Green-driven hardware mold manufacturing production line management method and platform
By conducting nearest neighbor search and processing quality analysis in the memory storage of the hardware mold manufacturing production line, the adaptability is identified and directional excellence is carried out, and the problem of repeated trials and mold repairs is solved, and the optimization of processing parameters and efficiency improvement is achieved.
Patent Information
- Application Number
- CN202411144269.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-20
- Publication Date
- 2024-11-22
- Estimated Expiration
- 2044-08-20
AI Technical Summary
The problem of repeated trial and mold repair in the existing hardware mold manufacturing production line management has led to an extension of the processing cycle, increasing production costs and delaying the product launch.
By searching nearest neighbors in the memory repository of the mold manufacturing production line, obtaining historical processing parameters sets, performing processing quality search and deviation analysis, identifying fitness, and generating target processing parameters sets through overall directional optimization to reduce mold trial repair Module times, optimize processing parameters.
It effectively reduces the number of trial mold repairs, shortens the processing cycle, improves processing efficiency, and realizes the optimization of processing parameters.
Smart Images

Figure CN118657444B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of data processing technology, and in particular to a green-driven hardware mold manufacturing production line management method and platform. Background Art
[0002] In the existing management of hardware mold manufacturing production lines, a significant problem is the cycle of repeated mold trials and mold repairs. For example, when a certain auto parts factory was manufacturing a complex structure engine cylinder mold, due to the lack of an accurate set of processing parameters in the early stage, it had to conduct multiple mold trials. After each mold trial, the mold had to be repaired according to the deviation. This process not only consumed a lot of manpower and material resources, but also extended the processing cycle to several weeks or even months. This repeated trial and error not only increased production costs, but also delayed product launch time and affected market competitiveness. Therefore, optimizing the management method of the hardware mold manufacturing production line and reducing the number of mold trials and mold repairs have become issues to be solved. Summary of the invention
[0003] The present application provides a green-driven hardware mold manufacturing production line management method and platform, which is used to solve the technical problem in the prior art that the hardware mold manufacturing production line management has repeated mold trials and mold repairs, resulting in a long processing cycle.
[0004] In view of the above problems, this application provides a green-driven hardware mold manufacturing production line management method and platform.
[0005] The first aspect of the present application provides a green-driven hardware mold manufacturing production line management method, the method comprising:
[0006] According to the mold structure information, mold material information and mold process parameter set of the target mold, a neighbor search is performed in the memory storage of the mold manufacturing production line to obtain multiple historical production line processing parameter sets;
[0007] Performing processing quality searches on the multiple historical production line processing parameter sets respectively to obtain multiple processing quality deviation sets;
[0008] Traversing the plurality of processing quality deviation sets to identify processing parameter fitness and generate a plurality of historical processing fitnesses;
[0009] With the goal of reducing processing deviation and the number of mold trials and mold repairs, the multiple historical production line processing parameter sets are optimized in an overall directional manner according to the multiple historical processing fitnesses to obtain a target processing parameter set;
[0010] Performing deviation mean analysis on the multiple processing quality deviation sets respectively to determine K deviation coefficients of K stamping processing nodes;
[0011] The K deviation coefficients and the target processing parameter set are transmitted to a production line management unit to perform production line processing management of the target mold.
[0012] The second aspect of the present application provides a green-driven hardware mold manufacturing production line management platform, the platform comprising:
[0013] A module for obtaining a set of historical production line processing parameters, which is used to perform a neighbor search in a memory repository of a mold manufacturing production line according to mold structure information, mold material information and mold process parameter set of a target mold, and obtain multiple sets of historical production line processing parameters;
[0014] A processing quality deviation set obtaining module, used to perform processing quality searches on the multiple historical production line processing parameter sets respectively to obtain multiple processing quality deviation sets;
[0015] A historical processing fitness generation module, used for traversing the plurality of processing quality deviation sets to identify processing parameter fitness and generate a plurality of historical processing fitnesses;
[0016] A target processing parameter set acquisition module is used to perform overall directional optimization on the multiple historical production line processing parameter sets according to the multiple historical processing fitnesses with the goal of reducing processing deviation and reducing the number of mold trials and mold repairs to obtain a target processing parameter set;
[0017] A deviation coefficient determination module, used to perform deviation mean analysis on the multiple processing quality deviation sets respectively, and determine K deviation coefficients of K stamping processing nodes;
[0018] The production line processing management module is used to transmit the K deviation coefficients and the target processing parameter set to the production line management unit to perform production line processing management of the target mold.
