Full-process quality control system for automobile die based on hierarchical quality gate

Abstract

The application provides a full-process quality control system for automobile molds based on hierarchical quality gates, and relates to the technical field of automobile molds.The system realizes phased, quantifiable and closed-loop quality control of the whole mold manufacturing process by constructing multiple progressive quality gates.The system sets up quality gates with data acquisition, intelligent judgment and feedback control functions at each key stage of assembly, initial adjustment, final adjustment and acceptance, and can automatically evaluate the current quality state at the end of each stage.The first stage ensures that the structure assembly meets the design requirements, the second stage controls the process release according to the precision compliance rate and problem closing rate, the third stage emphasizes the continuous improvement of precision and the complete closed loop of problems, and the fourth stage verifies whether the final product meets the standard of the host factory from the aspects of forming performance and appearance quality.This hierarchical and progressive control mechanism effectively realizes early detection, early interception and early rectification of quality problems, and improves the controllability and consistency of the mold manufacturing process.

Description

Technical Field

[0001] This application relates to the field of automotive mold technology, and in particular to a full-process quality control system for automotive molds based on graded quality gates. Background Technology

[0002] In the automotive mold manufacturing industry, ensuring that molds meet high-quality standards at every stage, from design to final acceptance, is crucial. However, existing technologies often lack a systematic approach to achieve end-to-end quality control, particularly in effectively transmitting quality information between different stages and ensuring timely correction of problems. Current practices typically rely on post-production inspections, lacking the ability to monitor and predict potential quality fluctuations during production in real time. This results in quality issues not being effectively controlled in the early stages, increasing the cost and difficulty of later rectification. Summary of the Invention

[0003] This application provides a quality control system for the entire process of automotive molds based on graded quality gates, in order to solve one or more technical problems existing in the prior art, and at least provide a beneficial option or create conditions.

[0004] On the one hand, this application provides a full-process quality control system for automotive molds based on graded quality gates, including multi-level progressive quality gates. Each quality gate includes a data acquisition terminal, a quality judgment module, and an execution feedback interface, which are used to perform phased technical verification and automatic control of the entire mold manufacturing process.

[0005] The first-level quality gate is used to detect the mold closing accuracy and the fit status of the guide components online through the sensing device during the mold assembly stage, and to verify the assembly conformity of the mold structural components in combination with the design model.

[0006] The second-level quality gate is used in the initial mold debugging stage to evaluate the accuracy based on the measurement data of key dimensions, and to determine whether the quality conditions for entering the next stage are met based on the degree of accuracy compliance and the completion rate of rectification of identified problems.

[0007] The third-level quality gate is used to dynamically evaluate the mold accuracy based on the retest data during the final mold debugging stage, and outputs control permission to enter the next manufacturing stage under the condition that the preset accuracy improvement target and quality problem closed-loop verification are met.

[0008] The fourth-level quality gate is used during the acceptance phase to comprehensively evaluate stamped parts through forming performance testing and surface quality analysis, and to generate an acceptance report.

[0009] Furthermore, the sensing device of the first-level quality gate includes a mold closing accuracy sensor and a guide component fit status detector, which are used to detect the mold closing accuracy, guide plate gap and standard component compatibility online, and compare the detection data with the design model to verify the assembly conformity of the mold structure components.

[0010] Furthermore, the accuracy evaluation criteria for the second-level quality gate include: the accuracy report compliance rate of key dimension measurement data is not lower than the preset first compliance rate threshold, and the rectification completion rate of identified quality problems is not lower than the preset completion rate threshold; the execution feedback interface outputs stage release or rectification instructions based on the evaluation results.

[0011] Furthermore, the dynamic evaluation indicators of the third-level quality gate include: the accuracy report compliance rate of the retest data is not lower than the preset second compliance rate threshold; the quality judgment module outputs the final adjustment stage release permission when the above conditions are met.

[0012] Furthermore, the forming performance test of the fourth-level quality gate includes forming limit curve testing, and the surface quality analysis includes light and shadow detection technology, which is used to evaluate the forming limit and surface defects of the stamped parts, and generate an acceptance report including the OEM's vehicle assembly standard compliance conclusion.

[0013] Furthermore, the data acquisition terminal includes a coordinate measuring machine, a 3D scanner, and a pressure sensor, used to collect the clamping force during the mold assembly stage, the travel of the blank holder and the amount of sheet metal flowing in during the initial and final adjustment stages, and the surface contour data of the stamped parts during the acceptance stage.

[0014] Furthermore, the machine learning model in the quality judgment module takes the key parameters acquired by the data acquisition terminal as input, trains and generates a quality score prediction model through historical quality data, and has the functions of anomaly detection, root cause analysis and optimization suggestion output; the key parameters include the pressure ring stroke, sheet material flow rate and guide plate gap.

[0015] Furthermore, the execution feedback interface is linked with the mold manufacturing execution system. When the quality judgment module detects that the parameter fluctuation exceeds the preset threshold, it automatically triggers process parameter adjustment instructions, including blank holder force optimization, drawbead structure adjustment, and standard part replacement prompts.

[0016] Furthermore, each quality gate also includes a problem tracing module. The problem tracing module uses root cause analysis to locate the root cause of problems that fail to pass the quality gate, and synchronizes the rectification plan and verification results to the technical database to form a knowledge base for preventing recurrence.

[0017] Furthermore, the judgment logic of the multi-level progressive quality gate adopts the statistical process control method to monitor the fluctuation trend of quality data at each stage in real time. When the process capability index is less than the preset capability threshold, a quality warning is automatically activated and the next stage of the production process is frozen.

[0018] This application provides a quality control system for the entire automotive mold manufacturing process based on hierarchical quality gates. By constructing multi-level progressive quality gates, it achieves phased, quantifiable, and closed-loop quality control throughout the mold manufacturing process. The system sets up quality gates with data acquisition, intelligent judgment, and feedback control functions at each key stage of assembly, initial adjustment, final adjustment, and acceptance. It can automatically assess the current quality status at the end of each stage: the first level ensures that the structural assembly meets design requirements; the second level controls the process release based on the accuracy compliance rate and problem closure rate; the third level emphasizes continuous improvement in accuracy and complete problem closure; and the fourth level verifies whether the final product meets the OEM's standards from the perspective of forming performance and appearance quality. This hierarchical, progressive control mechanism effectively enables early detection, early interception, and early rectification of quality problems, improving the controllability, consistency, and first-pass yield of the mold manufacturing process.

[0019] Other features and advantages of this application will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the application. The objectives and other advantages of this application may be realized and obtained by means of the structures particularly pointed out in the description, claims and drawings. Attached Figure Description

[0020] The accompanying drawings are provided to further understand the technical solutions of the present invention and constitute a part of the specification. They are used together with the embodiments of the present invention to explain the technical solutions of the present invention, and do not constitute a limitation on the technical solutions of the present invention.

