A detection method for data line production

Abstract

The invention discloses a detection method for data line production, which belongs to the technical field of data line detection, and specifically comprises: constructing a test system and performing intelligent parameter configuration, allocating an RFID unique identifier to each data line to be tested and entering information; during the test, the system automatically collects and records monitoring data of the data line, and uses a big data analysis method to perform in-depth analysis on the monitoring data, and combines an improved anomaly detection algorithm to determine whether the data line is qualified, and for unqualified samples, fault location and analysis will be performed, and improvement measures will be formulated; the measures will be fed back to the test system in real time, and at the same time, the test parameters and the anomaly detection algorithm will be optimized according to the fault case, so as to improve the test efficiency and accuracy, and realize the intelligent detection and optimization of the data line quality.

Description

Technical Field

[0001] The invention belongs to the technical field of data line detection, in particular to a detection method for data line production. Background Art

[0002] As a key component for transmitting data between connected devices, the quality and performance of data cables are of vital importance. With the rapid development of information technology, data cables not only need to meet basic data transmission needs, but also need to adapt to high-speed, large-capacity, and multi-functional transmission scenarios, as well as cope with various complex usage environments. Therefore, strict testing of data cables to ensure their stable quality and reliable performance is an indispensable part of the data cable production process.

[0003] In the early days of data cable development, due to technical limitations and relatively simple application scenarios, data cable testing mainly focused on basic electrical performance and physical properties, such as conductivity and insulation resistance. The detection methods were relatively simple and mainly relied on manual operation and basic testing instruments. With the increase in the transmission rate and complexity of data cables, the number of detection items has gradually increased, covering connection functions, charging functions, contact impedance, withstand voltage and other aspects, which has improved detection efficiency and accuracy, but is still not mature enough to cover the entire process from raw materials to finished products.

[0004] For example, the patent with the authorization announcement number CN216350993U discloses a conductivity detection device for data cable production, including: a box and a workbench, the box is fixedly installed at the bottom of the workbench, a fixing frame is provided on the top of the workbench, a horizontal board for placing the wire core is provided between the fixing frames, a multimeter and a detection contact are provided on the workbench, a battery is installed in the box, the detection contact and the multimeter are connected by a wire, and the multimeter and the battery are connected by a wire. The beneficial effect of this technical solution is: the conductivity detection device for data cable production places the wire core to be detected through the fixing frame and the horizontal board, one end of the detection contact contacts the wire core, and the other end is connected to the multimeter to display the resistance voltage, and the use cost is low and the operation is convenient.

[0005] For example, the patent with publication number CN115508750A discloses an automated detection equipment for data cable production, including: support legs, the support legs are arranged in two groups, a hexagonal cylinder and a hexagonal auxiliary cylinder are arranged between the two groups of support legs, and a motor is fixedly connected to the upper end side surface of one group of support legs, and the motor is used to drive the rotation of the hexagonal cylinder and the hexagonal auxiliary cylinder. It can be achieved by detecting the data cable and simulating manual shaking of the data cable during detection to avoid the problem of cold soldering or loose welding of the data cable. After the data cable detection is completed, the detected data cables are classified and put out to avoid manual observation of the line sequence of the data cables, which may cause workers to be easily dizzy. Manual workers only need to continuously plug in the data cables and no other work is required, which frees up a certain amount of labor for workers, thereby improving the efficiency of data cable detection.

[0006] The above existing technologies all have the following problems: 1) CN216350993U has a low degree of automation and intelligence, and can only display resistance and voltage through a multimeter, and is unable to collect, store and analyze large amounts of data. At the same time, when the basic conductivity detection device finds an unqualified wire core, it cannot automatically locate and analyze the fault, and requires manual troubleshooting; 2) In CN115508750A, although the functions of automatic detection and simulated swaying of the data line are realized, it is relatively weak in data analysis and intelligent judgment, and lacks the big data analysis capabilities and improved anomaly detection algorithms of advanced systems, and is unable to deeply mine and analyze test data. At the same time, the lack of real-time feedback and optimization mechanisms may result in the test process and parameters being unable to be adjusted and optimized in a timely manner according to actual conditions. Summary of the invention