[0019] One or more technical solutions provided in this application have at least the following technical effects or advantages:
[0020] This application obtains multiple historical production line processing parameter sets by performing neighbor retrieval in the memory storage of the mold manufacturing production line according to the mold structure information, mold material information and mold process parameter set of the target mold, and then performs processing quality search on multiple historical production line processing parameter sets respectively to obtain multiple processing quality deviation sets, and then traverses multiple processing quality deviation sets to identify processing parameter fitness, generate multiple historical processing fitness, and then, with the goal of reducing processing deviation and reducing the number of mold trials and mold repairs, perform overall directional optimization on multiple historical production line processing parameter sets according to multiple historical processing fitness, obtain target processing parameter sets, and determine K deviation coefficients of K stamping processing nodes by performing deviation mean analysis on multiple processing quality deviation sets respectively, and then transmit the K deviation coefficients and the target processing parameter set to the production line management unit to perform production line processing management of the target mold. The technical effect of optimizing processing parameters and improving processing efficiency is achieved. BRIEF DESCRIPTION OF THE DRAWINGS
[0021] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.
[0022] Figure 1 A schematic diagram of a process flow of a green-driven hardware mold manufacturing production line management method provided in an embodiment of the present application;
[0023] Figure 2 A schematic diagram of the structure of a green-driven hardware mold manufacturing production line management platform provided in an embodiment of the present application.
[0024] Explanation of the accompanying drawings: historical production line processing parameter set acquisition module 11, processing quality deviation set acquisition module 12, historical processing fitness generation module 13, target processing parameter set acquisition module 14, deviation coefficient determination module 15, production line processing management module 16. DETAILED DESCRIPTION
[0025] The present application provides a green-driven hardware mold manufacturing production line management method and platform to solve the technical problem in the prior art that the hardware mold manufacturing production line management has repeated mold trials and mold repairs, resulting in a long processing cycle.
[0026] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.
[0027] It should be noted that the terms "including" and "having" are intended to cover non-exclusive inclusions. For example, a process, method, platform, product or server that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or modules that are not explicitly listed or inherent to these processes, methods, products or devices.
[0028] Embodiment 1, as Figure 1 As shown, the present application provides a green-driven hardware mold manufacturing production line management method, wherein the method includes:
[0029] Step S100: performing a neighbor search in a memory storage of a mold manufacturing production line according to mold structure information, mold material information and mold process parameter set of a target mold, and obtaining a plurality of historical production line processing parameter sets;
[0030] In one possible embodiment, the target mold is any hardware mold that needs to be stamped. The mold manufacturing production line is a production line for stamping the target mold, including multiple stamping processing nodes such as stretching, punching, trimming, shaping, blanking, and stacking. The mold structure information includes the mold's geometry, size, structural complexity, etc., which are crucial for determining the processing method, processing path, and processing parameters. The mold material information refers to the material type, hardness, heat treatment state, etc. used in mold manufacturing. Different materials require different processing strategies and parameters. The mold process parameter set includes expected processing accuracy, surface roughness, processing time, and other requirements. These parameters are directly related to the final processing quality and production efficiency.
[0031] The memory repository is a summary of the processing records of the mold manufacturing production line in the historical time. By using advanced algorithms (such as K-nearest neighbor algorithm, support vector machine, etc.), the historical records that are most similar to the mold structure information, mold material information and mold process parameter set of the target mold are searched in the memory repository. This process involves a comprehensive consideration of the mold structure information, material information and process parameter set to find multiple related historical production line processing parameter sets. Among them, each historical production line processing parameter set in the multiple historical production line processing parameter sets includes the processing parameters of multiple stamping processing nodes. As a result, the technical effect of providing reliable data support for the subsequent reduction of mold repair time and the improvement of the degree of green manufacturing is achieved.
[0032] Furthermore, step S100 in the embodiment of the present application further includes:
[0033] Calling a memory storage library of the mold manufacturing production line, wherein the memory storage library includes a plurality of historical processing records, and each historical processing record includes a historical production line processing parameter set;
[0034] Randomly extracting M historical processing records from the plurality of historical processing records as M processing cluster centers, wherein the similarity between any two processing cluster centers of the M processing cluster centers is lower than a preset similarity threshold;
[0035] Performing cluster searches on the M processing cluster centers in the memory storage respectively, and performing iterative updates of the centers after each cluster search is completed, to obtain M updated processing cluster centers;
[0036] Analyzing the clustering qualities of the M update processing cluster centers respectively to obtain a memory score of the memory storage repository;
[0037] It is determined whether the memory score is greater than or equal to a preset score threshold. If not, a data supplement instruction is obtained to supplement the memory storage library to obtain a memory storage library with updated data.