[0021] Figure 1 This is a structural diagram of the automotive mold end-to-end quality control system based on graded quality gates provided in this application;

[0022] Figure 2 This is a structural diagram of the quality gate provided in this application. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0024] The present application will be further described below with reference to the accompanying drawings and specific embodiments. The described embodiments should not be considered as limitations on the present application, and all other embodiments obtained by those skilled in the art without inventive effort are within the scope of protection of the present application.

[0025] In the following description, references are made to “some embodiments,” which describe a subset of all possible embodiments. However, it is understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict.

[0026] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only and is not intended to limit this application.

[0027] In the automotive manufacturing supply chain, molds, as core process equipment in stamping production, directly determine the dimensional accuracy and surface quality of body panels, as well as the assembly and matching performance of the entire vehicle. With the automotive industry's ever-increasing demands for lightweighting, high safety, and aesthetic quality, the design and manufacturing of automotive molds face even greater technical challenges. This is especially true for large and complex body panel molds (such as side panels, doors, and hoods), which have complex structures, high precision requirements, and long manufacturing cycles, involving multiple stages including design, processing, assembly, debugging, trial molding, and acceptance. Quality deviations at any stage can lead to serious subsequent quality problems, resulting in frequent rework, increased costs, delivery delays, and even impacting the overall vehicle project schedule.

[0028] Currently, in the automotive mold manufacturing industry, a control method combining phased quality inspection and manual review is commonly used. For example, after mold assembly, a mold closing inspection is conducted; during the initial debugging stage, a first-piece trial stamp is performed and key dimensions are measured; during the final debugging stage, multiple mold trials are conducted for optimization; and finally, the OEM or a third-party organization conducts vehicle assembly review and acceptance. These processes typically rely on experienced technicians for visual inspection, manual measurement (such as using calipers, micrometers, coordinate measuring machines, etc.), and subjective judgment, lacking unified, quantifiable, and traceable quality judgment standards. Although some companies have introduced information management systems (such as MES and PLM) to record process flows and problem lists, most systems only realize data storage and process flow, failing to deeply integrate with the physical parameters of the actual production process, and also failing to achieve automated quality assessment and process intervention. Therefore, the main shortcomings of the existing technology are reflected in the following aspects.

[0029] First, quality control lacks systematicity and progression, making it difficult to form a closed-loop management system. Traditional quality inspections are mostly "point-based" or "fragmented" operations, with inconsistent quality requirements between stages and a lack of clear release standards. For example, whether the assembly stage is qualified is often judged solely by whether the mold closing gap is uniform, without establishing a connection with subsequent forming performance; after problems are discovered in the initial debugging stage, the rectification process lacks a tracking mechanism, easily leading to situations where "problems are recorded but not closed" or "superficially closed but actually recurring." Due to the lack of a progressive, layered quality gate control mechanism, hidden dangers in lower-level stages are carried over to higher-level stages, ultimately erupting during acceptance, resulting in a large amount of rework and wasted resources.

[0030] Secondly, quality assessment relies heavily on human experience, resulting in low standardization and high subjectivity. Currently, many mold-making companies still primarily rely on engineers' experience-based judgment for quality evaluation, lacking data-driven objective evaluation models. For example, issues such as whether the guide plate clearance is reasonable, whether the blank holder movement is smooth, and whether the drawing process is sufficient are often judged by "feel" or "visual inspection," lacking real-time, online sensing and monitoring methods. Even when using coordinate measuring machines to acquire data, it is mostly offline sampling inspection, with delayed feedback, making dynamic control during the process impossible. This human-driven model is not only inefficient but also exhibits significant variations among different personnel and across different projects, affecting the consistency and repeatability of product quality.

[0031] Third, process data acquisition is isolated, failing to achieve effective integration and intelligent analysis. Although modern mold workshops are equipped with advanced testing tools, the data generated by these devices is often scattered across different systems, failing to be effectively integrated with quality judgment logic. For example, key forming parameters such as blank holder stroke, sheet metal flow rate, and clamping force can be collected, but are usually only used for post-process analysis and not involved in real-time quality assessment and process adjustment. Furthermore, the lack of a systematic attribution analysis and knowledge accumulation mechanism for historical quality problems leads to the recurrence of the same types of problems in different projects, failing to achieve "recurrence prevention."

[0032] Fourth, the lack of automatic feedback and control capabilities prevents dynamic optimization. Most existing systems remain in a passive response mode, lacking the ability to automatically translate detection results into process adjustment commands. For example, when uneven sheet metal flow is detected in a certain area, the system cannot automatically suggest adjusting the drawbead height or optimizing the blank holder force distribution; when the mold closing accuracy deviation exceeds a threshold, it cannot trigger an early warning or freeze subsequent processes. This static and lagging control method is ill-suited to the high-precision, fast-paced demands of modern mold manufacturing.

[0033] Fifth, the acceptance criteria are out of touch with the requirements of the vehicle manufacturer, and the risk forward shift is insufficient. The mold ultimately needs to meet the vehicle loading standards of the vehicle manufacturer, including dimensional matching, surface appearance, forming stability, etc. However, many current mold enterprises do not introduce evaluation methods such as forming limit test (FLC) and light and shadow inspection in the early stage, which can truly reflect the forming performance and surface quality of stamping parts. As a result, the mold seems qualified during in-factory debugging, but still exposes problems such as cracking, wrinkling, and excessive springback on the actual production line of the vehicle manufacturer. This shows that the existing quality verification system fails to shift the terminal usage scenario forward to the manufacturing process and lacks the comprehensive prediction ability for the performance of the final product.

[0034] To sum up, the existing automotive mold quality control technologies generally have problems such as systematic deficiencies, vague standards, data fragmentation, response lags, and lack of closed-loop. It is difficult to meet the growing high-quality, high-efficiency, and high-consistency manufacturing requirements of high-end molds.

[0035] This application proposes a full-process quality control system for automotive molds based on hierarchical quality gates. The core lies in constructing a multi-level progressive quality control system covering the entire life cycle of mold manufacturing. By setting integrated quality gates at key process stages, the transformation from passive inspection to active control is achieved. Each quality gate integrates a data acquisition terminal, a quality judgment module, and an execution feedback interface, forming an integrated closed-loop control mechanism: In the assembly stage, a sensing device is used to online monitor key parameters such as mold closing accuracy and guide plate clearance, and automatically verify the compliance of structural part assembly in combination with the design model; In the initial adjustment and final adjustment stages, based on the key dimension data obtained by coordinate measuring machines, 3D scanning, etc., by setting quantitative indicators such as accuracy compliance rate and problem rectification completion rate, the mold performance is dynamically evaluated and a decision is made whether to release it into the next stage; In the acceptance stage, evaluation methods such as forming limit test (FLC) and light and shadow inspection are introduced to comprehensively evaluate the forming ability and surface quality of stamping parts, ensuring that the vehicle loading requirements of the vehicle manufacturer are met. The system deeply integrates data acquisition and intelligent analysis technologies. The quality judgment module is embedded with a machine learning model, which can predict quality trends, identify anomalies, conduct root cause analysis, and provide optimization suggestions based on historical data. At the same time, it combines statistical process control (SPC) to monitor the process capability in real time. Once the fluctuation exceeds the limit, it automatically triggers an alarm or freezes the process. In addition, the system is also equipped with a problem traceability module, which uses root cause analysis methods to locate the root cause of quality problems and incorporates the rectification measures into the knowledge base to promote the continuous update of standards and effectively prevent the recurrence of similar problems. The entire system realizes the standardization, automation, dataization, and intelligence of quality control, significantly improving the controllability, consistency, and delivery quality of the mold manufacturing process.

