A method and system for intelligent evaluation of gasket assembly fitment

Abstract

This invention discloses an intelligent evaluation method and system for gasket assembly compatibility, relating to the field of gasket assembly quality evaluation technology. The method includes: acquiring contour data of the gasket to be assembled, structural data of the assembly groove, and process timing response data during gasket assembly; constructing an initial bonding distribution model; extracting dynamic response features and determining corresponding assembly interface regions; correcting bonding state parameters of corresponding regions to generate a virtual bonding map; determining the regional bonding risk distribution; and generating gasket assembly compatibility evaluation results. This invention achieves a joint evaluation of regional bonding risk distribution and overall compatibility level, enabling the identification and location of hidden defects in the assembly process, providing a basis for selective rework and process optimization.

Description

Technical Field

[0001] This invention relates to the field of gasket assembly quality assessment technology, specifically to an intelligent assessment method and system for gasket assembly compatibility. Background Technology

[0002] As a core component for sealing in industrial equipment, the assembly quality of gaskets directly determines the sealing reliability and operational safety of the equipment throughout its service life. In fields with stringent leakage control requirements, such as automotive powertrains, hydraulic and pneumatic systems, precision instruments, and aerospace, the assessment technology for gasket assembly suitability has evolved from manual judgment and offline sampling to online dimensional monitoring. In recent years, with the in-depth application of intelligent manufacturing technologies, techniques such as machine vision-based geometric parameter measurement, sensor network-based assembly process monitoring, and computer-aided analysis-based virtual assembly simulation have been introduced into the field of gasket assembly quality assessment, driving the development of assessment methods from single static detection to multi-source information fusion. However, current technologies still face significant challenges in terms of the systematic nature, accuracy, and predictability of the assessment, especially in the lack of effective technological breakthroughs in the correlation modeling between the dynamic behavior of the assembly process and the final fit state.

[0003] Existing gasket assembly evaluation technologies have several limitations. First, mainstream evaluation methods focus on dimensional tolerance inspection of a single object, such as the gasket or assembly groove. For example, they extract the edge contour of the gasket or measure the key dimensions of the groove using a vision system. While these methods can determine whether the parts themselves meet design tolerance requirements, they struggle to characterize the spatial matching relationship and fit trend at the assembly interface, and cannot predict localized poor fit caused by form and position deviations. Second, some technical solutions attempt to introduce process parameter monitoring during assembly, such as real-time acquisition of pressing force or displacement data and plotting force-displacement curves. By judging the overall shape of the curves, they identify whether there are global anomalies in the assembly process. However, these methods lack a mechanism to establish a correspondence between local anomaly features in the time-series data and specific spatial areas of the assembly interface. This makes it difficult to accurately capture and locate latent defects such as local interference, short-term jamming, insufficient compression transmission, or inconsistent rebound that occur during assembly. These defects are often invisible in static appearance inspection after assembly, but they constitute important causes of subsequent seal failure. Furthermore, existing technologies generally lack a corrective mapping model between dynamic information during the assembly process and the static bonding state after assembly. Even if mechanical response data during the assembly process can be obtained, it cannot be transformed into a quantitative basis for correcting the final bonding state. This results in evaluation results that only reflect the theoretical bonding under ideal geometric conditions, leading to a systematic deviation from the actual interface state of the sealing gasket after undergoing the dynamic assembly process. In addition, the output of existing evaluation methods is mostly in the form of an overall pass / fail judgment or a single score, lacking an expression of the risk level distribution of various local areas of the assembly interface. This makes it difficult to provide spatially oriented decision support for selective rework, local process adjustments, or precise optimization of assembly parameters. Summary of the Invention

[0004] In view of the above-mentioned problems, the present invention is proposed.

[0005] Therefore, the technical problem solved by this invention is that existing gasket assembly evaluation methods have the following problems: they are difficult to characterize the spatial matching relationship and fitting trend between the gasket and the assembly groove at the assembly interface; they are unable to establish a correspondence between local abnormal features in the timing response data of the assembly process and specific spatial areas of the assembly interface; they have not established a correction mapping model between the dynamic information of the assembly process and the static fitting state after assembly; and they are concerned about how to achieve intelligent evaluation of gasket assembly adaptability by integrating the dynamic response features of the assembly process and the static geometric matching relationship.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: In a first aspect, embodiments of the present invention provide a smart evaluation method for the fit of a sealing gasket assembly, comprising the following steps: S1: Obtain the contour data of the gasket to be assembled, the structural data of the assembly groove, and the process timing response data during the gasket assembly process, and construct an initial bonding distribution model between the gasket and the assembly groove based on the contour data and the structural data of the assembly groove. S2: Analyze and process the process timing response data, extract dynamic response features that characterize the state changes during the assembly process of the sealing gasket, and determine the assembly interface area corresponding to the dynamic response features. S3: Based on the dynamic response characteristics, the bonding state parameters of the corresponding assembly interface area in the initial bonding distribution model are corrected to generate a virtual bonding map characterizing the actual bonding state of the sealing gasket assembly interface. S4: Determine the regional bonding risk distribution of the sealing gasket in the assembly interface based on the virtual bonding map; S5: Generate the gasket assembly compatibility assessment results based on the regional fit risk distribution.

[0007] As a preferred embodiment of the intelligent evaluation method for the fit of sealing gaskets described in this invention, step S1 specifically includes: Obtain the contour data of the sealing gasket to be assembled. The contour data is used to characterize the outer edge boundary, inner edge boundary, cross-sectional thickness distribution, corner transition shape, local raised parts, local recessed parts, and the continuous state of the joint. Obtain the structural data of the assembly groove, which includes the groove opening boundary position, groove bottom shape, groove wall contour direction, local groove width variation, local groove depth variation, corner structure transition state, and interference boundary. Acquire process timing response data during the assembly of the sealing gasket, including force changes, displacement changes, propulsion speed changes, springback state changes, local vibration changes, and temperature changes in the assembly area; Based on the contour data and the structural data of the assembly groove, an initial fit distribution model between the sealing gasket and the assembly groove is constructed. The initial fit distribution model is a basic fit state expression of each local area of ​​the assembly interface, which is used for subsequent dynamic correction and adaptability evaluation.

[0008] As a preferred embodiment of the intelligent evaluation method for the fit of sealing gaskets described in this invention, step S2 specifically includes: The process timing response data is preprocessed, including time alignment, noise suppression, burr removal, sampling interval unification, and stage boundary calibration of the raw data. The assembly process is divided into the initial entry stage, continuous advancement stage, local confinement stage, transition fitting stage, and clamping and stabilization stage according to the proportion of cumulative displacement to total propulsion displacement. Dynamic response features characterizing the state changes during the assembly of the sealing gasket are extracted from the pre-processed process timing response data. These dynamic response features include local force surge, staged force stagnation, enhanced force fluctuation, inconsistent rebound after unloading, decreased propulsion speed, interrupted propulsion rhythm, local motion stagnation, enhanced vibration amplitude, local frequency band energy concentration, and abnormal local temperature rise. The assembly interface region corresponding to the dynamic response feature is determined by determining the advancing position of the gasket in the assembly path when the dynamic response feature appears, based on the time sequence, displacement change, and advancing speed change corresponding to the process timing response data. Combining the correspondence between contour segments and structural segments established in the initial bonding distribution model, the advancing position is mapped to the corresponding region in the straight edge area, corner area, joint area, or local constraint enhancement area of ​​the assembly interface.

[0009] As a preferred embodiment of the intelligent evaluation method for the fit of sealing gaskets described in this invention, step S3 includes: Based on the dynamic response feature type output in step S2, the correction category for the corresponding assembly interface area is determined. If the dynamic response feature is that the force increase exceeds the preset increase range and the propulsion speed is lower than the preset speed range, the local gap state, local contact continuity state parameters, and local compression state parameters are corrected. If the force increase rate is lower than the preset increase rate, the local compression state parameters and local fit reliability state parameters are corrected. If the rebound deviation exceeds the preset deviation range, the fit stability state parameters, local suspension risk state parameters, and parameter reliability level are corrected. If the vibration amplitude exceeds the preset reference range and the energy ratio within the preset frequency band exceeds the preset ratio threshold, the boundary stability state parameters and local contact continuity state parameters are corrected. In the initial bonding distribution model, the bonding state parameters of the corresponding assembly interface area are read, and the area is determined as a first-level disturbed area, a second-level disturbed area, or a third-level disturbed area based on whether the dynamic response characteristics exceed the corresponding preset judgment range, duration, occurrence stage, and whether they appear in combination with other dynamic response characteristics. First-level correction is performed for the first-level disturbed area, second-level correction is performed for the second-level disturbed area, and third-level correction is performed for the third-level disturbed area. When performing regional parameter correction, for local gap state parameters, if there is local interference, jamming, or rebound exceeding the preset range in the corresponding region, the correction is made in the direction of increasing gap. For local compression state parameters, if the corresponding region has compression below the preset compression range, force growth rate below the preset growth rate, or propulsion speed below the preset speed range, the correction is made in the direction of decreasing compression. For local contact continuity state parameters and local bonding reliability state parameters, if the corresponding region has vibration amplitude exceeding the preset reference range, energy proportion within the preset frequency band exceeding the preset ratio threshold, or two or more dynamic response characteristics appearing simultaneously, the correction direction is determined based on the dominant dynamic response characteristics, and the correction is made in the direction of decreasing local contact continuity state parameters and decreasing local bonding reliability state parameters.

[0010] As a preferred embodiment of the intelligent evaluation method for the fit of sealing gaskets described in this invention, step S3 further includes: Centered on the region directly corresponding to the dynamic response characteristics, influence diffusion correction is performed on the adjacent regions before and after along the assembly path. Specifically, the central disturbed region is defined as the main correction region, and the adjacent regional units before and after it are defined as auxiliary correction regions. The main correction region is fully corrected, and the auxiliary correction region is attenuated. The magnitude of the attenuation correction is smaller than the correction magnitude of the main correction region.

[0011] As a preferred embodiment of the intelligent evaluation method for the fit of sealing gaskets described in this invention, step S4 includes: Read the set of bonding state parameters corresponding to each assembly interface area from the virtual bonding map, perform unified processing on each bonding state parameter, and establish a regional risk judgment unit. The regional risk judgment unit consists of one regional unit or multiple adjacent regional units. The sources of bonding risks in each regional risk assessment unit are classified and identified. These sources include gap mismatch risk, insufficient compression risk, contact interruption risk, boundary instability risk, and low-confidence prediction risk. Gap mismatch risk corresponds to the situation where the local gap state parameter exceeds the upper limit of the designed bonding gap in that region. Insufficient compression risk corresponds to the situation where the local compression state parameter is lower than the lower limit of the target compression range in that region. Contact interruption risk corresponds to the situation where the local contact continuity state parameter is lower than the preset contact continuity threshold. Boundary instability risk corresponds to the situation where the boundary stability state parameter deviates from the average level of adjacent regions by more than a preset proportion. Low-confidence prediction risk corresponds to the situation where the confidence level of the parameter is low. Based on the degree to which the fitting status parameters corresponding to each risk source deviate from the normal range, the risk levels are divided into low-level risk, medium-level risk, and high-level risk, and risk marking is performed on each regional-level risk judgment unit according to the risk levels.