[0038] Furthermore, step S100 in the embodiment of the present application further includes:
[0039] Calculating the similarity between the plurality of historical processing records and the M processing cluster centers, respectively adding the plurality of historical processing record data to the cluster clusters of the processing cluster center corresponding to the maximum similarity value, to obtain M processing cluster clusters;
[0040] Iteratively updating the processing cluster centers of the M processing clusters according to the weighted center iteration formula until a preset number of iterations is met to obtain the M updated processing cluster centers;
[0041] Calculating offset vectors from the M processing clusters to the M updated processing cluster centers to obtain M offset vector sets;
[0042] A clustering quality analysis is performed based on the M offset vector sets to determine the memory score.
[0043] Furthermore, the weighted center iteration formula is:
[0044] ;
[0045] in, To update the processing cluster center, To process multiple historical processing records in the cluster, is the i-th historical processing record in the processing cluster, It is the processing cluster center.
[0046] In the embodiment of the present application, the memory storage library is related to the optimization and adjustment of the processing parameters of the subsequent production line, so the quality of the memory storage library has an important impact on the quality of production line management. When the same type of data in the memory storage library is relatively scattered, it indicates that the stability of the corresponding mold manufacturing production line is poor, and it is necessary to call cloud data to supplement the historical production line processing parameter set, so as to obtain a target processing parameter set that meets the target mold processing requirements.
[0047] In a possible embodiment, a memory storage library of the mold manufacturing production line is called, wherein the memory storage library includes a plurality of historical processing records, each of which includes a set of historical production line processing parameters. M historical processing records are randomly selected from the plurality of historical processing records as M processing cluster centers, wherein the similarity of any two processing cluster centers of the M processing cluster centers is lower than a preset similarity threshold.
[0048] Optionally, in cluster analysis, the cluster center is used to represent the "center" or "representative" of a certain type of historical processing records. M historical processing records are randomly selected from the multiple historical processing records as M processing cluster centers, and the multiple historical processing records are classified with the M processing cluster centers as clustering targets.
[0049] The cosine similarity is used to calculate the similarity of any two processing cluster center points. If the similarity of any two processing cluster centers is higher than the preset similarity threshold (the maximum similarity of two processing cluster centers set by technicians in this field), two historical processing records are randomly selected from the multiple historical processing records to replace them, thereby achieving the goal of avoiding the processing cluster centers from being too similar and reducing the reliability of clustering.
[0050] In a possible embodiment, the M processing cluster centers are respectively used as clustering targets, and cluster searches are performed on multiple historical processing records in the memory storage repository. After each cluster search is completed, the processing cluster center in each processing cluster cluster is iteratively updated to determine a more representative center in each processing cluster cluster, and obtain the M updated processing cluster centers.
[0051] The clustering quality of the M updated processing cluster centers is analyzed respectively to determine the degree of clustering of the processing cluster clusters corresponding to each updated processing cluster center. If the degree of clustering is lower, it indicates that the processing cluster clusters are more dispersed, and the clustering quality is lower, reflecting that the processing volatility of the mold manufacturing production line is larger. After analysis, the memory score of the memory storage is obtained, and the memory score reflects the degree of clustering of the memory storage. The larger the memory score, the higher the degree of clustering and the better the quality of the memory storage.
[0052] Determine whether the memory score is greater than or equal to a preset score threshold. If not, it indicates that the aggregation level of the memory storage is low. At this time, it is necessary to obtain a data supplement instruction, call cloud data to supplement the memory storage, and obtain the memory storage after data update. The data supplement instruction is a command to supplement the memory storage. The preset score threshold is the minimum memory score that the memory storage needs to meet set by a technician in this field.
[0053] In a possible embodiment, the cosine similarity calculation formula is used to calculate the similarities between the multiple historical processing records and the M processing cluster centers, and the multiple historical processing record data are respectively added to the cluster clusters of the processing cluster center corresponding to the maximum similarity value to obtain M processing cluster clusters.
[0054] The weighted center iteration formula is used to iteratively calculate the M processing clusters and the M processing cluster centers, and the processing cluster centers of the M processing clusters are iteratively updated according to the calculation results until a preset number of iterations (the number of iterations set by technicians in this field) is met, and the M updated processing cluster centers are obtained.
