A metal product production control method and system based on model analysis

Abstract

The present invention discloses a metal product production control method and system based on model analysis, which relates to the field of intelligent manufacturing. The method comprises: interactively obtaining the initial parameter setting of the metal product, and conducting a trial production of the first sample. Collecting the surface image, calling the model analysis to obtain the first defect detection result. If it is a surface treatment defect, the process requirements are updated according to the result to obtain K updated process requirements and parameters, so as to schedule the steps from the assembly area of ​​the prefabricated process, assemble and adjust the parameters to generate the surface post-treatment process. The die-casting process is adjusted with the initial parameters, and batch production control is performed in combination with the surface post-treatment process. The technical problem that the existing metal product generation relies on fixed processing parameters and manual inspection, and the diversified defects occurring in the production process cannot be discovered and handled in a timely manner is solved, and the technical effect of avoiding omissions and inaccuracies in manual inspection and improving the efficiency and accuracy of quality inspection is achieved.

Description

Technical Field

[0001] The present application relates to the field of intelligent manufacturing, and in particular to a metal product production control method and system based on model analysis. Background Art

[0002] In the production scenario of metal products, the requirements for product quality and production efficiency are increasing day by day. The timely discovery and handling of various defects in the production process are particularly prominent. Achieving efficient and accurate production control has become a crucial link in improving the production level of metal products. Traditional metal product production often relies more on fixed processing parameters and manual inspection, and lacks overall control over the production process. It lacks the ability to timely identify and handle various defects, making it difficult to comprehensively and accurately evaluate the quality of products. In the production process, there are unreasonable situations, which lead to some defects that cannot be discovered and handled in time, and some unnecessary links waste resources. The control of the production process is relatively simple and fixed, and cannot cope well with the ever-changing actual situation in the production of metal products.

[0003] In the current relevant technologies, the production of metal products relies on fixed processing parameters and manual inspection, which is unable to timely discover and deal with the diverse defects that occur in the production process. Summary of the invention

[0004] The present application provides a metal product production control method and system based on model analysis, which uses interaction to obtain the initial parameter settings of the metal product and conduct trial production of the first sample. The surface image is collected, and the model analysis is called to obtain the first defect detection result. If it is a surface treatment defect, the process requirements are updated according to the results to obtain K updated process requirements and parameters, and the updated process execution sequence is updated by sequence dependency analysis, so as to schedule the steps from the assembly area of ​​the prefabricated process, assemble and adjust the parameters to generate the surface post-treatment process. Finally, the die-casting process is adjusted with the initial parameters, and batch production control is carried out in combination with the surface post-treatment process, so as to avoid the omissions and inaccuracies of manual inspection and improve the efficiency and accuracy of quality inspection.

[0005] The present application provides a metal product production control method based on model analysis, comprising:

[0006] Interactively obtain the initial parameter settings of the metal product to be processed, and use the initial parameter settings to run the die-casting process steps to conduct trial production of the metal product to obtain a first sample metal product; after collecting the surface image of the first sample metal product to obtain a first surface image set, call the pre-built surface defect inspection model to analyze the first surface image set to obtain a first defect detection result; if the defect type of the first defect detection result is a surface treatment defect, perform process requirement scheduling update according to the first defect detection result to obtain K updated process requirements and K updated process parameters; perform sequence dependency analysis on the K updated process requirements to obtain an updated process execution sequence; use the K updated process requirements as scheduling constraints to obtain K updated process steps from the assembly area of ​​the prefabricated process; assemble the K updated process steps with the updated process execution sequence, and use the K updated process parameter mapping to perform parameter setting of the K updated process steps to generate a surface post-treatment process; after using the initial parameter settings to perform parameter setting of the die-casting process steps, perform batch production control of the metal product to be processed based on the die-casting process steps and the surface post-treatment process.

[0007] The present application also provides a metal product production control system based on model analysis, including:

[0008] A first sample metal product acquisition module, the first sample metal product acquisition module is used to interactively obtain the initial parameter settings of the metal product to be processed, and use the initial parameter settings to run the die-casting process steps to conduct trial production of the metal product to obtain the first sample metal product; a first defect detection result acquisition module, the first defect detection result acquisition module is used to collect the appearance image of the first sample metal product, obtain the first appearance image set, call the pre-built appearance defect inspection model to analyze the first appearance image set, and obtain the first defect detection result; a scheduling update execution module, the scheduling update execution module is used to execute the process requirement scheduling update according to the first defect detection result if the defect type of the first defect detection result is a surface treatment defect, and obtain K updated process requirements and K updated process parameters; a process execution sequence update module The process execution sequence update module is used to perform sequence dependency analysis on the K update process requirements to obtain an update process execution sequence; the update process step acquisition module is used to use the K update process requirements as scheduling constraints to obtain K update process steps from the assembly area of ​​the prefabricated process; the surface post-treatment process generation module is used to assemble the K update process steps with the update process execution sequence, and use the K update process parameter mappings to set the parameters of the K update process steps to generate a surface post-treatment process; the production control module is used to use the initial parameter setting to set the parameters of the die-casting process steps, and then perform batch production control of the metal products to be processed based on the die-casting process steps and the surface post-treatment process.

[0009] The metal product production control method and system based on model analysis proposed in this application first interactively obtains the initial parameter settings of the metal product and conducts the first sample of trial production. Collect its surface image, call the model analysis to obtain the first defect detection result. If it is a surface treatment defect, update the process requirements according to the results to obtain K updated process requirements and parameters, and perform a sequence dependency analysis to update the process execution sequence, so as to schedule the steps from the assembly area of ​​the prefabricated process, assemble and adjust the parameters to generate the surface post-treatment process. Finally, the die-casting process is adjusted with the initial parameters, and batch production control is carried out in combination with the surface post-treatment process, so as to achieve the technical effect of avoiding omissions and inaccuracies in manual inspection and improving the efficiency and accuracy of quality inspection. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] In order to more clearly illustrate the technical solution of the embodiment of the present invention, the accompanying drawings of the embodiment of the present invention will be briefly introduced below. A flow chart is used in the present application to illustrate the operations performed by the system according to the embodiment of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, various steps can be processed in reverse order or simultaneously as needed. At the same time, other operations can also be added to these processes, or one or more operations can be removed from these processes.

[0011] Figure 1 A schematic diagram of a process flow of a metal product production control method based on model analysis provided in an embodiment of the present application;

[0012] Figure 2 A schematic structural diagram of a metal product production control system based on model analysis provided in an embodiment of the present application.

[0013] Explanation of the accompanying drawings: first sample metal product acquisition module 10, first defect detection result acquisition module 20, scheduling update execution module 30, process execution sequence update module 40, update process step acquisition module 50, surface post-treatment process generation module 60, production control module 70. DETAILED DESCRIPTION

[0014] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below.

