Circuit board sensor nondestructive testing method based on big data

Abstract

The invention relates to the technical field of circuit board detection, in particular to a circuit board sensor nondestructive testing method based on big data, which comprises the following steps of: acquiring temperature, pressure, current and voltage data of key positions of a circuit board sensor, sending the data to a central database through real-time communication for summarizing, and synchronously performing data preliminary screening; and checking the integrity and availability of the data to form sensor summarized data. According to the method, the difference value between each data point and the adjacent data point is calculated, and the variation amplitude of the continuous data points is monitored, so that tiny abnormal fluctuation can be more sensitively captured, potential fault signals can be marked, an abnormal mode can be compared with a recorded fault case, and the fault mode and the frequent occurrence area of the fault mode can be identified; the fault mode is directly associated with the operation condition, the accuracy of fault prediction is enhanced, the dynamic adaptability of the detection system is ensured through monitoring adjustment of detailed fault analysis, and the efficiency of fault response is remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of circuit board detection, and in particular to a nondestructive detection method of a circuit board sensor based on big data. Background Art

[0002] The field of circuit board inspection technology focuses on using various inspection technologies to analyze and ensure the functional integrity and performance of circuit boards and their components. It covers a variety of technologies from micro-inspection to macro-analysis, including electrical testing (such as resistance, voltage and current measurement), optical inspection (using high-resolution camera systems), X-ray imaging technology (used to view the internal structure of the board), infrared thermal imaging (detecting hot spots or cold spots to identify potential problems) and ultrasonic inspection, etc., which are not only used to detect manufacturing defects such as solder joint problems, short circuits or open circuits, but also to monitor aging or wear during the product life cycle, thereby ensuring the reliability and safety of the circuit board in its application environment.

[0003] Among them, the circuit board sensor nondestructive testing method involves the use of nondestructive testing technology to evaluate the integrity of the circuit board sensor without causing any damage to it. The main purpose is to ensure the quality and performance of the sensor and circuit board during the production process and use to avoid potential failures. Nondestructive testing can provide instant feedback and accurate fault analysis, which is crucial to maintaining high quality standards of electronic manufacturing and reducing maintenance costs, especially in areas with extremely high reliability and safety requirements such as aviation, medical equipment and advanced automotive electronic systems.

[0004] Although existing circuit board inspection technologies include a variety of methods, their capabilities in real-time data analysis and fault prediction are often limited. Traditional technologies such as electrical testing and optical inspection rely on fixed parameters and operating environments, and lack the ability to adapt to rapidly changing or unknown new fault types. Especially in the stage of small defects or early failures on circuit boards, they often cannot provide sufficient sensitivity, resulting in delayed diagnosis of potential problems. In addition, although technologies such as X-ray imaging and infrared thermal imaging can detect internal structural problems, their effects are not ideal in early fault warning and identification of small changes. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings existing in the prior art and to propose a non-destructive testing method for circuit board sensors based on big data.

[0006] In order to achieve the above object, the present invention adopts the following technical solution: a circuit board sensor nondestructive testing method based on big data, comprising the following steps: S1: Collect temperature, pressure, current and voltage data of key positions of circuit board sensors, send the data to the central database through real-time communication for aggregation, perform preliminary data screening simultaneously, verify the integrity and availability of the data, and form sensor summary data; S2: Based on the sensor summary data, the difference between each data point and the adjacent data points is calculated, and the change amplitude of the continuous data points is monitored. According to the change amplitude monitoring results, statistical analysis is performed to identify abnormal fluctuations in the data, mark the abnormal fluctuations and determine them as potential fault signals, and form an abnormal pattern recognition result; S3: using the abnormal pattern recognition result, comparing with the recorded fault cases, checking similar data patterns, determining the frequent occurrence area of ​​each fault mode, correlating the fault mode with the operating conditions according to the frequent occurrence area determination result, determining the fault type and actual cause, and generating fault location and cause analysis results; S4: Based on the fault location and cause analysis results, dynamically adjust the monitoring process, adjust the key areas of data collection according to the characteristics of the identified faults, perform system testing based on the key area adjustment results, confirm the effectiveness of the new settings for fault detection, and form a fault response detection plan.

[0007] As a further solution of the present invention, the steps of acquiring the sensor summary data are: S111: Perform real-time monitoring at key positions of circuit board sensors, collect and record temperature, pressure, current and voltage data, and generate real-time monitoring data records; S112: Based on the real-time monitoring data record, the data is checked and verified for correctness, and a preliminary format and integrity check is performed to obtain a verified data report; S113: Analyze and summarize the verified data reports, perform multi-dimensional data fusion, and form sensor summary data.