[0036] First, the full-process quality control system for automotive molds based on hierarchical quality gates provided by the embodiments of this application will be elaborated in detail below with reference to the accompanying drawings.

[0037] Reference Figure 1 and Figure 2 The automotive mold full-process quality control system based on hierarchical quality gates provided in this application includes multi-level progressive quality gates. Each quality gate includes a data acquisition terminal, a quality judgment module, an execution feedback interface, and a problem traceability module, which are used to perform phased technical verification and automatic control of the entire mold manufacturing process.

[0038] In some embodiments of this application, a first-level quality gate is used to perform online detection of mold closing accuracy and guide component fit status through sensing devices during the mold assembly stage, and to verify the assembly conformity of mold structural components in conjunction with the design model. The sensing devices of the first-level quality gate include a mold closing accuracy sensor and a guide component fit status detector, which are used to detect mold closing accuracy, guide plate clearance and standard component compatibility online, and to compare the detection data with the design model to verify the assembly conformity of mold structural components.

[0039] Specifically, using the first-level quality gate, during the mold assembly stage, the mold closing accuracy and the fit status of the guide components are detected online through sensing devices, and the assembly conformity of the mold structural components is verified in conjunction with the design model, including the following implementation steps.

[0040] Step S110: Obtain the design model data and structural component process parameters for the mold assembly stage.

[0041] In step S110, the 3D design model of the mold is retrieved from the PLM or CAD system to obtain key information such as the spatial position, fit tolerances, material specifications, and assembly sequence of each structural component (e.g., upper and lower mold bases, mold inserts, guide pillars and bushings, pressure plates, etc.). Simultaneously, a process parameter database is integrated to clarify process specifications such as standard part models, tightening torques, and fit clearance requirements. This data not only guides the placement and calibration of sensors but also provides a theoretical basis for subsequent automatic comparison and conformity judgment, ensuring that the inspection process is based on evidence and achieving precise alignment between design intent and manufacturing execution.

[0042] Step S120: Based on the design model data, deploy mold closing accuracy sensors and guide component fit status detectors to collect online data on mold guide plate clearance, standard part interface compatibility, and mold closing surface contact status.

[0043] In step S120, after clarifying the design requirements, based on the location of key mating areas in the model, appropriate mold-closing accuracy sensors (such as pressure-sensitive membranes and displacement sensors) and guide component mating status detectors (such as gap sensors and vibration sensors) are arranged to achieve real-time monitoring of the assembly process. During mold closing, the system automatically collects physical parameters such as guide plate gap values, standard component installation status (e.g., whether bolts are in place), and pressure distribution and fit in various areas of the mold-closing surface. This step transforms the process from "manual sampling inspection" to "full-process online monitoring," capturing instantaneous deviations in dynamic assembly, improving the comprehensiveness and accuracy of the inspection, and providing high-quality data input for subsequent precise analysis.

[0044] Step S130: Compare the online data with the theoretical parameters of the design model to generate the mold closing accuracy deviation value and the fitting clearance of the guide components.

[0045] In step S130, the actual data collected in step S120 is digitally compared with the theoretical values ​​in the design model provided in step S110, and the degree of deviation of each indicator is calculated using an algorithm. For example, local void areas on the mold parting surface are identified by the pressure distribution diagram, and the deviation value of uneven contact is quantified. By comparing the measured guide plate clearance with the design tolerance zone, the direction and magnitude of the deviation are determined. The standard part models are checked against the drawing list to determine whether there are any misuses or omissions. This step transforms subjective judgment into objective data output, generating a structured deviation report, providing a clear and quantifiable basis for the next step of conformity assessment.

[0046] Step S140: Based on the deviation value and clearance amount, determine the assembly conformity of the mold structure components, specifically including: the contact area of ​​the mold closing surface ≥98%, the guide plate clearance ≤0.05mm, and the standard part model is 100% consistent with the design drawings.

[0047] In step S140, the comparison results from the previous step are automatically evaluated based on preset process standards. The system determines whether key thresholds are met: the contact area of ​​the mold parting surface is not less than 98% to ensure closure stability; the guide plate gap is controlled within 0.05mm to ensure guiding accuracy; and all standard parts are completely consistent with the design drawings to prevent assembly errors. Only when all indicators meet the standards is the mold considered to have passed the first-level quality gate. This quantitative judgment mechanism improves the consistency and authority of quality control, avoids interference from human factors, and ensures that each mold has a reliable assembly foundation before entering the debugging stage.

[0048] Step S150: If the above conditions are met, generate an assembly stage quality acceptance report and output a permission instruction to proceed to the next quality gate. If not, trigger a structural component adjustment prompt through the feedback interface and return to the data acquisition step for re-inspection.

[0049] In step S150, when the judgment result is qualified, the system automatically generates a standardized assembly quality report and sends a release signal to the Manufacturing Execution System (MES) through the execution feedback interface, allowing the mold to enter the initial debugging stage. If any indicator fails to meet the standard, the system immediately issues a structural component adjustment prompt (such as "left guide plate gap exceeds tolerance by 0.03mm, requires repair") through the human-machine interface or alarm device, guiding on-site personnel to rework. After adjustment, S120 and subsequent steps are re-executed until all conditions are met. This closed-loop mechanism of "inspection-judgment-feedback-re-inspection" effectively prevents unqualified assemblies from flowing into the next stage, significantly improving the controllability and first-pass yield of the overall manufacturing process.

[0050] In some embodiments of this application, a second-level quality gate is used to perform accuracy assessment based on measurement data of key dimensions during the initial mold debugging stage, and to determine whether the quality conditions for entering the next stage are met based on the degree of accuracy compliance and the rectification completion rate of identified problems.

[0051] The accuracy assessment criteria for the second-level quality gate include: the accuracy reporting compliance rate of key dimension measurement data is not lower than the preset first compliance rate threshold, and the rectification completion rate of identified quality problems is not lower than the preset completion rate threshold. The execution feedback interface outputs stage release or rectification instructions based on the assessment results.

