An intelligent scheduling method for automobile sealing production resources based on big data analysis

Abstract

The application discloses a kind of based on big data analysis's automobile sealing production resource intelligent scheduling method, specifically related to production resource scheduling field, for solving the problem that it is difficult to match between resource allocation and assembly quality in the process of existing multi-sealing assembly production scheduling;Scheme by constructing sealing assembly coupling topological relation network, identifying coupling sealing assembly, establishing component collaborative size distribution domain and assembly stable sub-domain, constructing size collaborative deviation index, and combining historical production resource allocation to build associated prediction model, on this basis, with the minimum size collaborative deviation cumulative amount as the goal, using hybrid genetic algorithm to jointly optimize equipment resources, material resources and process path resources, generate production scheduling strategy for assembly stability, so as to realize the systematization, refinement and intelligent scheduling of sealing production resources.

Description

Technical Field

[0001] This invention relates to the field of production resource scheduling technology, and more specifically, to an intelligent scheduling method for automotive sealing component production resources based on big data analysis. Background Technology

[0002] In the current automotive manufacturing process, seals are crucial assembly components for key parts such as doors, windows, engine compartments, trunks, and sunroofs. Their assembly quality directly affects the sealing performance, NVH performance, and long-term reliability of the entire vehicle. With the development of vehicle platform and modular design, the same vehicle model often contains seals of various specifications and structural forms. There are significant geometric coupling and functional synergy relationships between different seals during the assembly process. Their dimensional matching depends not only on the manufacturing precision of individual seals but also on the combined influence of dimensional fluctuations in other related seals.

[0003] In actual production, the production of seals involves multiple resource elements, including equipment resources, material resources, and process path resources. Different combinations of resources can cause fluctuations in material properties, drift in processing dimensions, and changes in assembly consistency, leading to systematic matching deviations in the assembly stage of the same batch of orders. Most existing production scheduling methods focus on equipment utilization, delivery cycle, or production takt time as primary objectives, ignoring the assembly coupling relationship and dimensional synergy effect between seals. They typically allocate resources from a single process or single equipment perspective, making it difficult to comprehensively consider the collaborative matching of multiple seals at the level of the entire vehicle assembly system. This results in a lack of effective linkage between production scheduling strategies and actual assembly quality.

[0004] When production orders are complex, vehicle models change frequently, or resource status fluctuates significantly, traditional scheduling methods struggle to identify potential systemic assembly risks in a timely manner. This can easily lead to batch assembly anomalies, rework, and quality fluctuations, thereby increasing production costs and impacting delivery stability. Therefore, a method is needed that combines historical assembly data with production resource configuration characteristics, starting from the coupling structure of the vehicle sealing system assembly, to achieve intelligent production resource scheduling oriented towards assembly stability, thereby improving the scientific rigor and reliability of sealing component production scheduling. Summary of the Invention

[0005] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide an intelligent scheduling method for automotive sealing component production resources based on big data analysis to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] A method for intelligent scheduling of automotive sealing component production resources based on big data analysis includes the following steps:

[0008] S1. Construct the assembly coupling topology between seals to generate a sealing system coupling relationship network based on the entire vehicle.

[0009] S2. Perform community detection analysis on the coupling relationship network of the sealing system to identify coupled sealing components;

[0010] S3. Based on historical orders for vehicle assembly procurement, statistical analysis is performed on the size combinations of each seal within the coupling sealing assembly to establish a component collaborative size distribution domain. From the collaborative size distribution domain, assembly stability subdomains of different coupling sealing assemblies are extracted.

[0011] S4. Extract non-compliant orders from historical orders, and in the component coordination dimension distribution domain, quantify the offset of the coupling sealing component dimension combination corresponding to the non-compliant order relative to the assembly stability subdomain as a dimension coordination deviation index.

[0012] S5. Integrate historical production resource allocation and establish a correlation prediction model between production resource allocation and size coordination deviation index;

[0013] S6. For the production plan corresponding to the current order, group the seals included in the production plan according to their respective coupling sealing components to form the current production task component set;

[0014] S7. With minimizing the cumulative size coordination deviation of the current production task component set as the global objective, a hybrid genetic algorithm is used to jointly optimize the current production resource configuration and generate a production scheduling strategy.

[0015] As a further aspect of the present invention, in step S1, generating a sealing system coupling relationship network based on the entire vehicle specifically includes:

[0016] Based on the design and installation positions and dimensional tolerances of each seal in the vehicle, a seal association diagram is constructed. The nodes in the diagram represent seals. If two seals are physically adjacent in the vehicle model, a connection edge is established between the corresponding nodes.

[0017] Based on the fit clearance of adjacent seals in historical assembly records, the actual size matching dispersion between each pair of associated seals is calculated as the edge weight of the corresponding connection edge. Based on all seal nodes, connection edges and edge weights, a coupling relationship network of the sealing system is constructed.