[0015] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings. The described embodiments should not be regarded as limiting the present application. All other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of this application.

[0016] In the following description, reference is made to "some embodiments", which describe a subset of all possible embodiments, but it is understood that "some embodiments" may be the same subset or different subsets of all possible embodiments, and may be combined with each other without conflict, and the terms "first\second" involved are merely to distinguish similar objects and do not represent a specific ordering of objects. The terms "including" and "having" and any variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or modules that are not clearly listed or inherent to these processes, methods, products, or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those generally understood by technicians in the technical field of this application. The terms used herein are for the purpose of describing the embodiments of the present application only.

[0017] The present application embodiment provides a metal product production control method based on model analysis, such as Figure 1 As shown, the method includes:

[0018] Step S100, interactively obtain the initial parameter settings of the metal product to be processed, and use the initial parameter settings to run the die-casting process steps to conduct trial production of the metal product to obtain the first sample metal product. Specifically, first determine the interactive object, collect parameter information covering die-casting temperature, pressure, time, mold size and shape, metal material type characteristics and injection speed from the design department, production experience database and process engineers, and verify the rationality of the information. If there is any unreasonable information, communicate and adjust it in time. Check and prepare the die-casting machine, mold and other equipment according to the initial parameter requirements, debug the equipment according to the parameter settings, input the parameters into the control system and calibrate and confirm, adjust the mold cooling system, conduct trial production, inject molten metal into the mold, and die-cast at the set pressure and time. Closely observe the equipment operation status and product molding and record key data. After the trial production is completed, take out the first sample metal product, conduct preliminary appearance inspection and key dimension measurement, evaluate the trial production effect, and lay the foundation for subsequent quality inspection and process optimization.

[0019] In a possible implementation, the initial parameter setting of the metal product to be processed is interactively obtained, and the die-casting process steps are run using the initial parameter setting to conduct trial production of the metal product to obtain a first sample metal product, and step S100 further includes step S110, calling the die-casting mold of the metal product to be processed. Specifically, according to the design requirements and specifications of the metal product to be processed, the required die-casting mold type is determined, and the shape, size, complexity and other factors of the product are considered to select a suitable mold structure and material, and the required die-casting mold is retrieved in the mold repository or production management system, and the corresponding mold is accurately found through a specific code or identification, and it is called out to be installed on the die-casting machine.

[0020] Step S120, after the die-casting mold is installed on the die-casting machine, the initial parameter setting is used to perform parameter setting of the die-casting process step, wherein the die-casting process step includes an injection module, a cooling module, a robotic arm and the die-casting machine. Specifically, after the die-casting mold is installed on the die-casting machine, the parameters of the die-casting process step are set according to the initial parameter setting. For the injection module, the injection / filling speed of the molten metal is determined according to the initial parameter setting to ensure that the speed meets the requirements of the current metal product to be processed, and no additional parameter adjustment operation is performed. For the cooling module, the cooling rate is strictly controlled in accordance with the initial parameter setting, including the flow rate, temperature and cooling time of the cooling medium, to maintain the stability of the parameters. For the robotic arm, the action speed, gripping force and position accuracy are adjusted according to the initial parameter setting to ensure that the formed metal product can be accurately and safely removed. For the die-casting machine, the mold temperature, the melt metal temperature and the pressure applied during the die-casting process are controlled according to the initial parameter setting to provide good process conditions for the die-casting of metal products, and no parameter adjustment is performed.

[0021] Step S130, run the die-casting process step with the initial parameter setting to conduct trial production of the metal product to be processed, and obtain the first sample metal product. Specifically, after completing the adjustment of all parameters, start the die-casting process step, and start the injection module, cooling module, robotic arm and die-casting machine in sequence in a predetermined order to ensure the coordinated operation between the modules. During the trial production process, closely monitor the changes in various process parameters and the molding of the product, use sensors and monitoring equipment to collect temperature, pressure, speed and other data in real time, and display and analyze them through the control system. After the trial production is completed, take out the molded metal product, that is, the first sample metal product, from the die-casting machine, conduct detailed inspection and measurement of the sample, evaluate the effect of the trial production, and inspect the appearance quality of the product, including surface finish, defects, etc.; measure key dimensions, compare with design requirements, judge the dimensional accuracy of the product, and further adjust the process parameters according to the inspection and measurement results to prepare for subsequent mass production.

[0022] In a possible implementation, the die-casting process steps are run with the initial parameter settings to conduct trial production of the metal product to be processed, and the first sample metal product is obtained. Step S130 further includes step S131, interactively obtaining multiple sample die-casting defect image sets of multiple die-casting surface defects. Specifically, it is clear from which channels to obtain sample image sets of multiple die-casting surface defects, including defective product image libraries in actual production, experimental environments specifically for defect collection, databases shared with other companies or research institutions, etc., and metal products with different die-casting surface defects are photographed under fixed environmental conditions (such as lighting, angle, etc.) through specific image acquisition equipment, ensuring that the distance between the image acquisition equipment and the surface of the metal product is fixed to ensure the accuracy of defect size recognition, and collecting multiple image samples containing different types of die-casting surface defects to form multiple sample die-casting defect image sets.

[0023] Step S132, after the die-casting defects of the plurality of sample die-casting defect image sets are marked and a plurality of sample first marked image sets are obtained, the plurality of sample die-casting defect image sets and the plurality of sample first marked image sets are used to construct a plurality of die-casting defect recognition branches. Specifically, the defects in each sample die-casting defect image set are marked in detail, and the marking content includes the die-casting defect type (such as flow marks, cold shuts, shrinkage, shrinkage cavities, cracks, pinholes, pores, etc.) and size information. The position and characteristics of the die-casting defects are accurately marked on the image by professional image analysis software or manual annotation. After marking, a plurality of sample first marked image sets are obtained. For each specific die-casting defect type, the corresponding sample die-casting defect image set and the sample first marked image set are used to train and construct a die-casting defect recognition branch. One branch is specifically used to accurately identify a specific die-casting defect type among multiple die-casting defects. Through the convolutional neural network in deep learning, the model learns the characteristics and patterns of the die-casting defect type, so that the die-casting defect can be accurately identified in subsequent detection.

[0024] Step S133, interactively obtain multiple sample post-processing defect image sets of multiple post-processing surface defects. Specifically, just like obtaining a die-casting surface defect image set, determine the source of obtaining multiple post-processing surface defect sample image sets, which can be defective product images found in the post-processing process, images collected by a specially set detection link, etc. Use an image acquisition device to shoot metal products that have undergone post-processing but still have surface defects, and also keep the distance between the image acquisition device and the surface of the metal product fixed to ensure the accuracy of image quality and size recognition. Collect multiple image samples containing different types of post-processing surface defects to form multiple sample post-processing defect image sets.