[0008] As a further solution of the present invention, the step of monitoring the variation range is: S211: extracting time series data according to the sensor summary data, performing data synchronization and alignment, verifying data continuity, and generating preprocessed continuous data point sequence data; S212: According to the pre-processed continuous data point sequence data, the formula: ; Calculate the difference between adjacent data points , get the data point difference value sequence data, where, represents the current data point in the sequence, Represents the next data point in the sequence; S213: Based on the data point difference value sequence data, perform a change range analysis, identify key change trends, and generate continuous data point change range analysis results.

[0009] As a further solution of the present invention, the step of obtaining the abnormal pattern recognition result is: S221: Based on the analysis results of the change range of the continuous data points, statistical analysis is performed using the formula: and ; Calculate the average and standard deviation , generates a summary of statistical data, where Represents the magnitude of change of each data point. is the total number of data points; S222: Based on the statistical data summary, apply the mean value and standard deviation to define anomaly detection threshold as , and use the threshold to filter data points that exceed the normal range of change, mark them as abnormal fluctuations, and generate an abnormal fluctuation data set; S223: According to the abnormal fluctuation data set, an abnormal pattern corresponding to a potential fault signal is identified, the abnormal data is classified through cluster analysis, and a potential fault pattern is determined to form an abnormal pattern recognition result.

[0010] As a further solution of the present invention, the step of determining the frequent occurrence area is: S311: Based on the abnormal pattern recognition result, compared with historical failure cases, new patterns in the data similar to known failure patterns are revealed through pattern matching, and failure pattern similarity analysis results are generated; S312: Using the failure mode similarity analysis result, adopt the formula: ; Calculate the deviation between the occurrence frequency of the fault instance and the overall average frequency , measure the frequency of occurrence of failure modes in different operating environments and time periods, and obtain the frequent regional distribution data, where Indicates The frequency of occurrence of a fault instance, represents the average frequency of all fault instances. is the total number of failure instances; S313: Using the frequent region distribution data, mapping the relationship between the fault mode and the corresponding frequent region is performed to establish a corresponding relationship table between the fault mode and the frequent region.

[0011] As a further solution of the present invention, the steps of obtaining the fault location and cause analysis results are: S321: using the corresponding relationship table between the fault mode and the frequent region, by collecting and sorting the operation data in the frequent region, performing association rule mining, identifying the correlation between the frequent region and the fault mode, and generating the fault mode and operation condition correlation analysis result; S322: Based on the analysis result of the association between the fault mode and the operating condition, a causal analysis is performed on the data, potential factors are associated with the fault instance, the cause of the fault is accurately located, and the direct cause and indirect cause of the fault are distinguished to obtain a fault cause analysis result; S323: Integrate the fault mode and operating condition association analysis results and the fault cause analysis results, record the type and cause of each fault, display it through a dynamic dashboard, reflect the latest fault response information in real time, and generate fault location and cause analysis results.

[0012] As a further solution of the present invention, the step of adjusting the key area is: S411: Based on the fault location and cause analysis results, analyze the characteristics and impact range of key faults, identify blind spots by comparing fault characteristics with the existing coverage capabilities of the monitoring system, and generate a key fault characteristic analysis report; S412: redesigning the data collection node layout and parameter settings according to the key fault characteristic analysis report, verifying the integrity and accuracy of data collection through simulation tests, and obtaining a data collection area adjustment plan; S413: Implement the data collection area adjustment plan, update the system's monitoring configuration file, and perform on-site debugging to confirm the validity of each monitoring node for the target fault and obtain dynamic adjustment records of the monitoring process.

[0013] As a further solution of the present invention, the steps of obtaining the fault response detection solution are: S421: dynamically adjust the records according to the monitoring process, perform comprehensive testing, inject simulated fault data into the monitoring node, verify the response capability of the new monitoring to the predetermined fault, collect and evaluate the response data of the monitoring system, and generate a monitoring test report; S422: Analyze the monitoring test report, compare the performance of the new and old system configurations in terms of fault detection accuracy and response time, and use the formula: ; Calculate the percentage increase in success rate , verify the improvement effect of the new configuration and obtain performance improvement data, among which, is the successful detection rate of the new configuration, is the successful detection rate of the old configuration; S423: Based on the performance improvement data, formulate targeted response measures for target category faults, record the operation steps and expected effects under different fault conditions, and form a fault response detection plan.

