Optimized scheduling method for hardware mold production

Abstract

The invention relates to the technical field of production scheduling, in particular to an optimal scheduling method for hardware mold production. The method comprises the following steps: collecting hardware mold parameters and carrying out topological structure analysis to generate a hardware topological structure and a skeleton, then simulating a skeleton assembly process and recording a movement track, carrying out time sequence arrangement on the track by using mold production historical data, identifying a conflict track and a coordination track, and finally, carrying out hardware mold assembly. Resource requirements are calculated according to mold parameters, parallel production scheduling is carried out, meanwhile, the dependency relation between the steps, conflict track rearrangement steps, coordination and conflict rearrangement production plans are deduced, scheduling optimization is carried out, and intelligent scheduling of hardware mold production is achieved. The optimal scheduling method for hardware mold production is more efficient.

Description

Technical Field

[0001] The invention relates to the technical field of production scheduling, and in particular to an optimization scheduling method for hardware mold production. Background Art

[0002] In modern manufacturing, the production efficiency and quality of hardware molds directly affect the competitiveness of products. However, traditional production scheduling methods are often unable to flexibly respond to complex production environments and rapidly changing market demands, resulting in waste of resources and time delays in the production process, which in turn affects the overall production efficiency. As the market demand for hardware molds continues to increase, how to optimize production scheduling has become a key issue that needs to be solved urgently. In the production process of hardware molds, the complexity and diversity of mold parameters make production scheduling face many challenges, especially in mold assembly and production processes. There are various potential conflicts and inconsistencies, which not only increases the difficulty of production management, but also affects the accuracy and efficiency of production. Existing scheduling systems often lack effective use of historical data and cannot adjust and optimize production processes in real time. Summary of the invention

[0003] Based on this, it is necessary to provide an optimized scheduling method for hardware mold production to solve at least one of the above technical problems.

[0004] To achieve the above purpose, the optimization scheduling method for hardware mold production includes the following steps:

[0005] Step S1: Collecting hardware mold parameters; performing topological structural analysis on the hardware mold parameters to generate a hardware topological structure; reconstructing a mold skeleton on the mold point cloud data to generate a hardware mold skeleton;

[0006] Step S2: performing skeleton assembly simulation on the skeleton of the hardware mold based on the hardware topological structure to obtain a simulated skeleton assembly process; performing skeleton movement capture on the simulated skeleton assembly process to generate a skeleton assembly movement trajectory;

[0007] Step S3: Acquire mold production history data; arrange the skeleton assembly movement trajectory in time sequence according to the mold production history data to generate a trajectory occurrence sequence; perform conflict identification on the skeleton movement trajectory based on the trajectory occurrence sequence to obtain conflicting trajectory steps and coordinated trajectory steps;

[0008] Step S4: Calculate the running resources of the coordination trajectory step according to the hardware mold parameters to obtain the required resources of the coordination step; perform parallel production scheduling on the coordination trajectory step based on the required resources of the coordination step to generate a coordinated parallel production plan;

[0009] Step S5: Derivation of dependency relationships of conflicting trajectory steps based on the trajectory occurrence sequence to obtain dependencies between steps; and rearrangement of the conflicting trajectory steps based on the dependencies between steps to generate a conflict rearrangement production plan;

[0010] Step S6: Perform data fusion on the coordinated parallel production plan and the conflict rearrangement production plan to obtain a fused production scheduling plan; perform scheduling optimization based on the fused production scheduling plan to execute intelligent optimization scheduling for hardware mold production.

[0011] The present invention provides basic data for subsequent analysis through the collection of hardware mold parameters, the topological structure analysis effectively reveals the structural characteristics of the mold, the generation of hardware topological structure lays the foundation for mold design and optimization, the mold skeleton reconstruction ensures the accurate expression of the mold shape, the implementation of skeleton assembly simulation provides visualization support for the production process, the skeleton movement capture effectively records the dynamic changes in the assembly process, the generated movement trajectory provides important information for subsequent analysis, the acquisition of mold production history data provides a historical background for trajectory analysis, the implementation of time series arrangement ensures the logic of the trajectory, the execution of conflict identification can effectively reveal potential problems in the production process, and the obtained conflict trajectory steps and The coordination trajectory step provides a key basis for subsequent scheduling, the implementation of operating resource calculation ensures the rational allocation of resources, the generation of coordination step demand resources provides data support for parallel production scheduling, the successful implementation of parallel production scheduling improves production efficiency, the execution of dependency derivation provides a theoretical basis for conflict management, the implementation of step sequence rearrangement optimizes the production process, the generated conflict rearrangement production plan improves production flexibility, the implementation of data fusion ensures the comprehensiveness of the scheduling plan, and the execution of scheduling optimization provides an intelligent solution for hardware mold production, which significantly improves the efficiency and scientificity of hardware mold production as a whole, and provides strong technical support for production management.

[0012] Preferably, step S1 comprises the following steps:

[0013] Step S11: collecting hardware mold parameters; performing multi-dimensional normalization processing on the hardware mold parameters to obtain normalized hardware data;

[0014] Step S12: performing point cloud mapping processing on the normalized hardware data to obtain mold point cloud data; performing morphological contour analysis on the mold point cloud data to obtain hardware morphological features;

[0015] Step S13: performing topological structural analysis on the hardware morphological features to generate a hardware topological structure; performing assembly constraint modeling on the hardware topological structure to obtain constraint modeling data;

[0016] Step S14: performing local parameter fitting on the hardware topological structure based on the constraint modeling data to generate a hardware mold skeleton.

[0017] The present invention improves the consistency of data through multi-dimensional normalization processing of hardware mold parameters. The generation of normalized hardware data provides a basis for subsequent analysis. The point cloud mapping processing effectively converts the mold morphological information. The formation of mold point cloud data lays a foundation for morphological feature extraction. The implementation of morphological contour analysis reveals the characteristics of hardware molds. The generation of hardware morphological features provides necessary information for topological construction. The topological structure analysis effectively identifies the complexity of the mold structure. The generated hardware topological structure provides support for assembly constraint modeling. The formation of constraint modeling data ensures the rationality of the mold assembly process. The implementation of local parameter fitting improves the accuracy of the mold skeleton. The generation of the hardware mold skeleton provides a key basis for subsequent optimization scheduling. Overall, the efficiency and accuracy of hardware mold production are significantly improved, providing strong technical support for the scientific production management.

[0018] Preferably, step S2 comprises the following steps:

[0019] Step S21: mapping the hardware topological structure to assembly constraints to obtain mold constraint data; performing assembly rule conversion on the mold constraint data to generate assembly sequence data;

[0020] Step S22: performing skeleton assembly simulation on the hardware mold skeleton based on the assembly sequence data to obtain a simulated skeleton assembly process;

[0021] Step S23: performing time-series frame processing on the simulated skeleton assembly process to obtain assembly key frame data; performing skeleton posture recognition on the assembly key frame data to generate skeleton spatial position data;

[0022] Step S24: Smoothing the skeleton spatial position data to generate continuous path data; supplementing the missing values ​​of the continuous path data to generate a skeleton assembly movement trajectory.

[0023] The present invention ensures the stability of the mold assembly process by mapping the assembly constraints of the hardware topological structure. The generation of mold constraint data provides a basis for the standardization of assembly. The implementation of assembly rule conversion makes the assembly sequence data more reasonable. The execution of skeleton assembly simulation provides visualization support for actual production. The formation of simulated skeleton assembly process provides detailed dynamic information for subsequent analysis. The time series frame processing effectively extracts the key stages in the assembly process. The generation of assembly key frame data helps to deeply understand the assembly process. The implementation of skeleton posture recognition ensures the accurate positioning of the skeleton in space. The formation of skeleton spatial position data provides a basis for subsequent path planning. The trajectory smoothing processing improves the continuity of the moving trajectory. The implementation of missing value supplementation ensures the integrity of the skeleton assembly moving trajectory. Overall, the production scheduling process of hardware molds is significantly optimized, the production efficiency and accuracy are improved, and reliable technical support is provided for production management.

[0024] Preferably, step S3 comprises the following steps:

[0025] Step S31: Acquire mold production history data; perform time dimension encoding on the mold production history data to obtain a history sequence code;

[0026] Step S32: Perform spectrum conversion on the historical sequence code to generate time series frequency domain features; perform periodic extraction on the time series frequency domain features to obtain the trajectory time series pattern;

[0027] Step S33: performing time-space mapping processing on the skeleton assembly movement trajectory in the trajectory timing mode to generate the trajectory occurrence timing;

[0028] Step S34: performing gradient detection on the trajectory occurrence sequence to obtain trajectory overlap data; performing spatiotemporal mapping processing on the skeleton assembly movement trajectory based on the trajectory overlap data to generate the trajectory occurrence sequence.

[0029] The present invention provides rich background information for analysis by acquiring historical data of mold production. Time dimension encoding improves the degree of structuring of historical data. The generation of historical sequence encoding lays the foundation for subsequent analysis. Spectral conversion effectively extracts frequency domain features in the data. The generation of time-series frequency domain features reveals the potential laws in the production process. Periodic extraction ensures the accuracy of trajectory timing patterns. The formation of trajectory timing patterns provides a basis for the analysis of skeleton assembly movement trajectories. Space-time mapping processing improves the visualization effect of trajectory data. Gradient detection effectively identifies trajectory overlapping phenomena. The generation of trajectory overlapping data provides an important reference for subsequent optimization. Space-time mapping processing based on overlapping data ensures the accuracy of trajectory occurrence timing, which improves the intelligent level of hardware mold production scheduling as a whole, optimizes the production process, and improves resource utilization efficiency.