[0018] As a further aspect of the present invention, in S2, community detection analysis is performed on the coupling relationship network of the sealing system to identify coupled sealing components, specifically including:

[0019] The global modularity of the entire network is calculated based on the edge weights of all connecting edges in the coupling relationship network of the sealing system. Each sealing node is treated as an independent community. Each node in the network is traversed, and the change in modularity gain caused by moving to each adjacent community is calculated.

[0020] Each sealing node is moved to the neighboring community that can generate the maximum positive modularity gain. After all nodes have been moved, the network modularity is recalculated. The process terminates when the network modularity cannot be improved after multiple iterations. The set of nodes corresponding to each community is defined as a coupled sealing component and labeled with the corresponding number.

[0021] As a further aspect of the present invention, in step S3, establishing a component cooperative size distribution domain and extracting assembly stability subdomains of different coupled sealing components from the cooperative size distribution domain specifically includes:

[0022] Retrieve all historical orders for seals from the vehicle assembly procurement database, assign corresponding historical order records to each coupling sealing component number, and extract the qualified order set from the historical order records whose data feedback meets the preset standards;

[0023] Extract the measured dimensional data of each seal from the qualified order set, and combine the dimensional data of all seals into a set of data points based on the coupling sealing component number;

[0024] The data point set corresponding to each coupling sealing component number is divided, and the cooperative size probability distribution of each coupling sealing component is generated by fitting the multivariate kernel density estimation method. The connected regions with probability density higher than the preset confidence level are extracted and defined as the assembly stability subdomain of the current coupling sealing component.

[0025] As a further aspect of the present invention, in step S4, the offset of the coupling sealing component size combination corresponding to the substandard order relative to the assembly stability subdomain is quantified into a size coordination deviation index, specifically including:

[0026] Filter the set of non-conforming orders in the historical orders corresponding to each coupling sealing component number whose data feedback does not meet the preset standard, extract the actual measured values ​​of the sealing component size corresponding to the non-conforming order set, integrate them into a set of data points divided based on the coupling sealing component number, and map them to the corresponding component collaborative size distribution domain;

[0027] For the set of data points corresponding to the non-conforming order set, calculate the multivariate statistical distance to all data points in the assembly stability subdomain corresponding to the same coupling sealing component number. The multivariate statistical distance is calculated using Mahalanobis distance, and the calculation result is used as the dimensional coordination deviation index of the coupling sealing component corresponding to the non-conforming order set.

[0028] As a further aspect of the present invention, in step S5, integrating historical production resource allocation and establishing a correlation prediction model between production resource allocation and size coordination deviation index specifically includes:

[0029] Based on the historical order records corresponding to the number of each coupling sealing component during vehicle assembly, obtain the production batch number corresponding to each historical order record, and extract the production resource configuration used in the production stage. The production resource configuration includes the equipment resources, material resources, and process path resources allocated to the production tasks of each seal in the coupling sealing component involved in the historical order records.

[0030] The production resource configuration of all historical orders is transformed into feature inputs. At the same time, the size coordination deviation index corresponding to the number of each coupled sealing component is used as a supervision label to construct a training sample set. Gradient boosting regression tree and training sample set are used for training to establish a correlation prediction model between production resource configuration and size coordination deviation index.

[0031] As a further aspect of the present invention, in step S6, for the production plan corresponding to the current order, the sealing components included in the production plan are grouped according to their respective coupling sealing assemblies to form the current production task component set, specifically including:

[0032] Obtain the production plan corresponding to the current vehicle assembly procurement order requirements, iterate through each coupling sealing component number, and check whether the current production plan includes all member seals corresponding to that coupling sealing component number.

[0033] All member seals that exist in the current production plan are integrated into the current production task component set.

[0034] As a further aspect of the present invention, in step S7, with minimizing the cumulative size coordination deviation of the current production task component set as the global objective, a hybrid genetic algorithm is used to jointly optimize the current production resource configuration to generate a production scheduling strategy, specifically including:

[0035] The available equipment resources, material resources, and process path resources of the current sealing production line are combined and encoded to construct a resource allocation scheme that represents the complete production resource allocation structure. A hybrid genetic evolution process is constructed to perform selection, crossover, and mutation operations on the current resource allocation scheme to form a continuously iteratively updated set of candidate resource allocation schemes.

[0036] In each iteration, the cumulative size coordination deviation is output based on the correlation prediction model between production resource allocation and size coordination deviation index. The iteration ends after the preset number of iterations is reached, and the candidate resource allocation scheme with the smallest cumulative size coordination deviation during the iteration process is selected as the production scheduling strategy.

[0037] The technical effects and advantages of the intelligent scheduling method for automotive sealing component production resources based on big data analysis proposed in this invention are as follows:

[0038] This invention constructs an assembly coupling topology network between seals and introduces community discovery analysis to achieve automatic identification of strongly coupled sealing components in the vehicle sealing system. This enables production scheduling to be upgraded from a single seal level to a multi-seal collaboration level, effectively characterizing the complex assembly association structure between seals.