[0025] Step S134, after post-processing defect identification is performed on the multiple sample post-processing defect image sets to obtain multiple sample second identification image sets, the multiple sample post-processing defect image sets and the multiple sample second identification image sets are used to construct multiple post-processing defect recognition branches. Specifically, the defects in each sample post-processing defect image set are identified, and the identification content also includes defect type and size information. A method similar to die-casting surface defect identification is used to accurately mark the position and characteristics of the post-processing surface defects on the image through software or manual annotation to obtain multiple sample second identification image sets. For each specific post-processing surface defect type, the corresponding sample post-processing defect image set and the sample second identification image set are used to train and construct a post-processing defect recognition branch. One branch is specifically used to accurately identify a specific defect type among multiple post-processing surface defects, and a machine learning algorithm is used to allow the model to learn the characteristics and patterns of the defect type.

[0026] Step S135, connect the multiple die-casting defect recognition branches in parallel to obtain a die-casting defect recognition channel, and connect the multiple post-processing defect recognition branches in parallel to obtain a post-processing defect recognition channel. Specifically, connect the multiple die-casting defect recognition branches in parallel so that different branches analyze the input image at the same time, and each recognizes a specific type of die-casting surface defect. By connecting in parallel, a die-casting defect recognition channel is formed, which can comprehensively detect and identify a variety of die-casting surface defects. Similarly, connect the multiple post-processing defect recognition branches in parallel to form a post-processing defect recognition channel, which can efficiently detect and identify a variety of post-processing surface defects.

[0027] Step S136, connect the die-casting defect recognition channel and the post-processing defect recognition channel in parallel to complete the construction of the apparent defect inspection model. Specifically, the die-casting defect recognition channel and the post-processing defect recognition channel are connected in parallel, and the constructed apparent defect inspection model can simultaneously detect and identify die-casting surface defects and post-processing surface defects of the input metal product image, and further optimize and adjust the constructed apparent defect inspection model, by adding more training data, adjusting model parameters, and performing cross-validation, etc., to improve the accuracy and stability of the model, and finally complete an apparent defect inspection model that can accurately detect and identify various defects on the surface of metal products.

[0028] Step S200, after acquiring the surface image of the first sample metal product and obtaining the first surface image set, call the pre-built surface defect inspection model to analyze the first surface image set to obtain the first defect detection result. Specifically, select a high-resolution image acquisition device and set a suitable acquisition environment to ensure uniform lighting and stable environment, place the first sample metal product in a fixed position, shoot from different angles and distances to obtain a comprehensive surface image, organize it into a first surface image set, call the pre-built surface defect inspection model, input the first surface image set into it, and the model determines whether there are defects on the surface of the metal product through feature extraction and pattern recognition, determines the defect type and location, and comprehensively obtains the first defect detection result, providing a basis for subsequent process adjustment and processing.

[0029] In a possible implementation, after collecting the surface image of the first sample metal product and obtaining the first surface image set, the pre-built surface defect inspection model is called to analyze the first surface image set to obtain the first defect detection result, and step S200 further includes step S210, synchronizing F surface images in the first surface image set to the surface defect inspection model. Specifically, the basis for selecting F surface images from the first surface image set is clarified, so as to determine the specific value of F according to the processing capacity, computing resources and requirements for detection accuracy of the model, and select more images to obtain more comprehensive information; if the computing resources are limited, the number of images can be appropriately reduced to increase the processing speed, and the selected F surface images are synchronously transmitted to the surface defect inspection model through the data transmission channel to ensure the accuracy and integrity of the image transmission, avoid data loss or damage, and use a high-speed data transmission interface or network connection to achieve fast synchronization of images.

[0030] Step S220, image recognition is performed synchronously through the die-casting defect recognition channel and the post-processing defect recognition channel of the apparent defect inspection model to obtain die-casting defect recognition results and post-processing defect recognition results. Specifically, the die-casting defect recognition channel and the post-processing defect recognition channel in the apparent defect inspection model are activated. The channels are responsible for identifying die-casting surface defects and post-processing surface defects respectively, ensuring the normal operation of the channels and correct parameter settings to improve the accuracy of recognition. The F synchronized apparent images are simultaneously input into the die-casting defect recognition channel and the post-processing defect recognition channel for image recognition. The two channels detect and classify the defects in the images according to their trained models respectively. Through parallel processing, the recognition efficiency can be improved and the detection time can be reduced. After image recognition, the die-casting defect recognition results and the post-processing defect recognition results are obtained from the die-casting defect recognition channel and the post-processing defect recognition channel respectively. The results may include information such as whether there are defects of the corresponding type, the specific location and characteristics of the defects, etc.

[0031] Step S230: If the die-casting defect recognition result is an empty set and the post-processing defect recognition result is a non-empty set, the first defect detection result is a surface treatment defect. Specifically, the die-casting defect recognition result and the post-processing defect recognition result are analyzed and judged. If the die-casting defect recognition result is an empty set, it means that no die-casting surface defects are detected in these F apparent images, and the post-processing defect recognition result is a non-empty set, which means that there are post-processing surface defects. According to the above judgment results, it is determined that the first defect detection result is a surface treatment defect, which means that the metal product may not have obvious problems in the die-casting process, but defects have occurred in the subsequent post-processing link, and the post-processing process needs to be adjusted and optimized.

[0032] Step S240, if the die-casting defect identification result is a non-empty set, the first defect detection result is a die-casting process defect. Specifically, if the die-casting defect identification result is a non-empty set, it means that a die-casting surface defect is detected in the F surface images. At this time, the first defect detection result is determined to be a die-casting process defect, indicating that there may be a problem in the initial parameter setting of the metal product in the die-casting process, and the parameters of the die-casting process need to be optimized and adjusted to reduce or eliminate the die-casting surface defects.

[0033] In one possible implementation, if the die-casting defect identification result is a non-empty set, the first defect detection result is a die-casting processing defect, and step S240 further includes step S241. If the die-casting defect identification result is a non-empty set, the M types of measured die-casting defects in the die-casting defect identification result are used to traverse the defect index association table to locate the M parameter adjustment directions of the M types of associated indicators. Specifically, when the die-casting defect recognition result is a non-empty set, it indicates that die-casting surface defects are found in the detected image, and it is necessary to further determine the specific M measured die-casting defect types, which may include different defects such as pores, shrinkage holes, flow marks, etc. According to the determined M measured die-casting defects, the pre-constructed defect index association table is traversed. The table is established based on the analysis and summary of the association relationship between different die-casting defects and corresponding process parameter indicators. For example, for flow mark defects, the cause is the formation of stripes or ripples on the casting surface during the flow of molten metal in the die-casting stage, so the flow mark defect is associated with the molten metal injection speed, that is, the flow mark defect corresponds to the adjustment of the molten metal injection speed of the die-casting process. By traversing the table, the associated index corresponding to each measured die-casting defect can be found. For each associated index found for each measured die-casting defect, the corresponding parameter adjustment direction is determined. For example, if a defect is associated with the molten metal temperature, and the defect is currently detected to indicate that the molten metal temperature is too high, then the parameter adjustment direction is to reduce the molten metal temperature, and M parameter adjustment directions of M associated indicators are determined.