[0030] Preferably, step S34 includes the following steps:

[0031] Perform time window segmentation on the trajectory occurrence sequence to obtain time window segment data; perform local gradient extraction on the time window segment data to obtain local gradient data;

[0032] Perform threshold screening on local gradient data to generate a critical threshold; locate the boundary gradient based on the critical threshold to obtain critical gradient data;

[0033] Perform boundary enhancement processing on critical gradient data to obtain boundary feature data; perform region marking on boundary feature data to generate trajectory overlap data;

[0034] Extracting spatial coordinates of the trajectory overlap data to generate overlap coordinate points; performing trajectory projection processing on the skeleton assembly movement trajectory based on the overlap coordinate points to obtain projection feature data;

[0035] Performing spatiotemporal synchronization on the projection feature data to obtain synchronized feature data; performing trajectory reconstruction based on the synchronized feature data to generate reconstructed trajectory data;

[0036] The reconstructed trajectory data is integrated in a time-optimized manner to generate the trajectory occurrence time sequence.

[0037] The present invention improves the precision of data processing by segmenting the trajectory occurrence sequence into time windows. The generation of time window fragment data is convenient for local analysis. Local gradient extraction reveals the subtle features of trajectory changes. Threshold screening ensures the validity and accuracy of data. The generation of critical threshold provides a standard for subsequent boundary identification. Boundary gradient positioning effectively identifies the key change points of the trajectory. The formation of critical gradient data lays the foundation for boundary enhancement processing. The generation of boundary feature data improves the accuracy of trajectory analysis. Regional marking ensures the clear identification of overlapping parts. The generation of trajectory overlapping data provides a basis for spatial analysis. Spatial coordinate extraction ensures the accurate positioning of overlapping information. The generation of overlapping coordinate points provides a reference for subsequent trajectory projection. Trajectory projection processing improves the visualization effect of data. Time-space synchronization ensures the consistency of data in different time periods. The generation of synchronous feature data provides support for trajectory reconstruction. The implementation of trajectory reconstruction improves the integrity of the overall data. Timing optimization and integration ensure the rationality of trajectory occurrence timing. Overall, the intelligence and efficiency of hardware mold production scheduling are significantly improved.

[0038] Preferably, step S4 comprises the following steps:

[0039] Step S41: Decompose the coordination trajectory steps to obtain a coordination basic process set;

[0040] Step S42: Calculate the equipment requirements for the coordinated basic process set based on the hardware mold parameters to obtain equipment requirement data; perform load evaluation on the equipment requirement data to obtain the coordination step requirement resources;

[0041] Step S43: divide the coordination step demand resources into time windows to obtain scheduling time data;

[0042] Step S44: Optimize resource allocation for the coordinated trajectory steps based on the scheduling time data to generate a parallel task sequence; perform scheduling rule mapping on the parallel task sequence to generate a coordinated parallel production plan.

[0043] The present invention improves the operability of the production process by decomposing the coordination trajectory steps, refines the production tasks by coordinating the generation of the basic process set, ensures the scientific nature of resource allocation by calculating the equipment demand based on the hardware mold parameters, and provides a basis for load evaluation. The load evaluation accurately identifies the resources required for the coordination steps, and the generation of the coordination step demand resources provides a basis for subsequent scheduling. The time window division improves the flexibility of scheduling, the generation of scheduling time data ensures the timeliness of each task, and the resource allocation optimization improves the production efficiency. The generated parallel task sequence maximizes the use of resources, the scheduling rule mapping ensures the rationality of the production plan, and the formation of the coordinated parallel production plan improves the overall production synergy. Overall, the efficiency and flexibility of the hardware mold production are significantly improved, providing reliable technical support for production management.

[0044] Preferably, step S5 comprises the following steps:

[0045] Step S51: extracting the predecessor relationship of the conflicting trajectory step to obtain the conflicting step predecessor chain; performing directed graph projection on the conflicting step predecessor chain to generate a step predecessor directed graph;

[0046] Step S52: Derivation of dependency relationships of conflicting trajectory steps according to the step-prefix directed graph to obtain dependency relationships between steps;

[0047] Step S53: Modeling the dependency relationship between steps by graph structure to obtain a dependency node graph; hierarchically grading the dependency node graph to obtain a multi-level step tree;

[0048] Step S54: assigning weights to the multi-level step tree according to preset step weight values ​​to obtain a step priority matrix; evaluating the parallelism of the conflicting trajectory steps based on the step priority matrix to obtain a parallel feasible domain;

[0049] Step S55: Rearrange the conflicting trajectory steps in sequence according to the parallel feasible domain to generate a conflict-rearranged production plan.

[0050] The present invention clarifies the logical relationship between each step by extracting the predecessor relationship of the conflict trajectory steps. The generated conflict step predecessor chain provides a basis for subsequent analysis. The directed graph projection improves the visualization of the relationship between the steps. The formation of the step predecessor directed graph ensures the accurate identification of the dependency relationship. The dependency relationship derivation reveals the mutual influence between the steps. The modeling of the dependency relationship between the steps provides structured information for optimization. The generation of the dependency node graph helps to identify the key steps. The hierarchical grading improves the organization and manageability of the steps. The formation of the multi-level step tree provides a clear task hierarchy. The step weight assignment ensures that the importance of the task can be quantified. The generation of the step priority matrix provides a basis for resource allocation. The parallelism evaluation effectively identifies the feasible parallel operation range. The generated parallel feasible domain provides flexibility for production scheduling. The step sequence rearrangement improves the production efficiency. The formation of the conflict rearrangement production plan minimizes the resource conflict. Overall, the scheduling efficiency and flexibility of the hardware mold production are significantly optimized, and reliable technical support is provided for production management.

[0051] Preferably, step S51 includes the following steps:

[0052] Arrange the conflicting trajectory steps in time sequence to obtain step time sequence data; quantify the correlation of the conflicting trajectory steps based on the step time sequence data to generate a step correlation matrix;

[0053] According to the step association matrix, the preceding links of the conflicting trajectory steps are traced back to generate the conflicting preceding steps; the paths of the conflicting preceding steps are merged to obtain the preceding chains of the conflicting steps;

[0054] Perform node mapping processing on the preceding chain of the conflicting step to obtain mapping nodes; perform graph structure modeling on the mapping nodes to generate a conflicting node graph structure;

[0055] Directional analysis is performed on the conflict node graph structure to obtain directional data; and direction annotation is performed on the conflict node graph structure based on the directional data to generate a step-prefix directed graph.

[0056] The time sequence arrangement of the conflict trajectory steps improves the structured degree of the data. The generation of step time sequence data provides a basis for subsequent analysis. The association quantification ensures the accuracy of the relationship between steps. The formation of the step association matrix reveals the correlation of each step. The predecessor link tracing effectively identifies the order of the conflict steps. The generation of the conflict predecessor step provides key path information for optimization. The path merging process improves the connectivity between steps. The obtained conflict step predecessor chain provides support for the overall analysis. The node mapping process ensures the clear expression of information. The graph structure modeling improves the visualization of the conflict nodes. The generated conflict node graph structure provides structured information for subsequent analysis. The directional analysis reveals the flow relationship between steps. The generation of directional data provides a basis for subsequent annotation. The directional annotation ensures the accuracy of the step predecessor directed graph. Overall, it significantly improves the scientificity and efficiency of the hardware mold production scheduling and provides strong technical support for production management.

[0057] Preferably, step S52 includes the following steps:

[0058] Perform deep traversal marking on the step preceding directed graph to obtain a node access sequence; perform path tracing on the node access sequence to obtain a link path set;

[0059] The link path set is determined for reachability and a reachability matrix is ​​generated; the link path set is evaluated for connectivity based on the reachability matrix and node connectivity is generated;

[0060] Based on the node connectivity, the link path set is divided into strong and weak relationships to obtain a path relationship strength table;

[0061] Performing adjacency matrix conversion on the path relationship strength table to obtain a strength adjacency matrix; performing transfer expansion on the strength adjacency matrix to generate extended relationship data;

[0062] Based on the extended relationship data, the link path set is merged to obtain a transfer path set; the transfer path set is closed to obtain a complete dependency chain;

[0063] The complete dependency chain is tested for duplicate paths to obtain a duplicate edge set. Based on the duplicate edge set, the complete dependency chain is subjected to redundancy elimination to obtain the dependency relationship between steps.

