Blow molding quality detection method and system based on blow molding mold

Abstract

The present invention discloses a blow molding process quality inspection method and system based on a blow molding mold, belonging to the field of intelligent manufacturing, wherein the method includes: receiving blow molding task information; matching mold models according to the task information, performing mold allocation in combination with the mold models, and obtaining a mold position number set; performing image acquisition and dimension detection based on the mold position number set, and constructing a mold grid model set; performing batch dimension verification on the mold grid model set, and obtaining a dimension deviation grid; obtaining mold molding process information, performing blow molding process analysis in combination with the dimension deviation grid, and generating blow molding product molding defect predictions; and performing blow molding process quality inspection based on the predicted defects. This application solves the technical problems of low blow molding process quality inspection accuracy and efficiency, and high inspection error rate in the prior art, and achieves the technical effects of improving the precision of blow molding process quality inspection, improving inspection efficiency, and reducing inspection error rate.

Description

Technical Field

[0001] The present invention relates to the field of intelligent manufacturing, and in particular to a blow molding process quality detection method and system based on a blow molding mold. Background Art

[0002] With the improvement of automated production and quality control standards in the manufacturing industry, higher requirements are being placed on the accuracy and efficiency of blow molding product quality inspection. However, traditional manual spot check inspection methods cannot meet the needs of modern industrial production for real-time and full-process monitoring of product quality, and their inspection error rate is also difficult to meet the precision requirements of quality control. Summary of the Invention

[0003] This application provides a blow molding process quality detection method and system based on a blow molding mold, aiming to solve the technical problems of low blow molding process quality detection accuracy and detection efficiency, and high detection error rate in the existing technology.

[0004] In view of the above problems, the present application provides a blow molding process quality detection method and system based on a blow molding mold.

[0005] The first aspect disclosed in the present application provides a blow molding processing quality detection method based on a blow molding mold, the method comprising: receiving blow molding task information uploaded by a user end, the blow molding task information including blow molding product model, blow molding quantity information and blow molding time information; matching the parison mold model and the molding mold model according to the blow molding product model; performing mold allocation in combination with the parison mold model and the molding mold model according to the blow molding quantity information and the blow molding time information, and obtaining a parison mold position number set and a molding mold position number set; performing image acquisition and dimension detection based on the parison mold position number set and the molding mold position number set by an image acquisition device and a dimension detection device, and constructing a parison mold grid model set and a molding mold grid model set; performing batch dimension verification on the parison mold grid model set and the molding mold grid model set to obtain a dimension deviation grid; obtaining mold molding process information, performing blow molding processing analysis in combination with the dimension deviation grid, and generating blow molding product molding prediction defects; and performing blow molding processing quality detection based on the blow molding product molding prediction defects.

[0006] Another aspect disclosed in the present application provides a blow molding processing quality detection system based on a blow molding mold, the system comprising: a task information receiving module for receiving blow molding task information uploaded by a user end, the blow molding task information including the blow molding product model, blow molding quantity information and blow molding time information; a mold model matching module for matching the parison mold model and the molding mold model according to the blow molding product model; a mold position number acquisition module for performing mold allocation based on the blow molding quantity information and the blow molding time information in combination with the parison mold model and the molding mold model, and obtaining the parison mold position number set and the molding mold position number set; a mold grid construction module A block is used to perform image acquisition and dimension detection based on the parison mold position number set and the molding mold position number set through an image acquisition device and a dimension detection device, and to construct a parison mold grid model set and a molding mold grid model set; a batch dimension verification module is used to perform batch dimension verification on the parison mold grid model set and the molding mold grid model set to obtain dimension deviation grids; a blow molding processing analysis module is used to obtain mold molding process information, perform blow molding processing analysis in combination with the dimension deviation grid, and generate blow molding product molding defect prediction; a processing quality detection module is used to perform blow molding processing quality detection based on the blow molding product molding defect prediction.

[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0008] The blow molding product model, blow molding quantity information and blow molding time information uploaded by the user end are received, providing product and process information basis for quality inspection; according to the blow molding product model, the mold model is matched to provide mold information for image acquisition and size detection; according to the blow molding quantity information and blow molding time information, the mold model is combined to perform mold allocation, obtain the mold position information that needs to be detected, and provide an operation basis for image acquisition and size detection; through the image acquisition device and the size detection device, image acquisition and size detection are performed based on the mold position set, and a mold grid model set is constructed to provide data support for quality prediction and detection; the mold grid is The model set performs batch dimension verification, obtains dimension deviation grid, detects mold dimension accuracy, and provides a basis for quality prediction; obtains mold molding process information, combines the dimension deviation grid to perform blow molding processing analysis, generates blow molding product molding prediction defects, realizes predictive quality inspection, and improves inspection efficiency; performs blow molding processing quality inspection based on blow molding product molding prediction defects, realizes automated inspection technical solutions, solves the technical problems of low blow molding processing quality inspection accuracy and inspection efficiency, and high inspection error rate in the existing technology, and achieves the technical effect of improving the precision of blow molding processing quality inspection, improving inspection efficiency, and reducing inspection error rate.

[0009] 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. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1 A possible flow chart of a blow molding process quality inspection method based on a blow molding mold is provided for the embodiment of the present application;

[0011] Figure 2 A schematic diagram of a possible process for obtaining a parison mold grid model set and a forming mold grid model set in a blow molding process quality detection method based on a blow mold is provided for an embodiment of the present application;

[0012] Figure 3 A schematic diagram of a possible flow chart for setting a size deviation grid in a blow molding process quality inspection method based on a blow mold is provided for an embodiment of the present application;

[0013] Figure 4 A possible structural schematic diagram of a blow molding process quality detection system based on a blow molding mold is provided for the embodiment of the present application.