[0034] Step S242, preset the parameter adjustment step, and use the M parameter adjustment directions and the parameter adjustment step as constraints, update the initial parameter settings in multiple rounds, until the defect type of the Nth defect detection result is a surface treatment defect, and output the Nth updated parameter setting and the Nth sample metal product. Specifically, the parameter adjustment step is preset, and it is stipulated here that each parameter adjustment amount is 1 / 10 of the current parameter to ensure that the amplitude of the parameter adjustment is moderate, avoid excessive adjustment and cause new problems, and gradually approach the optimal parameter setting. With the M parameter adjustment directions and the preset parameter adjustment step as constraints, the initial parameter setting is updated in multiple rounds. In each round of update, the corresponding process parameters are adjusted according to the determined parameter adjustment direction and step. For example, if a certain parameter adjustment direction is to reduce the temperature of the molten metal, then the current molten metal temperature is reduced by a certain proportion according to the parameter adjustment step. Then, the die-casting process is trial-produced using the updated parameter settings to obtain new sample metal products, and surface image acquisition and defect detection are performed to obtain new defect detection results. The parameter update, trial production, and detection processes are repeated, and the parameters are continuously adjusted until the defect type of the Nth defect detection result obtained is a surface treatment defect. After multiple rounds of parameter adjustment, the die-casting surface defects have been effectively controlled. At this time, the Nth updated parameter settings and the corresponding Nth sample metal product are output.

[0035] Step S243, using the Nth updated parameter setting to replace the initial parameter setting. Specifically, when the Nth updated parameter setting that meets the conditions is obtained, it is used to replace the original initial parameter setting, and the updated parameters can be used in the subsequent production process to perform the die casting process to improve product quality.

[0036] Step S244, according to the Nth defect detection result, the process demand scheduling is updated to obtain H updated process demands and H updated process parameters. Specifically, according to the Nth defect detection result, if it is determined to be a surface treatment defect, it means that although the die-casting surface defects are controlled, there may still be some problems to be solved in the subsequent processing process. Based on the type and degree of the surface treatment defect, the process demand scheduling is updated, including analyzing which specific post-processing process steps need to be performed, and determining the process parameters required for each step. For example, if oxidation color is detected on the surface of the metal product, chemical treatment may be required to remove the oxide layer, which generates corresponding updated process demands and process parameters. In this way, H updated process demands and H updated process parameters can be obtained to provide guidance for the subsequent surface post-processing process.

[0037] In a possible implementation, a parameter adjustment step is preset, and the M parameter adjustment directions and parameter adjustment step are used as constraints, and the initial parameter setting is updated in a directional manner for multiple rounds until the defect type of the obtained Nth defect detection result is a surface treatment defect, and the Nth updated parameter setting and the Nth sample metal product are output. Step S242 further includes step S2421, in which the initial parameter setting is updated in a directional manner with the M parameter adjustment directions and parameter adjustment step as constraints to obtain a first updated parameter setting. Specifically, it is clear that the M parameter adjustment directions are adjustment directions determined based on the previous analysis of die-casting defects and related indicators, and the parameter adjustment step is 1 / 10 of the current parameter. The constraint condition will guide the update of the initial parameter setting, and the initial parameter setting is updated in a directional manner according to the M parameter adjustment directions and parameter adjustment step. For example, if a certain parameter adjustment direction is to reduce the temperature of the molten metal, and the parameter adjustment step is 1 / 10 of the current temperature, then the initial molten metal temperature is reduced by a corresponding proportion to obtain the first updated parameter setting.

[0038] Step S2422, using the first updated parameter setting to run the die-casting process step to conduct trial production of metal products, calling the apparent defect inspection model to analyze the trial production results, and obtaining a second defect detection result. Specifically, using the first updated parameter setting to run the die-casting process step to conduct trial production of metal products, including applying the adjusted parameters to various modules of the die-casting machine, such as the injection module that controls the injection / filling speed, the cooling module that controls the cooling rate, etc., and performing die-casting molding of the metal product according to the new parameters. After the trial production is completed, the obtained metal product is subjected to apparent image acquisition to obtain a new apparent image set. Then, the apparent defect inspection model is called to analyze this image set to detect whether there are defects on the surface of the metal product. After model analysis, a second defect detection result is obtained to determine whether the metal product has die-casting processing defects or other types of defects.

[0039] Step S2423, if the second defect detection result is a die-casting process defect, the first updated parameter setting is updated with the M parameter adjustment directions and parameter adjustment step lengths as constraints to obtain the second updated parameter setting. Specifically, if the second defect detection result is a die-casting process defect, it means that although the initial parameters have been updated once, there is still a die-casting surface defect problem, and the parameters need to be further adjusted. The first updated parameter setting is updated again with the M parameter adjustment directions and parameter adjustment step lengths as constraints. In the same way, each process parameter is adjusted according to the parameter adjustment direction and step length to obtain the second updated parameter setting.

[0040] Step S2424, using the second updated parameter setting to run the die-casting process step to conduct a trial production of a metal product, calling the apparent defect inspection model to analyze the trial production results, and obtaining a third defect detection result. Specifically, the second updated parameter setting is used to conduct a trial production of a metal product again, and the apparent defect inspection model is called to analyze the trial production results, just like the previous step, and after analysis, the third defect detection result is obtained to determine the effect of this parameter adjustment.

[0041] Step S2425, and so on, update the initial parameter setting with the M parameter adjustment directions and parameter adjustment step lengths as constraints, until the defect type of the obtained Nth defect detection result is a surface treatment defect, and output the Nth updated parameter setting and the Nth sample metal product. Specifically, the above steps are repeated continuously, and the parameters are updated one by one with the M parameter adjustment directions and parameter adjustment step lengths as constraints, and trial production and defect detection are carried out. Each time, it is determined whether the parameters need to be adjusted further based on the new defect detection results. This process is continued until the defect type of the obtained Nth defect detection result is a surface treatment defect. After multiple parameter adjustments, the die-casting surface defects are effectively controlled and enter the stage where post-processing surface defects need to be processed. The Nth updated parameter setting and the corresponding Nth sample metal product are output, and the results will be used for subsequent process demand scheduling updates and batch production control.