[0039] This invention establishes a component collaborative dimension distribution domain and an assembly stability subdomain, transforming historical assembly data into a quantifiable collaborative dimension stability space. Furthermore, it constructs a dimension collaborative deviation index to achieve a systematic characterization of the assembly stability of multiple sealing components, enabling accurate quantification of the impact of production resource allocation on assembly quality. Based on this, a correlation prediction model between historical production resource allocation and dimension collaborative deviation can assess the potential impact of different resource allocation schemes on assembly stability in advance, thus introducing assembly quality constraints into the production scheduling decision-making process. Through a hybrid genetic joint optimization with the global objective of minimizing the cumulative dimension collaborative deviation, it achieves coordinated optimization of equipment resources, material resources, and process path resources. This allows the production scheduling strategy to meet capacity and delivery requirements while also considering assembly consistency, effectively reducing the probability of systemic assembly anomalies, minimizing rework and quality fluctuations, and improving the overall stability and reliability of sealing component production and vehicle assembly processes. This significantly enhances the refined scheduling capability and comprehensive operational level of the production system. Attached Figure Description

[0040] Figure 1 This is a schematic diagram of an intelligent scheduling method for automotive sealing component production resources based on big data analysis, according to the present invention. Detailed Implementation

[0041] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0042] Example 1

[0043] Figure 1 This invention presents an intelligent scheduling method for automotive sealing component production resources based on big data analysis, comprising the following steps:

[0044] S1. Construct the assembly coupling topology between seals to generate a sealing system coupling relationship network based on the entire vehicle.

[0045] S2. Perform community detection analysis on the coupling relationship network of the sealing system to identify coupled sealing components;

[0046] S3. Based on historical orders for vehicle assembly procurement, statistical analysis is performed on the size combinations of each seal within the coupling sealing assembly to establish a component collaborative size distribution domain. From the collaborative size distribution domain, assembly stability subdomains of different coupling sealing assemblies are extracted.

[0047] S4. Extract non-compliant orders from historical orders, and in the component coordination dimension distribution domain, quantify the offset of the coupling sealing component dimension combination corresponding to the non-compliant order relative to the assembly stability subdomain as a dimension coordination deviation index.

[0048] S5. Integrate historical production resource allocation and establish a correlation prediction model between production resource allocation and size coordination deviation index;

[0049] S6. For the production plan corresponding to the current order, group the seals included in the production plan according to their respective coupling sealing components to form the current production task component set;

[0050] S7. With minimizing the cumulative size coordination deviation of the current production task component set as the global objective, a hybrid genetic algorithm is used to jointly optimize the current production resource configuration and generate a production scheduling strategy.

[0051] In S1, a sealing system coupling relationship network is generated on a vehicle-by-vehicle basis.

[0052] The process involves acquiring a complete 3D digital model of the target vehicle and its corresponding sealing system design data. This design data includes the installation coordinates, installation direction, contact interface structure, and corresponding dimensional tolerances of various seals within the vehicle structure. Based on the spatial geometry presented in the vehicle model, a unified coordinate mapping is performed on the installation areas of all seals, enabling comparable spatial distribution structures for different seals within the same vehicle coordinate system. Subsequently, each seal is treated as an independent node, and its installation area in the vehicle model is represented by a 3D bounding box. A spatial neighborhood scan is performed on all nodes to detect whether any two seal installation areas have contact, nesting, overlapping, or clearance fit relationships. If two seals are detected to have direct physical adjacency in the assembly structure, a connection edge is established between the corresponding nodes, and the contact type, relative position direction, and associated assembly structure features corresponding to this adjacency relationship are recorded in the edge attributes. For different types of adjacency relationships, and in conjunction with the design dimensional tolerance zones, the association strength of the connecting edges is initially calibrated. For example, for seal pairs with direct press-fit mating, a high association level is calibrated; for seal pairs with lap or indirect contact mating, a medium association level is calibrated; and for seal pairs with only partial contact, a low association level is calibrated. Through this method, all seal nodes and their physical adjacency relationships in the entire vehicle are constructed into a seal association diagram with a clear structure and initially controllable edge weights.