[0055] Optionally, vector difference calculation is performed based on the M historical processing record sets corresponding to the M processing clusters and the M historical processing records corresponding to the M updated processing cluster centers to obtain the M offset vector sets, wherein the M offset vectors reflect the degree of discreteness of the M processing clusters.
[0056] A plurality of sample offset vector sets and a plurality of sample memory scores are obtained as training data, and the framework constructed based on the convolutional neural network is supervised and trained using the training data until the model reaches convergence, thereby obtaining a trained memory score network layer. The trained memory score network layer is used to perform clustering quality analysis on the M offset vector sets to determine the memory score.
[0057] The weighted center iteration formula: , used to iteratively update the processing cluster center of each processing cluster. Among them, To update the processing cluster center, To process multiple historical processing records in the cluster, is the i-th historical processing record in the processing cluster, It is the processing cluster center.
[0058] Step S200: performing processing quality searches on the multiple historical production line processing parameter sets respectively to obtain multiple processing quality deviation sets;
[0059] In a possible embodiment, the multiple historical production line processing parameter sets are used as indexes to search in the memory storage to obtain multiple historical processing quality sets. The multiple historical processing quality sets are the processing quality of each node when the mold is manufactured after the K stamping processing nodes of the mold manufacturing line are controlled by equipment parameters according to the multiple historical production line processing parameter sets.
[0060] Multiple historical standard processing quality sets corresponding to multiple historical production line processing parameter sets are called, and the difference calculation is performed with the multiple historical processing quality sets to obtain the multiple processing quality deviation sets. Among them, the multiple processing quality deviation sets reflect the historical processing deviations of the K stamping processing nodes of the mold manufacturing production line. The technical effect of providing a basis for the subsequent refined management of the production line is achieved.
[0061] Step S300: traversing the plurality of processing quality deviation sets to identify processing parameter fitness and generate a plurality of historical processing fitnesses;
[0062] Furthermore, step S300 of the embodiment of the present application further includes:
[0063] Construct a fitness recognition unit;
[0064] The fitness identification unit is used to perform fitness identification of processing parameters on the multiple processing quality deviation sets to obtain the multiple historical processing fitnesses.
[0065] The processing fitness of multiple historical production line processing parameter sets during manufacturing processing is determined according to the multiple processing quality deviation sets to generate the multiple historical processing fitnesses, wherein the multiple historical processing fitnesses reflect the processing quality of the multiple historical production line processing parameter sets.
[0066] Optionally, the fitness recognition unit is a unit for intelligently analyzing the degree of processing adaptability of a historical processing parameter set according to a processing quality deviation set. Multiple sample processing quality deviation sets, multiple sample historical processing parameter sets, and multiple sample processing fitness are obtained as recognition construction data. The recognition construction data is preprocessed, and then the recognition construction data after data preprocessing is transmitted to the input layer, and the convolution layer is used to extract features of the data transmitted by the input layer, and the pooling layer is used for down sampling, and then the features are combined through the fully connected layer to output the processing fitness. The recognition construction data is used to train a convolutional neural network to learn the mapping relationship between the sample processing quality deviation set, the sample historical processing parameter set, and the sample processing fitness, until the output converges, and the fitness recognition unit that has been trained is obtained.
[0067] Optionally, the trained fitness recognition unit is used to perform fitness recognition of processing parameters on the multiple processing quality deviation sets to obtain multiple historical processing fitnesses, thereby achieving the technical effect of improving the efficiency of fitness recognition of processing parameters and intelligently improving the speed of data analysis.
[0068] Step S400: with the goal of reducing processing deviation and the number of mold trials and mold repairs, the plurality of historical production line processing parameter sets are optimized in an overall directional manner according to the plurality of historical processing fitnesses to obtain a target processing parameter set;
[0069] Furthermore, step S400 in the embodiment of the present application further includes:
[0070] Comparing the plurality of historical processing fitnesses respectively with the sum of the plurality of historical processing fitnesses, and taking the difference between 1 and the comparison result as a plurality of optimization step lengths;
[0071] With the goal of reducing processing deviation and the number of mold trials and mold repairs, the multiple historical production line processing parameter sets are directional fine-tuned based on the multiple optimization step lengths to obtain multiple adjusted production line processing parameter sets, wherein the multiple adjusted production line processing parameter sets include multiple adjusted processing fitness levels;
[0072] The adjusted production line processing parameter set corresponding to the maximum value among the multiple adjusted processing fitness values is used as an optimization particle, and the remaining multiple historical production line processing parameter sets are used as multiple following particles;
[0073] Taking the optimal particle as a target, fine-tuning the plurality of follower particles according to the plurality of optimal step lengths to obtain a plurality of fine-tuned follower particles;
[0074] The optimizing particle is updated and iterated based on the multiple fine-tuning follower particles until a preset number of iterations is met, and the optimizing particle corresponding to the maximum processing fitness value is used as a target processing parameter set.