[0014] Compared with the prior art, the advantages and positive effects of the present invention are: In the present invention, by calculating the difference between each data point and its adjacent data points and monitoring the change amplitude of continuous data points, it is possible to more sensitively capture tiny abnormal fluctuations, mark potential fault signals, compare abnormal patterns with recorded fault cases, identify fault patterns and their frequent occurrence areas, directly associate fault patterns with operating conditions, enhance the accuracy of fault prediction, ensure the dynamic adaptability of the detection system through monitoring and adjustment of detailed fault analysis, and significantly improve the efficiency of fault response. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 It is a main step flow chart of the present invention; Figure 2 A flowchart for obtaining the sensor summary data of the present invention; Figure 3 It is a flow chart of monitoring the variation range of the present invention; Figure 4 The present invention is a flowchart for obtaining abnormal pattern recognition results; Figure 5 A flow chart for determining the frequent occurrence area of ​​the present invention; Figure 6 A flowchart for obtaining fault location and cause analysis results of the present invention; Figure 7 It is the adjustment flow chart of the key area of ​​the present invention; Figure 8 This is a flow chart for obtaining the fault response detection solution of the present invention. DETAILED DESCRIPTION

[0016] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0017] In the description of the present invention, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, in the description of the present invention, "multiple" means two or more, unless otherwise clearly and specifically defined.

[0018] See also Figure 1 , a circuit board sensor nondestructive testing method based on big data, comprising the following steps: S1: Collect temperature, pressure, current and voltage data of key positions of circuit board sensors, send the data to the central database through real-time communication for aggregation, perform preliminary data screening simultaneously, verify the integrity and availability of the data, and form sensor summary data; S2: Based on the sensor summary data, the difference between each data point and the adjacent data points is calculated, and the change amplitude of the continuous data points is monitored. According to the change amplitude monitoring results, statistical analysis is performed to identify abnormal fluctuations in the data, mark the abnormal fluctuations and determine them as potential fault signals, and form abnormal pattern recognition results; S3: Using the abnormal pattern recognition results, check the similar data patterns against the recorded fault cases, determine the frequent occurrence areas of each fault mode, associate the fault mode with the operating conditions according to the frequent occurrence area determination results, determine the fault type and actual cause, and generate the fault location and cause analysis results; S4: Based on the results of fault location and cause analysis, dynamically adjust the monitoring process, adjust the key areas of data collection according to the characteristics of the identified faults, conduct system tests based on the results of key area adjustments, confirm the effectiveness of the new settings for fault detection, and form a fault response detection plan.

[0019] Sensor summary data includes data timestamp and data source identification records; abnormal pattern recognition results include abnormal duration, abnormal intensity level and abnormal frequency; fault location and cause analysis results include operating environment, fault occurrence time and fault-related equipment component records; fault response detection plan includes monitoring strategy optimization records, parameter adjustment list and effect expectation verification records.

[0020] See also Figure 2 ,The steps to obtain sensor summary data are: S111: Perform real-time monitoring at key positions of circuit board sensors, collect and record temperature, pressure, current and voltage data, and generate real-time monitoring data records; Installing sensors at critical locations on a circuit board first requires determining which locations are most critical, a process that involves finely measuring the temperature distribution and current load of the circuit board sensors. Before installation, use an infrared thermal imager and a precision ammeter to scan the temperature and current of the circuit board sensors multiple times under normal operating conditions, record the data, and use this data to analyze hot spots and locations where current is concentrated. After determining these locations, select the appropriate sensor type, such as thermistors for temperature monitoring, pressure sensors for sensing physical pressure on the board, and current and voltage sensors installed at nodes with large current and voltage changes. Each sensor is equipped with a microprocessor that converts the monitored physical signals into digital signals in real time and formats this data into a standard report format, including key information such as timestamps, locations, and monitoring values ​​for subsequent transmission and analysis.

[0021] S112: Based on the real-time monitoring data records, the data is checked and verified for correctness, and a preliminary format and integrity check is performed to obtain a verified data report; During the transmission of monitoring data, the data is first encrypted and sent through the built-in wireless module of the sensor to ensure the security of the data when it is transmitted in the air. Before the data reaches the central processing unit, it is initially collected through specially established data receiving stations, which are equipped with high-performance receivers that can handle data streams from multiple sensors at the same time. The received data includes the sensor ID, the measured physical quantity, and the corresponding data value. The first screening of the data is automatic, and the system checks the integrity of each data item, such as whether the data packet is complete and whether the data values ​​are within the preset reasonable range. In addition, there is an error detection mechanism to identify and correct any possible transmission errors, such as data loss or damage. After a series of processing, the data forms a preliminary screened data report, which will then be stored in the database for further analysis.

