Quality control method and device for spring seat sealing production process

Abstract

The application discloses a quality control method and device for a spring seat sealing production process, relates to the technical field of industrial quality control, and comprises the following steps: constructing a reference data set for quality determination, preprocessing collected parameters, forming an end face quality characteristic parameter set, obtaining corresponding end face quality characteristic parameters, generating quality correlation characteristics reflecting end face lubrication state and sealing force distribution uniformity, obtaining a quality state identifier, outputting corresponding adjustment instructions, and updating and optimizing the reference data set. The application improves the accuracy and reliability of batch overall quality evaluation by quantitatively analyzing the single feature deviation of the sealing element and the combined deviation index, and combining feature correlation and engineering importance coefficients for weighted fusion, enhances the production process quality monitoring and abnormal identification capability, ensures the stability and consistency of batch production of the sealing element, and provides a basis for process optimization and quality improvement.

Description

Technical Field

[0001] This invention relates to the field of industrial quality control technology, and in particular to a quality control method and apparatus for the production process of spring seat seals. Background Technology

[0002] In the current production process of spring seat seals, traditional quality control mainly relies on manual inspection and simple dimensional measurements, which are insufficient to comprehensively reflect the end-face lubrication status and the uniformity of sealing force distribution. In existing technologies, some companies assess sealing performance using a single end-face flatness or surface roughness, but because the correlation between spring loading characteristics and end-face quality is not considered, accurate judgment of overall sealing quality cannot be achieved, leading to unstable sealing performance or premature failure in some production batches. Simultaneously, with the increasing demand for high-precision and high-reliability seals, companies face the technical challenge of quickly identifying quality deviations and adjusting processing techniques in real time based on historical production data and multiple quality indicators. Summary of the Invention

[0003] This invention provides a quality control method and apparatus for the production process of spring seat seals, which solves the problem that traditional quality inspection methods in the prior art rely on manual or single-parameter evaluation, making it difficult to accurately reflect the end face lubrication status and the uniformity of spring force distribution, thus affecting the performance stability and production efficiency of the seal. It realizes full-process quality evaluation and closed-loop process adjustment based on the correlation analysis of end face quality characteristics and spring loading characteristics, thereby improving the controllability of seal quality in production batches, the efficiency of processing technology optimization, and the reliability of products.

[0004] In a first aspect, the present invention provides a quality control method for a spring seat seal manufacturing process, wherein the quality control method for a spring seat seal manufacturing process includes: This process involves collecting historical production batches of spring seat sealing end face quality data, spring loading data, and sealing performance test data. It analyzes parameter variation patterns under different batches and operating conditions, determines a quality characteristic parameter system characterizing end face lubrication stability and sealing force distribution uniformity, and constructs a benchmark dataset for quality assessment. Defect parameters such as end face flatness, surface roughness, scratches, and pits are collected and preprocessed to form a set of end face quality characteristic parameters. Spring loading characteristic parameters of the spring seat under test under corresponding operating conditions are obtained, including spring preload, fit clearance, equivalent stiffness, and spring seat deformation, and corresponding end face quality characteristic parameters are acquired. Finally, correlation analysis is performed between the end face quality parameters and spring loading parameters in the current characteristic parameter set to generate... The system generates quality correlation features reflecting the lubrication status of the end face and the uniformity of the sealing force distribution. These features are then input into a quality evaluation function, which maps them to sealing quality evaluation values. Based on the corresponding quality judgment rules in the benchmark dataset, the sealing quality evaluation values ​​are calculated and graded to obtain a quality status identifier. When the quality status identifier is in a warning state, process adjustment suggestions for the end face machining and spring assembly processes are generated, and corresponding adjustment instructions are output. The system records the end face quality characteristic parameters, spring loading characteristic parameters, sealing quality evaluation values, comprehensive deviation indicators, and corresponding quality status identifiers for the current production batch. Based on accumulated historical production data, the benchmark dataset is updated and optimized.

[0005] Secondly, the present invention also provides a quality control device for the production process of spring seat seals, wherein the quality control device for the production process of spring seat seals includes: The benchmark data construction module collects historical production batches of spring seat sealing end face quality data, spring loading data, and sealing performance test data. It analyzes parameter variation patterns under different batches and operating conditions, determines a quality characteristic parameter system characterizing end face lubrication stability and sealing force distribution uniformity, and constructs a benchmark dataset for quality assessment. The end face feature extraction module collects defect parameters such as end face flatness, surface roughness, scratches, and pits, and preprocesses the collected parameters to form a set of end face quality characteristic parameters. The spring feature extraction module obtains the spring loading characteristic parameters of the spring seat under test under corresponding operating conditions, including spring preload, fit clearance, equivalent stiffness, and spring seat deformation, and obtains the corresponding end face quality characteristic parameters. The feature association calculation module performs a correlation calculation between the end face quality parameters and spring loading parameters in the current feature parameter set. The system employs a correlation analysis to generate quality correlation features reflecting the lubrication status of the end face and the uniformity of the sealing force distribution. A quality evaluation calculation module inputs these quality correlation features into a quality evaluation function, which maps them to sealing quality evaluation values. Based on the corresponding quality judgment rules in the benchmark dataset, the sealing quality evaluation values ​​are calculated and graded to obtain a quality status identifier. A process strategy generation module generates process adjustment suggestions for the end face machining and spring assembly processes when the quality status identifier is in a warning state, and outputs corresponding adjustment instructions. A data self-updating module records the end face quality characteristic parameters, spring loading characteristic parameters, sealing quality evaluation values, comprehensive deviation indicators, and corresponding quality status identifiers for the current production batch, and updates and optimizes the benchmark dataset based on accumulated historical production data.

[0006] This invention discloses a quality control method and apparatus for the production process of spring seat seals, comprising: collecting end face quality parameters, spring loading parameters, and sealing performance test data from historical production batches to construct a benchmark dataset for quality judgment; collecting flatness, surface roughness, and defect parameters of the end face of the spring seat to be tested, and obtaining spring preload, fit clearance, equivalent stiffness, and spring seat deformation to form end face quality characteristics and spring loading characteristics; performing correlation analysis on the above characteristics to generate quality correlation characteristics reflecting the end face lubrication state and the uniformity of sealing force distribution; inputting the quality correlation characteristics into a preset quality evaluation function, calculating the sealing quality evaluation value based on the benchmark dataset judgment rules, and outputting a quality status identifier; when the quality status is a warning, generating process adjustment suggestions for end face processing and spring assembly and outputting executable instructions; recording the current batch data and updating the benchmark dataset to achieve closed-loop quality control of the production process, improving sealing performance stability and production efficiency. The quality control method and apparatus for spring seat seal production process disclosed in this invention solves the technical problems of lack of correlation analysis between end face quality and spring loading characteristics in the prior art, difficulty in accurately judging sealing performance and lag in process adjustment. It realizes correlation analysis, closed-loop quality evaluation and process adjustment based on end face quality and spring loading characteristics, and can identify quality deviations in real time and output executable process instructions, thereby improving the technical effects of improving the stability of sealing performance, processing accuracy and production efficiency of production batches. Attached Figure Description

[0007] Figure 1 This is a flowchart illustrating the quality control method for the spring seat seal production process of the present invention. Figure 2 This is a schematic diagram of the quality control device for the spring seat seal production process of the present invention; Figure labeling: 11. Baseline data construction module, 12. End face feature extraction module, 13. Spring feature extraction module, 14. Feature association calculation module, 15. Quality evaluation calculation module, 16. Process strategy generation module, 17. Data self-updating module. Detailed Implementation

[0008] The above technical solutions will now be described in detail with reference to the accompanying drawings and specific embodiments to provide a better understanding of them. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. It should be understood that the present invention is not limited to the exemplary embodiments used only to explain the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention. Furthermore, it should be noted that, for ease of description, only the parts related to the present invention are shown in the drawings, not all of them.

[0009] Example 1, as Figure 1 This is a flowchart illustrating the quality control method for the spring seat seal manufacturing process of the present invention, wherein the quality control method for the spring seat seal manufacturing process includes: S100: Collect quality data of spring seat sealing end face, spring loading data and sealing performance test data of historical production batches, analyze the parameter variation law under different batches and working conditions, determine the quality characteristic parameter system characterizing end face lubrication stability and sealing force distribution uniformity, and construct a benchmark dataset for quality judgment.