[0052] Specifically, using the second-level quality gate, during the initial mold debugging stage, accuracy is evaluated based on the measurement data of key dimensions, and the quality conditions for entering the next stage are determined according to the degree of accuracy compliance and the completion rate of rectification of identified problems, including the following steps.

[0053] Step S210: Obtain key dimension measurement data during the initial mold debugging stage.

[0054] Key dimensions include the hole diameter, profile, position, and surface clearance of the stamped parts, which are collected by the data acquisition terminal and transmitted to the quality judgment module.

[0055] In step S210, the geometric accuracy information of the stamped parts actually produced by the mold after the first trial stamping is comprehensively acquired. The system uses a coordinate measuring machine (CMM) or 3D scanning equipment to perform high-precision measurements on the key features of the test mold parts, focusing on collecting dimensional elements that directly affect subsequent assembly and matching, such as the diameter and position of holes used for positioning and connection, the contour accuracy affecting the appearance matching, the positional accuracy determining the assembly relationship, and the surface clearance involving lap fits. This data is transmitted to the quality assessment module in real time through the industrial network to form a structured measurement report. This process realizes the transformation from "experience-based judgment" to "data-driven", providing real and traceable input basis for subsequent quantitative evaluation, ensuring that quality judgment is based on objective facts.

[0056] Step S220: Based on the key dimension measurement data, calculate the accuracy report compliance rate through the quality judgment module.

[0057] The compliance rate is the percentage of dimensional items whose measured values ​​meet the design tolerance requirements out of the total number of measured items.

[0058] In step S220, after acquiring complete measurement data, the quality assessment module automatically compares each measured value with the design tolerance zone in the product's digital model to determine whether it is within the allowable range and counts the number of qualified items. By calculating the comprehensive indicator of "compliance rate" (i.e., number of qualified dimensions / total number of measured dimensions × 100%), a quantitative evaluation of the overall forming accuracy of the mold is achieved. This indicator directly reflects the current manufacturing level of the mold, avoiding the one-sidedness of negating overall performance based solely on a few out-of-tolerance points, and preventing the neglect of systemic risks caused by the accumulation of numerous small deviations. As one of the core criteria, the compliance rate provides a clear and unified standard for whether to proceed to the next stage, improving the scientific nature and consistency of decision-making.

[0059] Step S230: Collect rectification records of quality problems identified during the initial debugging process, obtain the problem handling status through the execution feedback interface, and calculate the percentage of closed problems, that is, the ratio of the number of closed problems to the total number of problems.

[0060] The status of the issues can be categorized as unresolved, under rectification, or closed.

[0061] In step S230, the feedback interface is used to link with the issue management system (such as QMS) to automatically extract a list of all quality issues discovered during the initial investigation phase and their current processing status (unrectified, under rectification, closed). Based on this, the "percentage of closed issues" is calculated to measure the progress of issue rectification. This indicator reflects the team's response efficiency and execution capability to defects, preventing the phenomenon of "only checking but not rectifying" or "rectifying while moving forward." Incorporating the issue closure rate into the release criteria emphasizes that quality management is not only a technical issue but also a manifestation of process discipline, ensuring that potential problems in lower-level stages are effectively eliminated and do not transfer to subsequent stages.

[0062] Step S240: Based on the accuracy report compliance rate and the percentage of closed issues, determine whether the quality conditions for entering the next stage are met. The specific criteria are: accuracy report compliance rate > 90% and the percentage of closed issues > 80%.

[0063] In step S240, a dual-threshold control strategy is adopted: only when the mold's dimensional accuracy compliance rate exceeds 90% and the historical problem rectification and closure rate reaches over 80%, is the final debugging stage permitted. This composite judgment mechanism takes into account both "result quality" and "process control," ensuring that the mold has good basic forming capabilities while also ensuring that previously exposed problems have been adequately addressed. Setting the 90% and 80% thresholds is neither too lenient, allowing defective products to be transferred, nor too stringent, hindering normal progress. This reflects the rationality and feasibility of phased control and is a key node for achieving orderly and efficient debugging.

[0064] Step S250: If the above conditions are met simultaneously, a quality release notification for the initial adjustment stage is generated, and a control command to enter the third-level quality gate is output through the execution feedback interface. If the conditions are not met, a problem rectification list is generated through the quality judgment module, and the process returns to the initial adjustment and modeling stage to readjust parameters and collect data again until the judgment criteria are met.

[0065] The problem rectification list includes items with out-of-tolerance dimensions and details of unclosed issues.

[0066] In step S250, when the judgment result meets both requirements, the system automatically generates a standardized initial adjustment quality release notification and sends a permission signal to the production management system to enter the final adjustment stage through the execution feedback interface, pushing the process forward. If any indicator fails to meet the standard, the system immediately generates a detailed problem rectification list, clearly listing all out-of-tolerance dimensional items (such as "left front door lower edge contour deviation 0.3mm") and unresolved quality issues (such as "A-pillar area wrinkling without verified repair effect"), guiding on-site technicians to make targeted process adjustments (such as correcting the profile and optimizing the edge clamping force). After rectification, the mold needs to be retested and data collected, and the process re-enters the S210 cycle until the standard is fully met. This rigid control mechanism of "no release if not meeting the standard" effectively ensures the authority of the quality gate, prevents unqualified molds from flowing into the fine adjustment stage, and significantly reduces later rework costs and project delay risks.

[0067] In some embodiments of this application, the third-level quality gate is used to dynamically evaluate the mold accuracy based on retest data during the final mold debugging stage, and outputs control permission to enter the next manufacturing stage under the condition that the preset accuracy improvement target and quality problem closed-loop verification are met.

[0068] The dynamic evaluation indicators for the third-level quality gate include: the accuracy report compliance rate of the retest data is not lower than the preset second compliance rate threshold. The quality judgment module outputs final approval permission when the above conditions are met.

[0069] Specifically, a third-level quality gate is used to dynamically evaluate the mold accuracy based on retest data during the final mold debugging stage. Under the condition that the preset accuracy improvement target and the closed-loop verification of all quality issues are met, a control permission to enter the next manufacturing stage is output, including the following steps.

[0070] Step S310: Obtain mold accuracy retest data and quality problem rectification verification records during the final mold debugging stage.

[0071] The retest data includes key dimensions such as surface contour and hole position accuracy, collected by a coordinate measuring machine and a 3D scanner. The rectification and verification records include the problem rectification plan, implementation evidence, and effect verification report.

[0072] In step S310, the final forming capability and rectification effectiveness of the mold during the final debugging stage are comprehensively acquired. The system utilizes a high-precision coordinate measuring machine and 3D scanning equipment to remeasure the key geometric features of the stamped parts across the entire field or in key areas. It focuses on collecting key dimensions that directly affect the assembly and matching of the entire vehicle, such as the profile and hole position accuracy, ensuring the data is representative and comparable. Simultaneously, it retrieves the rectification records from the initial to final debugging process, including the responsible person for each problem, rectification measures, photos or videos of the implementation process, and the final effect verification report (such as remeasurement data and trial stamping results). This structured data provides complete and traceable information support for subsequent accuracy assessment and closed-loop verification, ensuring that the assessment results accurately reflect the actual state of the mold.