[0042] Step S300: If the defect type of the first defect detection result is a surface treatment defect, the process requirement scheduling update is performed according to the first defect detection result to obtain K updated process requirements and K updated process parameters. Specifically, after determining that the first defect detection result is a surface treatment defect, further identify the specific manifestation of the surface treatment defect, establish a process requirement analysis framework, associate the specific defect with the possible treatment process, perform process requirement scheduling update, determine K updated process requirements according to the complexity and diversity of the surface treatment defects, clarify the specific content and objectives of each requirement, and determine the corresponding updated process parameters for each requirement to provide specific guidance for subsequent surface post-treatment processes.

[0043] In a possible implementation, if the defect type of the first defect detection result is a surface treatment defect, a process requirement scheduling update is performed according to the first defect detection result to obtain K updated process requirements and K updated process parameters, and step S300 further includes step S310, extracting a first defect identification image set of a first measured post-processing defect from the post-processing defect recognition result, wherein the post-processing defect recognition result includes K defect identification image sets of K measured post-processing defects. Specifically, the post-processing defect recognition result includes K defect identification image sets of K measured post-processing defects, which are obtained by performing an apparent defect inspection on a first sample metal product, wherein information such as the type and location of various post-processing surface defects is identified, and the first defect identification image set of the first measured post-processing defect is extracted from this post-processing defect recognition result in order to further analyze and process a specific post-processing defect.

[0044] Step S320, after the metal product model is constructed according to the structural design information of the metal product to be processed, L defect identification images in the first defect identification image set are fitted to the metal product model to locate the first measured defect array, where L is a positive integer less than or equal to F. Specifically, according to the structural design information of the metal product to be processed, such as size, shape, material, etc., a three-dimensional model of the metal product is constructed, and the model will serve as the basis for subsequent analysis and processing. L defect identification images (L is a positive integer less than or equal to F) are selected from the first defect identification image set, and these images are fitted to the constructed metal product model. The two-dimensional defect image can be combined with the three-dimensional metal product model to more accurately locate the position and range of the defects on the metal product. After image fitting, the first measured defect array is determined, and the array contains the specific position and distribution of specific post-processing defects on the metal product, providing an accurate target area for subsequent processing.

[0045] Step S330, vertical image acquisition is performed on the first sample metal product according to the first measured defect array to obtain a first measured image array. Specifically, vertical image acquisition is performed on the first sample metal product according to the defect position and range determined by the first measured defect array. Vertical acquisition can ensure better identification of the size and shape of the defect and improve the accuracy of defect detection. A suitable image acquisition device is used to take multiple shots of the defect area on the metal product to obtain a series of vertical images. The images constitute the first measured image array, which provides basic data for further analysis of defect parameters.

[0046] Step S340, synchronizing the first measured image array to the processing defect recognition channel to obtain a first defect parameter array. Specifically, the first measured image array is synchronously transmitted to the processing defect recognition channel, which is a part of the previously constructed apparent defect inspection model and is specifically used to analyze post-processing surface defects. The first measured image array is analyzed through the processing defect recognition channel to identify various defect parameters, such as defect size, depth, shape, etc. These parameters constitute the first defect parameter array, which provides a basis for subsequent process parameter matching.

[0047] Step S350, after matching and obtaining the first updated process requirement of the first measured post-processing defect, matching process parameters is performed according to the first defect parameter array to obtain the first process parameter array. Specifically, according to the characteristics and requirements of the first measured post-processing defect, the corresponding first updated process requirement is matched. For example, if the defect is a scratch, a grinding or polishing process may be required; if it is an oxidized color, chemical treatment may be required, etc., and process parameters are matched according to the defect parameters included in the first defect parameter array. Determine the specific process parameters suitable for the defect, such as the grinding intensity, polishing time, concentration of chemical reagents, etc. The parameters constitute the first process parameter array.

[0048] Step S360, perform trajectory optimization according to the first measured defect array, locate the first post-processing trajectory, fit the first process parameter array to the first post-processing trajectory, and obtain the first updated process parameters. Specifically, according to the defect distribution determined by the first measured defect array, perform trajectory optimization and determine the best post-processing trajectory so that all defect areas can be covered more efficiently and accurately when processing defects, fit the first process parameter array to the optimized first post-processing trajectory, ensure that each position can use appropriate process parameters during post-processing operations, and improve the processing effect. After the above steps, the first updated process parameters are finally obtained, and the parameters will be used to perform specific processing operations on the first measured post-processing defects.

[0049] Step S370, and so on, obtain the K update process requirements and K update process parameters. Specifically, according to the above steps, analyze and process other post-processing defects one by one, and for each specific post-processing defect, repeat the process of extracting defect identification image set, building model, fitting image, collecting image, synchronous analysis, determining update process requirements, matching process parameters, optimizing trajectory and obtaining update process parameters. Through continuous repetition, finally obtain K update process requirements and K update process parameters, corresponding to K measured post-processing defects. These parameters will provide comprehensive guidance for the surface post-processing process of metal products.

[0050] Step S400, performing sequence dependency analysis on the K update process requirements to obtain the update process execution sequence. Specifically, first clarify the K update process requirements, which correspond to the specific process operations of the surface post-processing defects of metal products. Considering that the processing steps corresponding to each update process requirement will have an impact on the subsequent steps, and the order will vary depending on the specific product requirements and processing technology. Usually, mechanical trimming or grinding is performed first to remove burrs and sharp edges, followed by corrosion and oxidation color treatment, bonding and pinhole removal, wrinkle, scar and pattern smoothing treatment, but during the analysis process, if the sequence needs to be adjusted due to specific product requirements or processing technology, the impact needs to be carefully evaluated. By performing sequence dependency analysis on the K update process requirements, a reasonable update process execution sequence is determined to guide the execution of the surface post-processing process, and it is continuously adjusted and optimized according to actual conditions to adapt to different product requirements and processing technology changes.

[0051] Step S500, taking the K update process requirements as scheduling constraints, obtain K update process steps from the prefabricated process assembly area. Specifically, firstly, the K update process requirements are clarified as scheduling constraints. The requirements are for post-processing defects on the surface of metal products. The prefabricated process assembly area stores and manages various process steps, which are used as resource integration areas. According to the K update process requirements, demand matching is performed in the prefabricated process assembly area. The appropriate process steps are selected and called through a specific scheduling mechanism. After evaluation and screening, it is ensured that the requirements can be effectively met. Finally, K update process steps are obtained to form a specific process plan for post-processing of the surface of metal products.