[0064] The deep traversal marking of the step preceding directed graph improves the systematicness of node access, the generation of node access sequence provides a basis for path analysis, the formation of link path set reveals the connection relationship between steps, the reachability judgment ensures the validity of the path, the generated reachability matrix provides a basis for subsequent evaluation, the connectivity evaluation effectively identifies the degree of interaction between nodes, the calculation of node connectivity provides a quantitative basis for path relationship division, the generation of path relationship strength table provides a reference for path optimization, the adjacency matrix conversion improves the structuring of data, the generation of strength adjacency matrix provides support for path relationship analysis, the transfer extension ensures the comprehensiveness of path relationship, the formation of extended relationship data provides a basis for subsequent path merging, the path merging process improves the connectivity between steps, the generation of the transfer path collection provides a complete view for the overall dependency analysis, the path closure process ensures the integrity of the dependency chain, the duplicate path detection effectively identifies redundant information, and the redundant elimination process improves the clarity of the dependency relationship. Overall, the dependency relationship between steps in the production of hardware molds is significantly optimized, providing strong technical support for production scheduling.

[0065] Preferably, step S6 comprises the following steps:

[0066] Step S61: performing time axis alignment processing on the coordinated parallel production plan and the conflicting rearranged production plan to obtain synchronous scheduling data;

[0067] Step S62: performing conflict elimination processing on the synchronous scheduling data to generate a fusion production scheduling plan;

[0068] Step S63: performing bottleneck identification on the fusion production scheduling plan to obtain a scheduling bottleneck point; performing resource reallocation on the fusion production scheduling plan based on the scheduling bottleneck point to generate an optimized scheduling plan;

[0069] Step S64: Perform scheduling optimization based on the optimized scheduling plan to execute intelligent optimized scheduling of hardware mold production.

[0070] The timeline alignment of coordinated parallel production plans and conflict-rearranged production plans improves the accuracy of time management. The obtained synchronous scheduling data provides a basis for subsequent processing. Conflict elimination effectively reduces resource conflicts. The generation of integrated production scheduling plans ensures efficient use of resources. Bottleneck identification reveals the key limiting factors in the production process. The identification of scheduling bottlenecks provides a clear goal for optimization. Resource reallocation improves the flexibility and efficiency of resource utilization. The generated optimized scheduling plan provides a scientific basis for production. Scheduling optimization ensures the rationality of the production process. The intelligent optimization scheduling of hardware mold production improves the overall production efficiency, significantly reduces the time waste and idle resources in production, and provides reliable technical support for production management. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] Figure 1 A schematic diagram of the steps of the optimization scheduling method for hardware mold production;

[0072] Figure 2 Detailed implementation flow chart of step S2;

[0073] Figure 3 is a schematic diagram of a detailed implementation step flow of step S3;

[0074] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0075] The technical method of the present invention is described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by technicians in this field without creative work are within the scope of protection of the present invention.

[0076] In addition, the accompanying drawings are only schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor methods and / or microcontroller methods.

[0077] It should be understood that, although the terms "first", "second", etc. may be used herein to describe various units, these units should not be limited by these terms. These terms are used only to distinguish one unit from another unit. For example, without departing from the scope of the exemplary embodiments, the first unit may be referred to as the second unit, and similarly the second unit may be referred to as the first unit. The term "and / or" used herein includes any and all combinations of one or more of the listed associated items.

[0078] To achieve this, please refer to Figures 1 to 3 , an optimization scheduling method for hardware mold production, comprising the following steps:

[0079] Step S1: Collecting hardware mold parameters; performing topological structural analysis on the hardware mold parameters to generate a hardware topological structure; reconstructing a mold skeleton on the mold point cloud data to generate a hardware mold skeleton;

[0080] Step S2: performing skeleton assembly simulation on the skeleton of the hardware mold based on the hardware topological structure to obtain a simulated skeleton assembly process; performing skeleton movement capture on the simulated skeleton assembly process to generate a skeleton assembly movement trajectory;

[0081] Step S3: Acquire mold production history data; arrange the skeleton assembly movement trajectory in time sequence according to the mold production history data to generate a trajectory occurrence sequence; perform conflict identification on the skeleton movement trajectory based on the trajectory occurrence sequence to obtain conflicting trajectory steps and coordinated trajectory steps;

[0082] Step S4: Calculate the running resources of the coordination trajectory step according to the hardware mold parameters to obtain the required resources of the coordination step; perform parallel production scheduling on the coordination trajectory step based on the required resources of the coordination step to generate a coordinated parallel production plan;

[0083] Step S5: Derivation of dependency relationships of conflicting trajectory steps based on the trajectory occurrence sequence to obtain dependencies between steps; and rearrangement of the conflicting trajectory steps based on the dependencies between steps to generate a conflict rearrangement production plan;

[0084] Step S6: Perform data fusion on the coordinated parallel production plan and the conflict rearrangement production plan to obtain a fused production scheduling plan; perform scheduling optimization based on the fused production scheduling plan to execute intelligent optimization scheduling for hardware mold production.

[0085] The present invention provides basic data for subsequent analysis through the collection of hardware mold parameters, the topological structure analysis effectively reveals the structural characteristics of the mold, the generation of hardware topological structure lays the foundation for mold design and optimization, the mold skeleton reconstruction ensures the accurate expression of the mold shape, the implementation of skeleton assembly simulation provides visualization support for the production process, the skeleton movement capture effectively records the dynamic changes in the assembly process, the generated movement trajectory provides important information for subsequent analysis, the acquisition of mold production history data provides a historical background for trajectory analysis, the implementation of time series arrangement ensures the logic of the trajectory, the execution of conflict identification can effectively reveal potential problems in the production process, and the obtained conflict trajectory steps and The coordination trajectory step provides a key basis for subsequent scheduling, the implementation of operating resource calculation ensures the rational allocation of resources, the generation of coordination step demand resources provides data support for parallel production scheduling, the successful implementation of parallel production scheduling improves production efficiency, the execution of dependency derivation provides a theoretical basis for conflict management, the implementation of step sequence rearrangement optimizes the production process, the generated conflict rearrangement production plan improves production flexibility, the implementation of data fusion ensures the comprehensiveness of the scheduling plan, and the execution of scheduling optimization provides an intelligent solution for hardware mold production, which significantly improves the efficiency and scientificity of hardware mold production as a whole, and provides strong technical support for production management.

[0086] In an embodiment of the present invention, the optimization scheduling method for hardware mold production includes the following steps:

[0087] Step S1: Collecting hardware mold parameters; performing topological structural analysis on the hardware mold parameters to generate a hardware topological structure; reconstructing a mold skeleton on the mold point cloud data to generate a hardware mold skeleton;

[0088] In this embodiment, a high-precision laser scanner is used to obtain the geometric parameters of the hardware mold. During the scanning process, the mold is placed on a fixed workbench, and the angle and distance of the laser scanner are adjusted to cover all key structural areas on the mold surface. The collected mold surface point cloud data is processed through data cleaning and denoising, and then the point cloud data is topologically analyzed using reverse engineering software (such as Geomagic Design X) to extract the connection relationship, surface features and boundary contours of the mold, and construct a complete hardware topological structure model. Subsequently, based on a reconstruction algorithm (such as a surface reconstruction algorithm based on multi-resolution analysis), the mold point cloud data is processed into a spatial skeleton form to generate mold skeleton data.

[0089] Step S2: performing skeleton assembly simulation on the skeleton of the hardware mold based on the hardware topological structure to obtain a simulated skeleton assembly process; performing skeleton movement capture on the simulated skeleton assembly process to generate a skeleton assembly movement trajectory;

[0090] In this embodiment, based on the generated hardware topological structure data, a virtual assembly tool (such as Siemens NXAssembly module) is called to perform assembly simulation on the mold skeleton data. During the simulation, constraints are set, including assembly sequence, contact surface pairing and motion direction constraints. The motion trajectory of each skeleton movement is recorded, and the skeleton assembly movement trajectory is formed by capturing the starting and end position information of the skeleton movement, and the trajectory is output as a three-dimensional vector sequence.

[0091] Step S3: Acquire mold production history data; arrange the skeleton assembly movement trajectory in time sequence according to the mold production history data to generate a trajectory occurrence sequence; perform conflict identification on the skeleton movement trajectory based on the trajectory occurrence sequence to obtain conflicting trajectory steps and coordinated trajectory steps;

[0092] In this embodiment, the mold production history data is imported, the time nodes and occurrence sequence of the mold processing links in the data are analyzed, and a timing analysis script is written in Python. The skeleton assembly movement trajectory is arranged in time sequence according to the production node timestamp to generate a trajectory occurrence sequence that conforms to the timeline logic. At the same time, combined with the process conflict records in the historical data, the potential conflict positions in the trajectory are analyzed, and the conflict detection algorithm (such as a spatial conflict recognition algorithm based on convex hull detection) is used to locate the conflict trajectory steps and coordinate the trajectory steps.

[0093] Step S4: Calculate the running resources of the coordination trajectory step according to the hardware mold parameters to obtain the required resources of the coordination step; perform parallel production scheduling on the coordination trajectory step based on the required resources of the coordination step to generate a coordinated parallel production plan;

[0094] In this embodiment, for the coordination trajectory step, a resource scheduling algorithm (such as a genetic algorithm) is used in combination with mold geometric parameters to calculate the operating resources required for the coordination step, and the equipment capacity, time overhead and material requirements are considered to generate the coordination step demand resource data. Based on this resource data, a scheduling engine (such as a scheduling model based on a Gurobi optimizer) is called to perform parallel production scheduling for the coordination trajectory steps. By adjusting the order between steps and the resource allocation for parallel execution, a coordinated parallel production plan is generated.

