accuracy threshold, it indicates that the product production data has been accurately desensitized; when in other situations, it indicates that the desensitizing processing of the product production data is inaccurate, and the product production data is repeatedly desensitized to obtain the product production desensitized data until the desensitization processing accuracy is analyzed according to the product production desensitized data, until the desensitization processing accuracy indicates that the product production data has been accurately desensitized.

[0078] In this embodiment, historical experience is obtained and combined with relevant experience in desensitization processing accuracy evaluation to obtain the first desensitization processing accuracy threshold and the second desensitization processing accuracy threshold. The first desensitization processing accuracy threshold is ρ1, and the value range of ρ1 is [1.15, 1.20]. The second desensitization processing accuracy threshold is ρ2, and the value range of ρ2 is [1.21, 1.25]. When the desensitization processing accuracy evaluation index is greater than the first desensitization processing accuracy threshold and less than the second desensitization processing accuracy threshold, that is, when When , it means that the product production data has been accurately desensitized; when it is in other cases, that is, when or When the desensitization processing of the product production data is inaccurate, the desensitization processing of the product production data is repeated to obtain the product production desensitized data until the accuracy of the desensitization processing is analyzed based on the product production desensitized data, until the accurate desensitization processing situation indicates that the product production data has been accurately desensitized.

[0079] like Figure 3As shown, it is a structural schematic diagram of the product production line data collection and desensitization system based on multi-dimensional features provided in an embodiment of the present application. The product production line data collection and desensitization system based on multi-dimensional features provided in an embodiment of the present application includes: a monitoring module, a parameter acquisition module, a first analysis module, a second analysis module, a desensitization module and a third analysis module; monitoring module: used to acquire product production data by monitoring the product production line and performing data collection; parameter acquisition module: used to acquire data on the product production line and product production data to obtain product production monitoring parameter data; first analysis module: used to analyze the completeness of data collection according to the product production monitoring parameter data; second analysis module: used to analyze the accuracy of data collection according to the product production monitoring parameter data when the completeness of data collection is represented by the complete collection of product production data; desensitization module: used to desensitize the product production data to obtain product production desensitized data when the accuracy of data collection is represented by the accuracy of the content of the collected product production data; third analysis module: used to analyze the accuracy of desensitization processing according to the product production desensitized data.

[0080] The technical scheme in the above-mentioned embodiment of the present application has at least the following technical effects or advantages: product production monitoring parameter data is obtained by acquiring data from the product production line and product production data, and the completeness of data collection is analyzed according to the product production monitoring parameter data. When the completeness of data collection is represented by the complete collection of product production data, the accuracy of data collection is analyzed according to the product production monitoring parameter data. When the accuracy of data collection is represented by the accuracy of the content of the collected product production data, the product production data is desensitized to obtain product production desensitized data, and then the accuracy of desensitization processing is analyzed according to the product production desensitized data. The product production data is repeatedly desensitized to obtain product production desensitized data until the accuracy of desensitization processing is analyzed according to the product production desensitized data, until the accuracy of desensitization processing indicates that the product production data has been accurately desensitized, thereby realizing accurate desensitization of product production data, and effectively solving the problem of overly inefficient data desensitization in the prior art.