[0010] Furthermore, step S100 of this application also includes: S101: Filter the retrieved historical production data of the spring seat seal, removing records with abnormal operating conditions, detection failures, or missing data, and retaining valid historical data samples corresponding to the target operating conditions to form a set of valid historical data samples; S102: Extract quality characteristic parameters related to the lubrication stability and sealing force distribution characteristics of the sealing end face from the set of valid historical data samples. These quality characteristic parameters include end face flatness, surface roughness, end face defect characteristic parameters, as well as spring preload, fit clearance, equivalent stiffness, and spring seat deformation, forming a set of quality characteristic parameters; S103: Perform statistical analysis on the set of quality characteristic parameters to obtain the distribution range of each quality characteristic parameter under different production batches and target operating conditions. The mean and fluctuation range are used to characterize the typical quality characteristic distribution of the spring seat seal under stable production conditions; S104: Based on the typical quality characteristic distribution, determine the quality judgment benchmark range corresponding to each quality characteristic parameter, and classify the quality judgment benchmark range according to the engineering requirements of end face lubrication stability and sealing force distribution uniformity to form normal range, warning range and abnormal range; S105: Based on the end face quality characteristic parameters, spring loading characteristic parameters and their corresponding historical quality status identifiers of historical production batches, associate and integrate each quality characteristic parameter with its corresponding quality judgment benchmark range to construct a benchmark dataset for quality status judgment, and store the benchmark dataset for subsequent quality evaluation and production status judgment.

[0011] Specifically, this process involves collecting historical production batches of spring seat seal end-face quality data, spring loading data, and sealing performance test data. Analyzing parameter variation patterns under different batches and operating conditions, a quality characteristic parameter system characterizing end-face lubrication stability and sealing force distribution uniformity is established. A benchmark dataset for quality assessment is constructed, providing a reliable data foundation for subsequent quality evaluation and enabling quantitative analysis and predictive control of production status. Specifically, the retrieved historical production data for spring seat seals is screened, removing records with abnormal operating conditions, testing failures, or missing data. Valid historical data samples corresponding to the target operating conditions are retained, forming a set of valid historical data samples to ensure the accuracy and representativeness of subsequent analysis data. Quality characteristic parameters related to the sealing end-face lubrication stability and sealing force distribution characteristics are extracted from this set of valid historical data samples. These quality characteristic parameters include end-face flatness, surface roughness, end-face defect characteristic parameters, as well as spring preload, fit clearance, equivalent stiffness, and spring seat deformation, forming a set of quality characteristic parameters to achieve a comprehensive quantitative description of sealing quality. The quality characteristic parameters related to the lubrication stability and sealing force distribution characteristics of the sealing end face are extracted from the historical valid data sample set and determined by the following methods: First, the lubrication state and sealing force distribution uniformity of the end face are quantitatively characterized, for example, by obtaining performance evaluation values ​​through indicators such as lubrication film thickness, friction coefficient change, and end face pressure distribution uniformity; then, statistical correlation analysis is performed on the end face processing parameters (such as flatness, surface roughness, scratches, pits, etc.) and spring loading parameters (such as preload, fit clearance, equivalent stiffness, and spring seat deformation) recorded in historical production batches with the above performance evaluation values, including correlation coefficient calculation, partial correlation analysis, or regression analysis; by evaluating the degree of correlation between each parameter and performance indicators, parameters that show a significant correlation with the lubrication stability and sealing force distribution uniformity of the end face are selected and included in the analysis set as quality characteristic parameters; in addition, the statistical analysis results can be reasonably corrected by combining engineering experience and manufacturing tolerance requirements to ensure that the selected quality characteristic parameters can reflect actual performance changes and facilitate subsequent evaluation and process adjustment. Furthermore, statistical analysis is performed on the set of quality characteristic parameters to obtain the distribution range, mean, and fluctuation range of each quality characteristic parameter under different production batches and target operating conditions. This is used to characterize the typical quality characteristic distribution of the spring seat seal under stable production conditions, thereby providing a basis for determining abnormal states. Specifically, the statistical analysis of the set of quality characteristic parameters involves unifying the units and dimensions of the end face quality characteristic parameters and spring loading characteristic parameters, eliminating outliers, and normalizing or standardizing the parameters to eliminate the influence of different parameter scales on the analysis results.For each quality characteristic parameter, its minimum, maximum, mean, median, and standard deviation are statistically analyzed under different production batches and target operating conditions to obtain the parameter's distribution range and fluctuation characteristics, thus reflecting its typical value under stable production conditions. Based on the statistical distribution, the normal fluctuation range and deviation range of each parameter are calculated. For example, the normal range is determined by the method of mean ± k times standard deviation, while abnormal and warning ranges are identified to provide a basis for subsequent quality status judgment. The same parameter is compared between different batches to analyze its mean and fluctuation trend, identify possible systematic deviations between batches, and ensure that the statistical results reflect both the characteristics of a single batch and the overall production stability. The statistical results are organized into tables or datasets to form typical distribution data and judgment thresholds for each quality characteristic parameter, providing a direct basis for the construction of benchmark datasets and subsequent quality evaluation. Based on the typical quality characteristic distribution, the quality judgment benchmark range corresponding to each quality characteristic parameter is determined. Combined with the engineering requirements of end-face lubrication stability and sealing force distribution uniformity, the judgment benchmark range is classified into normal range, warning range, and abnormal range, thereby realizing graded judgment of quality status and facilitating risk warning and process control during production. For example, in a historical valid data sample, the statistical mean of the flatness of the end face of a certain type of spring seat seal is 0.006 mm, and the standard deviation is 0.0015 mm. Therefore, its judgment benchmark intervals can be divided as follows: the normal interval is the mean ± 2 times the standard deviation, i.e., 0.003 mm to 0.009 mm; the warning interval is the range deviating from the normal interval by 0.001 mm, i.e., 0.002 mm to 0.003 mm or 0.009 mm to 0.010 mm; and the abnormal interval is a value less than 0.002 mm or greater than 0.010 mm. Similarly, if the statistical analysis of the spring preload parameter shows a mean of 120 N and a standard deviation of 5 N, then the normal interval can be set to 110 N to 130 N, the warning interval to 105 N to 110 N or 130 N to 135 N, and values ​​exceeding these ranges are considered abnormal. This grading method enables the quantitative determination of various quality characteristic parameters, thereby assessing the quality status of end-face lubrication and the uniformity of sealing force distribution, providing a basis for early warning and process adjustment during production. Furthermore, this application constructs a benchmark dataset based on end-face quality characteristic parameters, spring loading characteristic parameters, and their corresponding historical quality status identifiers from historical production batches. These historical quality status identifiers originate from quality inspection records, sealing performance test results, and manual inspection judgments generated during historical production processes, ensuring objectivity and traceability. First, statistical analysis is performed on historical production data to calculate the distribution range, mean, and fluctuation range of each quality characteristic parameter under different quality states, thereby determining the corresponding normal, early warning, and abnormal intervals. These intervals are obtained through interval statistical analysis of historical samples, rather than being subjectively set manually.Subsequently, each quality characteristic parameter is mapped to its statistically determined quality judgment benchmark interval, and a correspondence is established between each benchmark interval and its corresponding historical quality status identifier, thus forming a structured benchmark dataset containing characteristic parameters, interval ranges, and status labels. This benchmark dataset, after being stored, is used for subsequent production batch quality evaluation and status judgment, achieving objective judgment based on historical data, and providing a data foundation and rule basis for subsequent quality early warning and process adjustment. For example, in one embodiment, taking the end face flatness parameter as an example, statistical analysis was performed on data from a cumulative total of 1200 historical production batches, of which 950 batches were judged to be in a normal state, 180 were in an early warning state, and 70 were in an abnormal state. Interval statistics were performed on the end face flatness values ​​of the normal state samples, revealing a concentrated distribution range of 0.012mm to 0.020mm; statistics were performed on the early warning state samples, revealing a distribution range of 0.020mm to 0.026mm; and statistics were performed on the abnormal state samples, revealing a distribution range greater than 0.026mm. Based on the above statistical results, the end face flatness parameter is divided into corresponding normal, warning, and abnormal intervals, and a correspondence is established with historical quality status identifiers. Similarly, statistical analysis is performed on the spring preload parameter to obtain the numerical distribution range under different quality states, and corresponding interval divisions are formed. Finally, the interval division results of each quality characteristic parameter and its corresponding status label are integrated to form a structured data table, which is stored as a benchmark dataset. After the inspection data of subsequent production batches is input, the corresponding quality status identifier can be output by matching the intervals with this benchmark dataset, thus achieving objective quality judgment based on historical statistical data.