[0012] As a preferred embodiment of the intelligent evaluation method for the fit of sealing gaskets described in this invention, step S4 further includes: Spatial continuity analysis is performed with high-level risk units as the center. If there are similar risk sources in adjacent regional units before and after the high-level risk unit, and the risk level of the adjacent units is not lower than that of the low-level units, then the continuous area formed by the high-level risk unit and the adjacent regional units that meet the conditions of similar risk sources and risk levels is defined as a risk zone. If different types of risk sources with coupling relationships appear in the adjacent regional units around the high-level risk unit, then it is defined as a composite risk zone. The assembly interface is divided into a regular area, a sensitive area, and a highly sensitive area. The regular area corresponds to the straight edge area, the sensitive area corresponds to the corner area or the adjacent area of ​​the joint, and the highly sensitive area corresponds to the joint center area, the local slot width abrupt change area, or the local boundary interference area. For risk assessment units located in the sensitive area, the risk level is increased by one level based on the original risk level, but not exceeding the higher level. For units located in the highly sensitive area and with two or more risk sources at the same time, they are directly marked as high-risk areas.

[0013] As a preferred embodiment of the intelligent evaluation method for the fit of sealing gaskets described in this invention, step S5 includes: The risk distribution of the region output in step S4 is subjected to risk summary processing to form a global risk overview. The global risk overview includes statistics on the number of high-risk areas, the number of early warning areas, the length of risk zones, the number of composite risk areas, the risk units located in sensitive areas, the risk units located in highly sensitive areas, and the risk category distribution. Based on the overall risk overview, the compatibility of the sealing gasket assembly is assessed, and the compatibility level is divided into four levels: compatibility pass, compatibility concern, compatibility warning, and compatibility failure. Compatibility pass corresponds to a situation where the number of high-risk areas is zero and there are no risk zones. Compatibility concern corresponds to a situation that does not fall into the categories of compatibility pass, compatibility warning, or compatibility failure. Compatibility warning corresponds to a situation where the number of high-risk areas reaches the second preset number but not the third preset number, or where there are composite risk areas but the number of composite risk areas does not reach the fourth preset number. Compatibility failure corresponds to a situation where the number of high-risk areas reaches the third preset number, or where there is a high-risk area located within a highly sensitive area, or where the number of composite risk areas reaches the fourth preset number. The second preset number is less than the third preset number. The second, third, and fourth preset numbers are determined based on the total path length of the assembly interface, the density of area unit divisions, and the sealing requirement level.

[0014] As a preferred embodiment of the intelligent evaluation method for the fit of sealing gaskets described in this invention, step S5 further includes: Extract the dominant mismatch factor. The dominant mismatch factor is a risk category that has a major impact on the overall suitability level, which is determined by a combination of factors including the risk level, the structural sensitivity level of the risk area, the length of continuous risk extension, and the number of compound risk sources. The dominant risk category in the high-risk area is selected as the primary dominant mismatch factor. If there is no high-risk area, the dominant risk category in the risk zone is selected as the primary dominant mismatch factor. Generate a list of weak links for adaptation, and include areas that meet one of the following conditions in the list of weak links for adaptation: areas marked as high-risk areas, areas located in sensitive areas with a risk level of warning or above, areas located in highly sensitive areas with a risk level of warning or above, areas that constitute risk zones, and areas that constitute composite risk zones. Simultaneously record the starting position, ending position, dominant risk category, and risk level information of the areas included in the list.

[0015] Secondly, embodiments of the present invention provide an intelligent evaluation system for the fit of a sealing gasket assembly, comprising: Data modeling module: acquires the contour data of the gasket to be assembled, the structural data of the assembly groove, and the process timing response data during the gasket assembly process, and constructs an initial fit distribution model between the gasket and the assembly groove based on the contour data and the structural data of the assembly groove. Feature extraction module: Analyzes and processes the process timing response data, extracts dynamic response features that characterize the state changes during the gasket assembly process, and determines the assembly interface area corresponding to the dynamic response features; Atlas generation module: Based on the dynamic response characteristics, the bonding state parameters of the corresponding assembly interface area in the initial bonding distribution model are corrected to generate a virtual bonding atlas characterizing the actual bonding state of the sealing gasket assembly interface. Risk analysis module: Determines the regional adhesion risk distribution of the sealing gasket in the assembly interface based on the virtual bonding map; Evaluation output module: Generates gasket assembly compatibility evaluation results based on the regional fit risk distribution.

[0016] The beneficial effects of this invention are as follows: By constructing an initial bonding distribution model and introducing assembly process timing response data, this invention establishes a spatial mapping relationship between dynamic response characteristics and assembly interface areas, performs directional correction on bonding state parameters, and generates a virtual bonding map that reflects the actual bonding state of the assembly interface. This enables a joint assessment of regional bonding risk distribution and overall compatibility level. Compared with existing technologies, this invention can identify and locate hidden defects occurring during assembly before assembly is completed, making the assessment results closer to the actual service state of the gasket. This provides a spatially directional decision-making basis for selective rework and precise process optimization. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort, wherein: Figure 1 This is an overall flowchart of a smart evaluation method for gasket assembly compatibility provided in the first embodiment of the present invention; Figure 2 This is a module connection diagram of an intelligent evaluation system for gasket assembly compatibility provided in the third embodiment of the present invention. Detailed Implementation

[0018] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0019] Example 1, referring to Figure 1 As an embodiment of the present invention, a smart evaluation method for the fit of a sealing gasket assembly is provided.

[0020] S1: Obtain the contour data of the gasket to be assembled, the structural data of the assembly groove, and the process timing response data during the gasket assembly process, and construct an initial bonding distribution model between the gasket and the assembly groove based on the contour data and the structural data of the assembly groove.

[0021] In this embodiment, step S1 is mainly used to acquire the basic data required for assembly compatibility assessment and establish an initial fit distribution model between the gasket and the assembly groove. It should be noted that the initial fit distribution model established in this step does not directly represent the final fit state of the gasket after actual assembly. Instead, it is a basic description of the initial fit trend of the assembly interface based on the spatial matching relationship between the gasket's own contour shape and the assembly groove's structural boundary, without introducing the influence of dynamic disturbances during the assembly process. The purpose of setting up this model is to provide a clear spatial basis for the process timing response characteristics extracted in subsequent steps, enabling subsequent dynamic corrections to act on specific areas of the assembly interface, rather than remaining at the level of independent data analysis detached from the assembly object.

[0022] In practice, the first step is to acquire the contour data of the gasket to be assembled. This contour data preferably reflects the outer and inner boundaries, cross-sectional thickness distribution, corner transitions, local bulges, local depressions, and the continuity of the joint area of ​​the gasket before assembly. For annular gaskets, the boundary contour information of each segment can be collected sequentially along its continuous extension direction; for irregularly shaped gaskets, they can be divided into multiple contour segments according to their actual geometric orientation, collected separately, and then spliced ​​together to ensure the continuity and integrity of the contour representation. It should also be noted that the acquisition of contour data is not limited to a specific device. The gasket boundary can be extracted through visual imaging, the actual geometric undulations of the gasket surface can be obtained through contour scanning, and compensation and correction can be performed by combining multi-view acquisition results under tooling positioning conditions. Using the above methods to acquire contour data can more realistically preserve the geometric shape characteristics of the gasket before it enters the assembly station, providing a basis for subsequently determining its theoretical fit with the assembly groove.

[0023] Simultaneously, structural data of the assembly groove also needs to be acquired. This structural data primarily characterizes the assembly constraints and local limiting conditions encountered during the installation of the sealing gasket. Therefore, it typically includes the groove opening boundary position, groove bottom shape, groove wall contour direction, local groove width variations, local groove depth variations, corner structure transition states, and interference boundaries that may affect the gasket's entry and fit. If the assembly groove itself has local dimensional deviations, stepped structures, inconsistent transition fillets, or discontinuous edges, this information can also be included in the structural data to improve the adaptability of subsequent modeling results to real-world conditions. Furthermore, structural data can be acquired by an independent inspection station before assembly, or extracted in real-time by the inspection unit on the assembly equipment after the workpiece enters the assembly area. For batches of workpieces with high structural consistency, corrections can be made by combining pre-stored standard structural information with the calibration information of the current workpiece. Through these processes, it can be ensured that the assembly groove structural data not only reflects the geometric boundaries under design conditions but also reflects the constraint differences under actual assembly conditions.

[0024] While acquiring static geometric information, this step also includes acquiring process timing response data during the gasket assembly process. It should be noted that this process timing response data differs from simple equipment operation records; its focus is on characterizing the process response of the gasket as it changes over time during actual insertion, advancement, compression, and bonding. Therefore, it should possess a clear temporal sequence and process continuity. This type of data can originate from assembly equipment, pressing mechanisms, positioning execution components, or station detection units, specifically reflecting one or more of the following during the pressing process: force changes, displacement changes, advancement speed changes, springback state changes, local vibration changes, and temperature changes in the assembly area. When acquiring this type of data, it is advisable to use a unified time reference for continuous recording and to keep it synchronized with the assembly execution trajectory. Furthermore, during the recording of process timing response data, displacement progress information, execution trajectory information, or stage position information during the assembly advancement process can also be recorded simultaneously, so as to establish a correspondence between process timing response characteristics and local areas of the assembly interface in subsequent steps. The significance of this approach is that subsequent steps can not only identify the state differences of the gasket at different assembly stages based on temporal changes, but also pinpoint the corresponding state differences to specific areas of the assembly interface.

[0025] After acquiring the aforementioned contour data, structural data, and process timing response data, an initial bonding distribution model between the gasket and the assembly groove is further constructed based on the contour and structural data. It should be noted that this construction process is not a simple comparison of the two types of geometric data. Instead, it first aligns the gasket contour data and the assembly groove structural data under a unified spatial reference, based on the assembly reference edge, positioning reference point, clamping reference surface, or preset assembly posture, enabling virtual assembly positioning under the same spatial reference relationship. Subsequently, according to the actual path sequence of the gasket entering the assembly groove, the gasket contour segments are matched one-to-one with the assembly groove structural segments to form a basic bonding state expression for each local area of ​​the assembly interface. Furthermore, to ensure that subsequent regional corrections have a clear target, the assembly interface can be divided into local areas based on the gasket contour direction, changes in the assembly groove structure, and the distribution of corners, joints, and locally reinforced areas, thus forming a regional bonding state basis.

[0026] Based on the aforementioned spatial alignment and regional correspondence, the initial gap trend, initial compression trend, initial contact establishment trend, and local interference tendency of each local region under ideal entry conditions can be further determined. The set of fitting state parameters corresponding to each local region is then used to form an initial fitting distribution model oriented towards the assembly interface. It should be noted that the initial contact establishment trend mentioned here refers to the degree of tendency for the local contour of the sealing gasket to form contact with the corresponding structure of the assembly groove, without considering the influence of dynamic disturbances during the assembly process. This can be comprehensively characterized based on the initial gap state and initial compression allowance of the region. It should also be noted that the technical function of this initial fitting distribution model is not to directly provide a final judgment on whether the assembly is qualified, but to convert the original geometric contour information and structural constraint information into a regionalized fitting state basis that can be expressed towards the assembly interface. In this way, the process timing response features extracted in subsequent steps are no longer independent process data detached from spatial objects, but can be applied to the corresponding regions in this initial fitting distribution model based on the positional correspondence during the assembly process, allowing for targeted correction of the local fitting state. Compared to the method of performing static matching judgment using only pre-assembly dimensional information, the initial fitting distribution model established in this step has introduced the spatial correspondence, local constraint relationship and region division basis between the gasket contour and the assembly groove structure. Therefore, it can lay the foundation for the subsequent generation of virtual fitting maps that are closer to the real assembly interface state.

[0027] S2: Analyze and process the process timing response data, extract dynamic response features that characterize the state changes during the assembly process of the sealing gasket, and determine the assembly interface area corresponding to the dynamic response features.