[0081] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0082] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0083] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0084] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0085] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0086] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A product production line data collection and desensitization method based on multi-dimensional features, characterized in that: The following steps are involved: Obtain product production data by monitoring the product production line and collecting data; Acquire product production lines and product production data to obtain product production monitoring parameter data; Analyze the completeness of data collection based on product production monitoring parameter data; When the completeness of data collection indicates that product production data is completely collected, the accuracy of data collection is analyzed based on the product production monitoring parameter data; When the accuracy of data collection indicates that the content of the collected product production data is accurate, the product production data is desensitized to obtain product production desensitized data; Analyze the accuracy of desensitization processing based on product production desensitization data; The specific analysis process of analyzing the accurate situation of desensitization processing based on the desensitized data of the product production is as follows: Obtain desensitization processing accuracy parameter data based on product production desensitization data; Obtaining a desensitization processing accuracy evaluation index according to the desensitization processing accuracy parameter data, wherein the desensitization processing accuracy evaluation index is used to reflect the desensitization status of the product production data; Analyze the accuracy of desensitization treatment according to the desensitization treatment accuracy evaluation index; The specific analysis process of analyzing the accuracy of desensitization treatment according to the desensitization treatment accuracy evaluation index is as follows: Obtain a desensitization processing accuracy evaluation index, obtain a desensitization processing accuracy first threshold and a desensitization processing accuracy second threshold. When the desensitization processing accuracy evaluation index is greater than the desensitization processing accuracy first threshold and less than the desensitization processing accuracy second threshold, it indicates that the product production data has been accurately desensitized. When it is in other situations, it means that the desensitization processing of the product production data is inaccurate, and the product production data is desensitized repeatedly to obtain the product production desensitized data until the accurate situation of the desensitization processing is analyzed according to the product production desensitized data, until the accurate situation of the desensitization processing indicates that the product production data has been accurately desensitized; The specific calculation formula for the desensitization treatment accuracy evaluation index is: Where, j0=1,2,...,j,j0 represents the number of data acquisition times, j represents the total number of data acquisition times, h0=1,2,...,h,h0 represents the number of data acquisition times, h represents the total number of data acquisition times, ρ j0 It is represented as the evaluation index of the accuracy of desensitization processing of product production data at the j0th data collection. It is represented by the maximum number of encryption times when the product production data is acquired at the h0th data collection time during the j0th data collection. It is expressed as the minimum number of encryption times when the product production data is acquired at the h0th data collection time during the j0th data collection. It is represented by the number of desensitized data at the time of the h0th data acquisition of the product production data at the j0th data collection. It is expressed as the maximum number of standard encryption times at the j0th data collection. It is expressed as the minimum number of standard encryption times at the j0th data collection. It is expressed as the standard value of the number of desensitized data at the j0th data collection, Y1 is expressed as the weight ratio of the maximum number of encryption times in the desensitization processing accuracy assessment index, Y2 is expressed as the weight ratio of the minimum number of encryption times in the desensitization processing accuracy assessment index, Y3 is expressed as the weight ratio of the number of desensitized data in the desensitization processing accuracy assessment index, and e is expressed as a natural constant.

2. The product production line data collection and desensitization method based on multi-dimensional features as claimed in claim 1, characterized in that: The specific process of obtaining product production data is as follows: The product production line is monitored by torque sensors to obtain the screw tightening processing torque; The product production line is monitored by a timer to obtain the screw tightening processing time; Product production data includes screw tightening processing torque and screw tightening processing time.

3. The product production line data collection and desensitization method based on multi-dimensional features as claimed in claim 1, characterized in that: The specific analysis process of analyzing the complete data collection situation based on the production monitoring parameter data of the product is as follows: Obtain data collection completeness parameter data based on product production monitoring parameter data; Preprocess the data collection completeness parameter data; Analyze the preprocessed data collection completeness parameter data to obtain a data collection completeness evaluation coefficient, wherein the data collection completeness evaluation coefficient is used to indicate the completeness of the collected product production data; The data collection completeness was analyzed based on the data collection completeness assessment coefficient.

4. The product production line data collection and desensitization method based on multi-dimensional features as claimed in claim 3, characterized in that: The specific method for obtaining the data collection completeness evaluation coefficient is: Obtain the number of data collection times and number the data collection times; Obtain the number of data acquisitions and number the number of data acquisitions; Data collection completeness parameter data; Construct a calculation formula for the data collection completeness evaluation coefficient based on the data collection completeness parameter data; The specific data collection completeness evaluation coefficient calculation formula is: In the formula, It is the data collection completeness evaluation coefficient of the product production data at the j0th data collection, j0=1,2,...,j, j0 is the number of data collection times, j is the total number of data collection times, It represents the type of data collected when the product production data is collected at the h0th data collection at the j0th data collection, h0=1,2,...,h, h0 represents the number of data acquisition times, h represents the total number of data acquisition times, It represents the memory space occupied by the data at the h0th data acquisition of the product production data at the j0th data acquisition. It is represented as the reference value of the collected data type at the j0th data collection. It is represented as the reference value of the memory space occupied by the data at the j0th data collection, D1 is represented as the weight ratio of the collected data type in the data collection completeness evaluation coefficient, and D2 is represented as the weight ratio of the memory space occupied by the data in the data collection completeness evaluation coefficient.

5. The product production line data collection and desensitization method based on multi-dimensional features as claimed in claim 3, characterized in that: The specific analysis process of analyzing the completeness of data collection according to the data collection completeness evaluation coefficient is as follows: Obtaining a data collection completeness evaluation coefficient, obtaining a first data collection completeness threshold and a second data collection completeness threshold, comparing the data collection completeness evaluation coefficient with the first data collection completeness threshold and the second data collection completeness threshold, when the data collection completeness evaluation coefficient is greater than the first data collection completeness threshold and less than the second data collection completeness threshold, it indicates that the product production data is completely collected; When in other situations, it means that the product production data is not collected completely, and the product production data is obtained by repeatedly monitoring the product production line and collecting data until the data collection completeness situation is analyzed according to the data collection completeness assessment coefficient, until the data collection completeness situation indicates that the product production data is collected completely.