[0052] Step S600, assemble the K update process steps with the update process execution sequence, and use the K update process parameter mapping to adjust the parameters of the K update process steps to generate a surface post-processing process. Specifically, the update process execution sequence is obtained by performing a sequence dependency analysis on the K update process requirements, and the execution order of the K update process steps is clarified. Then, assemble the K update process steps according to the sequence to ensure smooth connection of the steps. At the same time, K update process parameters are determined for each step, and they are mapped to the corresponding steps for parameter setting to optimize the execution effect. Finally, the process steps and parameters are integrated to generate a complete and feasible surface post-processing process, which effectively handles the surface post-processing defects of metal products and improves product quality and appearance.

[0053] Step S700, after the initial parameter setting is used to adjust the parameters of the die-casting process steps, the batch production control of the metal product to be processed is performed based on the die-casting process steps and the surface post-treatment process. Specifically, the initial parameter setting is first reviewed, and after the preliminary trial production and defect detection, the initial parameter setting is used to adjust the parameters of the die-casting process steps according to the analysis results. Then, the adjusted die-casting process steps are combined with the surface post-treatment process to form a complete production process, and a production control system including quality inspection, production progress monitoring and equipment maintenance management is established. During the batch production process, quality inspections are carried out regularly, production progress is closely monitored, and equipment is kept in good operating condition. At the same time, data and feedback information are continuously collected to continuously optimize the production process to ensure the production of high-quality metal products.

[0054] The embodiment of the present application uses the interactive method to obtain the initial parameter settings of the metal product and conduct the trial production of the first sample. The surface image is collected, and the model is called for analysis to obtain the first defect detection result. If it is a surface treatment defect, the process requirements are updated according to the results to obtain K updated process requirements and parameters, and the updated process execution sequence is updated by sequence dependency analysis, so as to schedule the steps from the assembly area of ​​the prefabricated process, assemble and adjust the parameters to generate the surface post-treatment process. Finally, the die-casting process is adjusted with the initial parameters, and batch production control is carried out in combination with the surface post-treatment process, so as to achieve the technical effect of avoiding omissions and inaccuracies in manual inspection and improving the efficiency and accuracy of quality inspection.

[0055] In the above, refer to Figure 1 A metal product production control method based on model analysis according to an embodiment of the present invention is described in detail. Figure 2 A metal product production control system based on model analysis according to an embodiment of the present invention is described.

[0056] A metal product production control system based on model analysis according to an embodiment of the present invention is used to solve the technical problem that the existing metal product generation relies on fixed processing parameters and manual inspection, and cannot timely discover and handle the diverse defects that occur in the production process, so as to avoid the omissions and inaccuracies of manual inspection and improve the efficiency and accuracy of quality inspection. A metal product production control system based on model analysis includes: a first sample metal product acquisition module 10, a first defect detection result acquisition module 20, a scheduling update execution module 30, a process execution sequence update module 40, an update process step acquisition module 50, a surface post-treatment process generation module 60, and a production control module 70.

[0057] The first sample metal product acquisition module 10 is used to interactively obtain the initial parameter settings of the metal product to be processed, and use the initial parameter settings to run the die-casting process steps to conduct trial production of the metal product to obtain the first sample metal product.

[0058] The first defect detection result acquisition module 20 is used to collect the surface image of the first sample metal product and obtain the first surface image set, and then call the pre-built surface defect inspection model to analyze the first surface image set to obtain the first defect detection result.

[0059] The scheduling update execution module 30 is used to execute the process requirement scheduling update according to the first defect detection result if the defect type of the first defect detection result is a surface treatment defect, and obtain K updated process requirements and K updated process parameters.

[0060] The process execution sequence updating module 40 is used to perform sequence dependency analysis on the K updated process requirements to obtain an updated process execution sequence.

[0061] The update process step acquisition module 50 is used to obtain K update process steps from the assembly area of ​​the prefabricated process by taking the K update process requirements as scheduling constraints.

[0062] The surface post-processing process generation module 60 is used to assemble the K update process steps in the update process execution sequence, and use the K update process parameter mappings to perform parameter setting of the K update process steps to generate a surface post-processing process.

[0063] The production control module 70 is used to perform batch production control of the metal product to be processed based on the die-casting process steps and the surface post-treatment process after adjusting the parameters of the die-casting process steps using the initial parameter settings.

[0064] The specific configuration of the first sample metal product acquisition module 10 will be described in detail below. As described above, the initial parameter settings of the metal product to be processed are interactively obtained, and the initial parameter settings are used to run the die-casting process steps to carry out trial production of the metal product to obtain the first sample metal product. The first sample metal product acquisition module 10 further includes: a die-casting mold unit, which is used to call the die-casting mold of the metal product to be processed; a parameter setting unit, which is used to use the initial parameter settings to perform parameter settings for the die-casting process steps after the die-casting mold is installed on the die-casting machine, wherein the die-casting process steps include an injection module, a cooling module, a robotic arm and the die-casting machine; a first sample metal product acquisition unit, which is used to run the die-casting process steps with the initial parameter settings to carry out trial production of the metal product to be processed to obtain the first sample metal product.

[0065] Wherein, the die-casting process steps are run with the initial parameter settings to carry out trial production of the metal product to be processed, and the first sample metal product is obtained. The first sample metal product acquisition unit further includes: a die-casting defect image set acquisition subunit, and the die-casting defect image set acquisition subunit is used to interactively obtain multiple sample die-casting defect image sets of various die-casting surface defects; an identification branch construction subunit, and the identification branch construction subunit is used to perform die-casting defect identification on the multiple sample die-casting defect image sets, and after obtaining multiple sample first identification image sets, use the multiple sample die-casting defect image sets and multiple sample first identification image sets to construct multiple die-casting defect identification branches; a post-processing defect image set acquisition subunit, and the post-processing defect image set acquisition subunit is used to interactively obtain multiple sample post-processing surface defects. A post-processing defect image set; a post-processing defect branch subunit, the post-processing defect branch subunit is used to perform post-processing defect identification on the multiple sample post-processing defect image sets, obtain multiple sample second identification image sets, and then use the multiple sample post-processing defect image sets and multiple sample second identification image sets to construct multiple post-processing defect identification branches; a post-processing defect identification channel acquisition subunit, the post-processing defect identification channel acquisition subunit is used to connect the multiple die-casting defect identification branches in parallel to obtain a die-casting defect identification channel, and connect the multiple post-processing defect identification branches in parallel to obtain a post-processing defect identification channel; an apparent defect inspection model construction subunit, the apparent defect inspection model construction subunit is used to connect the die-casting defect identification channel and the post-processing defect identification channel in parallel to complete the construction of the apparent defect inspection model.