[0095] Step S5: Derivation of dependency relationships of conflicting trajectory steps based on the trajectory occurrence sequence to obtain dependencies between steps; and rearrangement of the conflicting trajectory steps based on the dependencies between steps to generate a conflict rearrangement production plan;

[0096] In this embodiment, for the conflicting trajectory steps, based on the generated trajectory occurrence sequence, a dependency inference tool (such as Neo4j graph database) is called to analyze the dependency relationship between the conflicting steps, and the input and output resources and time constraints between the steps are converted into a directed acyclic graph structure. The step sequence of the conflicting trajectory steps is rearranged according to the dependency graph, and the optimal step execution order is planned using a heuristic algorithm (such as the A* algorithm), and the rearrangement result is output as a conflict rearrangement production plan.

[0097] Step S6: Perform data fusion on the coordinated parallel production plan and the conflict rearrangement production plan to obtain a fused production scheduling plan; perform scheduling optimization based on the fused production scheduling plan to execute intelligent optimization scheduling for hardware mold production.

[0098] In this embodiment, the coordinated parallel production plan and the conflicting rearranged production plan are used as input, and a data fusion model (such as a Transformer model based on deep learning) is called to fuse the two to generate an overall fused production scheduling plan. By analyzing the various scheduling indicators of the fusion plan, resource utilization and step execution sequence are optimized, and the optimization results are imported and executed using the MES (manufacturing execution system) platform to complete the intelligent scheduling optimization of hardware mold production.

[0099] Preferably, step S1 comprises the following steps:

[0100] Step S11: collecting hardware mold parameters; performing multi-dimensional normalization processing on the hardware mold parameters to obtain normalized hardware data;

[0101] Step S12: performing point cloud mapping processing on the normalized hardware data to obtain mold point cloud data; performing morphological contour analysis on the mold point cloud data to obtain hardware morphological features;

[0102] Step S13: performing topological structural analysis on the hardware morphological features to generate a hardware topological structure; performing assembly constraint modeling on the hardware topological structure to obtain constraint modeling data;

[0103] Step S14: performing local parameter fitting on the hardware topological structure based on the constraint modeling data to generate a hardware mold skeleton.

[0104] In this embodiment, a three-dimensional coordinate measuring machine (CMM) is used to collect the key dimensions of the hardware mold. During the measurement process, a contact probe is used to perform multi-point scanning on each characteristic area of ​​the mold. The dimensional data obtained include parameters such as length, width, height and aperture. After the data is collected, the MATLAB data processing tool is used to normalize the parameters. During the normalization process, the minimum-maximum normalization method is used to scale all measurement values ​​to the [0,1] interval. By analyzing the material properties and geometric characteristics of the hardware mold, the normalization conversion is performed according to the maximum and minimum values ​​to finally form a normalized hardware data matrix. During the normalization process, weights are set for data types of different dimensions. A three-dimensional laser scanning device is used to obtain point cloud information corresponding to the normalized hardware data. Different scanning angles and light source intensities are set during the scanning process to ensure complete coverage of the mold surface. The scanned point cloud data is imported into the MeshLab software for preprocessing. Through denoising, simplification and alignment operations, redundant points and abnormal data are eliminated. Subsequently, PCL (PointCloud Library,The voxel grid filtering algorithm in the point cloud library is used to downsample the point cloud data to reduce data redundancy and improve subsequent calculation efficiency. On this basis, the contour detection algorithm is used to analyze the contour morphology of the mold surface through layer-by-layer slicing, and the contour lines of each section are obtained. The convex hull algorithm is used to extract the overall contour features of the mold to obtain complete hardware morphological feature data. The acquired hardware morphological feature data is imported, and the topological analysis tool (such as OpenCASCADE geometric modeling tool) is called to perform topological structural analysis on the morphological features. In the process of topological construction, according to the spatial distribution relationship of the feature points, Dela The unay triangulation algorithm constructs a topological network, extracts connection relationships and boundary information, and identifies the topological structure characteristics of the mold by calculating the Euler characteristic number of the connection area. Based on the identified topological relationship, the B-Rep (boundary representation) method is used to generate a complete hardware topological structure model. On the basis of the topological structure, combined with the assembly requirements of the mold, the geometric constraints of the assembly interface are defined, and a constraint model based on the assembly relationship is established. In the constraint modeling process, the parameter constraint method is used to define the position relationship, angle relationship and matching tolerance of the assembly surface, and the constraint modeling data file is output. Based on the constraint modeling data, the parameter fitting algorithm based on the least squares method is called to optimize the local parameters of the hardware topological structure. During the optimization process, the coordinates of the characteristic points of the topological structure are fitted with the standard parameters in the constraint model with the minimum error to ensure that the accuracy of the fitted parameters meets the assembly requirements. After fitting, the skeleton line of the topological structure is smoothed using the Bezier curve fitting method to generate the skeleton data of the hardware mold. Finally, the skeleton data is imported into the CAD system (such as CATIA) for three-dimensional visualization verification. By comparing with the original point cloud data, the fitting accuracy is evaluated and the relevant parameters are adjusted to meet the accuracy requirements of the mold assembly.

[0105] Preferably, step S2 comprises the following steps:

[0106] Step S21: mapping the hardware topological structure to assembly constraints to obtain mold constraint data; performing assembly rule conversion on the mold constraint data to generate assembly sequence data;

[0107] Step S22: performing skeleton assembly simulation on the hardware mold skeleton based on the assembly sequence data to obtain a simulated skeleton assembly process;

[0108] Step S23: performing time-series frame processing on the simulated skeleton assembly process to obtain assembly key frame data; performing skeleton posture recognition on the assembly key frame data to generate skeleton spatial position data;

[0109] Step S24: Smoothing the skeleton spatial position data to generate continuous path data; supplementing the missing values ​​of the continuous path data to generate a skeleton assembly movement trajectory.

[0110] In this embodiment, the generated hardware topological structure data is used to call industrial simulation software (such as Siemens Process Simulate) to perform assembly constraint mapping on the topological structure. During the mapping process, based on the geometric features, connection relationships and assembly tolerances of the mold, a three-dimensional constraint solving algorithm is used to associate the relative positions, rotation angles and contact points between the various components, and map to form mold constraint data. After the mapping is completed, a rule-based conversion method is used to convert the constraint data into standardized assembly sequence data according to predefined assembly process specifications. During the conversion process, each assembly step is sorted according to the process route to ensure the rationality and executability of the assembly sequence, and a structured assembly sequence data file is output. Based on the generated assembly sequence data, a simulation engine (such as ABB In the process of virtual assembly of the hardware mold skeleton, the mold skeleton model and assembly sequence data are loaded. The rigid body dynamics simulation algorithm is used to apply gravity, friction and external constraints, and the assembly steps are executed step by step. The displacement, rotation angle and force of each stage are recorded. In the simulation process, real-time monitoring points are set to collect key parameters in the assembly process for subsequent optimization analysis. Finally, the complete simulated skeleton assembly process data is output. For the generated simulated skeleton assembly process data, computer vision analysis tools (such as OpenCV) are called to perform time series frame processing. In the frame segmentation process, a fixed time interval strategy is adopted to discretize the assembly process according to the set sampling frequency, and the continuous motion trajectory is decomposed into a series of assembly key frame data. The key frame data includes timestamp, pose matrix and skeleton pose information. After the frame segmentation is completed, the skeleton pose recognition is performed on the key frame data based on the deep learning algorithm (such as CNN convolutional neural network). In the recognition process, the geometric center, motion direction and local morphological features of the skeleton are extracted, and the standard skeleton pose template in the database is matched to generate accurate skeleton spatial position data. Based on the acquired skeleton spatial position data, the trajectory optimization tool (such as ROS MoveIt) is used to smooth the motion trajectory of the skeleton. During the smoothing process, the B-Spline interpolation algorithm is used to fit the discrete trajectory points to eliminate the mutation and discontinuity in the trajectory. At the same time, the smoothing parameters are adjusted to optimize the smoothness and execution efficiency of the motion trajectory. After smoothing, the integrity of the continuous path data is checked. During the detection process, the interpolation algorithm (such as Lagrange interpolation) is used to supplement the missing positions in the trajectory. After the supplement is completed, a complete skeleton assembly movement trajectory is generated, and the trajectory data is converted into a format for subsequent path execution and scheduling optimization.

[0111] Preferably, step S3 comprises the following steps:

[0112] Step S31: Acquire mold production history data; perform time dimension encoding on the mold production history data to obtain a history sequence code;

[0113] Step S32: Perform spectrum conversion on the historical sequence code to generate time series frequency domain features; perform periodic extraction on the time series frequency domain features to obtain the trajectory time series pattern;

[0114] Step S33: performing time-space mapping processing on the skeleton assembly movement trajectory in the trajectory timing mode to generate the trajectory occurrence timing;

[0115] Step S34: performing gradient detection on the trajectory occurrence sequence to obtain trajectory overlap data; performing spatiotemporal mapping processing on the skeleton assembly movement trajectory based on the trajectory overlap data to generate the trajectory occurrence sequence.

