Charging pile quality inspection data management method and system based on intelligent processing

Abstract

The present invention discloses a charging pile quality inspection data management method and system based on intelligent processing, which belongs to the field of detection and analysis technology. The system includes a data perception module, an association analysis module, a quality inspection and debugging module and a data storage module; the data perception module is used to collect the detection data of the charging pile to be inspected, as well as the quality inspection log of the charging pile of the same model; the association analysis module is used to parse the various indicators in the detection data and mine the association rules between the indicators; the quality inspection and debugging module generates a detection plan according to the association rules, and dynamically adjusts the test sequence in combination with the values ​​of the detected indicators during the quality inspection process; the data storage module is used to analyze the detection results of each indicator after the quality inspection is completed and provide them to the quality inspectors for confirmation, and store the generated quality inspection records in the quality inspection log. The present invention constructs an adaptive detection optimization mechanism by integrating design principles, historical data and dynamic association rules, solving the core problems of data islands, decision-making rigidity and resource waste in the existing technology.

Description

Technical Field

[0001] The present invention relates to the field of detection and analysis technology, and specifically to a charging pile quality inspection data management method and system based on intelligent processing. Background Art

[0002] With the rapid development of the new energy vehicle industry, charging piles, as a crucial component of electric vehicle infrastructure, have a direct impact on the popularity and user experience of electric vehicles. To ensure the safety, reliability, and efficient operation of charging piles, quality inspection is particularly important.

[0003] Currently, traditional charging pile quality inspection processes typically rely on manual testing or simple automated equipment. These processes face challenges such as low inspection efficiency, limited data processing capabilities, and information silos, making them difficult to meet the needs of large-scale, diversified charging pile management. On the one hand, existing inspection processes lack dynamic adaptability. Pre-set, fixed test sequences are unable to adjust the order of subsequent items based on real-time inspection results, resulting in a large number of redundant tests and significantly extending the quality inspection cycle for individual units. On the other hand, the lack of indicator correlation modeling and the disconnection between design schematics and historical quality inspection log data make it impossible to establish a transmission relationship model between indicators. Consequently, known test results cannot effectively derive the predicted values ​​of related indicators. These deficiencies collectively restrict quality inspection efficiency, making it difficult to meet the needs of large-scale industrial development. Therefore, a more intelligent and efficient charging pile quality inspection management technology solution is currently needed to address these issues. Summary of the Invention

[0004] The purpose of the present invention is to provide a charging pile quality inspection data management method and system based on intelligent processing to solve the problems raised in the above background technology.

[0005] In order to solve the above technical problems, the present invention provides a charging pile quality inspection data management method based on intelligent processing, comprising:

[0006] S100: Before starting quality inspection, collect inspection data of the charging pile to be inspected and quality inspection logs of charging piles of the same model.

[0007] The test data includes design schematics and test project libraries.

[0008] The design schematic provides the physical structure logic, such as the power transmission link in the circuit topology, the force transmission path of the mechanical transmission chain, and the thermal resistance network for thermal management.

[0009] The test item library includes different types of test items, each test item includes different types of indicators, and the value of each indicator is obtained by testing the test items one by one.

[0010] Quality inspection logs are records of each completed quality inspection. Each record includes the values ​​of various indicators. They store historical indicator values ​​and provide training samples for correlation analysis.

[0011] By integrating the static design data and dynamic historical data of charging piles, a complete quality inspection input foundation is built.

[0012] S200: Analyze the various indicators contained in the test data and mine the association rules between the indicators in combination with the quality inspection log. Specifically, it includes:

[0013] S201. Analyze the design schematic of the charging pile, extract the circuit topology, mechanical transmission chain, and thermal management path, and construct different types of impact paths, including electrical paths, mechanical paths, and thermal conduction paths.

[0014] After the impact path is constructed, the location and characteristics of each node are analyzed and indicators are labeled to the corresponding nodes. Different types of indicators are labeled on each node. Each node corresponds to at least one indicator, and each impact path is composed of several indicators.

[0015] The design schematic of the charging pile includes the circuit schematic, mechanical assembly diagram, heat source distribution and heat dissipation path.

[0016] Electrical path construction involves analyzing the circuit schematic and identifying key nodes in the power transmission chain, including the AC / DC rectifier module, DC / DC converter stage, output filter unit, and control feedback loop. The electrical signal transmission direction is defined, and the causal chain from control signal anomaly to final output failure is established to generate the electrical path.