[0022] S113: Analyze and summarize the verified data reports, perform multi-dimensional data fusion, and form sensor summary data; In the central processing unit, the verified data will be further analyzed. The process includes integrating the data from each sensor and classifying them according to the type and source of the data. For example, all temperature data are aggregated into a single data set for unified temperature analysis, including statistics of the average, maximum, minimum and fluctuation of the temperature. Similarly, current and voltage data are analyzed and integrated separately. In order to improve the efficiency and accuracy of data processing, data cleaning technology is used to eliminate abnormal data that obviously deviates from the normal range. Finally, through data fusion technology, all monitoring data are combined to form a comprehensive data report containing the comprehensive operating status of the circuit board. The report not only records the monitoring data in detail, but also provides the overall health status and operating trend of the circuit board through statistical analysis.

[0023] See also Figure 3 , the monitoring steps of the change range are: S211: extracting time series data according to the sensor summary data, performing data synchronization and alignment, verifying the continuity of the data, and generating pre-processed continuous data point sequence data; According to the standardized process of time synchronization and data alignment, the data generated by each sensor node is retrieved and collected, the consistency of each data timestamp and data integrity are analyzed, and the data is preprocessed to eliminate time deviation and data missing. The missing data points are supplemented by linear interpolation to ensure the continuity and consistency of the data on the time axis. Next, according to the time interval of the data sequence, the data sampling frequency is adjusted to make the data points denser, enhance the resolution and processing accuracy of the data, so that the subsequent analysis can more accurately reflect the physical phenomena monitored by the sensor, and finally generate a continuous data point sequence processed by time synchronization and data alignment, which provides an accurate data basis for the next step of difference value calculation and change amplitude monitoring.

[0024] S212: According to the pre-processed continuous data point sequence data, the formula: ; Calculate the difference between adjacent data points , get the data point difference value sequence data, where, represents the current data point in the sequence, Represents the next data point in the sequence; The data points read from the sensor are , , , then the difference value of the first data point , the difference value of the second data point The calculation process is implemented by simple arithmetic subtraction operation, calculating the difference between each pair of consecutive data points. This result shows that arrive , the data value has an increasing range of 5, from arrive Also increasing by 5, this continuous increasing trend reflects a consistent changing pattern of the data, which is crucial for identifying abnormal fluctuations in the data.

[0025] S213: performing change range analysis based on the data point difference value sequence data, identifying key change trends, and generating continuous data point change range analysis results; After calculating the difference value, the next step is to analyze the change amplitude of the data point difference value sequence. First, a threshold is set to determine whether the data point difference value exceeds this threshold to identify significant changes in the data. For example, by setting the change amplitude threshold to 10, all data point difference values ​​are checked and the relationship between each data point difference value and the threshold is compared. If the difference value exceeds the threshold, this point is marked as an abnormal change point and the location of the abnormal point is recorded. Such analysis can help identify potential abnormal fluctuations in the data sequence, thereby providing a basis for subsequent fault detection and ultimately obtaining the change amplitude results of continuous data points. The identified abnormal fluctuation locations will be used for further analysis and fault prediction.

[0026] See also Figure 4 , the steps to obtain the abnormal pattern recognition results are: S221: Based on the analysis results of the change range of continuous data points, statistical analysis is performed using the formula: and ; Calculate the average and standard deviation , generates a summary of statistical data, where Represents the magnitude of change of each data point. is the total number of data points; There are the following data points , first calculate the mean: ; Calculate the standard deviation by summing the squares of the deviations of each data point from the mean and taking the square root: ; ; Using the calculation results, the anomaly detection threshold can be set to obtain the monitoring standard for abnormal fluctuations, reflecting the direct relationship between the numerical value and the statistical analysis results.

[0027] S222: Based on the statistical data summary, apply the mean and standard deviation and define the anomaly detection threshold as , and use the threshold to filter data points that exceed the normal range of change, mark them as abnormal fluctuations, and generate an abnormal fluctuation data set; After calculating the mean and standard deviation of the data, the next step is to use these statistical indicators to set the anomaly detection threshold. Specifically, the threshold is set to the mean plus two times the standard deviation, that is, . This method is based on the rule of normal distribution, in which about 95% of the data values ​​should be within two standard deviations of the mean. Applying this statistical rule to anomaly detection can effectively identify extreme values ​​that are outside the normal fluctuation range. In practice, this threshold is used to traverse the entire data set and check each data point: if the change in the data point exceeds this threshold, the point is marked as anomaly. This process not only helps to identify obvious outliers, but can also be used in subsequent steps to analyze the abnormal fluctuation pattern of data points, laying the foundation for further fault analysis.