[0116] In this embodiment, a database management system (such as MySQL) is called to retrieve historical production data from a hardware mold production management system. The data content includes multiple dimensions such as production time, process parameters, equipment status, workpiece number, and operation log. To ensure data integrity, an ETL (extraction, transformation, and loading) tool is used to deduplicate and check the consistency of the data. Subsequently, based on a time series processing library (such as Pandas), the data is resampled according to the production timestamp to convert the production records into equally spaced time series data. On this basis, time coding technology is used to embed time features into production data using binary coding, time window segmentation, and sliding window technology, and finally a historical sequence code is generated. Based on the generated historical sequence code, a fast Fourier transform (FFT) is used to perform spectrum conversion on the time series data. During the conversion process, the sequence data is first normalized to eliminate the scale differences between different batches. Subsequently, in the frequency domain conversion process, a window function (such as a Hanning window) is used to smooth the data to reduce the influence of spectrum leakage. After FFT operation, the main frequency components are extracted and the power spectrum density (PSD) is used to calculate the power spectrum density. Spectral Density) is used to analyze the signal energy distribution to form the time-series frequency domain characteristics. On this basis, wavelet transform is used for multi-scale analysis to extract the periodic pattern in the production process. The frequency domain characteristics are clustered according to the periodic law to obtain the trajectory time series pattern. Based on the trajectory time series pattern, spatiotemporal data processing tools (such as PostGIS) are called to perform spatiotemporal mapping processing on the skeleton assembly movement trajectory. In the processing process, the trajectory data is first matched with the time series pattern, and the start and end times of the trajectory are synchronously adjusted according to the time index. Then, dynamic time warping (DTW) is used. The shape matching of the trajectory path is performed using the Warping algorithm. During the matching process, the Euclidean distance between the trajectory path and the timing pattern is calculated, and the time scale of the trajectory is adjusted to ensure that the trajectory is highly aligned with the historical production pattern. Finally, the trajectory occurrence sequence is generated. The edge detection algorithm (such as the Sobel operator) is used to perform gradient detection on the trajectory occurrence sequence. During the detection process, the position points on the trajectory time axis are differentially operated to calculate the trajectory change rate of adjacent time points. In the area where the change rate exceeds the set threshold, it is marked as a trajectory overlap area. Subsequently, the morphological expansion and corrosion processing method is used to optimize the trajectory overlap area, eliminate the false detection points and enhance the boundary features, and finally the trajectory overlap data is obtained. On this basis, based on the trajectory overlap data, the skeleton assembly movement trajectory is subjected to spatiotemporal mapping processing. During the mapping process, the Kalman filter algorithm is used to dynamically update the trajectory data, and the Bayesian filter is combined to predict the trajectory state, and finally an accurate trajectory occurrence sequence is generated.

[0117] Preferably, step S34 includes the following steps:

[0118] Perform time window segmentation on the trajectory occurrence sequence to obtain time window segment data; perform local gradient extraction on the time window segment data to obtain local gradient data;

[0119] Perform threshold screening on local gradient data to generate a critical threshold; locate the boundary gradient based on the critical threshold to obtain critical gradient data;

[0120] Perform boundary enhancement processing on critical gradient data to obtain boundary feature data; perform region marking on boundary feature data to generate trajectory overlap data;

[0121] Extracting spatial coordinates of the trajectory overlap data to generate overlap coordinate points; performing trajectory projection processing on the skeleton assembly movement trajectory based on the overlap coordinate points to obtain projection feature data;

[0122] Performing spatiotemporal synchronization on the projection feature data to obtain synchronized feature data; performing trajectory reconstruction based on the synchronized feature data to generate reconstructed trajectory data;

[0123] The reconstructed trajectory data is integrated in a time-optimized manner to generate the trajectory occurrence time sequence.

[0124] In this embodiment, based on the acquired trajectory occurrence time series data, the sliding window technology is used to divide the time series at equal intervals, the fixed window length is set to 0.5 seconds, and a non-overlapping segmentation strategy is adopted to ensure that each time window contains independent trajectory data. In the segmentation process, a time series processing tool (such as Pandas) is called to resample the data according to the timestamp, and duplicate sampling points are removed. Based on the time window segment data, the first-order difference operation is used to extract the local gradient of the trajectory, and the central difference method is used to calculate the gradient of the trajectory coordinate data. In the gradient calculation process, the adjacent coordinate points in each time window segment are numerically differentiated in turn, and the direction and amplitude of the trajectory change are calculated. The local gradient data is stored as a structured array. Based on the local gradient data, the gradient threshold range is set to screen out the critical gradient area. In the screening process, the histogram distribution analysis method is used to perform frequency statistics on the gradient data, determine the main distribution interval of the gradient change, and set the upper and lower threshold boundaries to exclude the exceeding Gradient data outside the threshold range is marked as abnormal data. Based on the screened critical threshold, the bilateral filtering algorithm is used to locate the boundary gradient of the trajectory data. First, local extreme value detection is performed on the time axis to identify the peak and valley points of the gradient change. Then, the gradient direction information is used to further optimize the positioning accuracy of the boundary points. After completing the boundary gradient positioning, the critical gradient data is output to the trajectory boundary feature extraction module. For the critical gradient data, the image processing library (such as OpenCV) is called for boundary enhancement processing. During the processing, the Gaussian filter is used to smooth the trajectory boundary to reduce noise interference. At the same time, the Canny edge detection algorithm is used to enhance the contrast of the trajectory edge to ensure the clarity of the boundary. Based on the boundary feature data, the connected region analysis method is used to mark the trajectory boundary, calculate the connected components of each region, and assign a unique identifier to each region. In the process of region marking, the depth first search (DFS, DepthFirst The spherical projection coordinate system (such as WGS84) is used to transform the coordinates. Based on the overlapping coordinate points, the spatial transformation matrix is ​​used to project the skeleton assembly movement trajectory. In the projection process, the trajectory coordinate system is first converted to ensure that all trajectory points are unified in the same coordinate reference system. Then, the affine transformation (AffineTransformation) is used to scale and rotate the trajectory to generate projection feature data. The timing analysis tool (such as TSFresh) is called for spatiotemporal synchronization processing. The dynamic time warping algorithm is used to stretch and compress the time axis of the trajectory to synchronize the time series of the trajectory with the standard production rhythm. At the same time, the trajectory data is interpolated to supplement the missing time points.Finally, the synchronous feature data is generated. According to the synchronous feature data, the trajectory reconstruction algorithm (such as Kalman filtering) is called to dynamically reconstruct the trajectory data. In the reconstruction process, the trajectory noise is first filtered, and then the trajectory is smoothly adjusted using the curve fitting method to ensure the spatial continuity of the trajectory and generate the reconstructed trajectory data. Based on the reconstructed trajectory data, the timing optimization algorithm is called to perform smoothing, redundant point removal and trajectory segment merging operations in sequence to finally generate the trajectory occurrence sequence.

[0125] Preferably, step S4 comprises the following steps:

[0126] Step S41: Decompose the coordination trajectory steps to obtain a coordination basic process set;

[0127] Step S42: Calculate the equipment requirements for the coordinated basic process set based on the hardware mold parameters to obtain equipment requirement data; perform load evaluation on the equipment requirement data to obtain the coordination step requirement resources;

[0128] Step S43: divide the coordination step demand resources into time windows to obtain scheduling time data;

[0129] Step S44: Optimize resource allocation for the coordinated trajectory steps based on the scheduling time data to generate a parallel task sequence; perform scheduling rule mapping on the parallel task sequence to generate a coordinated parallel production plan.

[0130] In this embodiment, a production process modeling tool (such as BPMN, business process modeling and annotation) is called to perform process modeling on the hardware mold production process. According to the dependency relationship of each operation link in the production process, a recursive decomposition method is used to disassemble the complex process step by step, and it is refined into basic process units. In the decomposition process, multi-dimensional classification is performed according to time, space and processing attributes to ensure the independence and executability of each basic process. Finally, the decomposed coordinated basic process set is stored in the database in the form of a directed acyclic graph (DAG). Based on the parameters of the hardware mold, a feature parameter extraction tool (such as FeatureXtract) is used to analyze each process in the coordinated basic process set, and the equipment type, processing accuracy and process requirements required for the execution of the process are extracted. According to the equipment capability data in the equipment database, a constraint matching algorithm (Constraint Matching Algorithm) is used. The equipment demand of each process is calculated by using a load assessment tool (such as LoadSim). During the equipment demand calculation process, the processing capacity, operating efficiency and equipment maintenance status are considered. After the calculation is completed, the equipment demand data is generated and stored in the scheduling system. Subsequently, the load assessment tool (such as LoadSim) is used to evaluate the equipment demand data. The load balancing algorithm based on historical work order data is used to calculate the load utilization of each device under a given process arrangement, and the coordination step demand resource data is generated, including key information such as equipment occupancy time, load distribution and spare equipment scheduling strategy. When dividing the coordination step demand resources into time windows, the timing analysis tool (such as TimeEval) is called. According to the statistical characteristics of the process execution time, the fixed window division method is used to divide the scheduling time data. The window length is set to 1 hour, and the time axis is divided according to the process dependency order to ensure that the high-priority process is executed in the early window. In the division process, a dynamic adjustment strategy is used to adjust the boundary of the time window in real time based on the execution time of the process and the resource load. The divided scheduling time data is stored as a time interval sequence and sorted by time weight. Based on the scheduling time data, a heuristic optimization algorithm (Heuristic Optimization) is used. Algorithm) optimizes resource allocation for coordinated trajectory steps. First, a resource-process allocation matrix is ​​constructed. Combined with factors such as time window, equipment load and processing priority, a simulated annealing algorithm is executed to perform a global search to generate a parallel task sequence. The parallel task sequence is optimized and allocated according to the principle of equipment load balancing to ensure maximum production efficiency. After the parallel task sequence is generated, a scheduling rule mapping tool (such as SchedMap) is called to map the scheduling rules for the task sequence. According to the processing priority, processing path and equipment operation status in the production rule library,The Rule Mapping Algorithm is used to convert the task sequence into specific production execution instructions, and finally a coordinated parallel production plan is generated and stored in the production management system.