[0073] Step S320: Based on the retest data, calculate the final adjustment stage accuracy report compliance rate through the quality judgment module, which is the percentage of the number of retested dimensions that meet the design tolerance requirements out of the total number of measured items. Combined with the rectification verification records, check the closed-loop status of all quality issues, including the problem rectification completion rate and the validity of the verification results.

[0074] In step S320, after acquiring the retest data and rectification records, the quality judgment module first quantitatively evaluates the dimensional accuracy and calculates the "accuracy report compliance rate," which serves as a core indicator for measuring the mold manufacturing level. Compared to the initial adjustment stage, this stage has higher accuracy requirements, reflecting the improvement from "basically usable" to "stable and reliable." Simultaneously, the system automatically checks the closed-loop status of all quality issues, not only calculating the rectification completion rate but also thoroughly reviewing the validity of the verification results to prevent "false closures" or "insufficient evidence." For example, even if a wrinkling issue is marked as "draw beads have been modified," it is considered not closed if there are no subsequent trial stamping images or measurement data to corroborate it. This step achieves dual verification of technical indicators and management processes, ensuring that the mold meets high-quality delivery standards before entering acceptance.

[0075] Step S330: Based on the accuracy report compliance rate and the closed-loop status of quality issues, determine the preset accuracy improvement target and closed-loop verification conditions, specifically: the accuracy report compliance rate is ≥95%, and all quality issues are completed 100% closed-loop through the "responsibility division - rectification verification - standard update" process.

[0076] Among these, quality issues include problems left over from the initial investigation stage and newly identified problems in the final investigation stage.

[0077] Step S330 establishes a rigid entry standard for the third-level quality gate, reflecting the stringent requirements of the final debugging stage. A precision compliance rate exceeding 95% signifies that the vast majority of key dimensions of the mold are under stable control, enabling mass production consistency. Simultaneously, all issues (whether legacy issues or newly discovered during final debugging) must achieve 100% closed-loop management, strictly adhering to the standardized process of "responsibility allocation - rectification verification - standard update." This ensures that each problem is not only resolved but its root cause is traced, corrective measures are verified, and scalable technical experience is formed. This mechanism promotes the upgrade of quality management from "post-event remediation" to "systematic prevention of recurrence," laying the foundation for subsequent knowledge accumulation and process optimization, and is crucial for ensuring the long-term stable operation of the mold.

[0078] Step S340: The retest data is dynamically evaluated through the machine learning model in the quality judgment module, including abnormal fluctuation detection (such as dimensional deviation trend analysis), root cause tracing (such as hole position deviation caused by wear of locating pins) and optimization suggestion generation (such as adjusting the mating clearance of guide components).

[0079] In step S340, intelligent analysis methods are introduced to enhance the depth and foresight of quality assessment. The machine learning model, trained on a historical quality database, can perform trend analysis on current retest data, identifying potential abnormal fluctuations (such as a slow drift in the size of a certain area), and providing early warnings even before deviations are exceeded. Simultaneously, the model combines process parameters and structural features to conduct root cause analysis, such as determining whether hole position deviations are caused by wear or loose installation of guide components. Based on this, the system automatically generates optimization suggestions, such as "repairing the right guide plate gap to within 0.03mm" or "increasing the frequency of regular inspections of locating pins." This predictive analysis capability shifts quality control from "passive response" to "proactive prevention," further improving the reliability and lifespan of the mold.

[0080] Step S350: If the accuracy report compliance rate is ≥95%, the quality problem is 100% closed loop, and the machine learning model evaluation has no potential risks, generate a final adjustment stage quality qualification certificate, and output the control permission to enter the fourth level quality gate (acceptance stage) through the execution feedback interface.

[0081] In step S350, when all technical indicators and management requirements are met, the system automatically generates a standardized final adjustment quality certificate, serving as official proof that the mold has passed the fine-tuning stage. This certificate includes core content such as a summary of accuracy data, a problem closure list, and machine learning evaluation conclusions, and is auditable. Subsequently, a control permission to enter the acceptance stage is sent to the manufacturing execution system through the execution feedback interface, allowing the mold to proceed to the final verification stage. This step signifies that the mold has essentially completed manufacturing optimization and is ready for delivery. It is a key milestone in the manufacturing process, ensuring that only molds that truly meet the standards can enter the customer acceptance stage, effectively controlling delivery risks.

[0082] If step S360 is not met, trigger targeted rectification instructions (such as adjusting the processing parameters for out-of-tolerance dimensional items or prioritizing unclosed-loop issues) through the feedback interface, and return to the final debugging stage to retest and verify until all conditions are met.

[0083] In step S360, if any condition is not met (e.g., accuracy is only 92%, a problem is not verified, or the model warns of potential failure), the system immediately issues specific rectification instructions through the execution feedback interface, clearly indicating the out-of-tolerance items that need adjustment (e.g., "rear fender R-angle profile out of tolerance by 0.15mm") and the priority of handling unclosed-loop issues. After on-site personnel perform mold repair or process optimization according to the instructions, they need to re-test the mold, collect data, and submit verification records, then re-enter the S310 process for evaluation. This closed-loop mechanism ensures that no problems are overlooked, rectification is based on evidence, and verification is traceable, guaranteeing that the mold fully meets the standards before acceptance, significantly improving the final delivery quality and customer satisfaction.

[0084] In some embodiments of this application, a fourth-level quality gate is used to comprehensively evaluate stamped parts during the acceptance phase through forming performance testing and surface quality analysis, and to generate an acceptance report. The forming performance testing of the fourth-level quality gate includes forming limit curve testing, and the surface quality analysis includes optical and shadow detection technology, used to evaluate the forming limits and surface defects of the stamped parts, and to generate an acceptance report including conclusions regarding compliance with OEM (Original Equipment Manufacturer) vehicle assembly standards.

[0085] Specifically, using the fourth-level quality gate, during the acceptance phase, the stamped parts are comprehensively evaluated through forming performance testing and surface quality analysis, and an acceptance report is generated, including the following steps.

[0086] Step S410: Obtain forming performance data and surface quality image data of the stamping part to be accepted. Collect the deformation amount and strain distribution curve of the sheet metal through the forming limit test (FLC) equipment integrated in the data acquisition terminal, and collect the three-dimensional contour and defect images (including dents, protrusions and scratches) of the stamping part surface through the light and shadow detection system.