[0017] Mechanical path construction involves decomposing the mechanical assembly diagram and extracting the sequence of key force transmission components, including the charging gun locking mechanism, plug-in spring assembly, and signal contact spring. Mechanical performance degradation transmission logic is then established to generate the mechanical path.

[0018] Thermal path construction involves identifying heat source distribution and heat dissipation paths, annotating the thermal resistance network from the power device junction temperature to the heat sink surface, and defining the cause-and-effect relationships of temperature conduction to generate the thermal path.

[0019] S202: Analyze all indicators in the test project library and map them to the impact paths. Combine all indicators in pairs to create indicator pairs, and mark the indicator pairs that exist on at least one impact path at the same time.

[0020] S203, filter out quality inspection records containing marking indicator pairs in the quality inspection log, fit the values ​​of the marking indicator pairs in these quality inspection records, and obtain the relational expressions corresponding to the marking indicator pairs. Specifically including:

[0021] S2031, statistics quality inspection log contains marking indicators The number of quality inspection records is m, and the two values ​​of the marking indicator pair in each quality inspection record are used as the dependent variable and the independent variable respectively, and are packaged as samples.

[0022] S2032, set intercept and the regression coefficient , and establish a linear regression model . Take the independent variables in these m samples as input values , the output value of each sample The difference between the value of the dependent variable and the dependent variable is taken as the gap coefficient.

[0023] S2033, adjust the intercept and regression coefficient until the sum of the gap coefficients of all samples is minimized. A relational expression.

[0024] S2034. Similarly, a linear regression model is established for each labeled indicator pair, and the relational expression is obtained through training.

[0025] S204: Analyze the relational expression to calculate the asynchronous coefficient of the corresponding marking indicator pair .

[0026] Statistical marker index pair The number of impact paths , calculate the standard deviation of the gap coefficient of all samples , substitute into the formula to calculate the marking index pair Asynchronous coefficient:

[0027] ;

[0028] Where, and is a constant, For the Marking indicator pairs on the impact path The number of indicators corresponding to the interval between two indicators, For the The number of all indicators on the impact path.

[0029] The higher the number of indicators between indicators, the weaker the impact of the spread effect, and vice versa. Normalized quantitative marking indicators influence the propagation effect on each impact path.

[0030] For the The gap coefficient of the sample, For the The asynchrony coefficient of each marked indicator pair is calculated in the same way.

[0031] The asynchrony coefficient is used to reveal hidden fault chains, and the analysis of strongly correlated indicators highlights the detection focus.

[0032] The asynchronous coefficient Less than threshold The marked indicator pair is used as the influencing object, and an association relationship is established between the two indicators within the influencing object.

[0033] Identify implicit dependencies between indicators, quantify fault propagation paths, and optimize detection priorities.

[0034] S300: Generate a test plan based on the association rules and adjust the test sequence in combination with the tested indicators during the quality inspection process. Specifically, it includes:

[0035] S301: Analyze various indicators in the test item library, mark indicators with related relationships, and calculate the priority index of each test item based on the asynchronous coefficient of the impact object to which the marked indicator belongs.

[0036] Priority Index The calculation formula is as follows:

[0037] ;

[0038] Where, The number of marking indicators in the detection project, To mark the number of affected objects of the indicator, For the test project The marker indicator belongs to The asynchronous coefficient of the affected object.

[0039] Projects with more association rules and lower AC values ​​have higher priority. Projects with higher PI values ​​are tested first, and projects with the same PI value are randomly sorted.

[0040] Each marker indicator belongs to at least one influencing object, and the asynchronous coefficient of each influencing object is less than the threshold .

[0041] S302. After all test items are sorted from largest to smallest according to priority index, a test plan is established, and quality inspectors conduct tests one by one according to the order in which the test items are arranged in the test plan.

[0042] For the test items with the same priority index, they are randomly sorted within a range, and the range means that they are after the test items with larger priority indexes and before the test items with smaller priority indexes.

[0043] S303, Real-time analysis of detected indicators during quality inspection Is it a marking indicator? If yes, it will be marked. Substitute the value of into the corresponding expression, calculate and mark the index Reference values ​​of other marker indicators that are correlated with each other.

[0044] S304: When there are at least e reference values ​​for the marker indicator, calculate the standard deviation of all reference values, take the marker indicator with a standard deviation less than the threshold as the detected indicator, and take the average of all reference values ​​as the predicted value of the corresponding marker indicator.