[0028] S223: identifying an abnormal pattern corresponding to a potential fault signal according to the abnormal fluctuation data set, classifying the abnormal data through cluster analysis, and determining the potential fault pattern to form an abnormal pattern recognition result; After marking the abnormal data points, cluster analysis is then used to identify and classify the patterns of these abnormal fluctuations. In this process, clustering algorithms are used to group similar abnormal fluctuations, which helps to better understand the abnormal behavior in the data and its potential failure modes. Specifically, a suitable clustering algorithm (such as K-means or DBSCAN) is selected to process the abnormal points. Cluster analysis is performed based on the amplitude and frequency of abnormal fluctuations, and each cluster group represents a specific type of abnormal pattern. In this way, seemingly random abnormal points can be converted into meaningful information, revealing the potential causes of equipment failures or operational errors, which is crucial for the maintenance of the system and the formulation of preventive measures. The final failure mode recognition results will be directly used to optimize the operation and maintenance strategy.

[0029] See also Figure 5 , the steps to determine the frequent occurrence area are: S311: Based on the abnormal pattern recognition results, compared with historical failure cases, new patterns in the data similar to known failure patterns are revealed through pattern matching, and failure pattern similarity analysis results are generated; Before performing the failure mode similarity analysis, the historical failure data is first extracted, including the time, location, failure type, and operating conditions at the time of the failure. The data is extracted from multiple databases and stored in a unified data analysis platform. Then, through a detailed analysis of these historical failure data, especially the comparison of failure types and operating conditions, cluster analysis technology is used to group the failure cases according to similarity. The failure cases in each group have similar characteristics, such as similar operating conditions and failure manifestations. Next, the grouped data is used to identify possible failure modes in the current operating data through a decision tree algorithm. The analysis results are displayed through a graphical interface, allowing the maintenance team to intuitively see the potential failure modes and their characteristics, so as to take corresponding preventive measures in advance.

[0030] S312: Using the failure mode similarity analysis results, use the formula: ; Calculate the deviation between the occurrence frequency of the fault instance and the overall average frequency , measure the frequency of occurrence of failure modes in different operating environments and time periods, and obtain the frequent regional distribution data, where Indicates The frequency of occurrence of a fault instance, represents the average frequency of all fault instances. is the total number of failure instances; The following fault instances occur frequently Datasets: [15, 20, 22, 17, 19].

[0031] First calculate the mean of the data .

[0032] ; Compute the absolute difference between each data point and the mean and sum them: ; ; Substitute the sum into the formula and calculate : ; Calculation results The value of is 1.51, which indicates the deviation degree of the average occurrence frequency of each fault instance from the overall average frequency. A value of indicates that the data points are concentrated, that is, the fault types are more consistent or the fault frequencies in a specific area are similar. Conversely, a higher The value indicates that the data points are widely distributed, that is, the failure mode is dispersed, involving multiple different failure types or multiple areas. Through this analysis, the specific mode and frequent area of ​​failure can be more accurately located, so as to adjust the maintenance strategy and optimize resource allocation in a targeted manner.

[0033] S313: Using the frequent region distribution data, mapping the relationship between the fault mode and the corresponding frequent region is performed to establish a corresponding relationship table between the fault mode and the frequent region; After determining the frequent occurrence areas of each failure mode, the next step is to build a detailed mapping relationship table. This table not only records each failure mode, but also lists in detail the places and times where these modes most often occur, so as to help the maintenance team make targeted prevention and maintenance in future operations and maintenance. First, through further statistical analysis of the distribution data of frequent areas, the frequency of failures in each area is calculated, and the distribution pattern of frequent areas is identified through scatter plots. Then, the data is correlated with the operation log for analysis to confirm the relationship between the failure mode and the specific operating conditions. Finally, using the analysis results, a mapping relationship table is generated through an automated script. When similar failures occur in the future, maintenance personnel can be automatically reminded to pay attention to the relevant frequent areas, so as to quickly locate and solve the problem.

[0034] See also Figure 6 , the steps to obtain the fault location and cause analysis results are: S321: using the corresponding relationship table between the fault mode and the frequent area, by collecting and sorting the operation data in the frequent area, performing association rule mining, identifying the correlation between the frequent area and the fault mode, and generating the fault mode and operation condition correlation analysis result; When analyzing the failure mode in each frequent area, we first collect equipment operation data through on-site sensors, including physical parameters such as temperature, pressure, speed, and equipment operation logs to record each operation of the operator. The collected data is first stored in the local data warehouse, and then invalid and outliers are removed through a custom data cleaning script to ensure data quality. Next, SQL query statements are used to extract key operating parameters from the cleaned data set, and Excel's pivot table function is used to analyze the relationship between different parameters and the frequency of failures. Through detailed analysis, it is possible to identify key operating conditions that affect the occurrence of failures, such as the increase in failure rate when the equipment is running under high load, which not only helps to understand the environmental conditions where failures occur, but also guides the formulation of more effective preventive measures.