6. The product production line data collection and desensitization method based on multi-dimensional features as claimed in claim 1, characterized in that: The specific analysis process of analyzing the accuracy of data collection based on the product production monitoring parameter data is as follows: Obtain data collection accuracy parameter data based on product production monitoring parameter data; Normalize the data collection accuracy parameter data; Obtaining a data collection accuracy evaluation coefficient according to the normalized data collection accuracy parameter data analysis, wherein the data collection accuracy evaluation coefficient is used to reflect the accuracy of the collected product production data; Analyze the accuracy of data collection based on the data collection accuracy assessment coefficient; The specific analysis process of analyzing the accuracy of data collection according to the data collection accuracy evaluation coefficient is as follows: Obtaining a data collection accuracy evaluation coefficient, obtaining a first data collection accuracy threshold and a second data collection accuracy threshold. When the data collection accuracy evaluation coefficient is greater than the first data collection accuracy threshold and less than the second data collection accuracy threshold, it indicates that the content of the collected product production data is accurate. When in other situations, it means that the collected product production data is inaccurate, and the product production line is repeatedly monitored and data collected to obtain product production data until the data is collected accurately based on product production monitoring parameter data analysis, until the data collection is accurate, indicating that the collected product production data is accurate.

7. The product production line data collection and desensitization method based on multi-dimensional features as claimed in claim 1, characterized in that: The specific analysis process of desensitizing the product production data to obtain the desensitized product production data is as follows: The product production data is encrypted and masked to obtain product production desensitized data; The data masking process includes productive masking process and random masking process.

8. Product production line data collection and desensitization system based on multi-dimensional features, characterized by: include: A monitoring module, a parameter acquisition module, a first analysis module, a second analysis module, a desensitization module and a third analysis module; The monitoring module is used to obtain product production data by monitoring the product production line and collecting data; The parameter acquisition module is used to acquire product production line and product production data to obtain product production monitoring parameter data; The first analysis module is used to analyze the completeness of data collection based on product production monitoring parameter data; The second analysis module is used to analyze the accuracy of data collection according to the product production monitoring parameter data when the complete data collection situation indicates that the product production data is completely collected; The desensitization module is used to desensitize the product production data to obtain product production desensitized data when the data collection accuracy indicates that the collected product production data is accurate; The third analysis module is used to analyze the accuracy of the desensitization processing according to the product production desensitization data; The specific analysis process of analyzing the accurate situation of desensitization processing based on the desensitized data of the product production is as follows: Obtain desensitization processing accuracy parameter data based on product production desensitization data; Obtaining a desensitization processing accuracy evaluation index according to the desensitization processing accuracy parameter data, wherein the desensitization processing accuracy evaluation index is used to reflect the desensitization status of the product production data; Analyze the accuracy of desensitization treatment according to the desensitization treatment accuracy evaluation index; The specific analysis process of analyzing the accuracy of desensitization treatment according to the desensitization treatment accuracy evaluation index is as follows: Obtain a desensitization processing accuracy evaluation index, obtain a desensitization processing accuracy first threshold and a desensitization processing accuracy second threshold. When the desensitization processing accuracy evaluation index is greater than the desensitization processing accuracy first threshold and less than the desensitization processing accuracy second threshold, it indicates that the product production data has been accurately desensitized. When it is in other situations, it means that the desensitization processing of the product production data is inaccurate, and the product production data is desensitized repeatedly to obtain the product production desensitized data until the accurate situation of the desensitization processing is analyzed according to the product production desensitized data, until the accurate situation of the desensitization processing indicates that the product production data has been accurately desensitized; The specific calculation formula for the desensitization treatment accuracy evaluation index is: Where, j0=1,2,...,j,j0 represents the number of data acquisition times, j represents the total number of data acquisition times, h0=1,2,...,h,h0 represents the number of data acquisition times, h represents the total number of data acquisition times, ρ j0 It is represented as the evaluation index of the accuracy of desensitization processing of product production data at the j0th data collection. It is represented by the maximum number of encryption times when the product production data is acquired at the h0th data collection time during the j0th data collection. It is expressed as the minimum number of encryption times when the product production data is acquired at the h0th data collection time during the j0th data collection. It is represented by the number of desensitized data at the time of the h0th data acquisition of the product production data at the j0th data collection. It is expressed as the maximum number of standard encryption times at the j0th data collection. It is expressed as the minimum number of standard encryption times at the j0th data collection. It is expressed as the standard value of the number of desensitized data at the j0th data collection, Y1 is expressed as the weight ratio of the maximum number of encryption times in the desensitization processing accuracy assessment index, Y2 is expressed as the weight ratio of the minimum number of encryption times in the desensitization processing accuracy assessment index, Y3 is expressed as the weight ratio of the number of desensitized data in the desensitization processing accuracy assessment index, and e is expressed as a natural constant.

Images