[0045] S305 , continue to substitute the predicted value of the marking indicator into the corresponding relationship expression, calculate the reference values ​​of other mutually related marking indicators, and iteratively set the detected indicators.

[0046] In real time, based on the values ​​of the detected marker indicators, the reference values ​​of other interrelated marker indicators are calculated, the detected indicators are set and the predicted values ​​are calculated, and the predicted values ​​of other marker indicators are further deduced from the predicted values, thus realizing the iterative cycle update of "prediction-setting". Specifically, it includes:

[0047] Step 1: Substitute the measured indicator values ​​into the regression model to predict the reference values ​​of the associated indicators.

[0048] Step 2: If the standard deviation of multiple reference values ​​is small, directly adopt the mean as the predicted value and mark it as "detected".

[0049] Step 3: Use the predicted value to continue to derive the next level of correlation indicators.

[0050] Solution update: Remove the measured / predicted indicators, recalculate the PI values ​​of the remaining projects and adjust the order.

[0051] S306: Remove the detected indicators from all test items and delete the test items that do not contain the indicators in the test plan. Dynamically calculate the priority index of each test item and adjust the order of the test items in the test plan in real time.

[0052] Achieve real-time optimization of the testing process and reduce redundant testing by replacing actual testing with prediction.

[0053] S400: After the quality inspection is completed, the test results of various indicators are analyzed, and a quality inspection record is generated and stored in the quality inspection log.

[0054] If there are no test items in the test plan, it means that the quality inspection is completed, and the value of the marked indicator is highlighted for the quality inspector to confirm.

[0055] Generate quality inspection records based on the values ​​of all indicators in this inspection plan and store them in the quality inspection log.

[0056] Highlight marked indicators: The indicators involved in the association rules need to be manually reviewed to avoid the accumulation of prediction errors.

[0057] Structured storage: Quality inspection records contain the measured / predicted values ​​of all indicators, stored in log format, and support historical data retrospective analysis.

[0058] Effect value: Build a closed-loop data pool to drive model iterative optimization.

[0059] Ensure that quality inspection results are traceable and improve the credibility of results through human-machine collaboration.

[0060] The present invention also provides a charging pile quality inspection data management system based on intelligent processing, including a data perception module, a correlation analysis module, a quality inspection and debugging module, and a data storage module.

[0061] The data perception module is used to collect the inspection data of the charging pile to be inspected, as well as the quality inspection logs of charging piles of the same model.

[0062] Before the quality inspection begins, the inspection data of the charging pile to be inspected is collected, including the design schematic and the inspection project library, and the historical quality inspection logs of the same model of charging pile are obtained.

[0063] By integrating the static design information of the current charging piles and the historical quality inspection dynamic data, a comprehensive input basis is provided for subsequent analysis, ensuring that the quality inspection process has reliable data support, reducing the risk of data missing, and improving the decision-making accuracy of subsequent modules.

[0064] The association analysis module is used to parse the various indicators and design schematics contained in the test data, and to mine the association rules between the indicators in combination with the quality inspection log.

[0065] Analyze the design schematic to construct the impact path, analyze the indicators in the inspection project library, establish indicator pairs and map them to the path, fit the relational expression based on the quality inspection log, and calculate the asynchronous coefficient AC to mine the association rules between indicators.

[0066] Intelligently identify dependencies and propagation effects between indicators, help discover potential failure modes, optimize detection plan generation, thereby improving the relevance and efficiency of quality inspection and reducing the burden of manual analysis.

[0067] The quality inspection and debugging module is used to generate a detection plan and set the test sequence according to the association rules, and dynamically adjust the test sequence in combination with the values ​​of the detected indicators during the quality inspection process.

[0068] Generate detection plans based on association rules, calculate the priority index PI of the detection items to sort the test order, use the values ​​of the detected indicators in real time to predict other indicators during the quality inspection process, and dynamically adjust the detection plan.

[0069] Dynamically optimize the test process, reduce redundant tests through prediction-setting iteration, shorten quality inspection time, improve resource utilization, and ensure the adaptability of the test sequence to cope with real-time data changes.

[0070] The data storage module is used to analyze the test results of various indicators after the quality inspection is completed and provide them to the quality inspectors for confirmation, and then generate quality inspection records and store them in the quality inspection log.

[0071] After the quality inspection is completed, the test results of all indicators are analyzed, the values ​​of the marked indicators are highlighted for confirmation by the quality inspectors, a structured quality inspection record is generated, and the record is stored in the quality inspection log.