[0012] S200: Collects defect parameters such as end face flatness, surface roughness, scratches, and pits, and preprocesses the collected parameters to form a set of end face quality characteristic parameters.

[0013] Specifically, the end-face quality characteristic parameters of the spring seat under test are obtained for subsequent quality evaluation and condition determination. Specifically, defect parameters such as flatness, surface roughness, scratches, and pits on the end face of the spring seat under test are collected. The collected raw parameters are preprocessed, including outlier removal, normalization, and dimensional unification, forming a set of end-face quality characteristic parameters that can be used for analysis. This step ensures the reliability and comparability of the end-face characteristic data used in subsequent analysis, while eliminating deviations caused by measurement errors or acquisition noise. Furthermore, this application uses statistical analysis methods to perform a preliminary evaluation of the collected end-face characteristic data, such as calculating the mean, standard deviation, and maximum deviation of flatness, to determine the preliminary distribution of processing quality. For example, the measured flatness of the spring seat end face in a certain batch ranges from 0.004 mm to 0.009 mm, with an average value of 0.006 mm. Preprocessing removes an outlier of 0.012 mm, and the remaining data is normalized to obtain the end face flatness feature vector. Similarly, surface roughness, scratch depth, and pit diameter are processed to form a complete set of end face quality characteristic parameters. Through this step of end face feature acquisition and preprocessing, accurate and stable end face quality data can be obtained, providing a reliable foundation for subsequent quality feature correlation analysis, evaluation function calculation, and production batch quality status determination, thereby improving the accuracy of sealing performance evaluation and the controllability of the production process.

[0014] S300: Obtain the spring loading characteristic parameters of the spring seat under test under the corresponding working conditions, including spring preload, fit clearance, equivalent stiffness and spring seat deformation, and obtain the corresponding end face quality characteristic parameters.

[0015] Furthermore, step S300 of this application also includes: S301: Based on actual working conditions, perform consistency verification on the acquired end-face quality characteristic parameters and spring loading characteristic parameters respectively, and correct parameter data that do not meet the target working conditions to ensure that the characteristic parameters are consistent with the working conditions of the benchmark dataset; S302: Perform stability and effectiveness analysis on the characteristic parameters after consistency verification, eliminate characteristic parameters with insufficient stability, and retain key characteristic parameters that have a significant impact on the lubrication state and sealing force distribution of the sealing end face; S303: Based on the structural correlation between the retained end-face quality characteristic parameters and spring loading characteristic parameters, construct derived characteristic parameters, which include combined force-sensitive parameters and correlation parameters reflecting the influence of spring loading changes on the uniformity of end-face contact; S304: Integrate the screened original characteristic parameters with the derived characteristic parameters in a structured manner to form a complete set of characteristic parameters for the current production batch, which will be used for subsequent quality correlation analysis and quality status evaluation.

[0016] Specifically, the spring loading characteristic parameters of the spring seat under test are acquired under corresponding working conditions, including spring preload, fit clearance, equivalent stiffness, and spring seat deformation, and the corresponding end face quality characteristic parameters are acquired simultaneously. By synchronously collecting end face machining quality and spring loading state, a data structure that can be correlated between end face contact state and loading state is formed, providing a complete and traceable data foundation for subsequent quality correlation analysis. Based on actual working conditions, the acquired end face quality characteristic parameters and spring loading characteristic parameters are subjected to consistency verification. The consistency verification includes matching judgment of temperature conditions, load level, speed range, machining stage, and detection sequence, and correction processing of parameter data that does not conform to the target working conditions, ensuring that the characteristic parameters are consistent with the target working conditions corresponding to the benchmark dataset. Through working condition consistency verification, systematic deviations caused by mixing data under different working conditions can be avoided, improving the accuracy and repeatability of subsequent quality analysis. Specifically, firstly, based on the statistical reference value of the target working condition in the benchmark dataset, the offset of the current characteristic parameter relative to the center value of the target working condition is calculated. For continuous numerical characteristic parameters, such as spring preload, fit clearance, and end face flatness, they are corrected according to the offset to bring them back to the statistical range corresponding to the target working condition. For characteristic parameters with different dimensions or large amplitude differences, proportional mapping or normalization mapping is used for interval alignment to ensure that the relative proportional relationship between parameters is not distorted. During the adjustment process, structural strength constraints and manufacturing tolerance limits are also considered to ensure that the corrected parameters do not exceed the allowable working force of the spring and the structural safety boundary. Through the above adjustments, the one-time disturbances of measurement errors, batch differences, and working condition fluctuations on characteristic parameters can be eliminated, making the current set of characteristic parameters consistent with the benchmark dataset in a statistical sense. After completing the consistency verification, the characteristic parameters are screened for stability and effectiveness. Specifically, this includes fluctuation analysis of historical batches and current sampling data, calculating standard deviation, coefficient of variation, and time series stability index; parameters with fluctuations exceeding the allowable range or poor repeatability are judged to be unstable and are removed. At the same time, outlier detection, moving average, and filtering are combined to identify noise interference, and characteristic parameters with significant noise contributions are removed or smoothed. Further combining structural mechanics principles with sealing contact mechanism analysis, key parameters that significantly affect the end-face lubrication state and sealing force distribution are identified through correlation testing, such as end-face flatness, surface roughness, and spring preload. This screening process reduces redundant information interference and improves the sensitivity and stability of quality evaluation. Based on this, derived characteristic parameters are constructed according to the structural coupling relationship between end-face quality characteristic parameters and spring loading characteristic parameters. First, parameter matching is performed according to the structural contact relationship, and the correlation index between the two types of parameters is calculated to identify parameter pairs that significantly affect end-face quality due to changes in loading.For identified key parameter pairs, the combined weights are automatically determined based on their correlation, constructing combined stress-sensitive parameters to reflect the overall impact of loading conditions on the uniformity of end-face contact. Simultaneously, the proportional relationship between the rate of change of loading parameters and the rate of change of end-face mass is analyzed to form a correlation parameter reflecting the degree of influence of loading fluctuations on the contact state, thereby quantifying the contribution of loading disturbances to sealing performance. For example, if the flatness of the end-face of a batch of spring seats is 0.006 mm and the spring preload is 120 N, correlation analysis determines that the combined weights of the two are proportionally allocated, thus constructing combined stress-sensitive parameters to characterize the overall stress balance state. When a 5 N change in preload causes a 0.001 mm change in end-face flatness, the influence coefficient of the loading change on contact uniformity can be calculated to assess the sensitivity to loading fluctuations. These derived features can make structural coupling information explicit, improving the ability to identify potential quality risks. Finally, the verified and screened original feature parameters and the constructed derived feature parameters are structurally integrated to form a complete set of current feature parameters, which are then bound and stored with corresponding operating condition labels. This set of feature parameters provides a unified input for subsequent quality correlation analysis, quality evaluation function calculation, and production status determination, realizing a closed-loop processing flow from data collection, verification, and screening to feature enhancement, thereby improving the scientific, stable, and refined level of quality control in the production process.

[0017] S400: Perform correlation analysis on the end face quality parameters and spring loading parameters in the current feature parameter set to generate quality correlation features that reflect the end face lubrication status and the uniformity of sealing force distribution.