[0035] S322: Based on the results of the correlation analysis between the fault mode and the operating condition, a causal analysis is performed on the data to associate the potential factors with the fault instance, accurately locate the cause of the fault, and distinguish between the direct cause and the indirect cause of the fault to obtain the fault cause analysis result; In the process of determining the cause of the fault, the operation data in normal and faulty states were compared. First, the equipment logs before and after the fault were extracted from the maintenance database, and the logs were analyzed in time series to find out the trend of parameter changes before and after the fault. To this end, a Python script was written to use the Pandas library to process the time series data, calculate the rate of change of each parameter, and draw a trend chart through the Matplotlib library. Through the trend chart, it was found that the fault usually occurred after some sudden changes in parameters, such as a sharp increase in temperature or a sudden drop in pressure. Logistic regression was further used to analyze the relationship between these changes and the occurrence of the fault, and determine which parameter changes were the direct cause of the fault. The process helped to accurately locate the cause of the fault and provided a basis for formulating specific maintenance measures.

[0036] S323: Comprehensively analyze the association between the fault mode and the operating conditions and the fault cause analysis results, record the type and cause of each fault, display it on a dynamic dashboard, reflect the latest fault response information in real time, and generate fault location and cause analysis results; The process of generating a fault location and cause analysis report is to integrate the analysis results of the first two steps. Use Tableau to create a dynamic dashboard containing various fault data and analysis results. First, import all the analysis data into Tableau and set up multiple views to show different types of faults and their causes. Each view can show the frequency of occurrence, influencing factors and preventive measures of a specific type of fault. In addition, the dashboard also includes a real-time monitoring view that can display the current status of the equipment and any abnormal alarms. Through this dashboard, the maintenance team can instantly obtain the latest fault information and response guidelines, so as to quickly respond to and resolve faults and effectively reduce equipment downtime.

[0037] See also Figure 7 , the adjustment steps of the key areas are: S411: Based on the fault location and cause analysis results, analyze the characteristics and impact range of key faults, identify blind spots by comparing fault characteristics with the existing coverage capabilities of the monitoring system, and generate a key fault characteristic analysis report; Based on the results of fault location and cause analysis, a comprehensive system assessment is conducted to identify multiple types of faults that occur frequently and have serious impacts. The type analysis is obtained through the collection and statistics of historical fault data. The data is classified and cluster analysis methods are used to determine the mode and frequency of faults. In this way, the data collection blind spots in the monitoring system can be accurately identified. These blind spots are places where monitoring coverage was insufficient before. Through this process, a key fault characteristic analysis report is obtained, which lists in detail the specific characteristics of each fault type and the degree of its impact on the system, providing a scientific basis for the next step of monitoring process adjustment.

[0038] S412: Based on the key fault characteristic analysis report, redesign the data collection node layout and parameter settings, verify the integrity and accuracy of data collection through simulation tests, and obtain the data collection area adjustment plan; In the process of redesigning the layout and parameters of the data collection nodes, the previously identified blind spots were taken into consideration, and the locations and collection parameters of the monitoring points were adjusted according to the specific conditions of each blind spot. The completeness and accuracy of the data collection of the new layout were verified through simulation tests. During the test, simulated data was collected and multiple rounds of verification were carried out. The data was cleaned and analyzed to ensure that the collected data can truly reflect the situation when the fault occurred. Through these steps, a detailed data collection area adjustment plan was obtained, which ensured the maximization of the effectiveness and coverage of the monitoring system.

[0039] S413: Implement the data collection area adjustment plan, update the system's monitoring configuration file, and conduct on-site debugging to confirm the validity of each monitoring node for the target fault and obtain the dynamic adjustment record of the monitoring process; In the above content, when redesigning the data collection node, the formula is: , calculate the probability of occurrence P of a specific fault type, where is the number of times this fault type occurs during the monitoring period, and is the total number of monitoring times in the same period. Assuming that in the past monitoring cycle, a total of 1000 monitoring times were performed, and a specific fault type occurred 100 times, then and , put it into the formula and get , that is, the probability of this type of failure is 10%. This result shows that this type of failure is relatively common, and special attention should be paid to and the monitoring and prevention of such failures should be strengthened. Based on this result, the monitoring strategy and configuration can be further adjusted to ensure that such failures can be effectively prevented and responded to quickly.