[0028] In this embodiment, step S2 is mainly used to analyze and process the process timing response data obtained in step S1, extract dynamic response features that characterize the state changes during the gasket assembly process, and determine the assembly interface area corresponding to the dynamic response features, providing a basis for local correction of the initial bonding distribution model in subsequent steps. It should be noted that this step is not merely a general anomaly screening of the process timing response data, nor is it simply outputting the equipment operating status at a certain sampling moment. Instead, it combines the changes in the working conditions of the gasket during actual insertion, advancement, compression, and bonding processes to identify process features corresponding to local interference, local jamming, uneven compression, abnormal rebound, and unstable contact, and then assigns these process features to specific areas of the assembly interface. In other words, the dynamic response features extracted in this step are not arbitrary fluctuations in the time series, but rather feature manifestations that correspond to specific physical events or state transitions occurring during the gasket assembly process.

[0029] In practice, the process timing response data obtained in step S1 can be preprocessed first. This preprocessing primarily eliminates non-realistic fluctuations introduced during the acquisition process due to equipment jitter, sensor drift, instantaneous interference, and inconsistent sampling intervals, and provides a unified analytical basis for different types of process timing response data. During preprocessing, the original data can undergo time alignment, noise suppression, glitch removal, sampling interval unification, and stage boundary calibration. Glitch data can be defined as instantaneous jumps occurring between two consecutive sampling points, with the jump amplitude exceeding three times the average change amplitude of 10 adjacent sampling points. Sampling interval unification can be defined as organizing the original sampling sequence into a sequence with fixed time intervals, such as 1 millisecond, 5 millisecond, or 10 millisecond intervals. The effective fluctuations retained after noise suppression are preferably limited to change segments with a duration greater than 20 milliseconds to avoid mistaking a single sampling pulse for a real assembly response. It should also be noted that, to facilitate subsequent trend judgment and stage analysis, the process timing response data can be segmented according to preset analysis windows, with the duration of each analysis window set to 20 to 50 milliseconds. If the acquired process timing response data includes multiple data such as force changes, displacement changes, propulsion speed changes, springback changes, local vibration changes, and temperature changes, they can also be synchronized under a unified time reference to ensure that the dynamic response features extracted later correspond to the actual state changes in the same assembly stage, rather than originating from false deviations between asynchronous data.

[0030] Furthermore, after preprocessing, the process timing response data can be divided into stages based on the assembly process of the sealing gasket. It should be noted that sealing gasket assembly is typically not a single, uniform process, but can be divided into an initial entry stage, a continuous advancement stage, a locally confined stage, a transitional fitting stage, and a compression stabilization stage. In different stages, the stress state, deformation state, and contact state exhibited by the sealing gasket differ. Directly analyzing the complete timing data as a whole can easily mask the true state changes in local stages. Therefore, the assembly process can be divided into stages based on the pressing displacement progress, speed change trend, force change slope, springback change, or actuator operating cycle. For example, when the cumulative displacement is between 0% and 15% of the total advancing displacement, it can be defined as the initial entry stage; between 15% and 70%, it can be defined as the continuous advancement stage; between 70% and 90% with an increasing force growth rate, it can be defined as the locally confined stage or the transitional fitting stage; and above 90% with a slowing displacement change, it can be defined as the compression stabilization stage. The aforementioned ratio range is not the only limitation, and those skilled in the art can make equivalent adjustments based on the gasket size, assembly path length, and pressing stroke. This stage division method allows the subsequently extracted dynamic response characteristics to more closely approximate the actual behavior of the gasket during specific assembly stages.

[0031] During the dynamic response feature extraction process, key features that characterize the state changes during the gasket assembly process can be identified. These dynamic response features may include those reflecting changes in propulsion resistance, local interference trends, uneven compression trends, rebound hysteresis, and local contact instability. Specifically, for force and displacement change data, the trend of force increasing with displacement during assembly propulsion can be analyzed to identify state changes such as sudden increases in local force values, staged force stagnation, enhanced force fluctuations, and inconsistent rebound after unloading. Specifically, a sudden increase in local force value can be defined as a force value increase that reaches 1.5 to 2.5 times the average force value increase of the corresponding stage within a continuous displacement range; a staged force value stagnation can be defined as a force value growth rate that is lower than 50% of the baseline growth rate for more than three consecutive analysis windows under continuous displacement, wherein the duration of the analysis window is 20 to 50 milliseconds; enhanced force value fluctuation can be defined as the standard deviation of the force value within a certain analysis window reaching more than twice the baseline standard deviation for that stage; and inconsistent springback after unloading can be defined as the difference in springback amount between adjacent areas being greater than 20% and not higher than 40% of the average springback amount for that stage. It should also be noted that the above characteristics do not only indicate changes in the data values ​​themselves, but can also correspond to different physical states during the assembly process. For example, a sudden increase in force often reflects additional interference or short-term jamming between the local area of ​​the sealing gasket and the groove wall during the advancement process; a phased stagnation in force often reflects that although a certain section of the sealing gasket continues to advance, the actual compression transmission is insufficient; inconsistent rebound after unloading can be used to characterize the inconsistency between the recovery state of a local area after being compressed and that of the adjacent area, thus indicating that there may be unstable adhesion or local suspension risk in that area.

[0032] For data on changes in propulsion speed, it is possible to identify speed decreases, interruptions in propulsion rhythm, and localized motion stagnation during the propulsion process. Specifically, speed decreases are defined as the actual propulsion speed falling below 15% to 25% of the lower limit of the target propulsion speed range for the current stage; interruptions in propulsion rhythm are defined as the propulsion speed dropping below 10% of the lower limit of the target propulsion speed range and lasting for 50 to 150 milliseconds without the assembly command being terminated; localized motion stagnation is defined as a combination of continuously decreasing displacement increments and continuously increasing force values, lasting for more than 30 milliseconds, or corresponding to a displacement stroke greater than 0.2 mm. It should be noted that the target propulsion speed range mentioned here can be determined based on the process settings of the assembly equipment, the statistical speed range of the corresponding stage in historical normal assembly samples, or the stage speed parameters in the standard assembly procedure, thus ensuring that the determination of abnormal propulsion speeds has a clear reference basis. It should also be noted that if a decrease in speed is detected during the continuous advance phase, accompanied by a sudden increase in local force, the section can be further identified as having a tendency to be locally obstructed; if the advance rhythm is interrupted during the corner transition phase, the corresponding dynamic response characteristics can be used to indicate the possibility of curling, overturning, or uneven edge overlap in the corner area.

[0033] Vibration change data can identify situations such as increased vibration amplitude and concentrated energy in local frequency bands during a certain assembly stage, reflecting potential rubbing, squeezing, jumping, or local flipping trends of the sealing gasket during installation. Increased vibration amplitude is defined as a vibration peak reaching 1.8 to 2.5 times the reference peak value for the corresponding stage; concentrated energy in local frequency bands is defined as the proportion of energy within a preset frequency band exceeding 40% but not exceeding 60% of the total vibration energy, lasting for more than 30 milliseconds. Temperature change data can help identify localized heat accumulation caused by increased local friction, continuous pressure, or abnormal contact. For example, during continuous assembly, if the temperature rise in a corresponding area reaches 2 to 5 degrees Celsius above the ambient reference temperature at a certain stage, and the heating rate is more than 1.5 times higher than the average heating rate of the previous stage, it can be determined that there is a possibility of abnormally intensified local contact at that stage.

[0034] It should also be noted that when extracting dynamic response features, it is necessary to pay attention not only to the amplitude change itself, but also to the duration of the change, the order of occurrence, and the transition relationship with adjacent stages. For example, if a sudden increase in force value lasting 20 to 80 milliseconds occurs during the continuous advancement stage, and its increase exceeds twice the average increase of that stage, it can be determined that there is a local interference trend at that location; if, in the same section, there is a simultaneous decrease in advancement speed of more than 15% and an increase in vibration peak value of more than 1.8 times, it is more likely to indicate local edge jamming or abnormal obstruction at corners; if, during the compaction stabilization stage, the rebound amount still deviates from the average rebound amount of that stage by more than 25% but not more than 40%, it may indicate that a stable fit has not been formed in the corresponding area. By combining the analysis of single features with their occurrence stages and combined relationships, misjudgments caused by occasional sampling fluctuations can be effectively reduced, and the ability of dynamic response features to represent the actual assembly state can be improved.

[0035] It should also be noted that the aforementioned range of force increase, propulsion speed, force growth rate, rebound deviation, and energy percentage threshold within the preset frequency band can all be determined based on the baseline response of the corresponding assembly stage, the statistical results of historical normal assembly samples, and the assembly process setting range. This provides a clear reference basis for determining the dynamic response characteristics in different assembly stages and ensures that the characteristic determination conditions used for subsequent assembly interface area correction match the actual assembly conditions.

[0036] After extracting the dynamic response features, it is necessary to further determine the assembly interface area corresponding to the dynamic response features. It should be noted that the "correspondence" in this step is not an abstract association, but rather a mapping of the state change segments appearing in the time-series data to the specific assembly interface area between the gasket and the assembly groove, based on the time sequence, displacement process, and execution trajectory information during the assembly process. In specific implementation, the displacement process information, execution trajectory information, or stage position information synchronously recorded in step S1 can be used to determine the advancing position of the gasket in the assembly path when the dynamic response feature appears. Then, combined with the established correspondence between contour segments and structural segments in the initial bonding distribution model, this advancing position can be mapped to the corresponding area of ​​the assembly interface. Furthermore, given the known press-fit starting position, press-fit advancing direction, and the relative geometric relationship between the gasket and the assembly groove, the cumulative displacement or trajectory segment corresponding to any feature time period can be converted into the gasket's entry depth information along the assembly path, thereby determining which area—the straight edge area, corner area, joint area, or local constraint enhancement area—the feature acts upon. For continuous gaskets, a correspondence between the temporal progression and interface segments can be established sequentially along their extension path. For irregularly shaped gaskets, corner gaskets, or gaskets with joints, the mapping results can be locally corrected by considering the boundary positions of the corner area, joint area, and local constraint reinforcement area. In this embodiment, the assembly interface can be divided into one region unit every 3 to 5 millimeters along the assembly path direction. Under this condition, the local correction can be defined as an offset correction of the initial mapped area along the assembly path direction by no more than the range of two adjacent region units, with a corresponding path length of no more than 6 to 10 millimeters, to improve the consistency between the region positioning result and the actual assembly position.

[0037] Furthermore, in some embodiments, the region correspondence of dynamic response characteristics can be verified multiple times by combining the assembly stage division results. For example, when a sudden increase in force and a decrease in the incremental displacement occur within a certain period, the initial corresponding region can be determined first based on the advancement position, and then it can be determined whether it is located in a corner transition zone, a local groove width change zone, or a boundary interference sensitive zone based on the assembly stage at that time. If the deviation between the initial corresponding region and the stage constraint relationship exceeds a preset range, the corresponding region can be re-checked. In this embodiment, when the assembly interface is divided into one region unit every 3 to 5 millimeters, the deviation exceeding the preset range can be defined as exceeding three adjacent region units, and the corresponding assembly path length is approximately 9 to 15 millimeters. In this way, the region correspondence deviation caused by a single time point or a single displacement value can be avoided, making the determined assembly interface region more consistent with the stress and fit state during the actual assembly process of the sealing gasket.

[0038] Through the above analysis and processing, this step ultimately yields two types of results: one is the dynamic response characteristics characterizing the state changes during the gasket assembly process, and the other is the corresponding region of the dynamic response characteristics in the assembly interface. The former reflects what state changes occurred during the assembly process, and the latter reflects where this change affects the assembly interface. This established relationship allows subsequent steps to move beyond simply correcting the entire initial bonding distribution model; instead, they enable targeted adjustments to the bonding state parameters of affected areas. Compared to existing technologies that rely solely on the overall process curve to determine assembly abnormalities or rely solely on static images and dimensional data for acceptance testing after assembly, this step not only identifies what state changes occurred during assembly but also determines the corresponding assembly interface location before assembly completion. For areas that do not show any abnormalities after assembly but experienced localized jamming, abnormal obstruction, or uneven pressure during assembly, this step can still identify and locate them in advance using the corresponding dynamic response characteristics, thus providing direct support for generating more spatially realistic virtual bonding maps.