[0053] The historical vehicle assembly record database is retrieved, and for each pair of adjacent seals with established connections, the actual mating clearance data in the corresponding assembly batch is extracted. This mating clearance data originates from measured records of the seal contact area during the assembly inspection process, including key assembly parameters such as the compression of the sealing lip after installation, the slot fitting distance, and the overlap thickness. For each pair of associated seals, multiple mating clearance data collected within the same order cycle are centrally processed to form a time-series sample set. Based on this, statistical dispersion calculation is performed on the sample set. Specifically, the mean of the sample set is first calculated to obtain the typical mating clearance benchmark value for the seal under stable assembly conditions. Then, the deviation of each measured clearance value from this benchmark value is calculated, and the mean is obtained after absolute value processing of all deviations, resulting in a dispersion index reflecting the stability of the seal's assembly matching. The larger the dispersion index value, the more significant the dimensional matching fluctuation of the seal in different assembly batches, and the higher its assembly coupling instability. The dispersion index is assigned as the edge weight of the corresponding connection edge in the seal association graph, and the weighting is updated sequentially for all connection edges. Finally, based on the dispersion weights matched with the actual dimensions of all seal nodes, connection edges, and their corresponding dimensions, a sealing system coupling relationship network reflecting the distribution characteristics of the actual assembly coupling strength is constructed. This allows the network to realistically characterize the tightness of the assembly association between various seals in the vehicle sealing system at the structural level.

[0054] In step S2, community detection analysis is performed on the coupling relationship network of the sealing system to identify coupled sealing components.

[0055] The system reads all node, edge, and edge weight data of the coupling relationship network of the sealing system, treating the edge weight as a quantitative expression of the assembly coupling strength between adjacent sealing components, and calculates the global modularity of the network accordingly. The calculation of global modularity follows the approach of comparing the intra-community connection strength with a random connection benchmark: the sum of all edge weights in the network is used as the overall benchmark for the network's coupling strength; then, using the current community partitioning state as input, the edge weights between node pairs within each community are summarized to obtain the total intra-community coupling strength; simultaneously, based on the weighted degree value of each node, the cumulative amount of connection weights between that node and other nodes in the entire network is calculated, and a benchmark for the expected strength of random connections is constructed according to the proportion of the node's weighted degree in the entire network, used to measure the theoretically expected connection strength within a community without considering structural coupling; finally, the actual intra-community coupling strength is compared with the expected random connection strength, and the results are summed over all communities to obtain the global modularity index.

[0056] After calculating the global modularity, each sealing node is initialized as an independent community, forming the initial community partition. Then, each node in the network is traversed, and the communities of its neighboring nodes are enumerated. The operation of moving a node from its current community to a target neighboring community is simulated one by one. During the simulation, the internal coupling strength and expected strength terms of the two affected communities are updated to obtain the modularity gain change caused by the movement operation. For each node, the modularity gain change results corresponding to all neighboring community movement schemes are retained as the basis for subsequent node movement decisions. A positive modularity gain change indicates that the movement increases the relative concentration of internal coupling within the community, making it a feasible candidate movement direction.

[0057] After evaluating the modularity gain changes of each node relative to its neighboring communities, the node migration iteration process begins. Specifically, each sealed node is processed sequentially according to a pre-defined node traversal order. For the current node, the modularity gain changes of all its neighboring communities are read, and candidate neighboring communities with positive gains are selected. The neighboring community with the largest modularity gain change is chosen as the node's migration target community. If the modularity gain changes of the node are not positive for any of its neighboring communities, the node remains in its original community. After each node migration is completed, the community internal coupling strength statistics, node weighted degree summary, and corresponding random connection expectation strength terms related to that node are immediately updated to ensure that subsequent nodes use the updated community partitioning state when calculating migration benefits. After all nodes have completed their migration decisions in the current round, the global modularity is recalculated as the modularity result at the end of this iteration, and this result is compared with the modularity result of the previous iteration. The termination condition is determined by the criterion that the modularity has not improved after multiple consecutive iterations. The "multiple consecutive iterations" are defined by a fixed number of iterations; for example, convergence is considered achieved and the process terminates if the modularity has not improved after three consecutive iterations. If the modularity is still improving, the next iteration begins, and the node migration process described above is repeated. After the iteration terminates, the current community partitioning state is the final community discovery result, and the set of sealing nodes contained within each community is defined as a coupled sealing component.

[0058] In step S3, a component collaborative size distribution domain is established, and assembly stability subdomains of different coupled sealing components are extracted from the collaborative size distribution domain.

[0059] All historical order records are retrieved from the vehicle assembly procurement database. These historical order records include at least the order number, vehicle assembly number, production batch number, seal model, quantity information, assembly inspection results, and after-sales quality feedback data. Based on the aforementioned coupling sealing component numbers, each historical order record is matched one by one. Each seal involved in the order is categorized and organized according to its corresponding coupling sealing component number, forming a set of historical order records indexed by the coupling sealing component number. On this basis, for each historical order record, its corresponding assembly inspection results and after-sales quality feedback information are extracted to construct data fields for determining the order's quality status. The preset standards are set using a combined judgment method, specifically including: the assembly process inspection record showing that all seal assembly inspection items have passed; no seal-related defect marks appearing in the vehicle's off-line quality inspection; and no returns, exchanges, or repairs due to seal failure occurring within the after-sales quality traceability period. Through the above multi-condition joint judgment, historical orders that meet all judgment conditions are marked as qualified orders, and these are summarized to form a qualified order set under the corresponding coupling sealing component number.