[0040] See also Figure 8 , the steps to obtain the fault response detection scheme are: S421: dynamically adjust the records according to the monitoring process, conduct comprehensive tests, inject simulated fault data into the monitoring nodes, verify the response capability of the new monitoring to the predetermined faults, collect and evaluate the response data of the monitoring system, and generate a monitoring test report; The reconfigured monitoring system is started for comprehensive testing. During this process, a series of simulated fault scenarios are specially designed, covering various typical faults that the system may encounter, in order to comprehensively test the responsiveness of the new monitoring strategy. For each fault injection point, the fault must be correctly identified and responded to within the specified time. The real-time data analysis platform plays a key role in real-time monitoring of data flow, analysis of data quality, identification of fault modes, recording of system response time and fault handling results. Through this series of refined operations, a comprehensive system test report can be obtained. The report evaluates in detail the performance of the new configuration in actual operation, including the accuracy of fault identification, the system's response speed and the ability to handle faults. A detailed system test report is generated to help verify the performance of the new system and to guide further optimization of the system configuration.

[0041] S422: Analyze the monitoring test report and compare the performance of the new and old system configurations in terms of fault detection accuracy and response time, using the formula: ; Calculate the percentage increase in success rate , verify the improvement effect of the new configuration and obtain performance improvement data, among which, is the successful detection rate of the new configuration, is the successful detection rate of the old configuration; Success rate before upgrade is 70%, that is , the success rate after upgrading is 85%, that is The calculation process is as follows: Calculate the change in success rate: ; Calculate the lift percentage: ; The calculation results show that the fault detection success rate of the upgraded system has increased by about 21.43%. This shows that the new configuration not only improves the system's fault detection capability, but also significantly enhances the system's response efficiency to faults. This improvement rate is obtained through data collection and analysis in actual operations, reflecting the actual changes in system performance before and after the upgrade. This quantitative improvement shows the direct benefits of the system upgrade and provides data support for further system optimization and fault handling strategies.

[0042] S423: Based on the performance improvement data, formulate targeted response measures for target category faults, record the operation steps and expected effects under different fault conditions, and form a fault response detection plan; According to the performance improvement data, the fault response strategy is optimized and adjusted. In the process, a set of fault response processes are customized first according to the data analysis results of different fault types. This includes the identification standards and response measures for faults of different levels. For example, for low-level faults, it will automatically reset or send an alarm to the operator. For high-level complex faults, the emergency response procedure will be initiated to notify engineers to intervene. In addition, each response measure has been tested repeatedly to ensure that it can be implemented quickly and effectively in practice. Through continuous monitoring and feedback adjustment, the processing procedures and expected effects of various faults are recorded in detail, as well as the expected processing efficiency of the monitoring system after implementation, to ensure that the highest efficiency and stability can be maintained in the face of various expected and unexpected faults, and finally a fault response detection plan is formed.

[0043] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A nondestructive testing method for circuit board sensors based on big data, characterized in that: The following steps are involved: S1: Collect temperature, pressure, current and voltage data of key positions of circuit board sensors, send the data to the central database through real-time communication for aggregation, perform preliminary data screening simultaneously, verify the integrity and availability of the data, and form sensor summary data; S2: Based on the sensor summary data, the difference between each data point and the adjacent data points is calculated, and the change amplitude of the continuous data points is monitored. According to the change amplitude monitoring results, statistical analysis is performed to identify abnormal fluctuations in the data, mark the abnormal fluctuations and determine them as potential fault signals, and form an abnormal pattern recognition result; S3: using the abnormal pattern recognition result, comparing with the recorded fault cases, checking similar data patterns, determining the frequent occurrence area of ​​each fault mode, correlating the fault mode with the operating conditions according to the frequent occurrence area determination result, determining the fault type and actual cause, and generating fault location and cause analysis results; S4: Based on the fault location and cause analysis results, dynamically adjust the monitoring process, adjust the key areas of data collection according to the characteristics of the identified faults, perform system testing based on the key area adjustment results, confirm the effectiveness of the new settings for fault detection, and form a fault response detection plan.

2. The circuit board sensor nondestructive testing method based on big data according to claim 1 is characterized in that: The steps for obtaining the sensor summary data are as follows: S111: Perform real-time monitoring at key positions of circuit board sensors, collect and record temperature, pressure, current and voltage data, and generate real-time monitoring data records; S112: Based on the real-time monitoring data record, the data is checked and verified for correctness, and a preliminary format and integrity check is performed to obtain a verified data report; S113: Analyze and summarize the verified data reports, perform multi-dimensional data fusion, and form sensor summary data.