[0039] S3: Based on the dynamic response characteristics, the bonding state parameters of the corresponding assembly interface area in the initial bonding distribution model are corrected to generate a virtual bonding map characterizing the actual bonding state of the sealing gasket assembly interface.

[0040] In this embodiment, step S3 is mainly used to apply the dynamic response features extracted in step S2 to the corresponding assembly interface area in the initial bonding distribution model established in step S1, and to correct the bonding state parameters in this area, thereby generating a virtual bonding map that can characterize the actual bonding state of the sealing gasket assembly interface. It should be noted that this step is not a complete reconstruction of the initial bonding distribution model, nor is it simply outputting a general abnormal conclusion based on the dynamic response features. Instead, based on the completed area positioning, the dynamic process information is transformed into a directional correction amount for the local bonding state, so that the static geometric matching result is combined with the actual force, resistance, rebound and contact changes during the assembly process, thereby obtaining an interface state expression that is closer to the actual assembly result. It should also be noted that the correspondence between the dynamic response features and the assembly interface regions determined in step S2 can be further used in this step as a corrective linkage between features and regions. This ensures that each assembly interface region identified as disturbed can receive a corresponding correction command, while regions where no dynamic response features are detected can maintain the bonding state parameters in the initial bonding distribution model, or undergo only a slight adjustment of no more than 5% of the initial value according to a preset stable assembly correction rule. Furthermore, the stable assembly correction rule can be determined based on the parameter fluctuation range of regions without dynamic abnormal features in historical normal assembly samples, and is used to correct slight deviations caused by sampling errors, local minor disturbances, or transitions at region boundaries.

[0041] In practice, the correction category for the corresponding assembly interface area can be determined first based on the dynamic response characteristic type output in step S2. It should be noted that different dynamic response characteristics reflect different physical states, therefore their corresponding correction directions should also be differentiated. For example, if the dynamic response characteristics of a certain region show a sudden increase in force accompanied by a decrease in propulsion speed, then that region is more likely to have local interference or edge jamming. In this case, it is advisable to correct the local gap state, local contact continuity state parameters, and local compression state parameters of that region. If the dynamic response characteristics of a certain region show a phased force stagnation, then it can be determined that there is a risk of insufficient compression transmission or inadequate adhesion establishment in that region. In this case, it is advisable to correct the local compression state parameters and local adhesion reliability state parameters of that region. If the dynamic response characteristics of a certain region show inconsistent rebound after unloading, it indicates that the recovery behavior of that region after being compressed deviates significantly from that of adjacent regions. In this case, it is advisable to correct the adhesion stability state parameters, local suspension risk state parameters, and parameter reliability levels of that region. If the dynamic response characteristics of a certain region show an increase in vibration amplitude accompanied by local frequency band energy concentration, then that region can be considered to have a tendency for local friction, jumping, or overturning, and thus the boundary stability state parameters and local contact continuity state parameters of that region should be corrected. By determining the correction direction according to the feature type, this method can avoid treating all dynamic response characteristics as the same type of disturbance, which would lead to correction distortion. It should also be noted that when two or more types of dynamic response features appear simultaneously in the same area, the dominant correction direction corresponding to the dominant dynamic response feature can be determined first, and then the correction influence of other dynamic response features on the fitting state parameters can be integrated to avoid excessive local correction caused by simple superposition of multiple feature corrections.

[0042] Furthermore, after determining the correction category, the bonding state parameters of the corresponding assembly interface region can be read from the initial bonding distribution model. Based on the intensity, duration, occurrence stage, and combination relationship with adjacent features of the dynamic response characteristics, the correction magnitude of the bonding state parameters for that region is determined. It should be noted that the bonding state parameters in this step may include one or more of the following: local gap state parameters, local compression state parameters, local contact continuity state parameters, local bonding reliability state parameters, boundary stability state parameters, and parameter reliability levels. Among these, the local bonding reliability state parameters are mainly used to characterize the reliability of the bonding state itself in that region, while the parameter reliability level is mainly used to characterize the reliability of the parameter estimation results after correction in that region. The former expresses the bonding state, while the latter expresses the reliability of parameter prediction. In this embodiment, they participate in subsequent analysis as information at different levels. In practical implementation, if the dynamic response characteristic is a single mild characteristic, such as a sudden increase in force value lasting 20 to 50 milliseconds and increasing by 1.5 to 2 times the average increase of the stage during the continuous advancement phase, the corresponding area can be identified as a mildly disturbed area, and the bonding state parameters of the area will be corrected at level one. If the dynamic response characteristic is a composite characteristic, such as a sudden increase in force value, a decrease in advancement speed of more than 15%, and an increase in vibration peak value of more than 1.8 times occurring simultaneously in the same area, the corresponding area can be identified as a moderately disturbed area, and the bonding state parameters of the area will be corrected at level two. If the rebound amount is still detected to deviate from the average rebound amount of the stage by more than 25% but not more than 40% during the clamping and stabilization stage, and the area has also experienced force value stagnation or interruption of advancement rhythm in the preceding advancement stage, the area can be identified as a severely disturbed area, and the bonding state parameters of the area will be corrected at level three. It should also be noted that the above-mentioned first-level correction, second-level correction and third-level correction are not limited to a fixed order of magnitude. Their essence lies in distinguishing the correction depth of different regions based on the comprehensive degree of dynamic response characteristics, so that the virtual fitting map has sufficient ability to distinguish different risk levels.

[0043] In the specific correction process, a regional parameter correction method can be used to convert the dynamic response characteristics into adjustments to the initial fit state. For local gap state parameters, if the corresponding region exhibits local interference, short-term jamming, or abnormal rebound, the estimated gap value for that region can be corrected in the direction of increasing gap to reflect potential misfitting or localized suspension after actual assembly. For example, in a slightly disturbed region, the estimated gap state can be incrementally corrected by 5% to 10% of the initial value for that region; in a moderately disturbed region, by 10% to 20%; and in a severely disturbed region, by 20% to 35%. For local compression state parameters, if the corresponding region exhibits insufficient compression transmission, staged force stagnation, or decreased propulsion speed, the estimated compression degree for that region can be corrected in the direction of insufficient compression. For example, in a slightly disturbed region, the compression state parameter can be reduced by 5% to 8%; in a moderately disturbed region, by 8% to 15%; and in a severely disturbed region, by 15% to 25%. For local contact continuity parameters and local fit reliability parameters, if the corresponding area exhibits enhanced vibration, concentrated local frequency band energy, or multi-feature coupling phenomena, the contact continuity and fit reliability of that area can be corrected towards discontinuity and lower reliability. The correction range for slightly disturbed areas can be controlled within 10%, for moderately disturbed areas within 10% to 20%, and for heavily disturbed areas within 20% to 35%. Regarding parameter reliability levels, initially, the reliability level of parameters corresponding to undisturbed areas can be set to high. When an area is associated with a single slightly dynamic response feature, its reliability level can be adjusted to medium; when an area is associated with a composite dynamic response feature or a single heavily dynamic response feature, its reliability level can be adjusted to low. The above correction ratios and reliability level classification methods are not unique. Those skilled in the art can adjust them based on the elastic recovery characteristics of the sealing gasket material, the size grade of the assembly groove, the assembly speed range, and historical calibration results, but it should be ensured that the correction range and parameter reliability levels between different levels of disturbed areas are distinguishable.

[0044] It should also be noted that when performing regional parameter correction, it is not advisable to isolate only the single region directly corresponding to the dynamic response characteristics, but rather to consider the continuous fit relationship between this region and adjacent regions. The gasket is a continuously deformable body during assembly; when a certain region is locally disturbed, its impact usually propagates to adjacent regions. Therefore, in this step, the region directly corresponding to the dynamic response characteristics can be used as the center, and the impact diffusion correction can be performed on adjacent regions along the assembly path. Specifically, the central disturbed region can be defined as the main correction region, and one to two region units before and after it can be defined as auxiliary correction regions. Under the condition that the assembly interface is divided into one region unit every 3 to 5 millimeters, the impact diffusion length corresponding to the auxiliary correction region is approximately 3 to 10 millimeters. For the main correction region, a complete correction can be performed according to the aforementioned correction range; for the auxiliary correction region, an attenuation correction can be performed according to 30% to 60% of the correction range of the main correction region. This approach better reflects the continuous stress and deformation laws of the gasket in actual assembly, avoiding abrupt boundary changes in the virtual fit pattern that do not conform to physical reality.

[0045] Furthermore, for situations where multiple dynamic response features act simultaneously on the same area, this step requires fusing the multi-feature correction relationships. It should be noted that multi-feature correction is not simply adding up the correction magnitudes; otherwise, it can easily lead to over-correction in certain areas. A more reasonable approach is to first determine the dominant correction direction corresponding to each dynamic response feature, and then determine the final correction result based on the feature combination relationship. For example, when a region simultaneously experiences a sudden increase in force and a decrease in propulsion speed, the primary correction should be the increase in gap corresponding to local interference and local obstruction, supplemented by the correction for insufficient compression. When a region simultaneously experiences abnormal rebound and increased vibration, the primary correction should be the decrease in fit stability and contact continuity, followed by secondary corrections to the local gap state. For situations where more than three types of dynamic response features appear in the same area, the dominant feature can be determined first based on the sequence and duration of each feature's occurrence, and then the correction effects of other features can be fused based on the dominant feature. This multi-feature fusion correction method ensures that the final correction result reflects the complexity of the dynamic response while avoiding unreasonable amplification of the local fit state.

[0046] After correcting the regional parameters, the state distribution result facing the entire assembly interface can be regenerated based on the corrected bonding state parameters to form a virtual bonding map. It should be noted that the virtual bonding map is not a graphical display result in the ordinary sense, but rather a regionalized expression of the actual bonding state of the gasket assembly interface. It can be organized according to the assembly path sequence, interface segment distribution, or local structural features to reflect the bonding degree, suspension trend, compression sufficiency, contact continuity, boundary stability, and state reliability level of different regions after assembly. In specific implementation, the corrected set of disturbed region parameters and the set of undisturbed region parameters can be written together into the corresponding region's data set, and then a complete interface state distribution can be formed based on the continuity relationship between regions. Furthermore, to avoid abrupt changes in adjacent regions that do not conform to the continuous deformation law due to region division boundaries, continuity processing can be performed on the corrected adjacent region parameters. This continuity processing can be achieved through weighted averaging of adjacent region parameters, boundary transition zone attenuation fusion, or weight allocation based on path distance. For example, for the boundary region adjacent to the main correction region, the parameters of the boundary region can be merged with the parameters of the main correction region with a transition weight of 20% to 40%, so that the virtual bonding pattern shows a continuous change in spatial distribution, which is more in line with the actual deformation shape of the sealing gasket as a continuous elastomer after assembly.

[0047] Furthermore, the corrected bonding state parameters can be transformed into visual stratified results. For example, the bonding state can be divided into stable bonding areas, slightly unstable areas, moderately unstable areas, and high-risk mismatch areas. Stable bonding areas correspond to parameter deviations no higher than 10% of the baseline value; slightly unstable areas correspond to parameter deviations greater than 10% but no higher than 20%; moderately unstable areas correspond to parameter deviations greater than 20% but no higher than 35%; and high-risk mismatch areas correspond to parameter deviations greater than 35%. This stratification method helps subsequent steps to further classify regional bonding risks, but it does not limit the generation method of the virtual bonding map itself.