[0018] Furthermore, step S400 of this application also includes: S401: Based on the operating conditions, the end face quality characteristic parameters and spring loading characteristic parameters are normalized to form a standardized set of end face quality characteristic parameters. With the standardized set of spring loading characteristic parameters S402: Construct lubrication stability characteristic quantities to characterize the lubrication state of the end face. The formula is: ;in, To normalize the flatness of the end face, To normalize the surface roughness, To normalize the characteristic parameters of end face defects, To normalize the end face hardness, For processing temperature, The target temperature for the process, The viscosity of the lubricating oil. To design the target viscosity, S403: Sensitivity parameter for the influence of end-face defect features on lubrication stability; S403: Constructing a sealing force distribution characteristic quantity to characterize the uniformity of sealing force distribution. It can be represented as: ;in: This is the spring preload. To accommodate the gap, To achieve the target fit gap, The equivalent stiffness of the spring. This is the deformation of the spring seat. For other environmental or process-related parameters, For the corresponding target value, Number of environmental / process factors considered; S404: Characteristics of lubrication stability With sealing force distribution characteristic quantity Perform combined analysis to generate quality correlation features that simultaneously reflect the lubrication status of the sealing end face and the uniformity of the sealing force distribution. Q The formula is ：Q= ;in, Q The larger the value, the more stable the sealing quality, the more complete the lubricating film, and the more uniform the sealing force distribution.

[0019] Specifically, based on the current set of characteristic parameters, the intrinsic correlation between the end-face quality state and the spring loading state of the spring seat seal is analyzed. This generates quality correlation features that comprehensively reflect the lubrication state and uniformity of the sealing force distribution on the sealing end face, providing unified and stable input indicators for subsequent quality status evaluation. Based on the actual working conditions corresponding to the spring seat seal under test, the end-face quality characteristic parameters and spring loading characteristic parameters in the current set of characteristic parameters are normalized to eliminate the influence of differences in the dimensions and value ranges of different parameters on the correlation analysis results. Normalization is then performed to form a standardized set of end-face quality characteristic parameters. With the standardized set of spring loading characteristic parameters ,in, The normalized end-face flatness is used to characterize the uniformity of contact pressure distribution per unit area within the contact region of the sealing end face. Its value is obtained by normalizing contact pressure measurement data or contact stress analysis results. The larger the value, the more uniform the pressure distribution at the end face. Normalized surface roughness, used to characterize the overall roughness of the microscopic surface morphology of the sealing end face, is obtained by standardizing the end face roughness measurement results. The smaller the value, the smoother the end face surface. These are normalized end-face defect characteristic parameters used to characterize the overall geometric flatness of the sealing end face. Their values ​​are calculated from end-face morphology measurement data. The smaller the value, the higher the overall geometric consistency of the end face. To normalize the end face hardness, which characterizes the distribution stability of the hardness of the sealing end face material within the contact area, its value is obtained by normalizing multi-point hardness measurement results. The larger the value, the stronger the end face's resistance to local deformation. The processing temperature is used to characterize the actual temperature state of the sealing end face under the current production or operating conditions. Its value is obtained by the temperature acquisition device and compared with the process target temperature. Lubricating oil viscosity is a characteristic viscosity parameter of the lubricating medium, used to characterize the flow properties of the lubricating medium under current operating temperature conditions. Its value is obtained by normalizing the lubricating medium's physical properties or online estimation results. A smaller value indicates higher end-face integrity. Through the synergistic characterization of end-face quality parameters, the contact capability and stability basis of the sealing end face can be reflected from multiple dimensions, including geometry, morphology, and material properties. The set of spring-loaded characteristic parameters is used to characterize the sealing load formation mechanism and its stability, and includes the following parameters: ,in, The spring preload is used to characterize the axial load level applied by the spring to the sealing end face under initial loading conditions. Its value is obtained by standardizing the spring loading measurement results. To fit the clearance, it is used to characterize the size of the assembly clearance between the spring seat and the adjacent structure, and its value is obtained by normalizing the assembly inspection results; The equivalent stiffness characteristic parameter of the spring is used to characterize the force-displacement response characteristics of the spring within its working stroke range. Its value is obtained by spring parameter calculation or experimental calibration. The deformation of the spring seat is used to characterize the degree of axial deformation of the spring seat structure under the action of sealing load. The axial displacement under the action of standard sealing load is recorded by installing a micro displacement sensor on the key section of the spring seat. These are environmental or process disturbance characteristic parameters used to characterize the influence of non-dominant factors such as temperature fluctuations and assembly deviations on the spring loading state. Temperature and displacement sensors are deployed in the assembly environment to record temperature fluctuations and deviations during assembly and curing. This processing ensures the comparability of characteristic parameters from different sources and with different physical meanings, providing a unified data foundation for subsequent feature construction. Based on the standardized end-face quality characteristic parameter set, and according to the influence mechanism of end-face flatness, surface roughness, and end-face defect characteristic parameters on lubrication film formation and stability, lubrication stability characteristic quantities are constructed to characterize the lubrication state of the sealing end face. ;in, To normalize the end face flatness, the end face flatness P is measured using a coordinate measuring machine and linearly normalized to the 0–1 interval. , To normalize the surface roughness, the end face roughness is measured using a roughness meter and then normalized. , To normalize the end-face defect feature parameters, the end-face defect feature D is obtained through 3D image recognition and then normalized with historical data intervals to form... , To normalize the end face hardness, the end face hardness H is measured and normalized according to the reference hardness range. The processing temperature is measured using a thermocouple. To ensure that the actual thermal conditions on end-face lubrication are considered during the evaluation, The target temperature is the process temperature specified in the process design, used for deviation calculations. To determine the viscosity of the lubricating oil, a rotational viscometer is used to measure the viscosity of the lubricating oil. The ideal lubricating oil viscosity value, determined by process design or lubrication scheme, is used to achieve the target viscosity. This is a sensitivity parameter for the influence of end-face defect features on lubrication stability. Its value can be determined through statistical analysis or empirical analysis of the end-face defect features and lubrication stability results in the benchmark dataset. Those skilled in the art can reasonably set this parameter based on the features and corresponding evaluation results in the benchmark data, according to the principles of statistical correlation or weight allocation. The range of values ​​is used for subsequent deviation weighting calculations and batch quality evaluation. This calculation method transforms the combined influence of end-face geometry, material properties, and process environment factors on lubricant film stability into a single quantitative index, thus avoiding the error accumulation problem caused by judging a single parameter. Based on the standardized spring loading characteristic parameter set, and according to the influence relationship between spring preload, fit clearance, equivalent stiffness, and spring seat deformation on the sealing pressure distribution, a sealing force distribution characteristic quantity is constructed to characterize the uniformity of the sealing force distribution. It can be represented as: ;in: The preload force of the spring is measured by a spring force sensor. Alternatively, it can be calculated based on the spring pre-compression and the spring's equivalent stiffness. To accommodate the gap, The target fit clearance, specified in the process standard, is used to calculate the sealing force deviation and is obtained by measuring the actual clearance after assembly. The equivalent stiffness of the spring is calculated based on the spring's design parameters. The deformation of the spring seat is obtained by measuring the deformation under assembled or loaded conditions using a displacement sensor or laser rangefinder. Other environmental or process influence parameters, and characteristic values ​​of process influence parameters, are used to characterize the disturbance impact of assembly environmental conditions or process states on the consistency of spring loading in the current production batch. The environmental or process parameters collected in the current production batch are normalized. These environmental or process-related parameters include assembly ambient temperature, assembly ambient humidity, assembly cycle time, spring assembly positioning deviation, and tooling fixture repeatability error. Therefore, these parameters do not directly determine the sealing performance, but they will disturb the spring loading and end face contact state. To correspond to the target value, when constructing the benchmark dataset, statistical analysis was performed on the environmental and process parameters corresponding to production batches that were in normal quality condition during historical production processes. The mean or median value of each parameter was extracted and used as the target value for the corresponding environmental or process parameter. , To account for the number of environmental / process factors considered, each For each identified source of disturbance, This indicates the total number of these disturbance sources. Specifically, during the construction of the quality characteristic parameter system and benchmark dataset, environmental and process parameters that are significantly correlated with the uniformity of sealing force distribution in historical production data are analyzed to select the set of parameters participating in the evaluation of sealing force distribution. The number of these parameters is [number missing]. By constructing this characteristic quantity, the influence of multiple spring loading-related parameters on the sealing specific pressure distribution can be uniformly mapped to the same evaluation scale, thereby accurately reflecting the uniformity of the sealing force distribution. (This relates to the lubrication stability characteristic quantity.) With sealing force distribution characteristic quantity Perform combined analysis to generate quality correlation features that simultaneously reflect the lubrication status of the sealing end face and the uniformity of the sealing force distribution. Q The formula is ：Q= ;in, Q A larger value indicates a more stable seal quality, a more complete lubrication film, and a more uniform sealing force distribution. This quality correlation feature serves as input data for subsequent quality status evaluation, used to determine the current production quality status of the spring seat seal. Among these features... Q The numerical value is used to comprehensively characterize the sealing quality level of the spring seat seal under current production conditions. Q A higher value indicates a more stable end-face lubrication state, a more complete lubrication film, and a more uniform distribution of sealing force. (The aforementioned quality-related characteristic...) Q As input data for subsequent quality status evaluation steps, it is used to determine whether the current spring seat seal is in a normal, warning, or abnormal production quality status.