[0075] Furthermore, step S400 in the embodiment of the present application further includes:
[0076] Determine whether there is a fine-tuning follower particle whose adjustment processing fitness is greater than or equal to the optimization fitness of the optimization particle among the multiple fine-tuning follower particles, and if so, update the fine-tuning follower particle corresponding to the maximum adjustment fitness among the multiple fine-tuning follower particles as the optimization particle;
[0077] If not, a random number generation function is called to obtain a first random number. When the first random number is less than or equal to a preset random value, the fine-tuning follower particle corresponding to the maximum value of the adjusted fitness among the multiple fine-tuning follower particles is accepted to update the optimization particle.
[0078] In one possible embodiment, with the goal of reducing processing deviation and the number of mold trials and mold repairs, the multiple historical production line processing parameter sets are optimized in an overall directional manner according to the multiple historical processing fitness, and the target processing parameter set that is more suitable for the target mold is obtained by optimizing and adjusting the multiple historical production line processing parameter sets. The technical effect of optimizing the processing parameters, reducing the number of mold repairs, and shortening the processing cycle is achieved, and the goal of green manufacturing is achieved.
[0079] In a possible embodiment, the plurality of historical processing fitness values are respectively compared with the sum of the plurality of historical processing fitness values, and the difference between 1 and the comparison result is used as a plurality of optimization step lengths. The optimization step length is the amplitude of a single parameter adjustment when performing overall directional optimization.
[0080] Furthermore, with the goal of reducing processing deviations and the number of mold trials and mold repairs, the multiple historical production line processing parameter sets are directional fine-tuned based on the multiple optimization step lengths, that is, the multiple historical production line processing parameter sets are adjusted in the direction of reducing processing deviations and the number of mold trials and mold repairs with the adjustment range of multiple optimization step lengths, thereby obtaining multiple adjusted production line processing parameter sets, wherein the multiple adjusted production line processing parameter sets include multiple adjusted processing fitnesses. The fitness identification unit performs fitness identification on the multiple adjusted production line processing parameter sets to determine the multiple adjusted processing fitnesses.
[0081] Optionally, the set of adjusted production line processing parameters corresponding to the maximum value of the multiple adjusted processing fitness is used as the optimization particle, that is, the stage target of the optimization, and then the remaining multiple historical production line processing parameter sets are used as multiple follower particles. Then, with the optimization particle as the target, the multiple follower particles are fine-tuned according to the multiple optimization step lengths to obtain multiple fine-tuned follower particles.
[0082] Determine whether there is a fine-tuning follower particle whose adjustment processing fitness is greater than or equal to the optimization fitness of the optimization particle among the multiple fine-tuning follower particles, and if so, update the fine-tuning follower particle corresponding to the maximum adjustment fitness among the multiple fine-tuning follower particles as the optimization particle;
[0083] If not, a random number generation function (such as the rand function) is called to obtain a first random number. When the first random number is less than or equal to a preset random value (a random value set by a person skilled in the art, and the random value belongs to [0,1]), the fine-tuning follower particle corresponding to the maximum value of the adjusted fitness among the multiple fine-tuning follower particles is accepted to update the optimization particle. The technical effect of avoiding falling into a local optimal solution is achieved.
[0084] Then, the optimization particle is updated by using the fine-tuning follower particle corresponding to the maximum value of the fitness adjustment among the multiple fine-tuning follower particles until the preset number of iterations is met, and the optimization particle corresponding to the maximum value of the processing fitness is used as the target processing parameter set. The technical effect of directional optimization of processing parameters and improving the accuracy and reliability of production line management is achieved.
[0085] Step S500: performing deviation mean analysis on the multiple processing quality deviation sets respectively to determine K deviation coefficients of K stamping processing nodes;
[0086] Step S600: transmitting the K deviation coefficients and the target processing parameter set to a production line management unit to perform production line processing management of the target mold.
[0087] In a possible embodiment, the K stamping processing nodes are used as indexes to retrieve the multiple processing quality deviation sets to obtain K node processing quality deviation sets. The K node processing quality deviation sets are respectively averaged to obtain K node processing deviations.