[0131] Preferably, step S5 comprises the following steps:

[0132] Step S51: extracting the predecessor relationship of the conflicting trajectory step to obtain the conflicting step predecessor chain; performing directed graph projection on the conflicting step predecessor chain to generate a step predecessor directed graph;

[0133] Step S52: Derivation of dependency relationships of conflicting trajectory steps according to the step-prefix directed graph to obtain dependency relationships between steps;

[0134] Step S53: Modeling the dependency relationship between steps by graph structure to obtain a dependency node graph; hierarchically grading the dependency node graph to obtain a multi-level step tree;

[0135] Step S54: assigning weights to the multi-level step tree according to preset step weight values ​​to obtain a step priority matrix; evaluating the parallelism of the conflicting trajectory steps based on the step priority matrix to obtain a parallel feasible domain;

[0136] Step S55: Rearrange the conflicting trajectory steps in sequence according to the parallel feasible domain to generate a conflict-rearranged production plan.

[0137] In this embodiment, a dependency analysis tool is used to analyze the input and output data streams of the conflicting trajectory steps. According to the process flow of hardware mold production, the input materials, processing resources and time constraints of each step are determined. The conflicting trajectory steps are traversed using a depth first search (DFS) algorithm to extract the direct predecessor dependency between each step and generate a predecessor chain of the conflicting steps. In the process of extracting the predecessor chain, a branch-and-merge method is used to merge multiple predecessor paths in view of multi-path dependency. When the predecessor chain of the conflicting steps is projected into a directed graph, a graph mapping tool is used to convert the predecessor chain into a directed acyclic graph (DAG). The topological sorting method is used to sort the predecessor chains. The predecessor relationship is hierarchically organized using the Sorting method to ensure that all predecessor steps are on the corresponding level. During the projection process, the dependency pruning technology is used to prune the infeasible paths for the existing circular dependencies to ensure that the generated step predecessor directed graph maintains order and traceability, and the directed graph is topologically stored. When the dependency relationship of the conflicting trajectory steps is derived based on the step predecessor directed graph, the relationship deduction tool (Relation Deduction Tool) is called. According to the hierarchical structure of the directed graph, the breadth first search (BFS) algorithm is used to traverse layer by layer from the top node to calculate the dependency depth and path weight of each step. The direct dependency relationship is converted into a global dependency relationship through matrix operations. During the derivation process, the historical production data is used to verify the derivation results, and redundant paths are screened out to obtain the final inter-step dependency relationship. When the inter-step dependency relationship is modeled as a graph structure, the graph modeling tool (Graph Modeling Tool) is used to perform a structural transformation on the step dependency data, and the adjacency matrix (Adjacency Matrix) stores dependent nodes and builds a dependent node graph. In the modeling process, the node attribute enhancement strategy is used to embed attributes such as process execution time and resource usage into the node data structure, and the hierarchical clustering algorithm is called to classify the dependent nodes. A multi-level step tree is built according to the node dependency and execution priority. When the multi-level step tree is weighted according to the preset step weight value, the weight assignment tool is called. According to the key path indicators in the production process specification, the hierarchical analysis method (AHP) is used to calculate the weight value of each step. In the weight assignment process, factors such as processing time, equipment load, and material consumption are considered.The step nodes at each level are assigned values ​​step by step, and finally the step priority matrix is ​​generated. Then, the parallelism evaluation tool is used to analyze the parallel feasibility of the conflict trajectory steps based on the step priority matrix. The task overlap detection algorithm is used to calculate the parallelism of the steps and generate parallel feasible domain data. When the step sequence of the conflict trajectory steps is rearranged according to the parallel feasible domain, the sequence optimization tool is called, and the improved genetic algorithm is used to search for the optimal execution order of the conflict trajectory steps in combination with the process dependency and parallelism constraints. During the search process, the step arrangement and combination are dynamically adjusted according to the set production objective function. After multiple iterative optimizations, the optimal conflict rearrangement production plan is generated.

[0138] Preferably, step S51 includes the following steps:

[0139] Arrange the conflicting trajectory steps in time sequence to obtain step time sequence data; quantify the correlation of the conflicting trajectory steps based on the step time sequence data to generate a step correlation matrix;

[0140] According to the step association matrix, the preceding links of the conflicting trajectory steps are traced back to generate the conflicting preceding steps; the paths of the conflicting preceding steps are merged to obtain the preceding chains of the conflicting steps;

[0141] Perform node mapping processing on the preceding chain of the conflicting step to obtain mapping nodes; perform graph structure modeling on the mapping nodes to generate a conflicting node graph structure;

[0142] Directional analysis is performed on the conflict node graph structure to obtain directional data; and direction annotation is performed on the conflict node graph structure based on the directional data to generate a step-prefix directed graph.

[0143] In this embodiment, a temporal sequencing tool is used to parse the execution timestamps of the conflicting trajectory steps. According to the standard operating procedures of the hardware mold production process, the steps are sorted in chronological order. A sorting algorithm based on time constraints is used to compare the weights of the steps in the same time interval to determine the priority execution order. In the temporal arrangement process, for steps with timing conflicts, an interpolation sorting method is used to adjust the steps to a suitable time position to ensure the order integrity of the steps, and finally generate step temporal data. When the conflicting trajectory steps are quantified based on the step temporal data, a correlation analysis tool is called to establish a correlation weight model between the steps by calculating the input and output materials, equipment occupancy, time dependency and other parameters between the steps. In the quantification process, a correlation coefficient calculation method is used to quantify the material flow frequency, resource sharing degree and other factors between the steps, and the values ​​of different dimensions are unified to the same scale through normalization processing, and finally a step correlation matrix is ​​generated. Each element of the matrix represents the correlation weight between the corresponding steps. When the conflicting trajectory steps are traced back to the front link according to the step correlation matrix, a link tracing tool is called. The method based on breadth first search (BFS, Breadth First Search) is used to extract the directly related steps of each step from the association matrix in turn, and sort them according to the association weight to screen out the predecessor dependency. In the tracing process, for the case of multi-path dependency, a dynamic backtracking strategy is adopted to backtrack the path layer by layer, merge the same predecessor steps, avoid redundant links, and finally generate conflicting predecessor steps. When merging the paths of conflicting predecessor steps, a path optimization tool (PathOptimization Tool) is used to identify the same paths in the conflicting predecessor steps using the path merging algorithm. By comparing the resource requirements, execution time and process similarity between the paths, the paths with the same execution logic are merged, and the path compression technology is used to delete the paths with redundant nodes. In the merging process, for the steps with a high path dependency depth, a critical path maintenance strategy is adopted to ensure the integrity of the critical process path, and finally the conflicting step predecessor chain is obtained. When the conflicting step predecessor chain is node mapped, the node mapping tool (Node Mapping Tool) is called. Tool), according to the equipment layout and resource distribution of hardware mold production, the physical equipment mapping is performed on the steps in the front chain, and the hash mapping method is used to correspond each step's unique identifier to the actual equipment information one by one. In the mapping process, combined with the constraints in the production scenario, the consistency check of the mapping nodes is performed, and unreasonable mapping relationships are eliminated. Finally, the mapping node data is obtained. When the graph structure modeling of the mapping nodes is performed,Use the Graph Modeling Tool to build a graph structure for mapping nodes according to the predecessor relationship, use the Adjacency List to store the node connection relationship, and call the hierarchical partitioning algorithm to divide the mapping nodes into different levels to reflect the execution priority of the process. In the modeling process, use the ring detection algorithm to detect and eliminate the existing ring dependencies to ensure that the generated conflict node graph structure maintains the acyclic characteristics. When performing directional analysis on the conflict node graph structure, call the Direction Analysis Tool to identify the connection direction between nodes based on the sequential dependency of the process execution, and use the path tracking algorithm to analyze the data flow between each node. In the analysis process, the directionality strength is determined by counting the in-degree and out-degree of the node, and the directionality is quantified using the weighted calculation method to finally obtain the directional data. When the conflict node graph structure is directional labeled based on the directional data, call the Direction Labeling Tool. Tool), based on the weight information in the directional data, the directional relationship of the nodes is annotated one by one in the conflict node graph structure, different arrow styles are used to represent the forward or feedback path of the process execution, and the directions are hierarchically annotated in combination with the execution order. During the annotation process, the node layout optimization algorithm is used to adjust the visual layout of the graph structure to ensure that the hierarchical structure of the annotated directed graph is clear, and finally a step-preceding directed graph is generated.