[0020] S500: Input the quality association features into the quality evaluation function, which maps the quality association features to the sealing quality evaluation value; calculate and classify the sealing quality evaluation value based on the corresponding quality judgment rules in the benchmark dataset to obtain the quality status identifier.

[0021] Furthermore, step S500 of this application also includes: S501: Vectorize the quality-related features to construct a quality-related feature vector for the current production batch; S502: Input the quality-related feature vector into a quality evaluation function for calculation to generate a corresponding sealing quality evaluation value; S503: Perform deviation analysis on the sealing quality evaluation value and the ideal state evaluation value in the benchmark dataset to generate a comprehensive deviation index reflecting the overall deviation degree of the current production batch; S504: Based on the correspondence between the comprehensive deviation index and a preset judgment threshold, classify and determine the current production status and output a quality status identifier.

[0022] Furthermore, step S502 of this application also includes: S502-1: Obtain the quality judgment benchmark interval matching the current batch operating conditions, including the normal interval, warning interval, and abnormal interval. If necessary, perform cross-validation based on historical batch judgment results to select the most robust judgment interval; S502-2: Convert the quality-related feature vector C= Input the quality evaluation function, and select an existing feature mapping model based on the benchmark dataset. M (·), perform function operations on each quality-related feature component to obtain the corresponding feature evaluation component E= ;in, = M (·), i= 1,2 ， n S502-3: Perform a combination operation on the feature evaluation component E to generate a sealing quality evaluation value that characterizes the sealing quality status of the current production batch. The formula is: = × ;in, These are the feature weight coefficients determined based on the benchmark dataset.

[0023] Furthermore, step S503 of this application also includes: S503-1. Based on the feature composition of the sealing quality evaluation value, quantify the deviation of each quality-related feature from the corresponding ideal state in the benchmark dataset to form a single-feature deviation set. On this basis, construct deviation correlation pairs of feature combinations and calculate the corresponding joint deviation index. S503-2. Combine the feature correlation index determined based on the benchmark dataset and the engineering importance coefficient, perform weighted fusion of the single-feature deviation and joint deviation index to generate a weighted deviation set reflecting different deviation contribution levels. S503-3. Perform multi-constraint summary analysis on the weighted deviation set and generate a comprehensive deviation index for the current production batch based on a preset fusion function. The fusion function at least comprehensively considers the single-feature deviation contribution, feature cross-deviation contribution, and correlation constraint terms to quantify the comprehensive deviation of the overall quality status of the current production batch from the benchmark state.

[0024] Specifically, the obtained quality correlation features are input into a quality evaluation function, which maps the quality correlation features into a sealing quality evaluation value that characterizes the overall quality state of the spring seat seal. Based on the pre-constructed quality judgment rules in the benchmark dataset, the sealing quality evaluation value is calculated, analyzed, and graded to obtain the quality evaluation result corresponding to the current production batch, and the corresponding quality status identifier is output. A quality judgment benchmark interval matching the operating conditions of the current production batch is obtained. This benchmark interval includes at least a normal interval, a warning interval, and an abnormal interval. In the case of multiple candidate benchmark intervals, based on the quality judgment results of historical production batches, cross-validation analysis is performed on the judgment stability of each candidate benchmark interval, and the benchmark interval with the highest judgment consistency is selected as the target judgment interval for the current production batch. Specifically, the historical batch dataset... : Contains the measured quality correlation feature vector for each batch The judgment at that time Candidate benchmark interval set Each interval contains the boundaries for normal / warning / abnormal. The cross-validation step includes A: simulating historical judgments, associating the quality of historical batches with feature vectors. Input to based Based on the judgment rules, simulated judgment results are obtained. The formula represents: By traversing all historical batches, each candidate interval is obtained. The corresponding set of simulation judgment results Step B involves calculating the consistency index of the judgment and simulating the judgment results. Results of historical fact determination Compare and calculate the consistency index. : ;in, The indicator function is used, with 1 indicating similarity and 0 indicating dissimilarity. The historical consistency percentage for each candidate interval is obtained. A higher value indicates greater stability of the candidate interval across historical batches. (This refers to all...) The comparison is performed, and the target judgment benchmark interval for the current production batch is output. When multiple candidate benchmark intervals exist, the quality correlation feature vectors of historical batches are input into the judgment rules corresponding to each candidate benchmark interval to obtain simulated judgment results. The judgment consistency index for each candidate interval is calculated based on the historical actual judgment results, and the interval with the highest judgment consistency is selected as the target judgment benchmark interval for the current production batch. The judgment consistency index can be expressed as the proportion of simulated judgment results that are the same as the actual judgment results in historical batches. The quality correlation feature vector C = Input the quality evaluation function, and select an existing feature mapping model based on the benchmark dataset. M (·), perform function operations on each quality-related feature component to obtain the corresponding feature evaluation component E= ;in, = M (·), i= 1,2 ， n S501-3: Perform a combination operation on the feature evaluation component E to generate a sealing quality evaluation value that characterizes the sealing quality status of the current production batch. The formula is: = × ;in, These are the feature weight coefficients determined based on the benchmark dataset. They are based on each feature parameter in the historical batches. Compared with the final sealing quality evaluation value The statistical correlation was calculated, its absolute value was normalized, and the weights of each feature were obtained. This is used to reflect the contribution of each feature to the overall sealing quality evaluation. The sealing quality evaluation value of the current production batch is compared with the benchmark dataset, and the deviation degree of each feature is calculated, forming a single-feature deviation set. The deviation degree can be achieved by calculating the difference between the feature value and the ideal value, the standardized residual, or the normalized distance. Subsequently, key features are paired to calculate a joint deviation index, which describes the deviation interaction between two features. The joint deviation index can be generated using weighted product, Euclidean distance, or correlation weighting methods. For example, if the end face flatness and spring preload both deviate from the target value, the joint deviation can be represented by the product or weighted sum of their deviation degrees, quantifying the impact of loading changes on the end face contact uniformity. The deviation degree of each quality-related feature relative to the corresponding ideal state in the benchmark dataset is quantified to form a single-feature deviation set. Based on this, any two quality-related features are grouped together to construct pairwise deviation correlation pairs, and the corresponding joint deviation indices are calculated to characterize the cooperative deviation relationship between features. To more accurately reflect the impact of various quality-related features of the current production batch on the overall quality status, in addition to quantifying the deviation of individual features, a joint deviation analysis of pairwise feature combinations is constructed. Specifically, for each quality-related feature collected in the current batch, the deviation magnitude of each feature relative to the ideal state corresponding to the benchmark dataset is first calculated to characterize the individual contribution of that feature to the deviation in production quality. Subsequently, all features are paired in pairs, with each pair forming a joint analysis pair to capture the synergistic deviation effect between features. In each feature combination, by considering the deviation magnitude of each of the two features and their statistical correlation, a corresponding joint deviation index is generated, thereby quantifying the overall deviation while reducing the interference of feature redundancy on the results. The calculation of the joint deviation index can be implemented using existing data analysis methods and functions. For example, it can be achieved by calculating the synergistic deviation magnitude of feature combinations, or by weighted fusion of joint deviations when the feature correlation is known. Through the joint analysis of pairwise feature combinations, this application can more comprehensively characterize the overall deviation of the current batch's production quality, improve the sensitivity to abnormal states and the reliability of judgment, thereby providing a scientific basis for production process adjustments. By combining the quality correlation feature correlation index and engineering importance parameter determined in the benchmark dataset, the single feature deviation degree and joint deviation index are weighted and fused to generate a weighted deviation degree set; wherein, the correlation index is used to weaken the repeated contribution of highly correlated features to the comprehensive deviation assessment, and the engineering importance parameter is used to highlight the key features that have a greater impact on sealing quality.A multi-constraint summary analysis is performed on the weighted deviation set to generate a comprehensive deviation index for the current production batch based on a preset fusion function. This fusion function considers single-feature deviation contributions, feature cross-deviation contributions, and correlation constraints during the summary process to quantify the overall deviation of the current production batch's quality status from the baseline state. The summary analysis of the weighted deviation set is completed through a preset fusion processing mechanism to generate the comprehensive deviation index for the current production batch. The core idea of ​​this fusion processing mechanism is to simultaneously consider single-feature deviation contributions, the joint deviation contributions of pairwise feature combinations, and the correlation constraints between features, thereby quantifying the deviation of the current batch's overall quality status from the baseline state. First, the single-feature deviation of each quality-related feature is combined with its corresponding weighting coefficient to reflect the independent impact of each feature on the overall quality status. Subsequently, statistical analysis is performed on the joint deviation index formed by pairwise feature combinations to capture the synergistic deviation effect between features, and by fusing with the single-feature deviation, the interference of redundant information on the comprehensive evaluation is eliminated. In this process, the correlation constraints between features are also considered, and deviation weights are reasonably allocated to moderately adjust the deviation contribution of highly correlated features, so as to prevent a single deviation from being repeatedly included in the comprehensive index. It is worth noting that the fusion processing mechanism described in this application is part of the technical solution, and its implementation can be achieved through existing data processing methods, such as weighted summarization, collaborative deviation quantification, and correlation adjustment, without being limited to specific formula forms. It can feasiblely quantify the deviation contribution of single features and cross-features in a unified manner in engineering, thereby obtaining a comprehensive deviation index that accurately reflects the overall quality status of the production batch. This mechanism not only ensures the reliability of the quantification results but also provides an operable scientific basis for quality control and process optimization in the production process. Based on the correspondence between the comprehensive deviation index and the target quality judgment benchmark interval, the current production status is graded and judged, and a quality status identifier of normal or warning status is output; the quality status identifier serves as the input basis for subsequent process parameter adjustment and production process control. For example, suppose that in the historical production process, a target quality judgment benchmark interval was established for a certain type of spring seat seal, including a normal interval, a warning interval, and an abnormal interval. The comprehensive deviation index range for each interval can be obtained through statistical analysis of historical batch data: Normal interval: comprehensive deviation index between 0 and 0.3; Warning interval: comprehensive deviation index between 0.3 and 0.6; Abnormal interval: comprehensive deviation index greater than 0.6. In the current production batch, by weighting and summing the various quality-related characteristics and their cross-deviation contributions, the comprehensive deviation index is obtained as 0.45. Based on the above baseline interval, this index falls into the warning interval; therefore, the current production status is determined to be "warning status".The judgment results can then be used to guide process adjustments, such as prompting operators to fine-tune end-face machining parameters, spring assembly forces, or lubrication conditions to reduce deviations and ensure that the next batch of production reaches a stable quality state. Simultaneously, this status indicator can serve as input for subsequent adjustments to process parameters in the automated control system, forming a closed-loop quality control mechanism. In this way, combining historical data statistics and benchmark interval mapping, it is possible to clearly describe how comprehensive deviation indicators are mapped to graded quality states, achieving a closed loop of production process control and optimization.