[0048] It should also be noted that the technical significance of generating the virtual bonding map in this step lies in transforming the representation of the gasket assembly interface from a theoretical bonding result based on static geometric relationships to an actual bonding result that incorporates the dynamic behavior of the assembly process. Compared with the initial bonding distribution model obtained solely based on contour and structural data, the virtual bonding map generated in this step takes into account process factors such as local interference, local jamming, insufficient compression transfer, abnormal rebound, and unstable contact that occur during assembly. Therefore, it is closer to the actual interface state of the gasket after assembly. For some areas where no obvious abnormalities are observed on the surface after assembly, but which experienced local obstruction and were forcibly pushed through during assembly, the initial bonding distribution model may still classify them as theoretically normal. However, this step can correct these areas by increasing the gap, decreasing the bonding confidence level, lowering the parameter confidence level, or decreasing the contact continuity. This makes these areas appear as potential risk areas in the virtual bonding map, thus providing a more realistic state basis for determining the regional bonding risk distribution in subsequent steps.

[0049] S4: Determine the regional bonding risk distribution of the sealing gasket in the assembly interface based on the virtual bonding map.

[0050] In this embodiment, step S4 is mainly used to identify, quantify, and distribute the regional bonding risks of the sealing gasket at the assembly interface based on the virtual bonding map generated in step S3, thereby forming a regional bonding risk distribution oriented towards the assembly interface. It should be noted that this step does not directly convert the parameters in the virtual bonding map into a single risk conclusion, nor does it perform threshold judgment based solely on a single bonding state parameter. Instead, it jointly discriminates the gap state, compression state, contact continuity state, boundary stability state, bonding credibility state, and parameter credibility level corresponding to different regions in the virtual bonding map to determine the bonding instability probability, local suspension probability, undercompression probability, and subsequent sealing failure tendency of each assembly interface region after assembly. It should also be noted that since the above-mentioned different bonding state parameters differ in dimensional form, numerical range, and sensitivity, before performing joint discrimination, different types of parameters can be standardized to convert them into standardized state quantities that can be used for risk judgment, thereby avoiding risk judgment bias caused by differences in parameter scales.

[0051] In practice, the set of bonding state parameters corresponding to each assembly interface region can be read from the virtual bonding map first, and then the bonding state parameters corresponding to each assembly interface region can be standardized. It should be noted that the standardization process does not necessarily require a fixed mathematical conversion method; its core is to ensure that different bonding state parameters can be converted into comparable state quantities reflecting the degree of deviation. For example, for local gap state parameters, a gap deviation state quantity can be formed based on its deviation from the upper limit of the designed bonding gap or the reference gap state of that region; for local compression state parameters, a compression deviation state quantity can be formed based on its insufficiency relative to the lower limit of the target compression range of the region; for local contact continuity state parameters and boundary stability state parameters, continuity deviation state quantities and stability deviation state quantities can be formed based on their decrease relative to the contact continuity reference level and the boundary stability reference level; for local bonding credibility state parameters and parameter credibility levels, bonding credibility deviation state quantities and predicted credibility deviation state quantities can be formed respectively. In this way, bonding state parameters with different physical meanings can all be entered into the same risk assessment framework for joint analysis.

[0052] After the standardization process is completed, regional-level risk assessment units are further established. Each regional-level risk assessment unit consists of one regional unit or multiple adjacent regional units. It should be noted that the regional-level risk assessment units in this step can directly correspond to the regional units formed in step S3, or they can be merged or subdivided based on the original regional units according to the needs of risk analysis. For example, for continuous straight-edge regions with gentle state changes and small parameter differences, two to three adjacent regional units can be merged into one risk assessment unit. For corner areas, junction areas, areas with enhanced local constraints, and areas identified as main correction areas in step S3, it is advisable to maintain the original unit division. If necessary, further subdivision can be performed according to a path length of 1 to 3 millimeters to improve the ability of the risk distribution results to distinguish key parts. This processing method avoids over-amplifying local fluctuations in low-sensitivity areas while ensuring sufficient spatial resolution in high-sensitivity areas.

[0053] Furthermore, after standardization, the sources of bonding risks for each regional risk assessment unit can be categorized and identified. It should be noted that bonding risk is not a single source; it can be decomposed into at least one or more of the following: gap mismatch risk, insufficient compression risk, contact interruption risk, boundary instability risk, and low-confidence prediction risk. Specifically, gap mismatch risk mainly corresponds to situations where the local gap state parameter in the virtual bonding map exceeds the upper limit of the designed bonding gap for that region; insufficient compression risk mainly corresponds to situations where the local compression state parameter is below the lower limit of the target compression range for that region; contact interruption risk mainly corresponds to situations where the local contact continuity state parameter is below the preset contact continuity threshold; boundary instability risk mainly corresponds to situations where the boundary stability state parameter deviates from the average level of adjacent regions by more than a preset proportion; and low-confidence prediction risk mainly corresponds to situations where the parameter confidence level is low or low-to-medium. Furthermore, for the same regional risk assessment unit, the dominant risk source can be identified first, followed by the secondary risk sources, to avoid overlapping between different sources and weakening the interpretability of the risk distribution results.

[0054] Furthermore, after identifying various risk sources, risk values ​​for each regional risk assessment unit can be generated first, and then classified into corresponding risk levels based on preset risk thresholds. It should be noted that the risk values ​​are not limited to a single scoring format; their essence lies in characterizing the comprehensive degree to which the fit in a certain region deviates from normal fit requirements. Specifically, the risk values ​​can be formed according to preset combination rules based on the magnitude of various deviations in the region, the strength of the dominant risk sources, and the degree to which the reliability level of parameters amplifies the risk outcome. This allows the risk values ​​to reflect both the degree of anomaly of a single parameter and the comprehensive risk level after multi-parameter coupling. In practice, regional risk values ​​can be generated based on the magnitude of various deviations and the importance of the risk sources. On this basis, regional risk values ​​can be further divided into low-level, medium-level, and high-level risks, or into different levels such as stable, watchful, warning, and high-risk, according to preset threshold ranges. For clearance mismatch risk, the local clearance state parameters can be compared with the baseline clearance state of the region. A deviation greater than 10% but not exceeding 20% ​​of the baseline value is classified as a low-level clearance mismatch risk; a deviation greater than 20% but not exceeding 35% is classified as a medium-level clearance mismatch risk; and a deviation greater than 35% is classified as a high-level clearance mismatch risk. For undercompression risk, the local compression state parameters can be compared with the target compression range of the region. An undercompression amount greater than 5% but not exceeding 10% of the target lower limit is classified as a low-level undercompression risk; an undercompression amount greater than 10% but not exceeding 20% ​​is classified as a medium-level undercompression risk; and an undercompression amount greater than 20% is classified as a high-level undercompression risk. For contact interruption risk, the local contact continuity parameters can be compared with the contact continuity baseline level. A decrease of more than 10% but not more than 20% indicates a low-level contact interruption risk; a decrease of more than 20% but not more than 35% indicates a medium-level contact interruption risk; and a decrease of more than 35% indicates a high-level contact interruption risk. For boundary instability risk, the average boundary stability parameters of adjacent areas can be used as a reference. A deviation of 15% to 25% from the average of adjacent areas indicates a low-level boundary instability risk; a deviation of 25% to 40% indicates a medium-level boundary instability risk; and a deviation of more than 40% indicates a high-level boundary instability risk. For low-confidence prediction risk, parameter confidence levels can be divided into high, medium, and low levels, with low-parameter confidence levels being directly treated as risk-enhancing areas.

[0055] It should also be noted that in this step, various risk sources should not be judged in isolation, but rather comprehensively assessed in conjunction with the coupling relationship between state parameters within the region. For example, if a region simultaneously exhibits an increase in gap state parameters and a decrease in compression state parameters, it can be preferentially classified as a misfit risk zone; if a region simultaneously exhibits a decrease in contact continuity and a shift in boundary stability, it can be preferentially classified as an edge instability risk zone; if the gap deviation of a region itself has not yet reached the high-level risk threshold, but its parameter confidence level is at a low level, and there are medium-level or higher risks in adjacent regions, then this region can be defined as a transitional risk zone. By introducing a multi-parameter coupling discrimination mechanism, the distribution of regional fit risk can be made not limited to threshold segmentation of a single parameter, but can more accurately reflect the true risk structure of the assembly interface.

[0056] In the specific implementation process, spatial continuity analysis can be further performed on the regional risk assessment results. It should be noted that the risk distribution in the gasket assembly interface usually exhibits a certain degree of continuity and transmissibility, especially in corner areas, joint areas, and areas where major corrections have occurred. Risks are often not limited to a single isolated unit. Therefore, in this step, the risk status of adjacent areas can be examined along the assembly path and interface boundary directions, centered on a high-level risk unit. If similar risk sources exist within one regional unit before and after a high-level risk unit, and the risk level of adjacent units is not lower than the lower level, then this continuous area can be defined as a risk zone. If different types of risk sources with coupling relationships appear within two regional units surrounding a high-level risk unit, such as the coexistence of gap mismatch risk and boundary instability risk, then it can be defined as a composite risk zone. Under the condition that the assembly interface is divided into regional units of 3 to 5 millimeters, the continuous length of the risk zone can preferably be defined as 6 to 15 millimeters, and the coverage area of ​​the composite risk zone can preferably be defined as a local area extending forward and backward by no more than 10 millimeters from the central risk unit. This method avoids too many isolated dot marks in the risk distribution results, making them more consistent with the actual morphology of continuous sealing gasket instability.

[0057] Furthermore, for certain structurally sensitive areas, this step can introduce a regional sensitivity coefficient to further adjust the risk distribution results. It should be noted that different assembly interface areas have varying tolerances to fitting deviations. For example, corner areas are more sensitive to local overturning and boundary instability, joint areas are more sensitive to contact interruption and local suspension, and areas with enhanced local constraints are more sensitive to insufficient compression and interference. Therefore, the assembly interface can be pre-divided into regular areas, sensitive areas, and highly sensitive areas. Regular areas correspond to straight-edge areas, sensitive areas correspond to corner areas or adjacent joint areas, and highly sensitive areas correspond to joint center areas, areas with abrupt changes in local slot width, or areas with local boundary interference. For risk assessment units located within sensitive areas, the risk level can be increased by one level from the original level, but not exceeding the highest level. For units located within highly sensitive areas and simultaneously possessing two or more risk sources, they can be directly marked as high-risk areas. It should also be noted that the high-risk area in this embodiment refers to the risk level marking result in the regional fitting risk distribution, used to characterize the corresponding area as being in a high-risk state. Furthermore, for areas with higher sealing requirements or greater fluctuations in subsequent service conditions, the corresponding risk threshold range can be appropriately tightened, so that deviations in the same degree of fit are judged as higher-level risks in high-requirement areas. It should be noted that these high-requirement areas may include the joint center area, corner transition area, local groove width abrupt change area, sealing area subjected to high medium pressure, or interface section located in a temperature-sensitive position. By applying differentiated risk threshold settings to these areas, the distribution of regional fit risk can be made more closely aligned with actual service requirements. This regional sensitivity correction method allows the risk distribution result to more closely reflect the sensitivity differences of the gasket to local mismatches during actual service.