[0088] Optionally, the ratio of the processing deviation of the K nodes to the sum of the processing deviation of the K nodes is calculated respectively to obtain K deviation coefficients. The K deviation coefficients are transmitted to the production line management unit. The larger the deviation coefficient, the more management resources are allocated to the corresponding stamping processing node. Furthermore, the target processing parameter set is transmitted to the production line management unit to configure the processing equipment parameters of the stamping processing node, and the production line processing management is performed according to the K deviation coefficients. The technical effect of fine and reliable management of the production line and shortening the processing cycle is achieved.
[0089] In summary, the embodiments of the present application have at least the following technical effects:
[0090] This application obtains multiple historical production line processing parameter sets by performing neighbor retrieval in the memory storage of the mold manufacturing production line according to the mold structure information, mold material information and mold process parameter set of the target mold, and then performs processing quality search on multiple historical production line processing parameter sets respectively to obtain multiple processing quality deviation sets, and then traverses multiple processing quality deviation sets to identify processing parameter fitness, generate multiple historical processing fitness, and then, with the goal of reducing processing deviation and reducing the number of mold trials and mold repairs, perform overall directional optimization on multiple historical production line processing parameter sets according to multiple historical processing fitness, obtain target processing parameter sets, and determine K deviation coefficients of K stamping processing nodes by performing deviation mean analysis on multiple processing quality deviation sets respectively, and then transmit the K deviation coefficients and the target processing parameter set to the production line management unit to perform production line processing management of the target mold. The technical effect of optimizing processing parameters and improving processing efficiency is achieved.
[0091] Embodiment 2 is based on the same inventive concept as the green-driven hardware mold manufacturing production line management method in the previous embodiment. Figure 2 As shown, the present application provides a hardware mold manufacturing production line management platform based on green drive, and the platform and method embodiments in the present application embodiment are based on the same inventive concept. Among them, the platform includes:
[0092] A historical production line processing parameter set acquisition module 11 is used to perform a neighbor search in a memory storage of a mold manufacturing production line according to mold structure information, mold material information and mold process parameter set of a target mold, and obtain multiple historical production line processing parameter sets;
[0093] A processing quality deviation set obtaining module 12 is used to perform processing quality searches on the multiple historical production line processing parameter sets respectively to obtain multiple processing quality deviation sets;
[0094] A historical processing fitness generation module 13 is used to traverse the multiple processing quality deviation sets to identify the processing parameter fitness and generate multiple historical processing fitnesses;
[0095] The target processing parameter set obtaining module 14 is used to perform overall directional optimization on the multiple historical production line processing parameter sets according to the multiple historical processing fitnesses with the goal of reducing processing deviation and reducing the number of mold trials and mold repairs to obtain a target processing parameter set;
[0096] A deviation coefficient determination module 15 is used to perform deviation mean analysis on the multiple processing quality deviation sets respectively to determine K deviation coefficients of K stamping processing nodes;
[0097] The production line processing management module 16 is used to transmit the K deviation coefficients and the target processing parameter set to the production line management unit to perform production line processing management of the target mold.
[0098] Furthermore, the historical production line processing parameter set acquisition module 11 is used to perform the following steps:
[0099] Calling a memory storage library of the mold manufacturing production line, wherein the memory storage library includes a plurality of historical processing records, and each historical processing record includes a historical production line processing parameter set;
[0100] Randomly extracting M historical processing records from the plurality of historical processing records as M processing cluster centers, wherein the similarity between any two processing cluster centers of the M processing cluster centers is lower than a preset similarity threshold;
[0101] Performing cluster searches on the M processing cluster centers in the memory storage respectively, and performing iterative updates of the centers after each cluster search is completed, to obtain M updated processing cluster centers;
[0102] Analyzing the clustering qualities of the M update processing cluster centers respectively to obtain a memory score of the memory storage repository;
[0103] It is determined whether the memory score is greater than or equal to a preset score threshold. If not, a data supplement instruction is obtained to supplement the memory storage library to obtain a memory storage library with updated data.
[0104] Furthermore, the historical production line processing parameter set acquisition module 11 is used to perform the following steps:
[0105] Calculating the similarity between the plurality of historical processing records and the M processing cluster centers, respectively adding the plurality of historical processing record data to the cluster clusters of the processing cluster center corresponding to the maximum similarity value, to obtain M processing cluster clusters;
[0106] Iteratively updating the processing cluster centers of the M processing clusters according to the weighted center iteration formula until a preset number of iterations is met to obtain the M updated processing cluster centers;
[0107] Calculating offset vectors from the M processing clusters to the M updated processing cluster centers to obtain M offset vector sets;
[0108] A clustering quality analysis is performed based on the M offset vector sets to determine the memory score.