[0025] S600: When the quality status is marked as a warning, generate process adjustment suggestions for end face machining and spring assembly, and output corresponding adjustment instructions.

[0026] Furthermore, step S600 of this application also includes: S601: Analyze the comprehensive deviation index to identify the deviation features that have a major impact on the current production status and their corresponding process steps, generating deviation source identification results; S602: Map the deviation source identification results to end face machining process parameters and spring assembly process parameters to determine the target process parameter set corresponding to the deviation features; S603: Based on the target process parameter set and the adjustment rules corresponding to the benchmark data, generate process adjustment suggestions for end face machining and spring assembly processes; S604: Convert the process adjustment suggestions into executable adjustment instructions and output them to the corresponding processing equipment or assembly control system to guide production process adjustments.

[0027] Specifically, when the quality status is identified as a warning state, process adjustment suggestions for end-face machining and spring assembly are generated, and corresponding adjustment instructions are output to achieve closed-loop quality control of the production process. This step can detect potential deviations in advance during production and guide the operation or automatic control system to make targeted adjustments, thereby reducing product quality fluctuations and ensuring the sealing performance and assembly reliability of the seals. The comprehensive deviation index of the current production batch is analyzed to determine the contribution of each quality-related feature, identifying the deviation features that have the main impact on the current production status and their corresponding process steps. By mapping the deviation features to the process steps, it can be determined which end-face machining parameters (such as end-face flatness adjustment, surface roughness control, machining temperature, or lubrication conditions) and spring assembly parameters (such as preload, fit clearance, or equivalent stiffness) play a dominant role in the current deviation state, thus generating deviation source identification results. These deviation source identification results are then mapped to end-face machining process parameters and spring assembly process parameters to determine the target process parameter set corresponding to the deviation features. Based on the engineering importance and relevance of each deviation characteristic, the system prioritizes adjusting the process parameters that have the greatest impact on the overall quality, forming an optimized adjustment plan for the current batch. Based on the target process parameter set and the corresponding adjustment rules in the benchmark dataset, process adjustment suggestions are generated for end-face machining and spring assembly. These suggestions may include parameter fine-tuning ranges, adjustment priorities, and executable steps, ensuring that the process optimization plan maximizes the reduction of quality deviations while maintaining production efficiency. The process adjustment suggestions are converted into executable adjustment instructions and output to the corresponding machining equipment or assembly control system to guide production process adjustments. These instructions can directly drive CNC machining equipment or assembly robots, or they can be manually intervened through operator prompts, thus forming a closed-loop feedback for quality control. For example, when the overall deviation index of a batch of spring seat seals is too high and falls into the warning range, S601 can identify the main deviation characteristics as high end face roughness and insufficient spring preload; S602 maps this deviation to process parameters such as end face machining speed, cutting depth and spring pressing force; S603 generates specific adjustment suggestions based on the adjustment rules in the benchmark dataset, such as reducing the machining speed by 10% and increasing the spring pressing force by 5N; S604 converts the adjustment suggestions into CNC machining programs or robot assembly instructions. After the system executes them, the end face roughness and preload of the next batch of products can be restored to the target range, thereby ensuring stable sealing performance.

[0028] S700: Records the end face quality characteristic parameters, spring loading characteristic parameters, sealing quality evaluation value, comprehensive deviation index and corresponding quality status identifier of the current production batch, and updates and optimizes the benchmark dataset based on the accumulated historical production data.