[0144] Preferably, step S52 includes the following steps:

[0145] Perform deep traversal marking on the step preceding directed graph to obtain a node access sequence; perform path tracing on the node access sequence to obtain a link path set;

[0146] The link path set is determined for reachability and a reachability matrix is ​​generated; the link path set is evaluated for connectivity based on the reachability matrix and node connectivity is generated;

[0147] Based on the node connectivity, the link path set is divided into strong and weak relationships to obtain a path relationship strength table;

[0148] Performing adjacency matrix conversion on the path relationship strength table to obtain a strength adjacency matrix; performing transfer expansion on the strength adjacency matrix to generate extended relationship data;

[0149] Based on the extended relationship data, the link path set is merged to obtain a transfer path set; the transfer path set is closed to obtain a complete dependency chain;

[0150] The complete dependency chain is tested for duplicate paths to obtain a duplicate edge set. Based on the duplicate edge set, the complete dependency chain is subjected to redundancy elimination to obtain the dependency relationship between steps.

[0151] In this embodiment, when deeply traversing and marking the step predecessor directed graph, a depth-first search (DFS) algorithm is used to recursively visit each node in the graph and mark the access order. Combined with the production process of hardware mold production, the execution time and equipment usage of each step are recorded during the visit of each node to ensure that each node is marked in the order of execution. For each edge in the graph, the predecessor dependency of the node is checked and traced to the child node during marking, and finally a node access sequence is generated. The node access sequence is stored in a list form, in which each element represents the node identifier visited. When the node access sequence is path traced, a path tracking tool is used to traverse the step predecessor directed graph according to the node access sequence. During the path tracing process, the out-degree edge of each node is traced until it can no longer be accessed to generate a complete path chain. Combined with the dependency between steps, the path is optimized to ensure that the path covers all predecessor steps and avoids redundant paths. After the path tracing is completed, all feasible path sets are collected to finally obtain a link path set. When the link path set is reachable, a reachability checking algorithm is applied. Algorithm), according to the structure of the directed graph in the preceding step, by traversing each path in the link path set, determine whether the path can reach all relevant nodes in the graph, based on the depth-first search algorithm, check whether each node in the path can reach the target node in turn, in the judgment process, by analyzing the connectivity in each path, mark each node in the path and record the visited nodes, generate a reachability matrix through reachability judgment, each element of the matrix represents the reachability of the node in the path, when evaluating the connectivity of the link path set according to the reachability matrix, use the connectivity evaluation algorithm (Connectivity Evaluation Algorithm), by analyzing the connectivity value of each node in the reachability matrix, analyze the paths in the link path set, identify the parts with weak connectivity or disconnected connectivity in the path, based on the connectivity evaluation algorithm, calculate the connection strength of each node in the path, distinguish strong connections from weak connections, and represent the degree of connectivity between nodes in the path through the node connectivity value, and finally generate node connectivity data, when dividing the link path set into strong and weak relationships based on the node connectivity, call the strength classification tool (Strength Classification Algorithm). Tool), divides the nodes according to their connectivity values, sets a preset threshold, and divides the paths with node connectivity higher than the threshold into strong connection paths, and divides the paths with node connectivity lower than the threshold into weak connection paths. In the division process, the connectivity of the paths is refined to ensure that the strength relationship of each path meets the actual production requirements, and finally generates a path relationship strength table.Each row of data in the table represents the strength of the path and the relationship between the paths. When the path relationship strength table is converted to an adjacency matrix, the adjacency matrix generator is used to convert each path relationship in the path relationship strength table into the adjacency matrix form in the graph. According to the strength of the path, the corresponding element in the matrix is ​​set to 1 (indicating that the path is connected) or 0 (indicating that the path is not connected). Each element in the adjacency matrix reflects the direct connection between the paths. The matrix can intuitively reflect the connection relationship and strength between the paths. When the strength adjacency matrix is ​​transitively extended, the transitive closure algorithm is used to iteratively extend the adjacency matrix to determine whether the paths in the matrix can be indirectly connected through other paths. For each pair of nodes, check whether there is a transitive path between them. During the extension process, the elements in the matrix are updated to mark the transitivity of the path to ensure that the matrix can reflect all direct and indirect path connection relationships. Finally, the extended relationship data is generated. The extended relationship data is stored in the matrix database for subsequent path merging. When the link path set is merged based on the extended relationship data, the path merging tool is called. The transfer paths in the path set are merged according to the transfer paths marked in the extended relationship data. For paths that can be connected by transfer paths, they are merged into a single path. During the merging process, the path optimization strategy is adopted to avoid redundant paths and ensure that the merged paths can reflect the actual dependencies between all steps. Finally, the collection of transfer paths is obtained. When the collection of transfer paths is closed, the closure processing tool is used to check whether there is a closed path for each path in the path collection. For paths with loops or self-loops, the cycle detection algorithm is used to perform loop detection to ensure the closure of the path and eliminate redundant or invalid closed paths. Finally, a complete dependency chain is obtained. Each path in the complete dependency chain represents a clear dependency relationship between the steps in the metal mold production scheduling. When the complete dependency chain is repeated, the redundant path detection tool is used. The Redundancy Removal Tool is used to traverse each path in the complete dependency chain, check whether there are duplicate nodes or duplicate edges in the path, and use graph algorithms to perform duplicate detection to ensure that each path in the chain is a unique path. Finally, a duplicate edge set is obtained. The duplicate edge set records all redundant path or node information. When the complete dependency chain is redundancy-removed based on the duplicate edge set, the redundant path removal tool (Redundancy Removal Tool) is used to delete the redundant paths or nodes in the chain according to the redundant paths in the duplicate edge set. During the removal process,Make sure the remaining path retains all necessary dependencies, and finally get the inter-step dependencies.

[0152] Preferably, step S6 comprises the following steps:

[0153] Step S61: performing time axis alignment processing on the coordinated parallel production plan and the conflicting rearranged production plan to obtain synchronous scheduling data;

[0154] Step S62: performing conflict elimination processing on the synchronous scheduling data to generate a fusion production scheduling plan;

[0155] Step S63: performing bottleneck identification on the fusion production scheduling plan to obtain a scheduling bottleneck point; performing resource reallocation on the fusion production scheduling plan based on the scheduling bottleneck point to generate an optimized scheduling plan;

[0156] Step S64: Perform scheduling optimization based on the optimized scheduling plan to execute intelligent optimized scheduling of hardware mold production.

[0157] In this embodiment, each production step in the parallel production plan is extracted, and the production steps are sorted in time according to the production resources (such as machines, workers, etc.). For each task in the production plan, its start time, end time, resource usage and other information are obtained, and the scheduling algorithm (such as heuristic scheduling algorithm) is used to align the time axis of the task to ensure that the tasks in the parallel production plan will not conflict with each other. Next, the execution time and resource occupancy of each task are checked, and the time period division technology (such as time window technology) is used to allocate the optimal time period for each task. By adjusting the time axis of each task, the synchronous scheduling data is finally obtained, the conflicts in the parallel plan are identified, and the resource occupancy table is established for the resource conflicts. Record the resource type and time period occupied by each production task, determine the conflicting tasks by comparing the resource occupancy of different tasks, and use conflict elimination algorithms (such as scheduling algorithms based on time windows) to adjust the conflicting tasks. During the adjustment process, give priority to the urgency of the task and its dependence on resources, and adjust the task order based on the priority of the production plan. In this way, resource conflicts in parallel plans are eliminated, and a fusion production scheduling plan is formed after adjustment. Finally, an integrated, conflict-free scheduling plan is obtained. Analyze all tasks in the production scheduling plan, identify the resources or links with the most sluggish production capacity based on indicators such as resource utilization rate and time consumption of each task, and use the critical path method (Critical Path Method) to optimize the scheduling process. PathMethod) analyzes the key links in the production process and determines the bottleneck points. Then, according to the duration and resource consumption of each task, calculates the potential bottleneck resources or equipment, and focuses on monitoring the key nodes in the production process. Finally, by calculating the load and resource utilization efficiency of each link, locates the scheduling bottleneck point. When reallocating resources for the fusion production scheduling plan based on the scheduling bottleneck point, a resource optimization algorithm (such as genetic algorithm, simulated annealing algorithm) is used to analyze the tasks where the bottleneck point is located, identify which resources are not fully utilized at the bottleneck point, or which resources will cause the bottleneck to appear, and reallocate task resources based on the optimization goal, and supplement resources for the tasks where the bottleneck point is located. If the task demand exceeds the resource tolerance, appropriate adjustments are made. The execution order or priority of tasks, in this way, balance resource allocation, reduce the occurrence of bottlenecks, generate optimized scheduling plans, ensure that resources in the entire production process are fully utilized, avoid production delays, and optimize scheduling based on the optimized scheduling plan. When optimizing scheduling, a dynamic scheduling optimization method (such as a dynamic scheduling model) is used. According to the real-time production data of hardware mold production, the tasks in the production process are continuously adjusted dynamically, considering factors such as production progress, resource status, and changes in the production environment. In the scheduling system, the task execution status is updated in real time. If new bottlenecks or conflicts occur, the system automatically readjusts the order of production tasks and resource allocation. Through this real-time optimization method, it is ensured that the production process always remains efficient and stable, and finally an intelligent optimized scheduling plan for the execution of hardware mold production is generated.Ensure that the scheduling plan can cope with complex production environments and improve production efficiency.