[0072] Ensure the complete storage and traceability of quality inspection data, facilitate historical query and subsequent analysis, support data-driven decision optimization, and at the same time, by highlighting key indicators, assist quality inspectors to quickly verify results and improve overall quality inspection quality.

[0073] Data perception module: solves the problem of heterogeneity of multi-source data.

[0074] Association analysis module: combines physical paths with data fitting to break through the limitations of traditional statistical association.

[0075] Quality inspection and debugging module: Dynamic PI mechanism and prediction iteration to achieve adaptive resource allocation.

[0076] Data storage module: Highlights indicators through tagging, balancing automation and manual intervention.

[0077] Compared with the prior art, the present invention has the following beneficial effects:

[0078] Dynamically adaptively optimize the testing process: Dynamically sort test items through the priority index PI, and iteratively predict related indicators in real time based on the values ​​of tested indicators during the quality inspection process. Remove tested items and recalculate priorities, breaking through the rigid mode of traditional fixed-order testing and significantly reducing redundant testing.

[0079] Multi-source data-driven association rule mining: This system integrates design schematics (circuit / mechanical / thermal paths) with historical quality inspection logs to build a dual analysis framework of physical conduction chains and statistical models (linear regression). This system addresses decision-making blind spots caused by data fragmentation by quantifying implicit associations between asynchronous coefficients (AC).

[0080] The prediction iteration mechanism replaces redundant actual measurements: "Prediction-setting" iteration is implemented based on relational expressions: the measured indicator values ​​are used to derive the reference values ​​of related indicators. When the reference values ​​are highly consistent, they are directly adopted as the predicted values ​​and marked as "detected", avoiding repetitive testing and improving resource utilization.

[0081] Human-machine collaborative result verification mechanism: Highlights strongly correlated indicators (marked indicator pairs with low AC values) in quality inspection records for manual review to ensure the credibility of key data. At the same time, complete test results are stored in a structured manner to support historical tracing and model iterative optimization.

[0082] Systematic collaboration improves decision-making efficiency: The data perception module integrates multi-source inputs, the correlation analysis module establishes physical-data fusion rules, the quality inspection and debugging module realizes dynamic resource allocation, and the data storage module completes closed-loop verification, forming a complete chain from data collection to decision optimization, comprehensively improving quality inspection efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0083] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0084] Figure 1 This is a flow chart of the charging pile quality inspection data management method based on intelligent processing of the present invention;

[0085] Figure 2 It is a structural diagram of the charging pile quality inspection data management system based on intelligent processing of the present invention. DETAILED DESCRIPTION

[0086] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0087] See also Figure 1 The present invention provides a charging pile quality inspection data management method based on intelligent processing, including:

[0088] S100: Before starting quality inspection, collect inspection data of the charging pile to be inspected and quality inspection logs of charging piles of the same model.

[0089] The test data includes design schematics and test project libraries.

[0090] Design schematics (circuit schematics, mechanical assembly drawings, heat source distribution) provide physical structure logic, such as power transmission links in circuit topology, force transmission paths in mechanical transmission chains, and thermal resistance networks for thermal management.

[0091] The test item library includes different types of test items, each test item includes different types of indicators, and the value of each indicator is obtained by testing the test items one by one.

[0092] Quality inspection logs are records of each completed quality inspection. Each record includes the values ​​of various indicators. They store historical indicator values ​​and provide training samples for correlation analysis.

[0093] By integrating the static design data (schematic diagrams, test project library) and dynamic historical data (quality inspection logs) of charging piles, a complete quality inspection input foundation is built.

[0094] S200: Analyze the various indicators contained in the test data and mine the association rules between the indicators in combination with the quality inspection log. Specifically, it includes:

[0095] S201. Analyze the design schematic of the charging pile, extract the circuit topology, mechanical transmission chain, and thermal management path, and construct different types of impact paths, including electrical paths, mechanical paths, and thermal conduction paths.

[0096] After constructing the impact path, analyze the location and characteristics of each node and assign indicators to the corresponding nodes (e.g., "output voltage fluctuation" is mapped to the output filter unit node in the electrical path). Different types of indicators are assigned to each node, with each node corresponding to at least one indicator. Each impact path is composed of several indicators.

[0097] The design schematic of the charging pile includes the circuit schematic, mechanical assembly diagram, heat source distribution and heat dissipation path.