[0029] Specifically, after completing the quality evaluation and process adjustment of the current production batch, the system records the end-face quality characteristic parameters, spring loading characteristic parameters, sealing quality evaluation value, comprehensive deviation index, and corresponding quality status identifier of the current production batch, and stores them in the production process data archive. Through the summary analysis of accumulated historical production data, the benchmark dataset is dynamically updated and optimized to achieve continuous adaptive adjustment of quality judgment and production control rules, thereby ensuring the accuracy of the judgment benchmark and the long-term stability of the production process. Specifically, the end-face quality characteristic parameters and spring loading characteristic parameters of the current batch are classified, standardized, and structured for storage, forming a dataset that can be directly used for statistical analysis. This dataset includes not only original characteristic parameters but also derived characteristic parameters, weighted deviation degrees, and comprehensive deviation indices, ensuring that subsequent analysis can cover all parameter dimensions that affect the quality status. The quality status judgment results of the current batch are compared and analyzed with those of historical batches to evaluate the rationality and deviation trend of the current judgment results. By statistically analyzing the distribution range, mean, and fluctuation range of each characteristic parameter in historical valid data samples, long-term deviations or systematic deviations are identified, providing a basis for optimizing the benchmark dataset. Based on the accumulated historical production data and current batch data, the judgment intervals of each quality characteristic parameter in the benchmark dataset are adaptively adjusted. The adjustment process may include updating the boundary values ​​of the normal, warning, and abnormal intervals, recalculating the engineering weight coefficients of each characteristic parameter, and optimizing the fusion function parameters to make the quality status judgment of subsequent batches more accurate and reliable. The optimized benchmark dataset is stored and applied to the next batch of production, achieving closed-loop quality control of the production process. This dynamic update mechanism can automatically adjust the judgment rules according to changes in production process conditions, material batches, or environment, thereby maintaining the consistency and stability of the seal end face quality and spring assembly. For example, when a batch is found to have end face flatness slightly higher than the historical average and spring preload slightly lower, S701–S702 will record these parameters and deviations; S703 updates the judgment interval according to the historical distribution to ensure that even if similar deviations occur in the next batch, the system can still accurately determine the quality status and generate optimized process suggestions; S704 applies the updated benchmark dataset to the next batch of processing and assembly to achieve stable control of end face quality and sealing force distribution, improve product consistency, and reduce rework rates.

[0030] In summary, the quality control method for spring seat seal production process provided by this invention has the following technical effects: By constructing a quality characteristic parameter system and a judgment benchmark dataset based on historical production data, the method achieves full-process monitoring and analysis of end-face quality characteristic parameters and spring loading characteristic parameters. The method can perform structured collection of production data from each batch, verify consistency of operating conditions, and screen key features to generate quality-related features reflecting the lubrication state of the sealing end face and the uniformity of sealing force distribution. Through comprehensive deviation analysis, it quantifies the degree of deviation in the overall production state, achieving accurate judgment and graded early warning of quality status. Simultaneously, by mapping deviation features to end-face machining and spring assembly process parameters, this method generates targeted process adjustment suggestions and outputs executable instructions, achieving closed-loop control of the production process. Furthermore, by recording the characteristic parameters and judgment results of each batch and dynamically updating the benchmark dataset based on historical data, the invention achieves adaptive optimization of judgment rules. This improves product consistency, reduces rework rates and production costs while ensuring the stability of end-face quality and sealing force distribution, enhancing the reliability and controllability of the production process. This provides a scientific and quantifiable quality control solution for the large-scale stable production of spring seat seals.

[0031] Example 2, as Figure 2 This is a schematic diagram of the quality control device for the spring seat seal production process of the present invention. For example, Figure 1 The flowchart of the quality control method for spring seat seal production process of the present invention can be seen as follows: Figure 2 The structure shown is implemented.

[0032] Based on the same concept as the quality control method for the spring seat seal production process described in the embodiments, the present invention also provides a quality control device for the spring seat seal production process, comprising: The benchmark data construction module 11 is used to collect quality data of spring seat sealing end face, spring loading data and sealing performance test data of historical production batches, analyze the parameter variation law under different batches and working conditions, determine the quality characteristic parameter system that characterizes the end face lubrication stability and sealing force distribution uniformity, and construct a benchmark dataset for quality judgment. The end face feature extraction module 12 is used to collect defect parameters such as end face flatness, surface roughness, scratches, and pits, and to preprocess the collected parameters to form a set of end face quality feature parameters. The spring feature extraction module 13 is used to obtain the spring loading feature parameters of the spring seat under test under the corresponding working conditions, including the spring preload, fit clearance, equivalent stiffness and spring seat deformation, and to obtain the corresponding end face quality feature parameters. The feature association calculation module 14 is used to perform association analysis on the end face quality parameters and spring loading parameters in the current feature parameter set, and generate quality association features that reflect the end face lubrication state and the uniformity of sealing force distribution. The quality evaluation calculation module 15 is used to input the quality correlation features into the quality evaluation function, which maps the quality correlation features to the sealing quality evaluation value; and to calculate and classify the sealing quality evaluation value based on the corresponding quality judgment rules in the benchmark dataset to obtain the quality status identifier. The process strategy generation module 16 is used to generate process adjustment suggestions for end face machining and spring assembly when the quality status is identified as a warning status, and output corresponding adjustment instructions. The data self-updating module 17 is used to record the end face quality characteristic parameters, spring loading characteristic parameters, sealing quality evaluation value, comprehensive deviation index and corresponding quality status identifier of the current production batch, and to update and optimize the benchmark dataset based on the accumulated historical production data.

[0033] It should be understood that the focus of the embodiments mentioned in this specification is their difference from other embodiments. The specific embodiments in the aforementioned Embodiment 1 are also applicable to the quality control device for the spring seat seal production process described in Embodiment 2. For the sake of brevity, they will not be elaborated further here.

[0034] It should be understood that the embodiments disclosed in this invention and the above description enable those skilled in the art to implement this invention. However, this invention is not limited to the embodiments mentioned above. It should be understood that those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this invention, and should all be included within the protection scope of this invention.

Claims

1. A quality control method for the production process of spring seat seals, characterized in that, include: S100. Collect quality data of spring seat sealing end face, spring loading data and sealing performance test data of historical production batches, analyze the parameter variation law under different batches and working conditions, determine the quality characteristic parameter system characterizing end face lubrication stability and sealing force distribution uniformity, and construct a benchmark dataset for quality judgment. S200: Collect defect parameters such as end face flatness, surface roughness, scratches, and pits, and preprocess the collected parameters to form a set of end face quality characteristic parameters. S300: Obtain the spring loading characteristic parameters of the spring seat under test under the corresponding working conditions, including spring preload, fit clearance, equivalent stiffness and spring seat deformation, and obtain the corresponding end face quality characteristic parameters. S400. Perform correlation analysis on the end face quality parameters and spring loading parameters in the current feature parameter set to generate quality correlation features reflecting the end face lubrication state and the uniformity of sealing force distribution. The generated quality correlation features reflecting the end face lubrication state and the uniformity of sealing force distribution include: S401. Based on the operating conditions, the end face quality characteristic parameters and spring loading characteristic parameters are normalized to form a standardized set of end face quality characteristic parameters. With the standardized set of spring loading characteristic parameters ; S402. Construct lubrication stability characteristic quantities to characterize the lubrication state of the end face. The formula is: ； in, To normalize the flatness of the end face, To normalize the surface roughness, To normalize the characteristic parameters of end face defects, To normalize the end face hardness, For processing temperature, The target temperature for the process, The viscosity of the lubricating oil. To design the target viscosity, This is a sensitivity parameter for the influence of end-face defect characteristics on lubrication stability; S403. Construct a sealing force distribution characteristic quantity to characterize the uniformity of sealing force distribution. It can be represented as: ； in: This is the spring preload. To accommodate the gap, To achieve the target fit gap, The equivalent stiffness of the spring. This is the deformation of the spring seat. For other environmental or process-related parameters, For the corresponding target value, To account for the number of environmental / process factors, These are characteristic parameters of environmental or process disturbances. S404, Lubrication stability characteristic quantity With sealing force distribution characteristic quantity Perform combined analysis to generate quality correlation features that simultaneously reflect the lubrication status of the sealing end face and the uniformity of the sealing force distribution. Q The formula is ： Q= ； in, Q The larger the value, the more stable the sealing quality, the more complete the lubricating film, and the more uniform the sealing force distribution. S500. Input the quality association features into the quality evaluation function, which maps the quality association features to a sealing quality evaluation value; calculate and classify the sealing quality evaluation value based on the corresponding quality judgment rules in the benchmark dataset to obtain a quality status identifier; S600. When the quality status is marked as a warning, generate process adjustment suggestions for end face machining and spring assembly, and output corresponding adjustment instructions. S700: Record the end face quality characteristic parameters, spring loading characteristic parameters, sealing quality evaluation value, comprehensive deviation index and corresponding quality status identifier of the current production batch, and update and optimize the benchmark dataset based on the accumulated historical production data.