[0087] In step S410, the mechanical properties and appearance quality of the stamped parts under actual stamping conditions are comprehensively acquired. By deploying a Form Limit Test (FLC) system, grid strain analysis technology is used to collect the principal and secondary strain values ​​of each region of the sheet metal during the stamping process, generating a strain distribution cloud map to assess the safety boundaries of material forming. Simultaneously, a high-precision optical and shadow detection system (such as a white or blue light scanner, an industrial camera combined with structured light) is used to perform non-contact 3D scanning of the stamped part surface, acquiring surface contour data and microscopic defect images. This accurately identifies quality issues affecting appearance perception, such as dents, protrusions, scratches, and orange peel texture. This data not only covers geometric dimensions but also delves into material deformation behavior and surface texture, providing comprehensive and objective input for subsequent integrated evaluation, ensuring that the acceptance results truly reflect the actual output capacity of the mold.

[0088] Step S420: Based on the OEM's vehicle assembly standards, determine the forming performance evaluation model and surface quality judgment threshold. The forming performance evaluation model adopts the forming limit curve (FLC) to determine whether the critical deformation area of ​​the stamped part is located within the safe area of ​​the FLC curve (no cracking, no excessive thinning). The surface quality judgment threshold includes: single defect depth ≤ 0.02 mm, number of defects per unit area (1 m²) ≤ 2, and surface gloss uniformity deviation ≤ 5%.

[0089] In step S420, the system establishes a forming performance evaluation model based on the vehicle assembly technical requirements provided by the OEM. The forming limit curve (FLC) is used as the core tool to determine whether the stamped part will crack or break, clearly defining the boundaries of the safe zone. Simultaneously, strict surface quality judgment thresholds are set, such as a single defect depth not exceeding 0.02mm, no more than 2 defects per square meter, and gloss fluctuation controlled within 5%, ensuring that the stamped part meets the stringent appearance requirements of high-end vehicles. These quantitative standards provide unified and authoritative criteria for subsequent evaluations, avoiding biases caused by subjective judgments, making the acceptance process more scientific and comparable, and effectively aligning with the actual usage scenarios of the OEM.

[0090] Step S430: Analyze the forming performance data to generate the FLC curve of the stamped part, and calculate the strain value of the critical section and the safety margin of the FLC curve (safety margin = (FLC curve ultimate strain - measured strain) / FLC curve ultimate strain × 100%). Perform grayscale value comparison and three-dimensional reconstruction on the surface quality image to identify the defect type, location, and geometric parameters (depth, area, length).

[0091] In step S430, the collected raw data undergoes in-depth processing and feature extraction to transform "data" into "information." Regarding forming performance, the system overlays measured strain data with the material's FLC curve to identify high-risk areas and calculates the key indicator of "safety margin," quantifying the robustness of the forming process—even if no cracks have yet appeared, a low safety margin still poses a risk of failure due to production fluctuations. Regarding surface quality, grayscale analysis and 3D reconstruction of the light and shadow images automatically identify defect types (such as indentations, roughening, and necking marks) and accurately measure their depth, area, length, and other geometric parameters to form a defect distribution map. This process achieves both qualitative and quantitative analysis of defects, providing structured data support for subsequent comprehensive scoring and root cause tracing.

[0092] In step S440, the quality judgment module inputs the FLC safety margin and surface defect parameters into the machine learning model, outputs a comprehensive quality score (out of 100 points, with a pass standard of ≥90 points), and automatically generates a root cause analysis report (such as "insufficient draw bead resistance leading to incomplete forming" or "scratches on the die cavity surface leading to indentation defects").

[0093] In step S440, an intelligent evaluation mechanism is introduced to enhance the scientific rigor and foresight of acceptance decisions. The machine learning model, trained based on historical acceptance data and quality issue cases, comprehensively calculates a weighted overall quality score (out of 100, with a passing score of 90) by considering multiple dimensions such as FLC safety margin, defect quantity, and severity. This allows for a quantitative rating of mold performance. More importantly, the model possesses root cause analysis capabilities, enabling it to trace possible process or mold-related causes based on current defect characteristics. For example, it can determine that "severe thinning in a certain area" is due to "unreasonable drawbead layout" or "insufficient blank holder force," and that "surface indentations" originate from "local damage to the mold cavity" or "excessive guide clearance leading to uneven loading." This function not only assists technicians in quickly locating problems but also provides a clear direction for subsequent rectification, significantly improving problem-solving efficiency.

[0094] Step S450: If the overall quality score is ≥90 points and the root cause analysis shows no systemic risks in the mold, generate an acceptance report that includes the FLC test report, surface quality inspection map, and OEM standard compliance conclusions, and output the mold acceptance certificate through the execution feedback interface.

[0095] In step S450, when all evaluation indicators meet the standards and there are no significant systemic risks, the system automatically generates a standardized and traceable acceptance report. The report is comprehensive, covering FLC test results, surface defect distribution maps, overall scoring, root cause analysis conclusions, and a clear conclusion that it "meets OEM assembly standards," possessing technical authority and legal validity. Through the execution feedback interface, the system simultaneously outputs a mold acceptance certificate as formal proof of project delivery and can trigger subsequent processes in conjunction with the materials or financial systems. This step marks the successful completion of the entire mold manufacturing process, ensuring that the delivered product not only meets technical specifications but also possesses a verifiable and auditable chain of quality evidence, enhancing customer trust and the company's brand image.

[0096] If step S460 is not met, trigger targeted rectification instructions (such as adjusting the blank holder force parameters, polishing the mold cavity surface, or replacing worn guide components) through the feedback interface, return to the final mold debugging stage for re-optimization, and then conduct acceptance testing again until all conditions are met.

[0097] In step S460, if the overall score is below 90 points or the machine learning model identifies systemic risks (such as insufficient safety margin in multiple areas or repetitive surface defects), the system immediately issues specific rectification instructions through the execution feedback interface, clearly indicating the technical measures to be taken, such as "increasing the blank holder force from 80t to 85t", "mirror polishing the cavity in the A-pillar area", and "replacing the worn guide pillars and bushings". After the mold returns to the final debugging stage for optimization, it needs to be re-tested and data collected, and then re-enter the S410 process for re-inspection. This closed-loop mechanism ensures that any problems affecting mass production stability and appearance quality are thoroughly resolved, preventing "low-score pass" or "delivery with problems", truly achieving the goal of "zero-defect" delivery, and ensuring the smooth operation of the OEM production line.

[0098] In some embodiments of this application, the data acquisition terminal includes a variety of sensors, such as a coordinate measuring machine, a 3D scanner, and a pressure sensor, for collecting the clamping force during the mold assembly stage, the travel of the blank holder and the amount of sheet metal flowing in during the initial and final adjustment stages, and the surface contour data of the stamped parts during the acceptance stage.