[0098] Electrical path construction involves analyzing the circuit schematic and identifying key nodes in the power transmission chain, including the AC / DC rectifier module, DC / DC converter stage, output filter unit, and control feedback loop. The electrical signal transmission direction is defined, and the causal chain from control signal anomaly to final output failure is established to generate the electrical path.

[0099] Mechanical path construction involves decomposing the mechanical assembly diagram and extracting the sequence of key force transmission components, including the charging gun locking mechanism, plug-in spring assembly, and signal contact spring. Mechanical performance degradation transmission logic is then established to generate the mechanical path.

[0100] Thermal path construction involves identifying heat source distribution and heat dissipation paths, annotating the thermal resistance network from the power device junction temperature to the heat sink surface, and defining the cause-and-effect relationships of temperature conduction to generate the thermal path.

[0101] S202: Analyze all indicators in the test project library and map them to the impact paths. Combine all indicators in pairs to create indicator pairs, and mark the indicator pairs that exist on at least one impact path at the same time.

[0102] S203, filter out quality inspection records containing marking indicator pairs in the quality inspection log, fit the values ​​of the marking indicator pairs in these quality inspection records, and obtain the relational expressions corresponding to the marking indicator pairs. Specifically including:

[0103] S2031, statistics quality inspection log contains marking indicators The number of quality inspection records is m, and the two values ​​of the marking indicator pair in each quality inspection record are used as the dependent variable and the independent variable respectively, and are packaged as samples.

[0104] S2032, set intercept and the regression coefficient , and establish a linear regression model . Take the independent variables in these m samples as input values , the output value of each sample The difference between the value of the dependent variable and the dependent variable is taken as the gap coefficient.

[0105] S2033, adjust the intercept and regression coefficient until the sum of the gap coefficients of all samples is minimized. A relational expression.

[0106] S2034. Similarly, a linear regression model is established for each labeled indicator pair, and the relational expression is obtained through training.

[0107] S204: Analyze the relational expression to calculate the asynchronous coefficient of the corresponding marking indicator pair .

[0108] Statistical marker index pair The number of impact paths , calculate the standard deviation of the gap coefficient of all samples , substitute into the formula to calculate the marking index pair Asynchronous coefficient:

[0109] ;

[0110] Where, and is a constant, For the Marking indicator pairs on the impact path The number of indicators corresponding to the interval between two indicators, For the The number of all indicators on the impact path.

[0111] The higher the number of indicators between indicators, the weaker the impact of the spread effect, and vice versa. Normalized quantitative marking indicators influence the propagation effect on each impact path.

[0112] For the The gap coefficient of the sample, For the The asynchrony coefficient of each marked indicator pair is calculated in the same way.

[0113] The asynchronous coefficient is used to reveal hidden fault chains (such as mechanical wear leading to increased contact resistance), and the analysis of strongly correlated indicators highlights the detection focus.

[0114] The asynchronous coefficient Less than threshold The marked indicator pair is used as the influencing object, and an association relationship is established between the two indicators within the influencing object.

[0115] Identify implicit dependencies between indicators, quantify fault propagation paths, and optimize detection priorities.

[0116] S300: Generate a test plan based on the association rules and adjust the test sequence in combination with the tested indicators during the quality inspection process. Specifically, it includes:

[0117] S301: Analyze various indicators in the test item library, mark indicators with related relationships, and calculate the priority index of each test item based on the asynchronous coefficient of the impact object to which the marked indicator belongs.

[0118] Priority Index The calculation formula is as follows:

[0119] ;

[0120] Where, The number of marking indicators in the detection project, To mark the number of affected objects of the indicator, For the test project The marker indicator belongs to The asynchronous coefficient of the affected object.

[0121] The more association rules there are ( Projects with higher PI values ​​will be tested first, and projects with the same PI value will be randomly sorted.

[0122] Each marker indicator belongs to at least one influencing object, and the asynchronous coefficient of each influencing object is less than the threshold .

[0123] S302. After all test items are sorted from largest to smallest according to priority index, a test plan is established, and quality inspectors conduct tests one by one according to the order in which the test items are arranged in the test plan.

[0124] For the test items with the same priority index, they are randomly sorted within a range, and the range means that they are after the test items with larger priority indexes and before the test items with smaller priority indexes.

[0125] S303, Real-time analysis of detected indicators during quality inspection Is it a marking indicator? If yes, it will be marked. Substitute the value of into the corresponding expression, calculate and mark the index Reference values ​​of other marker indicators that are correlated with each other.