[0007] In view of the shortcomings of the prior art, the present invention proposes a detection method for data cable production, constructs a test system and performs intelligent parameter configuration, assigns an RFID unique identifier to each data cable to be tested and enters information; during the test, the system automatically collects and records the monitoring data of the data cable, and uses a big data analysis method to perform in-depth analysis on the monitoring data, and combines an improved anomaly detection algorithm to determine whether the data cable is qualified, and for unqualified samples, fault location and analysis will be performed, and improvement measures will be formulated; the measures will be fed back to the test system in real time, and the test parameters and anomaly detection algorithms will be optimized according to the fault cases, thereby improving the test efficiency and accuracy, and realizing the intelligent detection and optimization of the data cable quality.

[0008] To achieve the above object, the present invention provides the following technical solutions:

[0009] A detection method for data line production, comprising:

[0010] Step S1: construct a test system and perform intelligent parameter configuration. At the same time, uniquely identify each data line to be tested with RFID and enter its information into the test system;

[0011] Step S2: Connect the data line to be tested to the test system, and the test system automatically collects and records monitoring data;

[0012] Step S3: Use big data analysis methods to analyze the monitoring data recorded during the test process. According to the data analysis results, combined with the improved anomaly detection algorithm, determine whether the data line is qualified. If unqualified, locate and analyze the unqualified samples. At the same time, formulate improvement measures based on the analysis results.

[0013] Step S4: Feedback the improvement measures to the test system in real time, and optimize the test parameters and anomaly detection algorithm based on the failure cases of the test data.

[0014] Specifically, the specific steps of step S3 include:

[0015] S3.1: Collect monitoring data recorded during the test from the test system , and preprocess the collected monitoring data to obtain the preprocessed monitoring data ,in, represents the Nth preprocessed monitoring data, and N represents the number of preprocessed monitoring data;

[0016] S3.2: Calculation using correlation analysis The correlation coefficient between the data , construct a correlation matrix based on the correlation coefficient, and at the same time, construct a distance function based on the correlation matrix Measure the similarity between data points and use clustering algorithms to classify them according to the distance function Grouping, where Indicates monitoring data and The correlation coefficient between ;

[0017] S3.3: Load the pre-trained prediction model and group the Input into the pre-trained prediction model to obtain the prediction results of performance trends. At the same time, use exploratory analysis methods to identify potential abnormal patterns based on the prediction results of performance trends.

[0018] Specifically, the specific steps of step S3 also include:

[0019] S3.4: Load the pre-trained anomaly detection model and group the The data is input into the trained anomaly detection model, and the anomaly detection results are output based on the performance trend prediction results and potential anomaly patterns. At the same time, the test samples are judged based on the anomaly detection results.

[0020] If all test samples can work normally under the specified current load and no abnormal phenomenon occurs, the batch of data cables is judged to be qualified;

[0021] If the current value exceeds the set range or the data line is overheated, the data line is judged as unqualified, and the fault of the unqualified sample is located, and the abnormal data point is marked.

[0022] Specifically, the specific steps of step S3 also include:

[0023] S3.5: Based on the marked abnormal data line, the output abnormal detection results are optimized using the improved distance-based anomaly detection algorithm to verify the accuracy of the abnormal data points. The formula of the improved distance-based anomaly detection algorithm is: ;

[0024] in, express Monitoring data The degree of abnormality relative to its neighbors, Represents the nearest neighbor monitoring data The nearest neighbor distance, Indicates monitoring data arrive The Euclidean distance of Indicates monitoring data Yudi j Nearest neighbor monitoring data, j Represents the coordinate index of the nearest neighbor monitoring data;

[0025] S3.6: Analyze the causes of the anomalies based on the optimized anomaly detection results and anomaly verification results, and formulate improvement measures and plans based on the causes of the anomalies. At the same time, output the improvement measures and plans in the form of reports and charts.

[0026] Specifically, the calculation formula of the correlation coefficient in S3.2 is: ;

[0027] in, Indicates monitoring data The weight of Indicates monitoring data The weight of Indicates monitoring data The weighted mean of Indicates monitoring data The weighted mean of .