[0099] Among them, the coordinate measuring machine is used to accurately collect the three-dimensional coordinate data of key feature points of stamped parts during the initial and final adjustment stages. By comparing with the theoretical digital model, the dimensional deviation is quantified, providing an accurate basis for mold accuracy assessment and springback compensation. The 3D scanner is used to quickly acquire the full-field surface contour data of the mold surface and stamped parts, which is especially suitable for deformation analysis of complex curved surfaces. It improves the inspection efficiency and data integrity in assembly verification, mold trial optimization and acceptance comparison. The pressure sensor is integrated into the mold or press to monitor the contact force distribution in real time during the mold closing process. It is used to assess the uniformity of mold stress, identify the risk of off-center loading during the assembly stage, and provide feedback for the optimization of blank holder force during the debugging stage.

[0100] In some embodiments of this application, the machine learning model in the quality assessment module takes key parameters acquired by the data acquisition terminal as input, trains on historical quality data to generate a quality score prediction model, and has functions for anomaly detection, root cause analysis, and optimization suggestion output. Key parameters include the pressure ring stroke, sheet material flow rate, and guide plate gap.

[0101] The machine learning model takes the blank holder stroke as input, analyzes the deviation between its motion stability and the target stroke, predicts forming stability, identifies abnormal fluctuations, and combines historical data to determine whether they are caused by guide plate wear or guide offset, and then suggests adjusting the guide clearance or lubrication strategy; taking the sheet material inflow rate as input, the model evaluates the uniformity of material flow in each area, predicts the risk of wrinkling or cracking, and achieves early anomaly detection; it performs root cause analysis by associating process parameters and outputs optimization suggestions such as adjusting the drawbead height or blank holder force distribution; taking the guide plate clearance as input, the model combines the clamping force and motion state data to determine the matching status of the guide system, predict the wear trend under long-term operation, discover potential interference or loosening problems in the assembly stage, and suggest replacing or repairing standard parts to ensure the accuracy of mold operation.

[0102] In some embodiments of this application, the execution feedback interface is linked with the mold manufacturing execution system. When the quality judgment module detects that the parameter fluctuation exceeds the preset threshold, it automatically triggers process parameter adjustment instructions, including blank holder force optimization, drawbead structure adjustment, and standard part replacement prompts.

[0103] When the quality judgment module detects abnormalities in blank holder force-related parameters, the execution feedback interface automatically sends a blank holder force optimization command to the manufacturing execution system to adjust press parameters in real time, ensuring sheet metal forming stability and preventing cracking or wrinkling defects caused by improper blank holder force. When analysis reveals that uneven sheet metal flow is related to poor drawbead matching, the execution feedback interface triggers a drawbead structure adjustment command to guide on-site correction of drawbead height or shape, improving material flow control accuracy. When guide plate clearance or mold clamping force data exceeds the threshold, indicating wear or assembly problems in the guide components, the execution feedback interface issues a standard parts replacement prompt, linking the material system to prepare spare parts, ensuring timely problem handling and avoiding mold damage and precision degradation.

[0104] In some embodiments of this application, the problem tracing module locates the root cause of problems that fail to pass the quality gate based on root cause analysis, and synchronizes the rectification plan and verification results to the technical database to form a knowledge base for preventing recurrence.

[0105] Specifically, the problem tracing module is activated when the mold fails to pass any quality gate. Based on root cause analysis methods such as 5W1H, it systematically traces the defects to pinpoint the root cause, avoiding superficial treatment. Through structured analysis of the problem's occurrence, responsible processes, and influencing factors, it clarifies rectification responsibilities and optimization directions, improving the pertinence and effectiveness of problem solving. Confirmed rectification plans and subsequent verification results are automatically archived into the company's technical database, forming standardized cases and achieving knowledge accumulation. By accumulating experience in handling typical problems, a recurrence prevention knowledge base is built, providing early warnings and references for similar projects in the future, reducing the occurrence of recurring quality problems.

[0106] In some embodiments of this application, the judgment logic of the multi-level progressive quality gate adopts the statistical process control method to monitor the fluctuation trend of quality data at each stage in real time. When the process capability index is less than the preset capability threshold, a quality warning is automatically activated and the next stage of production process is frozen.

[0107] Statistical Process Control (SPC) is employed to monitor key parameters at each quality gate in real time. By analyzing data fluctuation trends, process stability is assessed, shifting the focus from result inspection to process control. When the distribution of critical dimensions or process parameters exceeds control limits, exhibiting abnormal fluctuations, the system automatically triggers a quality warning, alerting relevant personnel to intervene promptly and prevent defects from escalating. A process capability index (such as CPK) is used as a quantitative criterion. When this index falls below a preset threshold (such as 1.33), it indicates that the current process cannot consistently meet quality requirements. The system automatically freezes the next stage of production, preventing the flow of non-conforming products and ensuring that problems are not closed before entering subsequent stages, thus strengthening the rigid control role of quality gates.

[0108] In summary, the automotive mold end-to-end quality control system based on graded quality gates provided in this application has the following technical effects.

[0109] This system achieves phased, standardized, and closed-loop quality control throughout the entire mold manufacturing process by constructing a multi-level, progressive quality gate system. Each quality gate integrates a data acquisition terminal, a quality judgment module, and an execution feedback interface, transforming traditional discrete inspection relying on manual experience into a data-driven, automated, and intelligent control process. The system sets up targeted detection and judgment logic at key stages such as assembly, initial adjustment, final adjustment, and acceptance, progressively raising quality requirements from mold closing accuracy, dimensional compliance rate, and problem closure rate to forming performance and surface quality. This ensures early detection, early interception, and early rectification of problems, effectively preventing defects from being passed on to downstream processes and significantly improving the first-pass yield and process controllability of mold manufacturing.

[0110] This system deeply integrates sensing technology, statistical process control (SPC), and machine learning models to achieve real-time monitoring of quality status, anomaly warnings, and root cause analysis. It can also automatically trigger process optimization commands, forming a dynamic control mechanism of "detection-evaluation-feedback-optimization." Simultaneously, through a problem traceability module and the construction of a recurrence prevention knowledge base, it promotes the structured accumulation and reuse of quality experience, reducing the recurrence rate of similar problems. The final standardized acceptance report fully aligns with OEM assembly standards, ensuring mold delivery quality. The overall solution improves the intelligence and standardization of automotive mold manufacturing, effectively shortens the debugging cycle, reduces rework costs, and enhances the company's quality competitiveness and delivery reliability.

[0111] In some alternative embodiments, the functions / operations mentioned in the block diagrams may not occur in the order shown in the operation diagrams. For example, depending on the functions / operations involved, two consecutively shown blocks may actually be executed substantially simultaneously, or the blocks may sometimes be executed in reverse order. Furthermore, the embodiments presented and described in the flowcharts of this application are provided by way of example to provide a more comprehensive understanding of the technology. The disclosed methods are not limited to the operations and logic flows presented herein. Alternative embodiments are contemplated in which the order of various operations is changed and sub-operations described as part of a larger operation are executed independently.