[0126] S304: When there are at least e reference values ​​for the marker indicator, calculate the standard deviation of all reference values, take the marker indicator with a standard deviation less than the threshold as the detected indicator, and take the average of all reference values ​​as the predicted value of the corresponding marker indicator.

[0127] S305 , continue to substitute the predicted value of the marking indicator into the corresponding relationship expression, calculate the reference values ​​of other mutually related marking indicators, and iteratively set the detected indicators.

[0128] In real time, based on the values ​​of the detected marker indicators, the reference values ​​of other interrelated marker indicators are calculated, the detected indicators are set and the predicted values ​​are calculated, and the predicted values ​​of other marker indicators are further deduced from the predicted values, thus realizing the iterative cycle update of "prediction-setting". Specifically, it includes:

[0129] Step 1: Substitute the measured indicator value (such as "heat sink temperature") into the regression model to predict the reference value of the related indicator (such as "power device junction temperature").

[0130] Step 2: If the standard deviation of multiple reference values ​​is small (high consistency), directly adopt the mean as the predicted value and mark it as "detected".

[0131] Step 3: Use the predicted value to continue to derive the next level of related indicators (such as "junction temperature" to predict "efficiency loss").

[0132] Solution update: Remove the measured / predicted indicators, recalculate the PI values ​​of the remaining projects and adjust the order.

[0133] S306: Remove the detected indicators from all test items and delete the test items that do not contain the indicators in the test plan. Dynamically calculate the priority index of each test item and adjust the order of the test items in the test plan in real time.

[0134] Achieve real-time optimization of the testing process and reduce redundant testing by replacing actual testing with prediction.

[0135] S400: After the quality inspection is completed, the test results of various indicators are analyzed, and a quality inspection record is generated and stored in the quality inspection log.

[0136] If there are no test items in the test plan, it means that the quality inspection is completed, and the value of the marked indicator is highlighted for the quality inspector to confirm.

[0137] Generate quality inspection records based on the values ​​of all indicators in this inspection plan and store them in the quality inspection log.

[0138] Highlighting marked indicators: Indicators involved in association rules (such as the "temperature rise-current" pair with low AC value) need to be manually reviewed to avoid accumulation of prediction errors.

[0139] Structured storage: Quality inspection records contain the measured / predicted values ​​of all indicators, stored in log format, and support historical data retrospective analysis.

[0140] Effect value: Build a closed-loop data pool to drive model iterative optimization (such as updating regression model parameters).

[0141] Ensure that quality inspection results are traceable and improve the credibility of results through human-machine collaboration.

[0142] See also Figure 2 The present invention also provides a charging pile quality inspection data management system based on intelligent processing, including a data perception module, a correlation analysis module, a quality inspection and debugging module and a data storage module.

[0143] The data perception module is used to collect the inspection data of the charging pile to be inspected, as well as the quality inspection logs of charging piles of the same model.

[0144] Before the quality inspection begins, the inspection data of the charging pile to be inspected is collected, including the design schematics (circuit schematics, mechanical assembly drawings and heat source distribution) and the inspection project library (containing different types of inspection items, each item involving multiple indicator types). At the same time, the historical quality inspection logs of the same model of charging piles are obtained (each log records the value of each indicator).

[0145] By integrating the static design information of the current charging piles and the historical quality inspection dynamic data, a comprehensive input basis is provided for subsequent analysis, ensuring that the quality inspection process has reliable data support, reducing the risk of data missing, and improving the decision-making accuracy of subsequent modules.

[0146] The association analysis module is used to parse the various indicators and design schematics contained in the test data, and to mine the association rules between the indicators in combination with the quality inspection log.

[0147] Analyze the design schematic to construct the impact path (electrical path, mechanical path, and thermal conduction path), analyze the indicators in the inspection project library, establish indicator pairs and map them to the paths, fit the relationship expression (linear regression model) based on the quality inspection log, and calculate the asynchronous coefficient AC to explore the association rules between indicators (establish an association relationship when AC is less than the threshold).

[0148] Intelligently identify dependencies and propagation effects between indicators, help discover potential failure modes, optimize detection plan generation, thereby improving the relevance and efficiency of quality inspection and reducing the burden of manual analysis.

[0149] The quality inspection and debugging module is used to generate a detection plan and set the test sequence according to the association rules, and dynamically adjust the test sequence in combination with the values ​​of the detected indicators during the quality inspection process.