[0109] Furthermore, the weighted center iteration formula is:
[0110] ;
[0111] in, To update the processing cluster center, To process multiple historical processing records in the cluster, is the i-th historical processing record in the processing cluster, It is the processing cluster center.
[0112] Furthermore, the target processing parameter set obtaining module 14 is used to perform the following steps:
[0113] Comparing the plurality of historical processing fitnesses respectively with the sum of the plurality of historical processing fitnesses, and taking the difference between 1 and the comparison result as a plurality of optimization step lengths;
[0114] With the goal of reducing processing deviation and the number of mold trials and mold repairs, the multiple historical production line processing parameter sets are directional fine-tuned based on the multiple optimization step lengths to obtain multiple adjusted production line processing parameter sets, wherein the multiple adjusted production line processing parameter sets include multiple adjusted processing fitness levels;
[0115] The adjusted production line processing parameter set corresponding to the maximum value among the multiple adjusted processing fitness values is used as an optimization particle, and the remaining multiple historical production line processing parameter sets are used as multiple following particles;
[0116] Taking the optimal particle as a target, fine-tuning the plurality of follower particles according to the plurality of optimal step lengths to obtain a plurality of fine-tuned follower particles;
[0117] The optimizing particle is updated and iterated based on the multiple fine-tuning follower particles until a preset number of iterations is met, and the optimizing particle corresponding to the maximum processing fitness value is used as a target processing parameter set.
[0118] Furthermore, the target processing parameter set obtaining module 14 is used to perform the following steps:
[0119] Determine whether there is a fine-tuning follower particle whose adjustment processing fitness is greater than or equal to the optimization fitness of the optimization particle among the multiple fine-tuning follower particles, and if so, update the fine-tuning follower particle corresponding to the maximum adjustment fitness among the multiple fine-tuning follower particles as the optimization particle;
[0120] If not, a random number generation function is called to obtain a first random number. When the first random number is less than or equal to a preset random value, the fine-tuning follower particle corresponding to the maximum value of the adjusted fitness among the multiple fine-tuning follower particles is accepted to update the optimization particle.
[0121] Furthermore, the historical processing fitness generation module 13 is used to perform the following steps:
[0122] Construct a fitness recognition unit;
[0123] The fitness identification unit is used to perform fitness identification of processing parameters on the multiple processing quality deviation sets to obtain the multiple historical processing fitnesses.
[0124] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description and does not represent the advantages and disadvantages of the embodiments. And the above-mentioned specific embodiments of this specification are described. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
[0125] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application should be included in the protection scope of the present application.
[0126] This specification and drawings are merely exemplary illustrations of the present application and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, a person skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application intends to include these modifications and variations.
Claims
1. Green-driven hardware mold manufacturing production line management method, It is characterized in that The method comprises: According to the mold structure information, mold material information and mold process parameter set of the target mold, a neighbor search is performed in the memory storage of the mold manufacturing production line to obtain multiple historical production line processing parameter sets; Performing processing quality searches on the multiple historical production line processing parameter sets respectively to obtain multiple processing quality deviation sets; Traversing the plurality of processing quality deviation sets to identify processing parameter fitness and generate a plurality of historical processing fitnesses; With the goal of reducing processing deviation and the number of mold trials and mold repairs, the multiple historical production line processing parameter sets are optimized in an overall directional manner according to the multiple historical processing fitnesses to obtain a target processing parameter set; Performing deviation mean analysis on the multiple processing quality deviation sets respectively, calculating the ratio of the processing deviations of K nodes to the sum of the processing deviations of K nodes, and determining K deviation coefficients of the K stamping processing nodes; Transmitting the K deviation coefficients and the target processing parameter set to a production line management unit to perform production line processing management of the target mold; Calling a memory storage library of the mold manufacturing production line, wherein the memory storage library includes a plurality of historical processing records, and each historical processing record includes a historical production line processing parameter set; Randomly extracting M historical processing records from the plurality of historical processing records as M processing cluster centers, wherein the similarity between any two processing cluster centers of the M processing cluster centers is lower than a preset similarity threshold; Performing cluster searches on the M processing cluster centers in the memory storage respectively, and performing iterative updates of the centers after each cluster search is completed, to obtain M updated processing cluster centers; Analyzing the clustering qualities of the M update processing cluster centers respectively to obtain a memory score of