[0112] Furthermore, although this application is described in the context of functional modules, it should be understood that, unless otherwise stated, one or more of the functions and / or features may be integrated into a single physical device and / or software module, or one or more functions and / or features may be implemented in a separate physical device or software module. It is also understood that a detailed discussion of the actual implementation of each module is unnecessary for understanding this application. Rather, given the properties, functions, and internal relationships of the various functional modules in the apparatus disclosed herein, the actual implementation of the module will be understood within the scope of ordinary skill of an engineer. Therefore, those skilled in the art can implement the application set forth in the claims using ordinary skill. It is also understood that the specific concepts disclosed are merely illustrative and are not intended to limit the scope of this application, which is determined by the full scope of the appended claims and their equivalents.

[0113] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several programs to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0114] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequential list of executable programs for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, a program execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can retrieve and execute a program from or in conjunction with such a program execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can mean any means that can contain, store, communicate, propagate, or transmit a program for use by or in conjunction with a program execution system, apparatus, or device.

[0115] More specific examples (a non-exhaustive list) of computer-readable media include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Additionally, computer-readable media can even be paper or other suitable media on which programs can be printed, for example, by optically scanning the paper or other media, then editing, interpreting, or, if necessary, processing it in a suitable manner to obtain the program electronically, and then storing it in computer memory.

[0116] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable program execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0117] In the foregoing description of this specification, the reference to terms such as "one embodiment / implementation," "another embodiment / implementation," or "certain embodiments / implementations," etc., indicates that a specific feature, structure, material, or characteristic described in connection with an embodiment or example is included in an embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0118] Although embodiments of the invention have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.

[0119] The above is a detailed description of the preferred embodiments of the present invention. However, the present invention is not limited to the embodiments. Those skilled in the art can make various equivalent modifications or substitutions without departing from the spirit of the present invention. All such equivalent modifications or substitutions are included within the scope defined by the claims of the present invention.

Claims

1. A full-process quality control system for automotive molds based on graded quality gates, characterized in that, It includes a multi-level progressive quality gate, each of which includes a data acquisition terminal, a quality judgment module, and an execution feedback interface, used for phased technical verification and automatic control of the entire mold manufacturing process; The first-level quality gate is used during the mold assembly stage to conduct online detection of mold closing accuracy and guide component fit status through sensing devices, and to verify the assembly conformity of mold structural components in conjunction with the design model; the specific criteria for judging the assembly conformity are: mold closing surface contact area ≥98%, guide plate gap ≤0.05mm, and standard part model consistency with design drawings 100%; The second-level quality gate is used in the initial mold-making stage to evaluate the accuracy based on the measurement data of key dimensions. Based on the degree of accuracy compliance and the completion rate of rectification of identified problems, it is determined whether the quality conditions for entering the next stage are met. The criteria for determining the quality conditions are: accuracy report compliance rate > 90% and the proportion of closed problems > 80%. The third-level quality gate is used to dynamically evaluate the mold accuracy based on retest data during the final mold debugging stage. Under the condition that the preset accuracy improvement target and quality problem closed-loop verification are met, the gate outputs control permission to enter the next manufacturing stage. The accuracy improvement target is an accuracy report compliance rate of ≥95%, and the quality problem closed-loop verification requires all problems to complete 100% closure through the process of "responsibility division - rectification verification - standard update". The fourth-level quality gate is used to comprehensively evaluate stamped parts during the acceptance phase through forming performance testing and surface quality analysis, and to generate an acceptance report. The forming performance test of the fourth-level quality gate includes forming limit curve test, and the surface quality analysis includes light and shadow detection technology, which is used to evaluate the forming limit and surface defects of the stamped parts. The forming limit curve test is used to calculate the strain value of the critical section and the safety margin of the forming limit curve; the surface quality analysis is used to identify the defect type, location and geometric parameters; the comprehensive evaluation is based on the OEM's vehicle assembly standards, including: single defect depth ≤ 0.02 mm, number of defects per unit area ≤ 2 / ㎡, surface gloss uniformity deviation ≤ 5%, and generates an acceptance report containing the OEM's vehicle assembly standard compliance conclusion. The machine learning model in the quality assessment module takes the key parameters acquired by the data acquisition terminal as input, trains and generates a quality score prediction model through historical quality data, and has the functions of anomaly detection, root cause analysis and optimization suggestion output; the key parameters include the pressure ring stroke, sheet material flow rate and guide plate gap; The machine learning model is configured to: analyze motion stability based on the pressure ring stroke to predict forming stability; assess material flow uniformity based on sheet material inflow to predict wrinkling or cracking risk; and determine the guide system matching status based on guide plate gap combined with mold clamping force data to predict wear trend. Each quality gate also includes a problem tracing module. The problem tracing module uses root cause analysis to locate the root cause of problems that fail to pass the quality gate, and synchronizes the rectification plan and verification results to the technical database to form a knowledge base for preventing recurrence.

2. The automotive mold end-to-end quality control system based on graded quality gates according to claim 1, characterized in that, The sensing device of the first-level quality gate includes a mold closing accuracy sensor and a guide component fit status detector, which are used to detect mold closing accuracy, guide plate clearance and standard component compatibility online, and compare the detection data with the design model to verify the assembly conformity of the mold structure components.

3. The automotive mold end-to-end quality control system based on graded quality gates according to claim 1, characterized in that, The accuracy evaluation criteria for the second-level quality gate include: the accuracy report compliance rate of key dimension measurement data is not lower than the preset first compliance rate threshold, and the rectification completion rate of identified quality problems is not lower than the preset completion rate threshold; the execution feedback interface outputs stage release or rectification instructions based on the evaluation results.

4. The automotive mold end-to-end quality control system based on graded quality gates according to claim 1, characterized in that, The dynamic evaluation indicators of the third-level quality gate include: the accuracy report compliance rate of the retest data is not lower than the preset second compliance rate threshold; the quality judgment module outputs the final adjustment stage release permission when the above conditions are met.

5. The automotive mold end-to-end quality control system based on graded quality gates according to claim 1, characterized in that, The data acquisition terminal includes a coordinate measuring machine, a 3D scanner, and a pressure sensor, used to collect the clamping force during the mold assembly stage, the travel of the blank holder and the amount of sheet metal flowing in during the initial and final adjustment stages, and the surface contour data of the stamped parts during the acceptance stage.

6. The automotive mold end-to-end quality control system based on graded quality gates according to claim 1, characterized in that, The execution feedback interface is linked with the mold manufacturing execution system. When the quality judgment module detects that the parameter fluctuation exceeds the preset threshold, it automatically triggers process parameter adjustment instructions, including blank holder force optimization, drawbead structure adjustment, and standard part replacement prompts.

7. The automotive mold end-to-end quality control system based on graded quality gates according to claim 1, characterized in that, The multi-level progressive quality gate judgment logic adopts the statistical process control method to monitor the fluctuation trend of quality data at each stage in real time. When the process capability index is less than the preset capability threshold, a quality warning is automatically activated and the next stage of production process is frozen.

Images