[0058] After completing risk assessment, continuity analysis, and sensitivity correction for each region, the risk status of the entire assembly interface can be distributed to form a regional risk distribution. This regional risk distribution can be arranged according to the assembly path sequence, interface segment sequence, or local structural region sequence, and each risk assessment unit can be assigned a corresponding risk category and risk level identifier. In specific implementation, the risk distribution can be divided into stable zones, areas of concern, warning zones, and high-risk zones. Specifically, stable zones correspond to areas where no significant risk sources have been detected and the parameter confidence level is high; areas of concern correspond to areas with only a single low-level risk source or a declining parameter confidence level; warning zones correspond to areas with medium-level risk sources, continuously expanding risk bands, or located in sensitive areas; and high-risk zones correspond to areas with high-level risk sources, located in highly sensitive areas, and accompanied by complex risk sources. It should also be noted that the division into stable zones, areas of concern, warning zones, and high-risk zones is not merely for result display, but rather for forming a risk distribution expression with spatial order and hierarchical relationships at the assembly interface level, enabling subsequent steps to directly access the risk level and risk type information of different regions when generating assembly compatibility assessment results.

[0059] Furthermore, in some implementations, local consistency checks can be performed on the regional risk distribution. It should be noted that if a region is identified as a high-risk area, but its two adjacent regional units are both in stable regions, and the region itself is not in a corner, junction, or highly sensitive region, then the high-risk determination may stem from excessive local correction fluctuations or regional mapping errors. Based on this, consistency checks can be performed on such isolated high-risk units. If, after verification, it is confirmed that the parameter differences between its adjacent regions do not exceed 15%, and the parameter confidence level corresponding to the unit is not low, then its risk level can be lowered by one level; if the high-risk unit simultaneously has a low parameter confidence level or is located in a sensitive region, then the original determination remains unchanged. This process improves the stability of the risk distribution and avoids distortion of the entire risk distribution result due to single-point anomalies.

[0060] Through the above processing, this step ultimately yields the regional fit risk distribution result of the sealing gasket across the entire assembly interface. It should be noted that this risk distribution result is not based on surface judgment obtained directly from post-assembly visual inspection, nor on static results calculated from a single dimensional deviation. Instead, it is based on a virtual fit map, further transforming the interface state—after correction of the dynamic influences formed during assembly—into a risk structure with regionality, continuity, and hierarchy. This regional fit risk distribution provides direct input for generating assembly compatibility assessment results in subsequent steps, and transforms the assessment object from a single overall verification result into a structured judgment oriented towards specific areas of the assembly interface.

[0061] S5: Generate the gasket assembly compatibility assessment results based on the regional fit risk distribution.

[0062] In this embodiment, step S5 is mainly used to summarize and determine the overall fit status of the gasket assembly interface based on the regional fit risk distribution formed in step S4, and generate a gasket assembly fit assessment result. It should be noted that this step does not simply add up the risk levels of each region to give a single conclusion, nor does it perform a binary judgment based solely on the existence of high-risk areas. Instead, it structurally integrates the overall assembly fit status based on the spatial order, risk level, risk category, and continuity relationship of the regional fit risk distribution. This ensures that the final output result reflects whether the gasket as a whole meets the assembly requirements, and also shows the dominant risk areas, dominant risk types, and corresponding handling levels that affect the overall fit.

[0063] In specific implementation, the risk level identifier, risk category identifier, spatial position order, and continuous expansion relationship of each risk judgment unit can be read from the regional fitting risk distribution output in step S4, and an overall adaptability summary unit can be established. It should be noted that the overall adaptability summary unit in this step does not change the original regional risk distribution, but rather performs a higher-level organization of regional risk information for overall adaptability judgment. Furthermore, before performing overall adaptability judgment, risk summary processing can be performed on the regional risk distribution to form a global risk overview of the entire sealing gasket assembly interface. The global risk overview may include one or more of the following: statistics on the number of high-risk areas, statistics on the number of warning areas, statistics on the length of risk zones, statistics on the number of composite risk areas, statistics on risk units located in sensitive or highly sensitive areas, statistics on risk category distribution, and statistics on the cumulative path length of high-risk areas. Through the above processing, the overall adaptability assessment result can be based on a comprehensive analysis of the risk structure of the entire assembly interface, rather than relying on the isolated risk of a single region.

[0064] Furthermore, after the risk summary processing is completed, the compatibility level of the sealing gasket assembly can be determined. It should be noted that the compatibility level can be divided into four levels: compatibility pass, compatibility concern, compatibility warning, and compatibility mismatch, or it can be divided into three or five levels according to process control needs. In this embodiment, a four-level division is preferred. Specifically, compatibility pass corresponds to a situation where the number of high-risk areas is 0 and there are no risk zones; compatibility concern corresponds to situations that do not fall under compatibility pass, compatibility warning, or compatibility mismatch; compatibility warning corresponds to a situation where the number of high-risk areas reaches the second preset number but does not reach the third preset number, or there are composite risk areas but the number of composite risk areas does not reach the fourth preset number; compatibility mismatch corresponds to a situation where the number of high-risk areas reaches the third preset number, or there is a high-risk area located within a highly sensitive area, or the number of composite risk areas reaches the fourth preset number. Furthermore, the second and third preset quantities are used to determine the number of high-risk areas, and the fourth preset quantity is used to determine the number of complex risk areas. The second, third, and fourth preset quantities can be determined based on the total path length of the sealing gasket, the density of the area unit division, and the sealing requirement level, and the second preset quantity is less than the third preset quantity.

[0065] In the specific judgment process, it is not advisable to make a single statistical judgment based solely on the number of risk levels. Instead, the overall compatibility result should be corrected by combining the risk type and risk location. For example, if a sealing gasket has a low number of high-risk areas overall, but these high-risk areas are located in the joint center area, corner transition area, or local groove width abrupt change area, then the overall compatibility level can be lowered by one level because such areas have a significant impact on subsequent sealing integrity. If a sealing gasket has multiple areas of concern, but these areas of concern are all distributed in the straight edge sections of the regular area and are discontinuous with each other, and the corresponding risk category is a single low-level insufficient compression risk, then the overall compatibility level can remain unchanged. Furthermore, if an area is identified as a high-risk area, but its risk category is only low-confidence prediction risk, and the decrease in parameter confidence level is not caused by composite dynamic response characteristics but by local data sampling quality fluctuations, then this area can be marked separately as an area to be reviewed during the overall compatibility judgment, rather than being directly used as the sole basis for lowering the overall level. Through this overall correction mechanism involving both location factors and risk category factors, the assembly compatibility assessment results can be made more consistent with the actual engineering judgment logic.

[0066] It should also be noted that when generating the overall fit assessment results, the dominant mismatch factor can be extracted simultaneously. The dominant mismatch factor refers to the risk category or risk area that has a major impact on the overall fit level in the regional fit risk distribution. In practice, the dominant mismatch factor can be determined comprehensively based on one or more factors, including risk level, structural sensitivity level of the risk area, continuous risk extension length, number of composite risk sources, and frequency of occurrence of the corresponding risk category. Priority is given to selecting the dominant risk category in high-risk areas as the primary dominant mismatch factor; when no high-risk area exists, the dominant risk category in the risk zone is selected as the primary dominant mismatch factor, thus ensuring that the extracted dominant mismatch factor accurately reflects the main reasons for changes in the overall fit level. If multiple dominant risk sources of the same level exist simultaneously, the risk source with the greatest impact can be identified as the primary dominant mismatch factor, and the others as secondary dominant mismatch factors, based on the structural sensitivity level and continuous extension length of the corresponding region. Furthermore, for different types of gap mismatch risk, insufficient compression risk, contact interruption risk, boundary instability risk, and low-confidence prediction risk, corresponding dominant mismatch description items can be generated. For example, when the main reason for the overall fit level downgrade is the boundary instability risk in the corner area, this type of risk can be identified as the dominant mismatch description item; when the main reason for the overall fit level downgrade is the combined occurrence of contact interruption risk and gap mismatch risk in the junction area, both can be listed as dominant mismatch description items. In this way, the final assessment result not only provides the overall level but also explains the dominant risk factors causing that level.

[0067] Furthermore, after completing the overall level determination and extracting the dominant mismatch factors, a list of weak links in the adaptation process can be generated based on the regional risk distribution. It should be noted that the weak links refer to local areas in the assembly interface that have a significant impact on the overall adaptability results and require focused review or priority handling. Specifically, areas meeting one of the following conditions can be included in the list of weak links: areas marked as high-risk areas; areas located in sensitive or highly sensitive areas with a risk level of warning or higher; areas forming risk zones with a length greater than 6 mm; areas forming composite risk zones containing two or more dominant risk sources. For areas included in the list of weak links, their starting and ending positions, dominant risk categories, risk levels, and continuous extension lengths can be recorded simultaneously for structured output in subsequent adaptability assessment results.

[0068] Furthermore, in some implementations, suggested outputs corresponding to risk types can be generated based on the regional fit risk distribution and dominant mismatch factors. It should be noted that these suggested outputs do not directly issue control commands to the assembly equipment, but rather transform the assessment results into actionable information that can be accessed by process engineers and quality control personnel. For example, for weak links dominated by gap mismatch risk, a local gap review suggestion can be output; for weak links dominated by insufficient compression risk, a press-fit status review suggestion can be output; for weak links dominated by contact interruption risk, a local appearance review suggestion can be output; for weak links dominated by boundary instability risk, an adjacent structural boundary review suggestion can be output; and for weak links dominated by low-confidence prediction risk, a manual review suggestion can be output. Through this suggested output method corresponding to the dominant mismatch factor, the assessment results not only indicate where assembly fit problems exist, but also further point out the priority direction for subsequent review or action.

[0069] It should also be noted that, in some implementations, tiered handling conclusions can be generated based on regional risk distribution. These tiered handling conclusions do not directly issue control commands to the assembly equipment, but rather transform the evaluation results into handling level information that can be accessed by process engineers, quality inspectors, or automated management systems. Specifically, if the overall compatibility level is "Compatibility Passed," a normal workflow conclusion can be generated; if the overall compatibility level is "Compatibility Concern," a key re-inspection conclusion can be generated, and a list of corresponding concern areas or weak links in the compatibility can be output simultaneously; if the overall compatibility level is "Compatibility Warning," a partial rework or secondary inspection conclusion can be generated, and the dominant risk area, risk zone range, and dominant mismatch factor can be marked; if the overall compatibility level is "Compatibility Mismatch," a termination of workflow or overall rework conclusion can be generated, and high-risk areas, complex risk areas, and their corresponding dominant mismatch factors can be listed. It should also be noted that, in this embodiment, the key re-inspection conclusion preferably corresponds to 1 to 3 areas of concern or weak links, the partial rework or secondary inspection conclusion preferably corresponds to the situation where there is a risk zone or a complex risk area, and the overall rework conclusion preferably corresponds to the situation where the number of high-risk areas exceeds 3 or there are high-sensitivity areas and high-risk units.

[0070] When generating the gasket assembly compatibility assessment results, a compatibility result carrier can also be output simultaneously. This compatibility result carrier may include one or more of the following: overall compatibility level identifier, dominant mismatch factor identifier, risk area location index, risk zone length information, composite risk area identifier, list of weak compatibility links, and handling level identifier. In specific implementation, the above result carrier can be written into a preset assessment result data structure and stored in association according to the gasket number, assembly station number, batch number, or assembly time sequence. Furthermore, for scenarios using automated production lines, the compatibility assessment results can be bound to the assembly execution records to form a complete data chain for the same gasket, from assembly process response data, virtual bonding map, regional bonding risk distribution to the final compatibility conclusion. In this way, if subsequent traceability analysis of batch gasket assembly quality is required, the compatibility assessment result carrier can be directly called for retrieval and comparison.