[0066] The specific configuration of the first defect detection result acquisition module 20 will be described in detail below. As described above, after collecting the surface image of the first sample metal product and obtaining the first surface image set, the pre-built surface defect inspection model is called to analyze the first surface image set to obtain the first defect detection result. The first defect detection result acquisition module 20 further includes: an image synchronization unit, which is used to synchronize F surface images in the first surface image set to the surface defect inspection model; an image recognition unit, which is used to synchronously perform image recognition via the die-casting defect recognition channel and the post-processing defect recognition channel of the surface defect inspection model to obtain die-casting defect recognition results and post-processing defect recognition results; a surface treatment defect judgment unit, which is used to determine if the die-casting defect recognition result is an empty set and the post-processing defect recognition result is a non-empty set, then the first defect detection result is a surface treatment defect; a die-casting treatment defect judgment unit, which is used to determine if the die-casting defect recognition result is a non-empty set, then the first defect detection result is a die-casting treatment defect.

[0067] Among them, if the die-casting defect identification result is a non-empty set, the first defect detection result is a die-casting processing defect, and the die-casting processing defect judgment unit further includes: an association table traversal subunit, the association table traversal subunit is used to traverse the defect index association table if the die-casting defect identification result is a non-empty set, and locate the M parameter adjustment directions of the M associated indicators; an initial parameter setting update subunit, the initial parameter setting update subunit is used to preset the parameter adjustment step, and use the M parameter adjustment directions and parameter adjustment step as constraints to update the initial parameter setting in multiple rounds until the defect type of the Nth defect detection result is a surface treatment defect, and output the Nth updated parameter setting and the Nth sample metal product; an initial parameter setting replacement subunit, the initial parameter setting replacement subunit is used to replace the initial parameter setting with the Nth updated parameter setting; a demand scheduling update subunit, the demand scheduling update subunit is used to perform process demand scheduling update according to the Nth defect detection result, and obtain H updated process demands and H updated process parameters.

[0068] Among them, a parameter adjustment step is preset, and the M parameter adjustment directions and parameter adjustment step are used as constraints, and the initial parameter setting is updated in a directional manner for multiple rounds until the defect type of the Nth defect detection result is a surface treatment defect, and the Nth updated parameter setting and the Nth sample metal product are output. The initial parameter setting update subunit further includes: a first updated parameter setting acquisition microunit, the first updated parameter setting acquisition microunit is used to directional update the initial parameter setting with the M parameter adjustment directions and parameter adjustment step as constraints to obtain a first updated parameter setting; a trial production result analysis microunit, the trial production result analysis microunit is used to use the first updated parameter setting to run the die-casting process steps to carry out trial production of metal products, call the apparent defect inspection model to analyze the trial production results, and obtain a second defect detection result; a second updated parameter setting acquisition microunit, The second update parameter setting acquisition micro-unit is used for, if the second defect detection result is a die-casting treatment defect, to update the first update parameter setting with the M parameter adjustment directions and parameter adjustment step as constraints to obtain the second update parameter setting; the third defect detection result acquisition micro-unit, the third defect detection result acquisition micro-unit is used to use the second update parameter setting to run the die-casting process steps to carry out trial production of metal products, call the apparent defect inspection model to analyze the trial production results, and obtain the third defect detection result; the sample metal product output micro-unit, the sample metal product output micro-unit is used to, and so on, update the initial parameter setting with the M parameter adjustment directions and parameter adjustment step as constraints, until the defect type of the Nth defect detection result obtained is a surface treatment defect, and the Nth update parameter setting and the Nth sample metal product are output.

[0069] The specific configuration of the scheduling update execution module 30 will be described in detail below. As described above, if the defect type of the first defect detection result is a surface treatment defect, then the process requirement scheduling update is executed according to the first defect detection result to obtain K updated process requirements and K updated process parameters, and the scheduling update execution module 30 further includes: a first defect identification image set extraction unit, the first defect identification image set extraction unit is used to extract the first defect identification image set of the first measured post-processing defect from the post-processing defect recognition result, wherein the post-processing defect recognition result includes K defect identification image sets of K measured post-processing defects; a first measured defect array positioning unit, the first measured defect array positioning unit is used to fit L defect identification images in the first defect identification image set to the metal product model after constructing the metal product model according to the structural design information of the metal product to be processed, and locate the first measured defect array, wherein L is a positive integer less than or equal to F; a first measured image array acquisition micro unit, the first measured image array acquisition micro unit A micro-unit is used to perform vertical image acquisition of the first sample metal product according to the first measured defect array to obtain a first measured image array; a first defect parameter array acquisition micro-unit, the first defect parameter array acquisition micro-unit is used to synchronize the first measured image array to the processing defect identification channel to obtain a first defect parameter array; a first process parameter array acquisition micro-unit, the first process parameter array acquisition micro-unit is used to match the process parameters according to the first defect parameter array after matching the first updated process requirement of the first measured post-processing defect to obtain a first process parameter array; a trajectory optimization unit, the trajectory optimization unit is used to perform trajectory optimization according to the first measured defect array, after locating the first post-processing trajectory, fit the first process parameter array to the first post-processing trajectory to obtain the first updated process parameters; a process data acquisition unit, the process data acquisition unit is used to obtain the K updated process requirements and K updated process parameters by analogy.

[0070] A metal product production control system based on model analysis provided in an embodiment of the present invention can execute a metal product production control method based on model analysis provided in any embodiment of the present invention, and has functional modules and beneficial effects corresponding to the execution method.

[0071] Although the present application makes various references to certain modules in the system according to the embodiments of the present application, any number of different modules may be used and run on the user terminal and / or server, and the various units and modules included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the present invention.

[0072] The above specific implementation manner does not constitute a limitation to the protection scope of the present application. It should be understood by those skilled in the art that various modifications, combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present application should be included in the protection scope of the present application. In some cases, the actions or steps recorded in the present application can be performed in an order different from that in the embodiment and can still achieve the desired results. In addition, the process depicted in the accompanying drawings does 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.