2. The quality control method for the spring seat seal production process as described in claim 1, characterized in that, Construct a benchmark dataset for quality status determination, including: S101. Filter the retrieved historical production data of the spring seat seal, remove records with abnormal working conditions, detection failures or missing data, and retain valid historical data samples corresponding to the target working conditions to form a set of valid historical data samples. S102. Extract quality characteristic parameters related to the lubrication stability and sealing force distribution characteristics of the sealing end face from the historical valid data sample set. The quality characteristic parameters include end face flatness, surface roughness, end face defect characteristic parameters, as well as spring preload, fit clearance, equivalent stiffness and spring seat deformation, to form a set of quality characteristic parameters. S103. Perform statistical analysis on the set of quality characteristic parameters to obtain the distribution range, mean and fluctuation range of each quality characteristic parameter under different production batches and target working conditions, which are used to characterize the typical quality characteristic distribution of the spring seat seal under stable production conditions. S104. Based on the typical quality characteristic distribution, determine the quality judgment benchmark interval corresponding to each quality characteristic parameter, and classify the quality judgment benchmark interval according to the engineering requirements of end face lubrication stability and sealing force distribution uniformity to form normal interval, warning interval and abnormal interval. S105. Based on the end face quality characteristic parameters, spring loading characteristic parameters and their corresponding historical quality status identifiers of historical production batches, each quality characteristic parameter is associated and integrated with its corresponding quality judgment benchmark interval to construct a benchmark dataset for quality status judgment, and the benchmark dataset is stored for subsequent quality evaluation and production status judgment.

3. The quality control method for the spring seat seal production process as described in claim 1, characterized in that, Obtain the spring loading characteristic parameters of the spring seat under test under the corresponding working conditions, including spring preload, fit clearance, equivalent stiffness, and spring seat deformation, and obtain the corresponding end face quality characteristic parameters, including: S301. Based on actual working conditions, the obtained end face quality characteristic parameters and spring loading characteristic parameters are checked for consistency. Parameter data that do not meet the target working conditions are corrected to ensure that the characteristic parameters are consistent with the working conditions of the benchmark dataset. S302. Perform stability and effectiveness analysis on the characteristic parameters after consistency verification, remove characteristic parameters with insufficient stability, and retain key characteristic parameters that have a significant impact on the lubrication state of the sealing end face and the distribution of sealing force. S303. Based on the structural correlation between the retained end-face quality characteristic parameters and the spring loading characteristic parameters, derived characteristic parameters are constructed. The derived characteristic parameters include combined force-sensitive parameters and correlation parameters that reflect the influence of spring loading changes on the end-face contact uniformity. S304. The screened original feature parameters and the derived feature parameters are structurally integrated to form a complete set of feature parameters for the current production batch, which is used for subsequent quality correlation analysis and quality status evaluation.

4. The quality control method for the spring seat seal production process as described in claim 2, characterized in that, The quality-related features are input into a quality evaluation function, which is used to map the quality-related features into sealing quality evaluation values. Based on the quality judgment rules corresponding to the benchmark dataset, the sealing quality evaluation value is calculated and graded to obtain a quality status identifier, including: S501. The quality-related features are vectorized to construct the quality-related feature vector of the current production batch. S502. Input the quality-related feature vector into the quality evaluation function for calculation to generate the corresponding sealing quality evaluation value; S503. Perform deviation analysis on the sealing quality evaluation value and the central ideal state evaluation value of the benchmark dataset to generate a comprehensive deviation index that reflects the overall deviation of the current production batch. S504. Based on the correspondence between the comprehensive deviation index and the preset judgment threshold, the current production status is classified and judged, and a quality status identifier is output.

5. The quality control method for the spring seat seal production process as described in claim 4, characterized in that, The quality-related feature vector is input into the quality evaluation function for calculation to generate the corresponding sealing quality evaluation value, including: S502-1. Obtain the quality judgment benchmark range that matches the current batch operating conditions, including the normal range, warning range, and abnormal range. Perform cross-validation based on the judgment results of historical batches to select the most robust judgment range. S502-2, The quality-related feature vector C= Input the quality evaluation function, and select an existing feature mapping model based on the benchmark dataset. M (·), perform function operations on each quality-related feature component to obtain the corresponding feature evaluation component E= ; in, = M (·), i= 1,2 ， n ; S502-3. Perform a combination operation on the feature evaluation component E to generate a sealing quality evaluation value that characterizes the sealing quality status of the current production batch. The formula is: = × ； in, These are the feature weight coefficients determined based on the benchmark dataset.

6. The quality control method for the spring seat seal production process as described in claim 4, characterized in that, A deviation analysis is performed on the sealing quality evaluation value and the benchmark dataset's centralized ideal state evaluation value to generate a comprehensive deviation index reflecting the overall deviation of the current production batch, including: S503-1. Based on the feature composition of the sealing quality evaluation value, the deviation of each quality-related feature from the corresponding ideal state in the benchmark dataset is quantified to form a single feature deviation set, and on this basis, deviation correlation pairs of feature combinations are constructed to calculate the corresponding joint deviation index. S503-2. Combining the feature correlation index determined based on the benchmark dataset and the engineering importance coefficient, the single feature deviation degree and the joint deviation index are weighted and fused to generate a weighted deviation degree set that reflects the degree of contribution of different deviations. S503-3. Perform multi-constraint summary analysis on the weighted deviation set, and generate a comprehensive deviation index for the current production batch based on a preset fusion function; in, The fusion function comprehensively considers at least the single feature deviation contribution, the feature cross deviation contribution, and the correlation constraint term, and is used to quantify the overall deviation of the current production batch's overall quality status from the baseline status.

7. The quality control method for the spring seat seal production process as described in claim 6, characterized in that, When the quality status is marked as a warning, process adjustment suggestions are generated for the end face machining and spring assembly processes, and corresponding adjustment instructions are output, including: S601. Perform composition analysis on the comprehensive deviation index, identify the deviation characteristics that have a major impact on the current production status and their corresponding process links, and generate deviation source identification results; S602. Map the deviation source identification result to the end face processing parameters and spring assembly process parameters to determine the target process parameter set corresponding to the deviation feature; S603. Based on the target process parameter set and the corresponding adjustment rules in the benchmark dataset, generate process adjustment suggestions for end face machining and spring assembly. S604. Convert the process adjustment suggestions into executable adjustment instructions and output them to the corresponding processing equipment or assembly control system to guide production process adjustments.

8. A quality control device for the production process of spring seat seals, characterized in that, The quality control method for the spring seat seal production process according to any one of claims 1-7 includes: The benchmark data construction module is used to collect quality data of spring seat sealing end face, spring loading data and sealing performance test data of historical production batches, analyze the parameter variation law under different batches and working conditions, determine the quality characteristic parameter system that characterizes the end face lubrication stability and sealing force distribution uniformity, and construct a benchmark dataset for quality judgment. The end face feature extraction module is used to collect defect parameters such as end face flatness, surface roughness, scratches, and pits, and to preprocess the collected parameters to form a set of end face quality feature parameters. The spring feature extraction module is used to obtain the spring loading feature parameters of the spring seat under test under the corresponding working conditions, including spring preload, fit clearance, equivalent stiffness and spring seat deformation, and to obtain the corresponding end face quality feature parameters. The feature correlation calculation module is used to perform correlation analysis on the end face quality parameters and spring loading parameters in the current feature parameter set, and generate quality correlation features that reflect the end face lubrication state and the uniformity of sealing force distribution. The quality evaluation calculation module is used to input the quality correlation features into the quality evaluation function, which maps the quality correlation features to the sealing quality evaluation value; based on the corresponding quality judgment rules in the benchmark dataset, the sealing quality evaluation value is calculated and classified to obtain the quality status identifier; The process strategy generation module is used to generate process adjustment suggestions for end face machining and spring assembly when the quality status is marked as a warning, and output corresponding adjustment instructions. The data self-updating module is used to record the end face quality characteristic parameters, spring loading characteristic parameters, sealing quality evaluation value, comprehensive deviation index and corresponding quality status identifier of the current production batch, and to update and optimize the benchmark dataset based on the accumulated historical production data.

Images