[0060] After constructing the qualified order set, dimensional measurement data is extracted from each seal involved in the qualified order set. The extracted data objects are the measured values ​​of key functional dimensions of the seals that directly participate in the mating relationship during the assembly process, rather than all geometric dimensions of the seals. The key dimensions include, but are not limited to, structural parameters that directly affect the assembly clearance and clamping state, such as sealing lip height, sealing lip compression, groove fitting width, mounting base thickness, overlap area contact thickness, and functional distance from the assembly surface to the sealing surface. The dimensional data comes from online inspection records, offline sampling inspection records, and quality inspection data from the assembly station in the production inspection process. After consistency verification of data from different sources, a unified data record format is formed. Subsequently, each coupled sealing component is numbered, and the corresponding key dimension measurement data is combined and organized according to the order of the member seals in the component. The key dimension data of each seal in the same order are combined into a multi-dimensional data point, and each data point completely describes the collaborative dimensional state of the component in a single assembly. Perform the above combination operation on the order data in all qualified order sets to form a data point set structure divided by the coupling sealing component number, so that each coupling sealing component corresponds to an independent data point sample set, thereby completing the structured expression of the component-level collaborative size state at the data level.

[0061] After constructing the data point sets corresponding to the numbers of each coupled sealing component, collaborative size distribution modeling is performed on the data point sets of each component. Multidimensional statistical fitting is then performed on the data point sets corresponding to each component. A multivariate kernel density estimation method is used to fit the joint distribution characteristics of the multidimensional size data, resulting in a continuous probability distribution structure describing the collaborative size state space of the component. This probability distribution reflects the density and distribution trend of different size combinations in historical qualified orders. Based on the fitted collaborative size probability distribution, the probability density values ​​are hierarchically divided, and high probability density regions are selected as stable state determination regions. The pre-set confidence level is set using empirical statistical methods; for example, a high-density interval covering the main concentrated area of ​​the sample is selected as the standard for the stable interval. Connectivity analysis is performed on the dense regions of the probability distribution map to extract continuous high-density connected regions. This high-density connected region is defined as the assembly stability subdomain of the corresponding coupled sealing component. This stability subdomain expresses the collaborative size space range of the component under historical stable assembly conditions in the form of a multidimensional region.

[0062] In S4, the offset of the size combination of the coupling sealing components corresponding to the substandard order relative to the assembly stability subdomain is quantified as a size coordination deviation index.

[0063] All historical order records involving sealing component assembly are retrieved from the vehicle assembly procurement database and archived. Based on this, quality feedback information in the historical order records is systematically filtered, and a set of non-conforming orders is constructed according to clearly defined quality judgment criteria. These criteria are defined using a multi-dimensional quality anomaly joint judgment standard, including but not limited to: any sealing component showing an assembly non-conforming mark in the assembly inspection station records; abnormal inspection records related to the sealing system such as air leakage, water seepage, and abnormal noise during the final vehicle inspection; repair records of rework, replacement, or readjustment of sealing components in the after-sales quality traceability system; and feedback on issues directly related to sealing performance or assembly abnormalities in the return database. When a corresponding abnormal record appears in any of the above data sources for the same order, the order is determined to be a non-conforming order and included in the non-conforming order set. After constructing the non-conforming order set, for each coupled sealing component number, dimensional measurement data is extracted from the corresponding sealing component in the non-conforming order set. The extracted data consists of the measured values ​​of key functional dimensions of the sealing component that directly participate in the mating relationship during the assembly process. Subsequently, the measured data of the key dimensions of the seals of each member in the same order are combined according to the coupling sealing component number to form a data point set based on the component number, and the data point set is mapped to the corresponding cooperative dimension distribution domain of the coupling sealing component.

[0064] After mapping the set of non-conforming order data points, for each coupling sealing component number in the assembled vehicle, a dimensional coordination deviation quantification calculation is performed on the corresponding set of non-conforming order data points. Specifically, the constructed assembly stable subdomain for the coupling sealing component is retrieved, and statistical analysis is performed on the data points corresponding to all historical conforming orders within the stable subdomain to construct the central position and coordinated fluctuation structure representation of the stable subdomain in the multi-dimensional dimensional space. Subsequently, each non-conforming order data point is compared and analyzed with the set of data points corresponding to the stable subdomain to calculate its overall offset relative to the stable subdomain in the multi-dimensional dimensional space. This offset is quantified using a multivariate statistical distance method, where Mahalanobis distance is used as the calculation method to fully consider the correlation structure between different dimensional parameters, ensuring that the calculation results accurately reflect the overall deviation level of the multi-dimensional coordinated dimensional combination state. In the specific calculation process, for each non-conforming order data point, the statistical distance between it and all data points within the stable subdomain is calculated sequentially, and the obtained distance results are summarized to form the comprehensive deviation representation value of the non-conforming order in the coordinated dimensional space. The overall deviation is used as the dimensional coordination deviation index for the corresponding coupling sealing component under the order. The deviation indexes of all non-conforming orders are archived and stored to form a dimensional coordination deviation index database with the assembly vehicle number and the coupling sealing component number as the joint primary key.