[0028] Specifically, the specific steps of fault location and marking in S3.4 include:

[0029] S3.41: Set current threshold With temperature threshold and using a current sensor to monitor the current value flowing through the data line, and using a temperature sensor to monitor the temperature of the data line;

[0030] S3.42: The monitored current value Compared with the set current threshold,

[0031] like , then the data line is judged to be qualified;

[0032] like , then the data line is judged to be unqualified;

[0033] S3.43: The monitored temperature value With the set temperature threshold For comparison,

[0034] like , then the data line is judged to be qualified;

[0035] like , then the data line is judged to be unqualified;

[0036] S3.44: For data lines that are judged to be unqualified, use a fault indicator to locate the fault and record all current values ​​and temperature values ​​that exceed the set thresholds;

[0037] S3.45: Mark the data of current values ​​and temperature values ​​exceeding the set thresholds, and organize the monitoring results, judgment results, fault location information and abnormal data point information into a report.

[0038] Specifically, in step S1, the test system integrates a power supply, a load resistor, a high-precision current / voltage sensor, a temperature sensor and automatic control software.

[0039] Specifically, the intelligent configuration of parameters in step S1 is to automatically set the resistance value of the load resistor and the output voltage value and current limit value of the power supply by using the automation control software.

[0040] Compared with the prior art, the present invention has the following beneficial effects:

[0041] The present invention proposes a detection method for data cable production. By constructing a test system and performing intelligent parameter configuration, and providing an RFID unique identification for each data cable to be tested and entering the identification into the system, the automatic testing and precise management of the data cable are realized, which not only improves the test efficiency, but also ensures the accuracy and traceability of the test data. At the same time, by using big data analysis and an improved anomaly detection algorithm, the qualification of the data cable can be judged more accurately, and potential faults can be discovered in time.

[0042] The present invention proposes a detection method for data cable production, which forms a closed loop of continuous improvement by feeding back improvement measures to a test system in real time and optimizing test parameters and anomaly detection algorithms according to failure cases. This not only improves the accuracy and stability of the test system, but also makes the test process more intelligent and adaptive, reduces the failure rate of data cables, and improves product quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 This is a framework diagram of a detection method for data cable production according to the present invention;

[0044] Figure 2 This is an algorithm flow chart of a detection method for data line production according to the present invention;

[0045] Figure 3 The present invention is a data line abnormality detection flow chart of a detection method for data line production. DETAILED DESCRIPTION

[0046] Example 1: Please refer to Figure 1-Figure 3 , an embodiment of the present invention provides: a detection method for data line production, comprising the following steps:

[0047] Step S1: construct a test system and perform intelligent parameter configuration. At the same time, uniquely identify each data line to be tested with RFID and enter its information into the test system;

[0048] In step S1 , the test system integrates a power supply, a load resistor, a high-precision current / voltage sensor, a temperature sensor, and automated control software.

[0049] The intelligent configuration of parameters in step S1 is to automatically set the resistance value of the load resistor and the output voltage value and current limit value of the power supply by using the automatic control software.

[0050] Among them, each data line to be tested needs to be pre-processed before being uniquely identified by RFID. The pre-processing includes: 1) cleaning the raw materials through automatic cleaning equipment to remove surface dirt and oxide layer; 2) using automatic drying equipment to dry the cleaned raw materials to ensure that the raw materials are dry and free of moisture.

[0051] The specific steps of uniquely identifying each data line to be tested with RFID include:

[0052] (1) Select an RFID tag with a unique ID based on the size of the data cable to be tested, the usage environment, and the compatibility of the RFID reader / writer;

[0053] (2) Use an RFID reader to initialize the RFID tag, assign a unique identifier to each tag, and ensure that it is unique globally;

[0054] (3) Attach the RFID tag to the data cable, usually by pasting, embedding or hanging, ensuring that the tag is positioned so as not to interfere with the normal use of the data cable and is easily identified by the RFID reader;

[0055] (4) Enter the basic information of the data cable, such as model, specification, batch number, and unique ID of the RFID tag, into the test system to ensure the accuracy and completeness of the information for subsequent data tracking and analysis.

[0056] Step S2: Connect the data line to be tested to the test system, and the test system automatically collects and records monitoring data;

[0057] The monitoring data includes the current, voltage, transmission rate and impedance of the data line to be tested.