[0150] Generate detection plans based on association rules, calculate the priority index PI of the detection items to sort the test order, use the values ​​of the detected indicators in real time to predict other indicators during the quality inspection process (iteratively calculate the reference value and predicted value through relational expressions), and dynamically adjust the detection plan (remove the detected indicators and recalculate the priority index).

[0151] Dynamically optimize the test process, reduce redundant tests through prediction-setting iteration, shorten quality inspection time, improve resource utilization, and ensure the adaptability of the test sequence to cope with real-time data changes.

[0152] The data storage module is used to analyze the test results of various indicators after the quality inspection is completed and provide them to the quality inspectors for confirmation, and then generate quality inspection records and store them in the quality inspection log.

[0153] After the quality inspection is completed, the test results of all indicators are analyzed, the values ​​of the marked indicators are highlighted for confirmation by the quality inspector, a structured quality inspection record (including the values ​​of all indicators) is generated, and the record is stored in the quality inspection log.

[0154] Ensure the complete storage and traceability of quality inspection data, facilitate historical query and subsequent analysis, support data-driven decision optimization, and at the same time, by highlighting key indicators, assist quality inspectors to quickly verify results and improve overall quality inspection quality.

[0155] Data perception module: solves the heterogeneity problem of multi-source data (design drawings and logs).

[0156] Association analysis module: combines physical paths with data fitting to break through the limitations of traditional statistical association.

[0157] Quality inspection and debugging module: Dynamic PI mechanism and prediction iteration to achieve adaptive resource allocation.

[0158] Data storage module: Highlights indicators through tagging, balancing automation and manual intervention.

[0159] Example 1: Assume that there are marking indicators A1 and A2 in the detection project, and the asynchronous coefficient of the affected object to which each marking indicator belongs is as follows:

[0160] Marking index A1: Influence object G1: 0.35; Influence object G2: 0.42;

[0161] Marking index A2: Influence object G3: 0.25; Influence object G4: 0.76;

[0162] When the threshold of the asynchronous coefficient is 1, substitute the formula to calculate the priority index of the detection item:

[0163] ;

[0164] The priority index of the test item is 10.55.

[0165] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0166] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art will be able to modify the technical solutions described in the aforementioned embodiments or substitute equivalents for some of the technical features. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. A charging pile quality inspection data management method based on intelligent processing, characterized by: The method includes: S100. Before starting quality inspection, collect inspection data of the charging pile to be inspected and quality inspection logs of charging piles of the same model; S200, analyzing various indicators contained in the test data, and mining association rules between the indicators in combination with the quality inspection log; S300, generating a test plan based on the association rules, and adjusting the test sequence in combination with the tested indicators during the quality inspection process; S400: After the quality inspection is completed, the test results of various indicators are analyzed, and a quality inspection record is generated and stored in the quality inspection log; In S100, the test data includes the design schematics and the test project library; The test item library includes different types of test items, and each test item includes different types of indicators; The values ​​of various indicators are obtained by analyzing each test item one by one; The quality inspection log refers to the quality inspection record after each quality inspection is completed. Each quality inspection record includes the values ​​of various indicators; S200 includes: S201. Analyze the design schematic of the charging pile, extract the circuit topology, mechanical transmission chain, and thermal management path, and construct different types of influencing paths, including electrical paths, mechanical paths, and thermal conduction paths. S202: Analyze all indicators in the test project library and map them to the impact paths. Combine all indicators in pairs to create indicator pairs, and mark the indicator pairs that exist on at least one impact path at the same time. S203, screening out quality inspection records containing marked indicator pairs from the quality inspection log, fitting the values ​​of the marked indicator pairs in these quality inspection records, and obtaining a relational expression corresponding to the marked indicator pairs; S204: Analyze the relational expression to calculate the asynchronous coefficient of the corresponding marking indicator pair ; Set the asynchronous coefficient Less than threshold The marked indicator pair is used as the influencing object, and an association relationship is established between the two indicators within the influencing object.

2. The charging pile quality inspection data management method based on intelligent processing according to claim 1 is characterized in that: In S201, the design schematic diagram of the charging pile includes a circuit schematic diagram, a mechanical assembly diagram, a heat source distribution, and a heat dissipation path; The electrical path construction includes: parsing the circuit schematic and identifying the key nodes in the power transmission link, including the AC / DC rectifier module, DC / DC converter stage, output filter unit and control feedback loop; Define the direction of electrical signal conduction and establish a causal chain from control signal anomaly to final output failure, thereby generating an electrical path; Mechanical path construction includes: decomposing the mechanical assembly diagram and extracting the sequence of key force transmission components, including the charging gun locking mechanism, plug-in spring assembly, and signal contact spring; Establishing the logic of mechanical performance degradation transmission to generate mechanical pathways; The construction of heat conduction path includes: identifying the heat source distribution and heat dissipation path, marking the thermal resistance network from the junction temperature of the power device to the heat sink surface; Define the temperature conduction cause and effect relationship to generate the heat conduction path.