Claims

1. A metal product production control method based on model analysis, characterized in that: The method comprises: Interactively obtaining initial parameter settings of the metal product to be processed, and using the initial parameter settings to run the die-casting process steps to conduct trial production of the metal product to obtain a first sample metal product; After collecting surface images of the first sample metal product to obtain a first surface image set, calling a pre-built surface defect inspection model to analyze the first surface image set to obtain a first defect detection result; If the defect type of the first defect detection result is a surface treatment defect, performing a process requirement scheduling update according to the first defect detection result to obtain K updated process requirements and K updated process parameters; Performing sequence dependency analysis on the K update process requirements to obtain an update process execution sequence; Taking the K update process requirements as scheduling constraints, K update process steps are obtained from the assembly area of ​​the prefabricated process; Assembling the K update process steps in the update process execution sequence, and using the K update process parameter mappings to perform parameter setting of the K update process steps to generate a surface post-processing process; After the initial parameter setting is used to adjust the parameters of the die-casting process step, batch production control of the metal product to be processed is performed based on the die-casting process step and the surface post-treatment process; After collecting surface images of a first sample metal product to obtain a first surface image set, calling a pre-built surface defect inspection model to analyze the first surface image set to obtain a first defect detection result, the method includes: Synchronizing F apparent images in the first apparent image set to the apparent defect inspection model; The die-casting defect recognition channel and the post-processing defect recognition channel of the surface defect inspection model are used to synchronously perform image recognition to obtain a die-casting defect recognition result and a post-processing defect recognition result; If the die-casting defect recognition result is an empty set and the post-processing defect recognition result is a non-empty set, then the first defect detection result is a surface treatment defect; If the die-casting defect identification result is a non-empty set, then the first defect detection result is a die-casting process defect; If the die-casting defect recognition result is a non-empty set, the M types of measured die-casting defects in the die-casting defect recognition result are used to traverse the defect index association number table to locate M parameter adjustment directions of the M types of association indicators; Preset a parameter adjustment step length, and use the M parameter adjustment directions and parameter adjustment step lengths as constraints, update the initial parameter settings in multiple rounds until the defect type of the Nth defect detection result obtained is a surface treatment defect, and output the Nth updated parameter setting and the Nth sample metal product; replacing the initial parameter setting with the Nth updated parameter setting; Performing process requirement scheduling update according to the Nth defect detection result to obtain H updated process requirements and H updated process parameters; Directively updating the initial parameter settings with the M parameter adjustment directions and parameter adjustment step sizes as constraints to obtain a first updated parameter setting; The die-casting process step is run by using the first update parameter setting to conduct trial production of metal products, and the surface defect inspection model is called to analyze the trial production results to obtain a second defect detection result; If the second defect detection result is a die-casting process defect, the first update parameter setting is updated with the M parameter adjustment directions and parameter adjustment step sizes as constraints to obtain a second update parameter setting; The die-casting process step is run by using the second update parameter setting to conduct trial production of metal products, and the surface defect inspection model is called to analyze the trial production results to obtain a third defect detection result; By analogy, the initial parameter setting is updated directionally with the M parameter adjustment directions and parameter adjustment step sizes as constraints, until the defect type of the Nth defect detection result obtained is a surface treatment defect, and the Nth updated parameter setting and the Nth sample metal product are output; If the defect type of the first defect detection result is a surface treatment defect, performing a process requirement scheduling update according to the first defect detection result to obtain K updated process requirements and K updated process parameters, the method comprising: Extracting a first defect identification image set of a first measured post-processing defect from the post-processing defect identification result, wherein the post-processing defect identification result includes K defect identification image sets of K types of measured post-processing defects; After the metal product model is constructed according to the structural design information of the metal product to be processed, L defect identification images in the first defect identification image set are fitted to the metal product model to locate the first measured defect array, where L is a positive integer less than or equal to F; Performing vertical image acquisition on the first sample metal product according to the first measured defect array to obtain a first measured image array; Synchronizing the first measured image array to the processed defect identification channel to obtain a first defect parameter array; After obtaining the first updated process requirement of the first measured post-processing defect through matching, performing process parameter matching according to the first defect parameter array to obtain a first process parameter array; Performing trajectory optimization according to the first measured defect array, after locating a first post-processing trajectory, fitting the first process parameter array to the first post-processing trajectory to obtain first updated process parameters; And so on, the K updated process requirements and K updated process parameters are obtained.

2. A metal product production control method based on model analysis as claimed in claim 1, characterized in that: Interactively obtaining initial parameter settings of a metal product to be processed, and using the initial parameter settings to run a die-casting process step to conduct trial production of the metal product to obtain a first sample metal product, the method comprising: Calling the die-casting mold of the metal product to be processed; After the die-casting mold is installed on the die-casting machine, the initial parameter setting is used to perform parameter setting of the die-casting process step, wherein the die-casting process step includes an injection module, a cooling module, a robot arm and the die-casting machine; The die-casting process steps are run with the initial parameter settings to carry out trial production of the metal product to be processed, and the first sample metal product is obtained.

3. A metal product production control method based on model analysis as claimed in claim 2, characterized in that: The method comprises: Interactively obtain multiple sample die-casting defect image sets of various die-casting surface defects; After performing die-casting defect identification on the plurality of sample die-casting defect image sets to obtain a plurality of sample first identification image sets, a plurality of die-casting defect recognition branches are constructed using the plurality of sample die-casting defect image sets and the plurality of sample first identification image sets; Interactively obtain multiple sample post-processing defect image sets of various post-processing surface defects; After post-processing defect identification is performed on the multiple sample post-processing defect image sets to obtain multiple sample second identification image sets, multiple post-processing defect recognition branches are constructed using the multiple sample post-processing defect image sets and the multiple sample second identification image sets; The plurality of die-casting defect recognition branches are connected in parallel to obtain a die-casting defect recognition channel, and the plurality of post-processing defect recognition branches are connected in parallel to obtain a post-processing defect recognition channel; The die-casting defect recognition channel and the post-processing defect recognition channel are connected in parallel to complete the construction of the surface defect inspection model.

4. A metal product production control system based on model analysis, characterized in that: The system is used to implement a metal product production control method based on model analysis as described in any one of claims 1 to 3, and the system comprises: A first sample metal product acquisition module, the first sample metal product acquisition module is used to interactively obtain initial parameter settings of the metal product to be processed, and use the initial parameter settings to run the die-casting process steps to conduct trial production of the metal product to obtain the first sample metal product; A first defect detection result acquisition module, wherein the first defect detection result acquisition module is used to acquire an apparent image of a first sample metal product and obtain a first apparent image set, and then call a pre-built apparent defect inspection model to analyze the first apparent image set to obtain a first defect detection result; A scheduling update execution module, wherein if the defect type of the first defect detection result is a surface treatment defect, the scheduling update execution module is used to execute a process requirement scheduling update according to the first defect detection result to obtain K updated process requirements and K updated process parameters; A process execution sequence updating module, wherein the process execution sequence updating module is used to perform sequence dependency analysis on the K updated process requirements to obtain an updated process execution sequence; An update process step acquisition module, wherein the update process step acquisition module is used to obtain K update process steps from the assembly area of ​​the prefabricated process by taking the K update process requirements as scheduling constraints; A surface post-processing process generation module, the surface post-processing process generation module is used to assemble the K update process steps in the update process execution sequence, and use the K update process parameter mappings to perform parameter setting of the K update process steps to generate a surface post-processing process; A production control module, wherein the production control module is used to perform parameter setting of the die-casting process steps using the initial parameter setting, and then perform batch production control of the metal product to be processed based on the die-casting process steps and the surface post-treatment process.

Images