[0065] In step S5, historical production resource allocation is integrated to establish a correlation prediction model between production resource allocation and size coordination deviation index.

[0066] Based on the historical order records corresponding to each coupling sealing component number during the vehicle assembly stage, the production batch number of each component is retrieved from the production management system. This batch number is then used as an index to trace the actual production resource configuration used for the corresponding batch during the seal production stage. The production resource configuration consists of three elements: equipment resources, material resources, and process path resources. Equipment resources specifically include the molding equipment number, vulcanizing equipment number, cutting equipment number, and surface treatment equipment number responsible for each seal production task, along with their corresponding workstation location identifiers. Material resources include the batch number of the compound rubber formula, the batch number of the skeleton material, the batch number of the adhesive model, and batch information of related auxiliary materials. Process path resources include the actual process sequence structure executed during the production of each seal, the sequence of processes, the identifiers of key process nodes, and the configuration status of the process parameters corresponding to each process node. For multiple seal production tasks involved in the same historical order, they are aggregated and organized according to their respective coupling sealing component numbers. The equipment resource allocation structure, material batch combination structure, and process path execution structure corresponding to each seal within that component are unified and integrated to construct a production resource configuration record with the coupling sealing component as the basic unit. Furthermore, the above-mentioned production resource allocation records are subjected to standardized coding processing, which transforms discrete equipment numbers, material batch numbers, and process node identifiers into a unified numbering system, and records the combination status of various resources in the production process in a sequential structure, so that each historical order forms a production resource allocation description vector with a complete structure and clear hierarchy.

[0067] After structuring and organizing historical production resource allocation data, feature input construction processing is performed on each production resource allocation description vector. Specifically, for equipment resources, features such as equipment type distribution, key equipment combination patterns, and equipment switching sequence are extracted. Equipment numbers are mapped to unified equipment category codes, and a sequential feature structure is formed according to the process execution order. Subsequently, for material resources, features such as compound rubber formulation type, material batch consistency identifier, and key performance level are extracted, transforming the original batch information into a classification feature vector with engineering meaning. Then, for process path resources, features such as the arrangement structure of process nodes, key process segment combination relationships, and process switching rhythm are extracted, transforming the original process path records into sequential feature descriptions. After completing the above multi-type feature extraction, equipment features, material features, and process path features are concatenated and integrated to construct a unified production resource allocation feature input vector. Simultaneously, the dimensional coordination deviation index corresponding to each coupled sealing component number is used as a supervision label and matched one-to-one with the corresponding production resource allocation feature input vector to construct a complete training sample set. In the model construction process, a gradient boosting regression tree is used as the prediction model structure. This model is constructed by stacking multiple regression decision trees in an iterative manner. In each iteration, the prediction residual of the previous model is used as the learning target of the new tree model, gradually approximating the nonlinear mapping relationship between production resource allocation and size coordination deviation. During model training, a fixed-round iteration method is used for parameter updates. In the example, it is set to 100 consecutive iterations. After each iteration, the model output is evaluated for error based on the training sample set. Training is terminated early when the error decrease is lower than the preset convergence criterion for several consecutive rounds. Finally, a stable correlation prediction model between production resource allocation and size coordination deviation index is formed, realizing accurate prediction of the coordination deviation level under different resource allocation schemes.

[0068] In step S6, for the production plan corresponding to the current order, the seals included in the production plan are grouped according to their respective coupling sealing components to form the current production task component set.

[0069] The system retrieves current vehicle assembly order demand data from the vehicle assembly procurement management system. Based on the vehicle configuration, batch size, and delivery cycle information included in the order, a corresponding seal production plan is generated. This production plan is structured with seal model, quantity requirement, production batch arrangement, and planned production time as its core content. Subsequently, a mapping relationship is established for all seals involved in the current production plan, and each seal is categorized and organized according to the aforementioned coupled sealing component numbering system. For each coupled sealing component number, the current production plan is checked to see if it contains all member seal models defined within that component, and the corresponding quantity requirements and production batch arrangements are verified for consistency to ensure that all seals within the same component are fully covered within the current planning cycle. During the traversal process, components lacking any member seal are automatically marked and removed from the current round of component candidate sets, retaining only component records with complete member seals in the current production plan. After verifying the integrity of component members, the coupled sealing components that passed the verification are uniformly integrated to form the current production task component set. All coupled sealing components whose member seals are fully present in the current production plan are summarized according to component number. The seal model, quantity requirements, batch identifier, and planned production time window for each component are centrally organized to construct a component-level production task description structure. For multiple seal production tasks within the same component, their task order is uniformly marked according to the takt time requirements and delivery sequence given in the production plan, ensuring a consistent execution rhythm for each member task within the component. Simultaneously, the production task description structures of each component are uniformly written into the production task component set as constraints for production resource allocation, forming the current production task component set data structure with coupled sealing components as the basic scheduling unit.