[0058] Step S3: Use big data analysis methods to analyze the monitoring data recorded during the test process. According to the data analysis results, combined with the improved anomaly detection algorithm, determine whether the data line is qualified. If unqualified, locate and analyze the unqualified samples. At the same time, formulate improvement measures based on the analysis results.

[0059] Step S4: Feedback the improvement measures to the test system in real time, and optimize the test parameters and anomaly detection algorithm based on the failure cases of the test data.

[0060] The specific steps of step S3 include:

[0061] S3.1: Collect monitoring data recorded during the test from the test system , and preprocess the collected monitoring data to obtain the preprocessed monitoring data ,in, represents the Nth preprocessed monitoring data, and N represents the number of preprocessed monitoring data;

[0062] Among them, preprocessing includes: data cleaning, data integration, data transformation and data reduction; data cleaning includes removing duplicate, invalid or erroneous data; data integration includes integrating data from different sources; data transformation includes data standardization and normalization; data reduction refers to reducing the amount of data to improve analysis efficiency.

[0063] S3.2: Calculate using correlation analysis The correlation coefficient between the data , construct a correlation matrix based on the correlation coefficient At the same time, the distance function is constructed based on the correlation matrix Measure the similarity between data points and use clustering algorithms to classify them according to the distance function Grouping, where Indicates monitoring data and The correlation coefficient between , wherein, in the present invention, the clustering algorithm selects k-means clustering, and the calculation formula of the k-means clustering method is the prior art content in this field, which is not the inventive solution of the present application and will not be described in detail here;

[0064] S3.3: Load the pre-trained prediction model and group the Input into the pre-trained prediction model to obtain the prediction results of performance trends. At the same time, use exploratory analysis methods to identify potential abnormal patterns based on the prediction results of performance trends.

[0065] The specific steps of the exploratory analysis method include:

[0066] (1) Import , and use functions such as df.shape, df.head(), and df.tail() to Conduct preliminary observations to understand The number of rows, columns, and basic structure of the data;

[0067] (2) Use the df.info() function to view Detailed information about the monitoring data, including the data type and number of non-null values ​​in each column, and use the df.describe() function to obtain statistical summaries of numerical data, including mean, standard deviation, minimum, quartiles, and maximum values;

[0068] (3) Yes Cleaning is performed;

[0069] (4) Draw histograms, box plots, and QQ plots: Check the distribution of the data, including the symmetry, kurtosis, and skewness of the data;

[0070] (5) Draw scatter plots, line graphs, and heat maps: Explore the relationship between variables, including the relationship between continuous variables and continuous variables, between discrete variables and discrete variables, and between discrete variables and continuous variables;

[0071] (6) Use classification charts: such as bar charts, pie charts, etc. to view the distribution of classification data;

[0072] (7) Based on the results of data exploration, select features that have an important impact on the prediction results. If the number of features is too large, use the feature dimensionality reduction method of principal component analysis to reduce the number of features and improve the generalization ability of the model;

[0073] (8) After cleaning and exploring the data, a prediction model is constructed based on the selected features, and the performance of the model is evaluated using cross-validation metrics.

[0074] S3.4: Load the pre-trained anomaly detection model and group the The data is input into the trained anomaly detection model, and the anomaly detection results are output based on the performance trend prediction results and potential anomaly patterns. At the same time, the test samples are judged based on the anomaly detection results.

[0075] If all test samples can work normally under the specified current load and no abnormal phenomenon occurs, the batch of data cables is judged to be qualified;

[0076] If the current value exceeds the set range or the data line is overheated, the data line is judged as unqualified, and the fault of the unqualified sample is located, and the abnormal data point is marked;

[0077] S3.5: Based on the marked abnormal data line, the output abnormal detection results are optimized using the improved distance-based anomaly detection algorithm to verify the accuracy of the abnormal data points. The formula of the improved distance-based anomaly detection algorithm is: ;

[0078] in, express Monitoring data The degree of abnormality relative to its neighbors, Represents the nearest neighbor monitoring data The nearest neighbor distance, Indicates monitoring data arrive The Euclidean distance of Indicates monitoring data Yudi j Nearest neighbor monitoring data, j Represents the coordinate index of the nearest neighbor monitoring data;

[0079] S3.6: Analyze the causes of the anomalies based on the optimized anomaly detection results and anomaly verification results, and formulate improvement measures and plans based on the causes of the anomalies. At the same time, output the improvement measures and plans in the form of reports and charts.