the memory storage repository; Determine whether the memory score is greater than or equal to a preset score threshold, and if not, obtain a data supplement instruction to supplement the memory storage repository to obtain a memory storage repository with updated data; Comparing the plurality of historical processing fitnesses respectively with the sum of the plurality of historical processing fitnesses, and taking the difference between 1 and the comparison result as a plurality of optimization step lengths; With the goal of reducing processing deviation and the number of mold trials and mold repairs, the multiple historical production line processing parameter sets are directional fine-tuned based on the multiple optimization step lengths to obtain multiple adjusted production line processing parameter sets, wherein the multiple adjusted production line processing parameter sets include multiple adjusted processing fitness levels; The adjusted production line processing parameter set corresponding to the maximum value among the multiple adjusted processing fitness values is used as an optimization particle, and the remaining multiple historical production line processing parameter sets are used as multiple following particles; Taking the optimal particle as a target, fine-tuning the plurality of follower particles according to the plurality of optimal step lengths to obtain a plurality of fine-tuned follower particles; The optimizing particle is updated and iterated based on the multiple fine-tuning follower particles until a preset number of iterations is met, and the optimizing particle corresponding to the maximum processing fitness value is used as a target processing parameter set; Calculating the similarity between the plurality of historical processing records and the M processing cluster centers, respectively adding the plurality of historical processing record data to the cluster clusters of the processing cluster center corresponding to the maximum similarity value, to obtain M processing cluster clusters; Iteratively updating the processing cluster centers of the M processing clusters according to the weighted center iteration formula until a preset number of iterations is met to obtain the M updated processing cluster centers; Calculating offset vectors from the M processing clusters to the M updated processing cluster centers to obtain M offset vector sets; Performing cluster quality analysis based on the M offset vector sets to determine the memory score; The weighted center iteration formula is: ; in, To update the processing cluster center, To process multiple historical processing records in the cluster, is the i-th historical processing record in the processing cluster, is the processing cluster center, is the Gaussian kernel function.
2. The method according to claim 1, It is characterized in that include: Determine whether there is a fine-tuning follower particle whose adjustment processing fitness is greater than or equal to the optimization fitness of the optimization particle among the multiple fine-tuning follower particles, and if so, update the fine-tuning follower particle corresponding to the maximum adjustment fitness among the multiple fine-tuning follower particles as the optimization particle; If not, a random number generation function is called to obtain a first random number. When the first random number is less than or equal to a preset random value, the fine-tuning follower particle corresponding to the maximum value of the adjusted fitness among the multiple fine-tuning follower particles is accepted to update the optimization particle.
3. The method according to claim 1, It is characterized in that include: Construct a fitness recognition unit; The fitness identification unit is used to perform fitness identification of processing parameters on the multiple processing quality deviation sets to obtain the multiple historical processing fitnesses.
4. Green-driven hardware mold manufacturing production line management platform, It is characterized in that The platform is used to execute the method according to any one of claims 1 to 3, and the platform comprises: A module for obtaining a set of historical production line processing parameters, which is used to perform a neighbor search in a memory repository of a mold manufacturing production line according to mold structure information, mold material information and mold process parameter set of a target mold, and obtain multiple sets of historical production line processing parameters; A processing quality deviation set obtaining module, used to perform processing quality searches on the multiple historical production line processing parameter sets respectively to obtain multiple processing quality deviation sets; A historical processing fitness generation module, used for traversing the plurality of processing quality deviation sets to identify processing parameter fitness and generate a plurality of historical processing fitnesses; A target processing parameter set acquisition module is used to perform overall directional optimization on the multiple historical production line processing parameter sets according to the multiple historical processing fitnesses with the goal of reducing processing deviation and reducing the number of mold trials and mold repairs to obtain a target processing parameter set; A deviation coefficient determination module, used to perform deviation mean analysis on the multiple processing quality deviation sets respectively, and determine K deviation coefficients of K stamping processing nodes; The production line processing management module is used to transmit the K deviation coefficients and the target processing parameter set to the production line management unit to perform production line processing management of the target mold.
Citation Information
Patent Citations
Processing equipment process parameter optimization method and system
CN111597729A
Optimized polishing method and system for aluminum alloy die casting
CN116551474A
Intelligent control method for numerical control machine tool production line
CN117075534A
Process parameter control method and system for chip production
CN117234171A
Forging process optimization method for non-quenched and tempered steel automobile parts
CN118260337A