3. The charging pile quality inspection data management method based on intelligent processing according to claim 1 is characterized in that: S203 includes: S2031, statistics quality inspection log contains marking indicators The number of quality inspection records is m, and the two values ​​of the marking indicator pair in each quality inspection record are used as the dependent variable and the independent variable respectively, and are packaged as a sample; S2032, set intercept and the regression coefficient , and establish a linear regression model ; Take the independent variables in these m samples as input values , the output value of each sample The difference from the dependent variable is taken as the gap coefficient; S2033, adjust the intercept and regression coefficient until the sum of the gap coefficients of all samples is minimized; obtain the training completed marking indicator pair relational expressions; S2034. Similarly, a linear regression model is established for each labeled indicator pair, and the relational expression is obtained through training.

4. The charging pile quality inspection data management method based on intelligent processing according to claim 3 is characterized by: In S204, the statistical marking index is The number of impact paths , calculate the standard deviation of the gap coefficient of all samples ; Calculate the marking index pair according to the formula Asynchronous coefficient: ； Where, and is a constant, For the Marking indicator pairs on the impact path The number of indicators corresponding to the interval between two indicators, For the The number of all indicators on the impact path; For the The gap coefficient of the sample, For the The dependent variable of the sample; And so on, the asynchronous coefficient of each marking indicator pair is calculated separately.

5. The charging pile quality inspection data management method based on intelligent processing according to claim 1 is characterized in that: S300 includes: S301, analyzing various indicators in the test item library, marking indicators with associated relationships; calculating the priority index of each test item based on the asynchronous coefficient of the affected object to which the marked indicator belongs; S302. After all test items are sorted from largest to smallest according to their priority index, a test plan is established. The quality inspector then performs tests one by one according to the order in which the test items are arranged in the test plan. S303, Real-time analysis of detected indicators during quality inspection Is it a marking indicator? If yes, it will be marked. Substitute the value of into the corresponding expression, calculate and mark the index Reference values ​​of other interrelated marker indicators; S304. When there are at least e reference values ​​for the marker indicator, calculate the standard deviation of all reference values; take the marker indicator with a standard deviation less than the threshold as the detected indicator, and take the average of all reference values ​​as the predicted value of the corresponding marker indicator; S305, continue substituting the predicted value of the marking indicator into the corresponding relationship expression, calculate the reference value of other related marking indicators, and iteratively set the detected indicators; S306: Remove the detected indicators from all detection items and delete the detection items that do not contain indicators in the detection plan; dynamically calculate the priority index of each detection item and adjust the arrangement order of the detection items in the detection plan in real time.

6. The charging pile quality inspection data management method based on intelligent processing according to claim 5 is characterized by: In S301, priority index The calculation formula is as follows: ； Where, The number of marking indicators in the detection project, To mark the number of affected objects of the indicator, For the test project The marker indicator belongs to The asynchronous coefficient of the affected object.

7. The charging pile quality inspection data management method based on intelligent processing according to claim 5 is characterized by: In S400 , if there is no test item in the test plan, it means that the quality inspection is completed, and the value of the marked indicator is highlighted for the quality inspector to confirm; Generate quality inspection records based on the values ​​of all indicators in this inspection plan and store them in the quality inspection log.

8. A charging pile quality inspection data management system based on intelligent processing, applied to the charging pile quality inspection data management method based on intelligent processing as claimed in claim 1, characterized in that: The system includes data perception module, correlation analysis module, quality inspection and debugging module and data storage module; The data perception module is used to collect the test data of the charging pile to be inspected, as well as the quality inspection log of the charging pile of the same model; The association analysis module is used to analyze the various indicators and design schematics contained in the test data, and to mine the association rules between indicators in combination with the quality inspection log; The quality inspection and debugging module is used to generate inspection plans and set the test sequence based on association rules, and dynamically adjust the test sequence based on the values ​​of the inspected indicators during the quality inspection process; The data storage module is used to analyze the test results of various indicators after the quality inspection is completed and provide them to the quality inspectors for confirmation, and then generate quality inspection records and store them in the quality inspection log.

Images