[0071] It should also be noted that the gasket assembly compatibility assessment results generated in this step are not limited to a simple binary conclusion of "qualified" or "unqualified." Instead, they further integrate the regional fit risk distribution formed in step S4 into a comprehensive assessment result with hierarchical, regional, and remedial characteristics. Compared to the method of making a post-hoc judgment on the assembly status based solely on the final inspection results, the assessment results generated in this step can make a more structured judgment on the overall fit status of the gasket assembly interface after assembly is completed but before it enters subsequent service. For some cases that do not show overall mismatch but have formed risk zones in the joint area or corner area, this step can still determine them as fit warning or fit concern status, rather than simply classifying them as pass status. Therefore, the final output assembly compatibility assessment result is no longer a general conclusion detached from the spatial context, but a comprehensive judgment result based on the regional fit risk distribution, thus completing the transformation from regional risk expression to overall fit judgment.

[0072] Example 2 is the second embodiment of the present invention, which differs from the previous embodiment in that: If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art or the current technical solution, can be embodied in the form of a software product. This current computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0073] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-including system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0074] More specific examples of computer-readable media (a non-exhaustive list) include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which the program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.

[0075] Example 3, referring to Figure 2 As an embodiment of the present invention, a smart evaluation system for the compatibility of sealing gasket assembly is provided, which includes a data modeling module, a feature extraction module, a map generation module, a risk analysis module, and an evaluation output module; Data modeling module: acquires the contour data of the gasket to be assembled, the structural data of the assembly groove, and the process timing response data during the gasket assembly process, and constructs an initial fit distribution model between the gasket and the assembly groove based on the contour data and the structural data of the assembly groove. Feature extraction module: Analyzes and processes the process timing response data, extracts dynamic response features that characterize the state changes during the gasket assembly process, and determines the assembly interface area corresponding to the dynamic response features; Atlas generation module: Based on the dynamic response characteristics, the bonding state parameters of the corresponding assembly interface area in the initial bonding distribution model are corrected to generate a virtual bonding atlas characterizing the actual bonding state of the sealing gasket assembly interface. Risk analysis module: Determines the regional adhesion risk distribution of the sealing gasket in the assembly interface based on the virtual bonding map; Evaluation output module: Generates gasket assembly compatibility evaluation results based on the regional fit risk distribution.

[0076] Example 4 is an embodiment of the present invention, which provides a smart evaluation method for the fit of sealing gasket assembly. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through economic benefit calculation and simulation / comparative experiments.

[0077] This embodiment uses the automated pressing process of a cylinder head gasket for a certain type of engine as an application scenario to specifically illustrate the intelligent evaluation method for gasket assembly compatibility described in this invention. The effectiveness of this method in judging assembly compatibility is verified through comparative experiments. The test subjects were twenty fluororubber gaskets produced in the same batch, with ten representative samples selected as demonstration objects, designated SY-01 to SY-10. The nominal cross-sectional diameter of the gasket is 3.5 mm, and the theoretical total assembly path length is approximately 480 mm. The corresponding assembly groove is set on the surface of the aluminum alloy cylinder head, and is a continuous annular rectangular groove with a designed groove width of 3.6 mm and a designed groove depth of 2.8 mm.

[0078] The assembly equipment employs a servo press-fitting system with a press-fitting speed set at 25 mm / s and a target press-fitting force set at 1200 N. To acquire the multi-source data required for implementing the method of this invention, a laser profilometer, industrial vision unit, force sensor, displacement grating ruler, accelerometer, and infrared thermal imaging probe are integrated at the press-fitting station. The laser profilometer and industrial vision unit acquire the contour data of the gasket to be assembled, the online detection unit acquires the structural data of the assembly groove, and the force sensor, displacement grating ruler, accelerometer, and infrared thermal imaging probe acquire the process timing response data during the gasket assembly process. All sensor sampling signals are uniformly connected to the same data acquisition terminal and synchronously recorded using a unified time reference, with a sampling frequency set to 1000 Hz.

[0079] In practice, before pressing, the sealing gasket to be assembled is first scanned and imaged from multiple angles to obtain its outer edge boundary, inner edge boundary, cross-sectional thickness distribution, corner transition shape, local raised parts, local recessed parts, and the continuous state of the joint. Simultaneously, the assembly groove on the workpiece to be assembled is inspected online to obtain the groove opening boundary position, groove bottom shape, groove wall contour direction, local groove width variation, local groove depth variation, corner structural transition state, and interference boundary. Then, based on the contour and structural data, the position and direction alignment between the sealing gasket and the assembly groove are completed under a unified spatial reference. Following the actual path sequence of the sealing gasket entering the assembly groove, a correspondence is established between the sealing gasket contour segments and the assembly groove structural segments. Along the assembly path, a regional unit is divided every 4 millimeters, forming a total of 120 regional units. Based on this, the initial gap trend, initial compression trend, initial contact establishment trend, and local interference tendency of each regional unit under ideal entry conditions are determined, thereby constructing an initial fit distribution model between the sealing gasket and the assembly groove. After entering the pressing stage, data on force changes, displacement changes, propulsion speed changes, springback state changes, local vibration changes, and temperature changes in the assembly area are continuously collected. The process timing response data is then processed for time alignment, noise suppression, burr removal, sampling interval unification, and stage boundary calibration. Based on the proportion of cumulative displacement to total propulsion displacement, the assembly process is divided into the initial entry stage, continuous propulsion stage, locally restricted stage, transition fitting stage, and pressing and stabilizing stage.

[0080] Furthermore, dynamic response features characterizing changes in the assembly process state are extracted from the preprocessed process timing response data. These features include sudden increases in local force, inconsistent rebound after unloading, decreased propulsion speed, and increased local vibration amplitude. Based on the time sequence, displacement changes, and propulsion speed changes corresponding to the dynamic response features, the propulsion position of the gasket in the assembly path when an anomaly occurs is determined. Then, combining the correspondence between contour segments and structural segments in the initial bonding distribution model, the dynamic response features are mapped to the corresponding assembly interface regions. Subsequently, the bonding state parameters of the corresponding assembly interface regions in the initial bonding distribution model are corrected based on the dynamic response features, generating a virtual bonding map characterizing the actual bonding state of the gasket assembly interface. Based on the virtual bonding map, the bonding state parameters of each assembly interface region are standardized, bonding risk sources are classified and identified, risk levels are divided, spatial continuity analysis is performed, and sensitive area correction is applied to form a regional bonding risk distribution. Finally, a gasket assembly compatibility assessment result is generated.

[0081] To verify the technical effectiveness of the method of the present invention, a traditional static geometric evaluation method was set as a control. This traditional static geometric evaluation method calculates the theoretical gap value of each regional unit based solely on the pre-assembly contour data and assembly groove structure data, and uses whether the theoretical gap value exceeds the upper limit of the designed fitting gap as the main criterion. It does not incorporate process timing response data during assembly, does not extract dynamic response features, and does not generate virtual fitting maps or regional fitting risk distributions. The judgment results of both evaluation methods were compared with the airtightness test results after assembly. The airtightness test was conducted using a helium mass spectrometer leak detector; a leakage rate exceeding 1×10^-6 Pa·m³ / s was considered unqualified. Key recorded data are shown in Table 1. Table 1: Experimental Data Recording Table Sample number Theoretical maximum gap (mm) Number of regions where a sudden increase in force was detected (number of units) Number of units with inconsistent rebound detected. Number of high-risk areas after revision (number of areas) Maximum risk zone length (mm) Traditional method for determining results Adaptability level of the method of the present invention Leakage rate (Pa·m³ / s) in airtightness test Is it qualified? SY-01 0.142 0 0 0 0 qualified Adaptation passed 6.2e-07 qualified SY-02 0.138 1 0 1 7.2 qualified Adapt and follow 5.8e-07 qualified SY-03 0.156 0 1 0 0 Unqualified Adapt and follow 7.1e-07 qualified SY-04 0.148 2 1 2 11.5 qualified Adaptation warning 2.3e-06 Unqualified SY-05 0.163 0 0 0 0 Unqualified Adaptation passed 6.8e-07 qualified SY-06 0.141 3 2 3 16.8 qualified Mismatch 4.5e-06 Unqualified SY-07 0.152 1 1 1 8.3 Unqualified Adapt and follow 9.2e-07 qualified SY-08 0.137 0 0 0 0 qualified Adaptation passed 5.3e-07 qualified SY-09 0.161 2 0 2 9.6 Unqualified Adaptation warning 1.8e-06 Unqualified SY-10 0.145 4 1 4 19.2 qualified Mismatch 5.1e-06 Unqualified

[0082] As can be seen from Table 1, there are significant differences between the traditional static geometric evaluation method and the method of the present invention in the judgment results of multiple samples, and the method of the present invention shows higher consistency with the final airtightness test results.

[0083] Specifically, in SY-03, SY-05, and SY-07, the traditional static geometric evaluation method deemed them unqualified because the maximum theoretical gap exceeded the upper limit of the designed fit gap. However, the airtightness test results showed that the actual leakage rate of these samples was below the unqualified threshold, indicating that they were qualified samples in actual assembly. Among them, the maximum theoretical gap of SY-05 reached 0.163 mm, which was a significantly out-of-tolerance sample according to the traditional method. However, no dynamic abnormalities such as sudden force increases or inconsistent rebound were detected during the actual pressing process. After correction, the number of high-risk areas was 0, and the length of the maximum risk zone was 0. The method of this invention determined that it passed the fit test. The final airtightness test leakage rate was 6.8 × 10^-7 Pa·m³ / s, verifying that the sample still had effective sealing capability under actual assembly conditions. This shows that the traditional static geometric evaluation method only judges based on the geometric dimensional relationship before assembly and cannot reflect the elastic compression compensation behavior of the sealing gasket material during the actual assembly process. It is easy to misjudge samples with slightly out-of-tolerance theoretical gaps but which can still form an effective fit after pressing as unqualified, thus causing false scrapping.

[0084] In comparison, the method of this invention, by introducing process timing response data and combining dynamic response characteristics to perform regional correction on the initial bonding distribution model, can more closely approximate the bonding state of the actual assembly interface, thus demonstrating higher judgment accuracy on such samples. On the other hand, in SY-04, SY-06, and SY-10, the traditional static geometric evaluation method judged them as qualified, but the final airtightness test results showed that they were all unqualified samples. For example, the theoretical maximum gap of SY-06 was 0.141 mm, which is within the acceptable range of the traditional method. However, during the pressing process, three areas of sudden force increase and two areas of inconsistent springback were detected. After dynamic response feature mapping and bonding state parameter correction, three high-risk areas were formed, with the maximum risk zone length reaching 16.8 mm. Based on this, the method of this invention judged the sample as misfit, and the final airtightness test leakage rate reached 4.5 × 10^-6 Pa·m³ / s, which significantly exceeded the unqualified threshold. SY-10 exhibited a similar pattern, with a theoretical maximum gap of only 0.145 mm. However, during the assembly process, four units with sudden increases in force were detected, forming a continuous risk zone with a length of 19.2 mm. Ultimately, it was determined by the method of this invention to be a mismatch, and the airtightness test also proved its sealing failure.

[0085] This demonstrates that traditional static geometric evaluation methods cannot identify process-related latent defects during assembly, such as local interference, jamming, insufficient compression transfer, and abnormal springback. There is a significant risk of missing defects in samples that appear normal after assembly but have compromised internal fit. In contrast, the method of this invention can transform dynamic anomalies during assembly into spatially directional risk results. It can not only determine whether a sample has assembly fit issues but also identify the dominant risk area and its continuous expansion, thereby enabling proactive identification of latent assembly defects.