[0158] Therefore, the embodiments should be regarded as illustrative and non-restrictive from all points, and the scope of the present invention is limited by the appended claims rather than the above description, and it is therefore intended that all changes falling within the meaning and range of equivalent elements of the application documents are included in the present invention.

[0159] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. An optimization scheduling method for hardware mold production, characterized in that: The following steps are involved: Step S1: collecting hardware mold parameters; performing topological structural analysis on the hardware mold parameters to generate a hardware topological structure; Reconstruct the mold skeleton based on the mold point cloud data to generate the hardware mold skeleton; Step S2: performing skeleton assembly simulation on the skeleton of the hardware mold based on the hardware topological structure to obtain a simulated skeleton assembly process; performing skeleton movement capture on the simulated skeleton assembly process to generate a skeleton assembly movement trajectory; Step S3: Acquire mold production history data; arrange the skeleton assembly movement trajectory in time sequence according to the mold production history data to generate the trajectory occurrence time sequence; Based on the trajectory occurrence sequence, the conflict of the skeleton movement trajectory is identified to obtain the conflict trajectory steps and the coordination trajectory steps; Step S4: Calculate the running resources of the coordination trajectory step according to the hardware mold parameters to obtain the required resources of the coordination step; Perform parallel production scheduling for the coordinated trajectory steps based on the required resources of the coordinated steps to generate a coordinated parallel production plan; Step S5: Derivation of dependency relationships of conflicting trajectory steps based on the trajectory occurrence sequence to obtain dependency relationships between steps; Rearranging the conflicting trajectory steps based on the dependencies between the steps to generate a conflict-rearranged production plan; Step S6: Perform data fusion on the coordinated parallel production plan and the conflict rearrangement production plan to obtain a fused production scheduling plan; Scheduling optimization is performed based on the fusion production scheduling plan to implement intelligent optimization scheduling of hardware mold production.

2. The optimization scheduling method for hardware mold production according to claim 1 is characterized in that: Step S1 includes the following steps: Step S11: collecting hardware mold parameters; performing multi-dimensional normalization processing on the hardware mold parameters to obtain normalized hardware data; Step S12: performing point cloud mapping processing on the normalized hardware data to obtain mold point cloud data; performing morphological contour analysis on the mold point cloud data to obtain hardware morphological features; Step S13: performing topological structural analysis on the hardware morphological features to generate a hardware topological structure; performing assembly constraint modeling on the hardware topological structure to obtain constraint modeling data; Step S14: performing local parameter fitting on the hardware topological structure based on the constraint modeling data to generate a hardware mold skeleton.

3. The optimization scheduling method for hardware mold production according to claim 1 is characterized in that: Step S2 includes the following steps: Step S21: mapping the hardware topological structure to assembly constraints to obtain mold constraint data; performing assembly rule conversion on the mold constraint data to generate assembly sequence data; Step S22: performing skeleton assembly simulation on the hardware mold skeleton based on the assembly sequence data to obtain a simulated skeleton assembly process; Step S23: performing time-series frame processing on the simulated skeleton assembly process to obtain assembly key frame data; performing skeleton posture recognition on the assembly key frame data to generate skeleton spatial position data; Step S24: Smoothing the skeleton spatial position data to generate continuous path data; supplementing the missing values ​​of the continuous path data to generate a skeleton assembly movement trajectory.

4. The optimization scheduling method for hardware mold production according to claim 1 is characterized in that: Step S3 includes the following steps: Step S31: Acquire mold production history data; perform time dimension encoding on the mold production history data to obtain a history sequence code; Step S32: Perform spectrum conversion on the historical sequence code to generate time series frequency domain features; perform periodic extraction on the time series frequency domain features to obtain the trajectory time series pattern; Step S33: performing time-space mapping processing on the skeleton assembly movement trajectory in the trajectory timing mode to generate the trajectory occurrence timing; Step S34: performing gradient detection on the trajectory occurrence sequence to obtain trajectory overlap data; performing spatiotemporal mapping processing on the skeleton assembly movement trajectory based on the trajectory overlap data to generate the trajectory occurrence sequence.

5. The optimization scheduling method for hardware mold production according to claim 4 is characterized in that: Step S34 includes the following steps: Perform time window segmentation on the trajectory occurrence sequence to obtain time window segment data; perform local gradient extraction on the time window segment data to obtain local gradient data; Perform threshold screening on local gradient data to generate a critical threshold; locate the boundary gradient based on the critical threshold to obtain critical gradient data; Perform boundary enhancement processing on critical gradient data to obtain boundary feature data; perform region marking on boundary feature data to generate trajectory overlap data; Extracting spatial coordinates of the trajectory overlap data to generate overlap coordinate points; performing trajectory projection processing on the skeleton assembly movement trajectory based on the overlap coordinate points to obtain projection feature data; Performing spatiotemporal synchronization on the projection feature data to obtain synchronized feature data; performing trajectory reconstruction based on the synchronized feature data to generate reconstructed trajectory data; The reconstructed trajectory data is integrated in a time-optimized manner to generate the trajectory occurrence time sequence.

6. The optimization scheduling method for hardware mold production according to claim 1 is characterized in that: Step S4 includes the following steps: Step S41: Decompose the coordination trajectory steps to obtain a coordination basic process set; Step S42: Calculate the equipment requirements for the coordinated basic process set based on the hardware mold parameters to obtain equipment requirement data; perform load evaluation on the equipment requirement data to obtain the coordination step requirement resources; Step S43: divide the coordination step demand resources into time windows to obtain scheduling time data; Step S44: Optimize resource allocation for the coordinated trajectory steps based on the scheduling time data to generate a parallel task sequence; perform scheduling rule mapping on the parallel task sequence to generate a coordinated parallel production plan.

7. The optimization scheduling method for hardware mold production according to claim 1 is characterized in that: Step S5 includes the following steps: Step S51: extracting the predecessor relationship of the conflicting trajectory step to obtain the conflicting step predecessor chain; performing directed graph projection on the conflicting step predecessor chain to generate a step predecessor directed graph; Step S52: Derivation of dependency relationships of conflicting trajectory steps according to the step-prefix directed graph to obtain dependency relationships between steps; Step S53: Modeling the dependency relationship between steps by graph structure to obtain a dependency node graph; hierarchically grading the dependency node graph to obtain a multi-level step tree; Step S54: assigning weights to the multi-level step tree according to preset step weight values ​​to obtain a step priority matrix; evaluating the parallelism of the conflicting trajectory steps based on the step priority matrix to obtain a parallel feasible domain; Step S55: Rearrange the conflicting trajectory steps in sequence according to the parallel feasible domain to generate a conflict-rearranged production plan.

8. The optimization scheduling method for hardware mold production according to claim 7 is characterized in that: Step S51 includes the following steps: Arrange the conflicting trajectory steps in time sequence to obtain step time sequence data; quantify the correlation of the conflicting trajectory steps based on the step time sequence data to generate a step correlation matrix; According to the step association matrix, the preceding links of the conflicting trajectory steps are traced back to generate the conflicting preceding steps; the paths of the conflicting preceding steps are merged to obtain the preceding chains of the conflicting steps; Perform node mapping processing on the preceding chain of the conflicting step to obtain mapping nodes; perform graph structure modeling on the mapping nodes to generate a conflicting node graph structure; Directional analysis is performed on the conflict node graph structure to obtain directional data; and direction annotation is performed on the conflict node graph structure based on the directional data to generate a step-prefix directed graph.

9. The optimization scheduling method for hardware mold production according to claim 7, characterized in that: Step S52 includes the following steps: Perform deep traversal marking on the step preceding directed graph to obtain a node access sequence; perform path tracing on the node access sequence to obtain a link path set; The link path set is determined for reachability and a reachability matrix is ​​generated; the link path set is evaluated for connectivity based on the reachability matrix and node connectivity is generated; Based on the node connectivity, the link path set is divided into strong and weak relationships to obtain a path relationship strength table; Performing adjacency matrix conversion on the path relationship strength table to obtain a strength adjacency matrix; performing transfer expansion on the strength adjacency matrix to generate extended relationship data; Based on the extended relationship data, the link path set is merged to obtain a transfer path set; the transfer path set is closed to obtain a complete dependency chain; The complete dependency chain is tested for duplicate paths to obtain a duplicate edge set. Based on the duplicate edge set, the complete dependency chain is subjected to redundancy elimination to obtain the dependency relationship between steps.

10. The optimization scheduling method for hardware mold production according to claim 1, characterized in that: Step S6 includes the following steps: Step S61: performing time axis alignment processing on the coordinated parallel production plan and the conflicting rearranged production plan to obtain synchronous scheduling data; Step S62: performing conflict elimination processing on the synchronous scheduling data to generate a fusion production scheduling plan; Step S63: performing bottleneck identification on the fusion production scheduling plan to obtain a scheduling bottleneck point; performing resource reallocation on the fusion production scheduling plan based on the scheduling bottleneck point to generate an optimized scheduling plan; Step S64: Perform scheduling optimization based on the optimized scheduling plan to execute intelligent optimized scheduling of hardware mold production.

Images