[0080] Furthermore, the specific steps of S3.6 include:

[0081] (1) Obtain optimized anomaly detection results and verification results:

[0082] Obtain optimized anomaly detection results from quality inspection systems or related tools, and verify the anomaly detection results to ensure the accuracy and reliability of the results;

[0083] (2) Analyze the cause of the abnormality:

[0084] Based on the abnormality detection results, preliminarily determine the possible causes of the abnormality, such as raw material problems, production process defects, and equipment failures. At the same time, organize a technical team to conduct an in-depth analysis of the abnormality causes, which may include on-site investigations, data comparisons, and experimental verification methods. After determining the abnormality causes, make detailed records and classifications to facilitate the subsequent formulation of improvement measures;

[0085] (3) Formulate improvement measures and plans:

[0086] According to the causes of the abnormalities, formulate corresponding improvement measures and plans, among which the improvement measures include replacing raw materials, optimizing production processes, repairing or replacing equipment, and strengthening employee training. At the same time, formulate specific implementation plans and timetables to ensure the effective implementation of improvement measures.

[0087] (4) Output improvement measures and solutions in a visual form:

[0088] Organize the improvement measures and plans into a report, describing in detail the causes of the abnormalities, improvement measures, implementation plans and expected results, and use charts, flow charts and visualization tools to present the improvement measures and plans in an intuitive and easy-to-understand manner. At the same time, submit the report and visualization charts to the staff so that they can understand the improvement measures and plans and conduct subsequent tracking and evaluation.

[0089] The calculation formula of the correlation coefficient in S3.2 is: ;

[0090] in, Indicates monitoring data The weight of Indicates monitoring data The weight of Indicates monitoring data The weighted mean of Indicates monitoring data The weighted mean of .

[0091] The specific steps for fault location and marking in S3.4 include:

[0092] S3.41: Set current threshold With temperature threshold and using a current sensor to monitor the current value flowing through the data line, and using a temperature sensor to monitor the temperature of the data line;

[0093] S3.42: The monitored current value Compared with the set current threshold,

[0094] like , then the data line is judged to be qualified;

[0095] like , then the data line is judged to be unqualified;

[0096] S3.43: The monitored temperature value With the set temperature threshold For comparison,

[0097] like , then the data line is judged to be qualified;

[0098] like , then the data line is judged to be unqualified;

[0099] S3.44: For data lines that are judged to be unqualified, use a fault indicator to locate the fault and record all current values ​​and temperature values ​​that exceed the set thresholds;

[0100] S3.45: Mark the data of current values ​​and temperature values ​​exceeding the set thresholds, and organize the monitoring results, judgment results, fault location information and abnormal data point information into a report.

[0101] Embodiment 2: An electronic device comprises a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the steps of a detection method for data line production are implemented. For details, please refer to the above method embodiment, which will not be repeated here.

[0102] A computer-readable storage medium stores computer instructions, which, when executed, execute steps of a detection method for data line production, wherein the storage medium may be a volatile or non-volatile computer-readable storage medium.

[0103] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation modes, which are merely illustrative rather than restrictive. Under the guidance of the present invention, ordinary technicians in the field may also change, modify, replace and modify the above-mentioned embodiments without departing from the purpose and scope of protection of the present invention, and all of these are within the protection of the present invention.