[0086] In summary, this embodiment verifies that the method of the present invention has the following advantages compared with traditional static geometric evaluation methods: First, it can reduce the risk of false rejection of samples with theoretical geometric deviations, and improve the authenticity of assembly quality judgment; second, it can identify dynamic anomalies and hidden mismatch areas in the assembly process that are difficult to detect by traditional methods, reducing the probability of missed detection; third, it can output risk results and suitability levels with regional orientation, providing a clearer basis for subsequent re-inspection, rework, and assembly process adjustment. Therefore, the present invention has strong engineering application value in the scenario of gasket assembly suitability evaluation, and can demonstrate innovation and practicality compared with the prior art.

[0087] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for intelligent evaluation of the fit of a sealing gasket assembly, characterized in that, include: S1: Obtain the contour data of the gasket to be assembled, the structural data of the assembly groove, and the process timing response data during the gasket assembly process, and construct an initial bonding distribution model between the gasket and the assembly groove based on the contour data and the structural data of the assembly groove. S2: Analyze and process the process timing response data, extract dynamic response features that characterize the state changes during the assembly process of the sealing gasket, and determine the assembly interface area corresponding to the dynamic response features. S3: Based on the dynamic response characteristics, the bonding state parameters of the corresponding assembly interface area in the initial bonding distribution model are corrected to generate a virtual bonding map characterizing the actual bonding state of the sealing gasket assembly interface. S4: Determine the regional bonding risk distribution of the sealing gasket in the assembly interface based on the virtual bonding map; S5: Generate the gasket assembly compatibility assessment results based on the regional fit risk distribution.

2. The intelligent evaluation method for gasket assembly compatibility as described in claim 1, characterized in that, Step S1 specifically includes: Obtain the contour data of the sealing gasket to be assembled. The contour data is used to characterize the outer edge boundary, inner edge boundary, cross-sectional thickness distribution, corner transition shape, local raised parts, local recessed parts, and the continuous state of the joint. Obtain the structural data of the assembly groove, which includes the groove opening boundary position, groove bottom shape, groove wall contour direction, local groove width variation, local groove depth variation, corner structure transition state, and interference boundary. Acquire process timing response data during the assembly of the sealing gasket, including force changes, displacement changes, propulsion speed changes, springback state changes, local vibration changes, and temperature changes in the assembly area; Based on the contour data and the structural data of the assembly groove, an initial fit distribution model between the sealing gasket and the assembly groove is constructed. The initial fit distribution model is a basic fit state expression of each local area of ​​the assembly interface, which is used for subsequent dynamic correction and adaptability evaluation.

3. The intelligent evaluation method for gasket assembly compatibility as described in claim 2, characterized in that, Step S2 specifically includes: The process timing response data is preprocessed, including time alignment, noise suppression, burr removal, sampling interval unification, and stage boundary calibration of the raw data. The assembly process is divided into the initial entry stage, continuous advancement stage, local confinement stage, transition fitting stage, and clamping and stabilization stage according to the proportion of cumulative displacement to total propulsion displacement. Dynamic response features characterizing the state changes during the assembly of the sealing gasket are extracted from the pre-processed process timing response data. These dynamic response features include local force surge, staged force stagnation, enhanced force fluctuation, inconsistent rebound after unloading, decreased propulsion speed, interrupted propulsion rhythm, local motion stagnation, enhanced vibration amplitude, local frequency band energy concentration, and abnormal local temperature rise. The assembly interface region corresponding to the dynamic response feature is determined by determining the advancing position of the gasket in the assembly path when the dynamic response feature appears, based on the time sequence, displacement change, and advancing speed change corresponding to the process timing response data. Combining the correspondence between contour segments and structural segments established in the initial bonding distribution model, the advancing position is mapped to the corresponding region in the straight edge area, corner area, joint area, or local constraint enhancement area of ​​the assembly interface.

4. The intelligent evaluation method for gasket assembly compatibility as described in claim 3, characterized in that, Step S3 includes: Based on the dynamic response feature type output in step S2, the correction category for the corresponding assembly interface area is determined. If the dynamic response feature is that the force increase exceeds the preset increase range and the propulsion speed is lower than the preset speed range, the local gap state, local contact continuity state parameters, and local compression state parameters are corrected. If the force increase rate is lower than the preset increase rate, the local compression state parameters and local fit reliability state parameters are corrected. If the rebound deviation exceeds the preset deviation range, the fit stability state parameters, local suspension risk state parameters, and parameter reliability level are corrected. If the vibration amplitude exceeds the preset reference range and the energy ratio within the preset frequency band exceeds the preset ratio threshold, the boundary stability state parameters and local contact continuity state parameters are corrected. In the initial bonding distribution model, the bonding state parameters of the corresponding assembly interface area are read, and the area is determined as a first-level disturbed area, a second-level disturbed area, or a third-level disturbed area based on whether the dynamic response characteristics exceed the corresponding preset judgment range, duration, occurrence stage, and whether they appear in combination with other dynamic response characteristics. First-level correction is performed for the first-level disturbed area, second-level correction is performed for the second-level disturbed area, and third-level correction is performed for the third-level disturbed area. When performing regional parameter correction, for local gap state parameters, if there is local interference, jamming, or rebound exceeding the preset range in the corresponding region, the correction is made in the direction of increasing gap. For local compression state parameters, if the corresponding region has compression below the preset compression range, force growth rate below the preset growth rate, or propulsion speed below the preset speed range, the correction is made in the direction of decreasing compression. For local contact continuity state parameters and local bonding reliability state parameters, if the corresponding region has vibration amplitude exceeding the preset reference range, energy proportion within the preset frequency band exceeding the preset ratio threshold, or two or more dynamic response characteristics appearing simultaneously, the correction direction is determined based on the dominant dynamic response characteristics, and the correction is made in the direction of decreasing local contact continuity state parameters and decreasing local bonding reliability state parameters.

5. The intelligent evaluation method for gasket assembly compatibility as described in claim 4, characterized in that, Step S3 also includes: Centered on the region directly corresponding to the dynamic response characteristics, influence diffusion correction is performed on the adjacent regions before and after along the assembly path. Specifically, the central disturbed region is defined as the main correction region, and the adjacent regional units before and after it are defined as auxiliary correction regions. The main correction region is fully corrected, and the auxiliary correction region is attenuated. The magnitude of the attenuation correction is smaller than the correction magnitude of the main correction region.

6. The intelligent evaluation method for gasket assembly compatibility as described in claim 5, characterized in that, Step S4 includes: Read the set of bonding state parameters corresponding to each assembly interface area from the virtual bonding map, perform unified processing on each bonding state parameter, and establish a regional risk judgment unit. The regional risk judgment unit consists of one regional unit or multiple adjacent regional units. The sources of bonding risks in each regional risk assessment unit are classified and identified. These sources include gap mismatch risk, insufficient compression risk, contact interruption risk, boundary instability risk, and low-confidence prediction risk. Gap mismatch risk corresponds to the situation where the local gap state parameter exceeds the upper limit of the designed bonding gap in that region. Insufficient compression risk corresponds to the situation where the local compression state parameter is lower than the lower limit of the target compression range in that region. Contact interruption risk corresponds to the situation where the local contact continuity state parameter is lower than the preset contact continuity threshold. Boundary instability risk corresponds to the situation where the boundary stability state parameter deviates from the average level of adjacent regions by more than a preset proportion. Low-confidence prediction risk corresponds to the situation where the confidence level of the parameter is low. Based on the degree to which the fitting status parameters corresponding to each risk source deviate from the normal range, the risk levels are divided into low-level risk, medium-level risk, and high-level risk, and risk marking is performed on each regional-level risk judgment unit according to the risk levels.

7. The intelligent evaluation method for gasket assembly compatibility as described in claim 6, characterized in that, Step S4 also includes: Spatial continuity analysis is performed with high-level risk units as the center. If there are similar risk sources in adjacent regional units before and after the high-level risk unit, and the risk level of the adjacent units is not lower than that of the low-level units, then the continuous area formed by the high-level risk unit and the adjacent regional units that meet the conditions of similar risk sources and risk levels is defined as a risk zone. If different types of risk sources with coupling relationships appear in the adjacent regional units around the high-level risk unit, then it is defined as a composite risk zone. The assembly interface is divided into a regular area, a sensitive area, and a highly sensitive area. The regular area corresponds to the straight edge area, the sensitive area corresponds to the corner area or the adjacent area of ​​the joint, and the highly sensitive area corresponds to the joint center area, the local slot width abrupt change area, or the local boundary interference area. For risk assessment units located in the sensitive area, the risk level is increased by one level based on the original risk level, but not exceeding the higher level. For units located in the highly sensitive area and with two or more risk sources at the same time, they are directly marked as high-risk areas.

8. The intelligent evaluation method for gasket assembly compatibility as described in claim 7, characterized in that, Step S5 includes: The risk distribution of the region output in step S4 is subjected to risk summary processing to form a global risk overview. The global risk overview includes statistics on the number of high-risk areas, the number of early warning areas, the length of risk zones, the number of composite risk areas, the risk units located in sensitive areas, the risk units located in highly sensitive areas, and the risk category distribution. Based on the overall risk overview, the compatibility of the sealing gasket assembly is assessed, and the compatibility level is divided into four levels: compatibility pass, compatibility concern, compatibility warning, and compatibility failure. Compatibility pass corresponds to a situation where the number of high-risk areas is zero and there are no risk zones. Compatibility concern corresponds to a situation that does not fall into the categories of compatibility pass, compatibility warning, or compatibility failure. Compatibility warning corresponds to a situation where the number of high-risk areas reaches the second preset number but not the third preset number, or where there are composite risk areas but the number of composite risk areas does not reach the fourth preset number. Compatibility failure corresponds to a situation where the number of high-risk areas reaches the third preset number, or where there is a high-risk area located within a highly sensitive area, or where the number of composite risk areas reaches the fourth preset number. The second preset number is less than the third preset number. The second, third, and fourth preset numbers are determined based on the total path length of the assembly interface, the density of area unit divisions, and the sealing requirement level.

9. The intelligent evaluation method for gasket assembly compatibility as described in claim 8, characterized in that, Step S5 also includes: Extract the dominant mismatch factor. The dominant mismatch factor is a risk category that has a major impact on the overall suitability level, which is determined by a combination of factors including the risk level, the structural sensitivity level of the risk area, the length of continuous risk extension, and the number of compound risk sources. The dominant risk category in the high-risk area is selected as the primary dominant mismatch factor. If there is no high-risk area, the dominant risk category in the risk zone is selected as the primary dominant mismatch factor. Generate a list of weak links for adaptation, and include areas that meet one of the following conditions in the list of weak links for adaptation: areas marked as high-risk areas, areas located in sensitive areas with a risk level of warning or above, areas located in highly sensitive areas with a risk level of warning or above, areas that constitute risk zones, and areas that constitute composite risk zones. Simultaneously record the starting position, ending position, dominant risk category, and risk level information of the areas included in the list.

10. A smart gasket assembly fit assessment system, used to implement the smart gasket assembly fit assessment method as described in any one of claims 1 to 9, characterized in that, include: Data modeling module: acquires the contour data of the gasket to be assembled, the structural data of the assembly groove, and the process timing response data during the gasket assembly process, and constructs an initial fit distribution model between the gasket and the assembly groove based on the contour data and the structural data of the assembly groove. Feature extraction module: Analyzes and processes the process timing response data, extracts dynamic response features that characterize the state changes during the gasket assembly process, and determines the assembly interface area corresponding to the dynamic response features; Atlas generation module: Based on the dynamic response characteristics, the bonding state parameters of the corresponding assembly interface area in the initial bonding distribution model are corrected to generate a virtual bonding atlas characterizing the actual bonding state of the sealing gasket assembly interface. Risk analysis module: Determines the regional adhesion risk distribution of the sealing gasket in the assembly interface based on the virtual bonding map; Evaluation output module: Generates gasket assembly compatibility evaluation results based on the regional fit risk distribution.

Images