3. The circuit board sensor nondestructive testing method based on big data according to claim 2 is characterized in that: The monitoring steps of the variation range are: S211: extracting time series data according to the sensor summary data, performing data synchronization and alignment, verifying data continuity, and generating preprocessed continuous data point sequence data; S212: According to the pre-processed continuous data point sequence data, the formula: ； Calculate the difference between adjacent data points , get the data point difference value sequence data, where, represents the current data point in the sequence, Represents the next data point in the sequence; S213: Based on the data point difference value sequence data, perform a change range analysis, identify key change trends, and generate continuous data point change range analysis results.

4. The circuit board sensor nondestructive testing method based on big data according to claim 3 is characterized in that: The steps for obtaining the abnormal pattern recognition result are: S221: Based on the analysis results of the change range of the continuous data points, statistical analysis is performed using the formula: ； and ； Calculate the average and standard deviation , generates a summary of statistical data, where Represents the magnitude of change of each data point. is the total number of data points; S222: Based on the statistical data summary, apply the mean value and standard deviation to define anomaly detection threshold as , and use the threshold to filter data points that exceed the normal variation range, mark them as abnormal fluctuations, and generate an abnormal fluctuation data set; S223: According to the abnormal fluctuation data set, an abnormal pattern corresponding to a potential fault signal is identified, the abnormal data is classified through cluster analysis, and a potential fault pattern is determined to form an abnormal pattern recognition result.

5. The circuit board sensor nondestructive testing method based on big data according to claim 4 is characterized in that: The steps for determining the frequent occurrence area are: S311: Based on the abnormal pattern recognition result, compared with historical failure cases, new patterns in the data similar to known failure patterns are revealed through pattern matching, and failure pattern similarity analysis results are generated; S312: Using the failure mode similarity analysis result, adopt the formula: ； Calculate the deviation between the occurrence frequency of the fault instance and the overall average frequency , measure the frequency of occurrence of failure modes in different operating environments and time periods, and obtain the frequent regional distribution data, where Indicates The frequency of occurrence of a fault instance, represents the average frequency of all fault instances. is the total number of failure instances; S313: Using the frequent region distribution data, mapping the relationship between the fault mode and the corresponding frequent region is performed to establish a corresponding relationship table between the fault mode and the frequent region.

6. The circuit board sensor nondestructive testing method based on big data according to claim 5 is characterized in that: The steps for obtaining the fault location and cause analysis results are as follows: S321: using the corresponding relationship table between the fault mode and the frequent region, by collecting and sorting the operation data in the frequent region, performing association rule mining, identifying the correlation between the frequent region and the fault mode, and generating the fault mode and operation condition correlation analysis result; S322: Based on the analysis result of the association between the fault mode and the operating condition, a causal analysis is performed on the data, potential factors are associated with the fault instance, the cause of the fault is accurately located, and the direct cause and indirect cause of the fault are distinguished to obtain a fault cause analysis result; S323: Integrate the fault mode and operating condition association analysis results and the fault cause analysis results, record the type and cause of each fault, display it through a dynamic dashboard, reflect the latest fault response information in real time, and generate fault location and cause analysis results.

7. The circuit board sensor nondestructive testing method based on big data according to claim 6 is characterized in that: The adjustment steps of the key area are: S411: Based on the fault location and cause analysis results, analyze the characteristics and impact range of key faults, identify blind spots by comparing fault characteristics with the existing coverage capabilities of the monitoring system, and generate a key fault characteristic analysis report; S412: redesigning the data collection node layout and parameter settings according to the key fault characteristic analysis report, verifying the integrity and accuracy of data collection through simulation tests, and obtaining a data collection area adjustment plan; S413: Implement the data collection area adjustment plan, update the system's monitoring configuration file, and perform on-site debugging to confirm the validity of each monitoring node for the target fault and obtain dynamic adjustment records of the monitoring process.

8. The circuit board sensor nondestructive testing method based on big data according to claim 7 is characterized in that: The steps of obtaining the fault response detection scheme are: S421: dynamically adjust the records according to the monitoring process, perform comprehensive testing, inject simulated fault data into the monitoring node, verify the response capability of the new monitoring to the predetermined fault, collect and evaluate the response data of the monitoring system, and generate a monitoring test report; S422: Analyze the monitoring test report, compare the performance of the new and old system configurations in terms of fault detection accuracy and response time, and use the formula: ； Calculate the percentage increase in success rate , verify the improvement effect of the new configuration and obtain performance improvement data, among which, is the successful detection rate of the new configuration, is the successful detection rate of the old configuration; S423: Based on the performance improvement data, formulate targeted response measures for target category faults, record the operation steps and expected effects under different fault conditions, and form a fault response detection plan.

Images