Claims

1. A detection method for data line production, characterized in that: include: Step S1: construct a test system and perform intelligent parameter configuration. At the same time, uniquely identify each data line to be tested with RFID and enter its information into the test system; Step S2: Connect the data line to be tested to the test system, and the test system automatically collects and records monitoring data; Step S3: Use big data analysis methods to analyze the monitoring data recorded during the test process. According to the data analysis results, combined with the improved anomaly detection algorithm, determine whether the data line is qualified. If unqualified, locate and analyze the unqualified samples. At the same time, formulate improvement measures based on the analysis results. Step S4: Feedback the improvement measures to the test system in real time, and optimize the test parameters and anomaly detection algorithm based on the failure cases of the test data; The specific steps of step S3 include: S3.1: Collect the monitoring data recorded during the test from the test system, and preprocess the collected monitoring data to obtain the preprocessed monitoring data X = {x1,...,x N }, where x N represents the Nth preprocessed monitoring data, and N represents the number of preprocessed monitoring data; S3.2: Use the correlation analysis method to calculate X = {x1,...,x N The correlation coefficient C between the data in pq , construct the correlation matrix based on the correlation coefficient, and at the same time, construct the distance function d based on the correlation matrix pq =1-|C pq |Measure the similarity between data points and use clustering algorithm to cluster X={x1,...,x N } are grouped, where C pq Represents monitoring data x p and x q The correlation coefficient between them, and p,q∈N; S3.3: Load the pre-trained prediction model and group the grouped X={x1,...,x N } Input into the pre-trained prediction model to obtain the prediction results of performance trends. At the same time, use exploratory analysis methods to identify potential abnormal patterns based on the prediction results of performance trends; The specific steps of step S3 also include: S3.4: Load the pre-trained anomaly detection model and group the grouped X={x1,...,x N } Input into the trained anomaly detection model, combine the performance trend prediction results and potential anomaly patterns, output anomaly detection results, and at the same time, judge the test samples based on the anomaly detection results; If all test samples can work normally under the specified current load and no abnormal phenomenon occurs, the batch of data cables is judged to be qualified; If the current value exceeds the set range or the data line is overheated, the data line is judged as unqualified, and the fault of the unqualified sample is located, and the abnormal data point is marked; The specific steps of step S3 also include: S3.5: Based on the marked abnormal data line, the output abnormal detection results are optimized using the improved distance-based anomaly detection algorithm to verify the accuracy of the abnormal data points. The formula of the improved distance-based anomaly detection algorithm is: Among them, Y(x i ) means X={x1,...,x N }Preprocessed monitoring data x i The degree of abnormality relative to its neighbors, d(K j ) represents the nearest neighbor monitoring data K j The distance to the nearest neighbor, d(x i ,K j ) represents the preprocessed monitoring data x i To K j The Euclidean distance, K j Represents the preprocessed monitoring data x i The jth nearest neighbor monitoring data, j represents the coordinate index of the nearest neighbor monitoring data; S3.6: Analyze the causes of the anomalies based on the optimized anomaly detection results and anomaly verification results, and formulate improvement measures and plans based on the causes of the anomalies. At the same time, output the improvement measures and plans in the form of reports and charts.

2. A detection method for data line production as claimed in claim 1, characterized in that: The calculation formula of the correlation coefficient in S3.2 is: Among them, w p Represents the preprocessed monitoring data x p The weight, w q Represents the preprocessed monitoring data x q The weight of Represents the preprocessed monitoring data x p The weighted mean of Represents the preprocessed monitoring data x q The weighted mean of .

3. A detection method for data line production as claimed in claim 2, characterized in that: The specific steps of fault location and marking in S3.4 include: S3.41: Set a current threshold h1 and a temperature threshold h2, and use a current sensor to monitor the current value flowing through the data line, and use a temperature sensor to monitor the temperature of the data line; S3.42: The monitored current value I x Compared with the set current threshold, If I x ≤h1, the data line is considered qualified; If I x >h1, the data line is judged to be unqualified; S3.43: The monitored temperature value T x Compared with the set temperature threshold h2, If T x ≤h2, the data line is judged to be qualified; If T x >h2, the data line is judged to be unqualified; S3.44: For data lines that are judged to be unqualified, use a fault indicator to locate the fault and record all current values ​​and temperature values ​​that exceed the set thresholds; S3.45: Mark the data of current values ​​and temperature values ​​exceeding the set thresholds, and organize the monitoring results, judgment results, fault location information and abnormal data point information into a report.

4. A detection method for data line production as claimed in claim 3, characterized in that: In step S1, the test system integrates a power supply, a load resistor, a high-precision current / voltage sensor, a temperature sensor and automatic control software.

5. A detection method for data line production as claimed in claim 4, characterized in that: The intelligent configuration of parameters in step S1 is to automatically set the resistance value of the load resistor and the output voltage value and current limit value of the power supply by using the automation control software.

Images