[0070] In step S7, with the goal of minimizing the cumulative size coordination deviation of the current production task component set, a hybrid genetic algorithm is used to jointly optimize the current production resource configuration and generate a production scheduling strategy.

[0071] All available equipment, material, and process resources in the current sealing production line are systematically organized to construct a structured resource list. Equipment resources include the numbers, workstation locations, and executable process ranges of various molding, vulcanizing, cutting, and processing equipment. Material resources include the compound formulation type, material batch identifier, skeleton material type, and auxiliary material configuration scheme. Process resources include the sequence of processes, key process nodes, and corresponding execution constraints experienced by various sealing components from raw material input to finished product output. Based on this resource list, according to the production requirements of each coupled sealing component in the production task component set, equipment allocation combinations, material batch combinations, and process path combinations are systematically enumerated and combined with coding. Each feasible resource combination structure is transformed into a resource configuration coding sequence of uniform length to fully describe a set of production resource configuration schemes. Subsequently, an initial candidate resource configuration scheme set is constructed using a combination of random generation and rule constraints. Random generation covers a large search space, while rule constraints ensure that the generated schemes meet basic production constraints such as equipment process matching, material process compatibility, and logical coherence of the process path. Based on this, a hybrid genetic evolution process is constructed to perform continuous iterative update operations on the initial candidate resource allocation scheme set. Specifically, this includes the selection of the best scheme based on fitness ranking, the crossover and recombination operation based on resource structure recombination rules, and the mutation adjustment operation based on local perturbation mechanism. This allows the candidate schemes to gradually evolve towards a better resource allocation structure during the continuous iteration process, forming a continuously iteratively updated candidate resource allocation scheme set.

[0072] In the hybrid genetic evolution process, after each round of updating the candidate resource allocation scheme set, a size coordination deviation evaluation is performed on each resource allocation scheme obtained in the current iteration. Specifically, the equipment resource combination structure, material resource combination structure, and process path combination structure included in the candidate scheme are converted into a unified format feature input vector, which is then input into the constructed correlation prediction model between production resource allocation and size coordination deviation index. The model outputs the corresponding size coordination deviation prediction result. For all production task components involved under the same candidate resource allocation scheme, the predicted size coordination deviation values ​​corresponding to each component are summarized and accumulated to form the cumulative size coordination deviation of the candidate scheme under the current order conditions. After completing a round of prediction evaluation of all candidate schemes, the cumulative size coordination deviations are sorted, and the sorting results are fed back to the selection and recombination stage in the genetic evolution process, so that subsequent iterations prioritize retaining resource allocation structures with lower coordination deviation levels. The above prediction evaluation and genetic update process is executed cyclically for a preset number of iterations; the example is set to one hundred consecutive iterations, and the current optimal candidate scheme is recorded after each iteration. Once the iteration process is complete, the scheme with the smallest cumulative size coordination deviation is selected from the candidate resource configuration schemes generated during all iterations and used as the final production resource configuration structure. The equipment assignment scheme, material batch allocation scheme, and process path execution sequence corresponding to this resource configuration structure are then integrated to generate a production scheduling strategy, which is used to guide the actual scheduling and execution of the sealing component production task under the current order conditions.

[0073] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. The semiconductor medium can be a solid-state drive.

[0074] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0075] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and modules described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0076] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or modules may be electrical, mechanical, or other forms.

[0077] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical modules; they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0078] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0079] If the aforementioned functions are implemented as software functional modules 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 application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several 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 application. 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.

[0080] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0081] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for intelligent scheduling of automotive sealing component production resources based on big data analysis, characterized in that, Includes the following steps: S1. Construct the assembly coupling topology between seals to generate a sealing system coupling relationship network based on the entire vehicle. S2. Perform community detection analysis on the coupling relationship network of the sealing system to identify coupled sealing components, including: The global modularity of the entire network is calculated based on the edge weights of all connecting edges in the coupling relationship network of the sealing system. Each sealing node is treated as an independent community. Each node in the network is traversed, and the change in modularity gain caused by moving to each adjacent community is calculated. Each sealing node is moved to the neighboring community that can generate the maximum positive modularity gain. After all nodes have been moved, the network modularity is recalculated. The process terminates when the network modularity cannot be improved after multiple iterations. The set of nodes corresponding to each community is defined as a coupled sealing component and labeled with the corresponding number. S3. Based on historical orders for vehicle assembly procurement, statistical analysis is performed on the size combinations of each seal within the coupling sealing assembly to establish a component collaborative size distribution domain. From the collaborative size distribution domain, assembly stability subdomains of different coupling sealing assemblies are extracted. S4. Extract non-compliant orders from historical orders, and in the component coordination dimension distribution domain, quantify the offset of the coupling sealing component dimension combination corresponding to the non-compliant order relative to the assembly stability subdomain as a dimension coordination deviation index. S5. Integrate historical production resource allocation and establish a correlation prediction model between production resource allocation and dimensional coordination deviation index, including: Based on the historical order records corresponding to the number of each coupling sealing component during vehicle assembly, obtain the production batch number corresponding to each historical order record, and extract the production resource configuration used in the production stage. The production resource configuration includes the equipment resources, material resources, and process path resources allocated to the production tasks of each seal in the coupling sealing component involved in the historical order records. The production resource configuration of all historical orders is transformed into feature inputs. At the same time, the size coordination deviation index corresponding to the number of each coupled sealing component is used as a supervision label to construct a training sample set. Gradient boosting regression tree and training sample set are used for training to establish a correlation prediction model between production resource configuration and size coordination deviation index. S6. For the production plan corresponding to the current order, group the seals included in the production plan according to their respective coupling sealing components to form the current production task component set; S7. With minimizing the cumulative size coordination deviation of the current production task component set as the global objective, a hybrid genetic algorithm is used to jointly optimize the current production resource allocation and generate a production scheduling strategy, including: The available equipment resources, material resources, and process path resources of the current sealing production line are combined and encoded to construct a resource allocation scheme that represents the complete production resource allocation structure. A hybrid genetic evolution process is constructed to perform selection, crossover, and mutation operations on the current resource allocation scheme to form a continuously iteratively updated set of candidate resource allocation schemes. In each iteration, the cumulative size coordination deviation is output based on the correlation prediction model between production resource allocation and size coordination deviation index. The iteration ends after the preset number of iterations is reached, and the candidate resource allocation scheme with the smallest cumulative size coordination deviation during the iteration process is selected as the production scheduling strategy.

2. The intelligent scheduling method for automotive sealing component production resources based on big data analysis according to claim 1, characterized in that, In step S1, generating the sealing system coupling relationship network on a vehicle-by-vehicle basis specifically includes: Based on the design and installation positions and dimensional tolerances of the seals in the vehicle, a seal association diagram is constructed. The nodes in the diagram represent seals. If two seals are physically adjacent in the vehicle model, a connection edge is established between the corresponding nodes. Based on the fit clearance of adjacent seals in historical assembly records, the actual size matching dispersion between each pair of associated seals is calculated as the edge weight of the corresponding connection edge. Based on all seal nodes, connection edges and edge weights, a coupling relationship network of the sealing system is constructed.

3. The intelligent scheduling method for automotive sealing component production resources based on big data analysis according to claim 1, characterized in that, In step S3, establishing a component collaborative size distribution domain and extracting assembly stability subdomains of different coupled sealing components from the collaborative size distribution domain specifically includes: Retrieve all historical orders for seals from the vehicle assembly procurement database, assign corresponding historical order records to each coupling sealing component number, and extract the qualified order set from the historical order records whose data feedback meets the preset standards; Extract the measured dimensional data of each seal from the qualified order set, and combine the dimensional data of all seals into a set of data points based on the coupling sealing component number; The data point set corresponding to each coupling sealing component number is divided, and the cooperative size probability distribution of each coupling sealing component is generated by fitting the multivariate kernel density estimation method. The connected regions with probability density higher than the preset confidence level are extracted and defined as the assembly stability subdomain of the current coupling sealing component.

4. The intelligent scheduling method for automotive sealing component production resources based on big data analysis according to claim 1, characterized in that, In step S4, the offset of the coupling sealing component size combination corresponding to the substandard order relative to the assembly stability subdomain is quantified into a size coordination deviation index, specifically including: Filter the set of non-conforming orders in the historical orders corresponding to each coupling sealing component number whose data feedback does not meet the preset standard, extract the actual measured values ​​of the sealing component size corresponding to the non-conforming order set, integrate them into a set of data points divided based on the coupling sealing component number, and map them to the corresponding component collaborative size distribution domain; For the set of data points corresponding to the non-conforming order set, calculate the multivariate statistical distance to all data points in the assembly stability subdomain corresponding to the same coupling sealing component number. The multivariate statistical distance is calculated using Mahalanobis distance, and the calculation result is used as the dimensional coordination deviation index of the coupling sealing component corresponding to the non-conforming order set.

5. The intelligent scheduling method for automotive sealing component production resources based on big data analysis according to claim 1, characterized in that, In step S6, for the production plan corresponding to the current order, the sealing components included in the production plan are grouped according to their respective coupling sealing assemblies to form the current production task component set, specifically including: Obtain the production plan corresponding to the current vehicle assembly procurement order requirements, iterate through each coupling sealing component number, and check whether the current production plan includes all member seals corresponding to that coupling sealing component number. All member seals that exist in the current production plan are integrated into the current production task component set.

Images