[0014] Description of the accompanying drawings: task information receiving module 11, mold model matching module 12, mold position number acquisition module 13, mold grid construction module 14, batch size verification module 15, blow molding process analysis module 16, processing quality detection module 17. DETAILED DESCRIPTION

[0015] The overall idea of ​​the technical solution provided by this application is as follows:

[0016] The embodiment of the present application provides a blow molding processing quality detection method and system based on blow molding molds. First, the blow molding production information uploaded by the user is received, including product and process data; second, the corresponding mold is matched according to the product data, and the mold to be tested is allocated according to the output and process; then, the allocated mold is imaged and dimensionally detected, a high-precision digital mold model is constructed, and the model is dimensionally calibrated to detect the mold accuracy; then, the mold accuracy detection results and process parameters are combined to predict the quality changes and potential defects of the product during the molding process; finally, the product quality is automatically detected based on the quality prediction results, realizing fully automatic and precise quality detection, thereby achieving the technical effect of improving the accuracy of quality inspection, improving detection efficiency, and reducing the detection error rate.

[0017] After introducing the basic principles of the present application, various non-limiting implementation methods of the present application will be specifically introduced in conjunction with the drawings in the specification.

[0018] Example 1

[0019] like Figure 1 As shown, the embodiment of the present application provides a blow molding process quality detection method based on a blow molding mold. The method is applied to a blow molding process quality detection system based on a blow molding mold, including a server, an image acquisition device, and a size detection device in communication connection.

[0020] Specifically, the server is a computing device used to receive blow molding task information uploaded by users and store and process related data. The image acquisition device, such as an industrial lens or probe, is used to capture images of the blow mold. The dimension detection device, such as a 3D scanner or lidar, is used to obtain the dimensional parameters of the blow mold. Both the image acquisition device and dimension detection device communicate with the server via wired or wireless communication.

[0021] Blow molding quality inspection methods include:

[0022] Step S1000: receiving blow molding task information uploaded by a user terminal, wherein the blow molding task information includes blow molding product model, blow molding quantity information, and blow molding time information;

[0023] Specifically, the server receives blow molding task information uploaded by the user through the human-computer interaction interface, including blow molding product model, blow molding quantity information, and blow molding duration information. The blow molding product model refers to the product model of the blow molding product specified by the user, which is used by the server to match the corresponding parison mold model and forming mold model; the blow molding quantity information refers to the number of blow molding products specified by the user, which is used by the server to allocate molds and determine the number of molds required; the blow molding duration information refers to the blow molding process duration parameter specified by the user, which is also used by the server to allocate molds and determine the required molds.

[0024] First, the server provides an entry point for receiving blow molding task information through a human-computer interaction interface. Users upload blow molding tasks by entering information such as product model, product quantity, and blow molding duration. After receiving the uploaded blow molding task information, the server parses and extracts the information to obtain information such as the blow molding product model, blow molding quantity, and blow molding duration, thus enabling the blow molding process quality inspection system to be launched and operated.

[0025] Through the interaction between the server and the user, the blow molding task information uploaded by the user is received, providing the necessary input basis for the operation of subsequent steps, which helps the server to perform operations such as mold matching and allocation.

[0026] Step S2000: matching the parison mold model and the forming mold model according to the blow molded product model;

[0027] Specifically, the server matches the received blow molded product model with the corresponding parison mold model and forming mold model in the mold database. The parison mold model refers to the mold model used to form the parison of the blow molded product; the forming mold model refers to the mold model used to further form the parison into the final product.

[0028] The server stores parison mold models and forming mold models corresponding to various blow molding products in its mold database. Upon receiving a blow molding product model, the server searches the database for all parison mold models and forming mold models corresponding to that product model. It then selects a parison mold model and a forming mold model based on a priority algorithm, or provides multiple models for the user to choose from. The selected parison mold model and forming mold model are stored as the mold type parameters for this blow molding task.

[0029] By matching the corresponding parison mold model and molding mold model according to the blow molding product model, a basis is provided for subsequent mold allocation and detection analysis, ensuring the accuracy of mold allocation and detection.

[0030] Step S3000: performing mold allocation according to the blow molding quantity information and the blow molding time information, in combination with the parison mold model and the forming mold model, to obtain a parison mold bit number set and a forming mold bit number set;

[0031] Specifically, based on the received blow molding quantity and duration information, as well as the matched parison mold and forming mold models, the server selects a set of available molds from the mold library, determines the set of molds to be used, and obtains the parison mold bit number set and the forming mold bit number set. The parison mold bit number set uniquely identifies a set of parison molds selected from multiple parison molds based on the blow molding task and mold information; the forming mold bit number set uniquely identifies a set of forming molds selected from multiple forming molds based on the blow molding task and mold information. Each bit number corresponds to a parison mold and contains the corresponding mold spatial location and unique mold identification code.

[0032] The server calculates the total number of molds required based on the number of products, process duration, and mold model. It then queries the mold library, selects currently available molds, determines the spatial locations of matching molds, and generates a specific mold set to be used. The server then extracts the unique identifiers of these molds to form a parison mold number set and a forming mold number set. These two number sets contain information about all the molds required for this blow molding task.

[0033] By selecting a group of idle molds from multiple molds based on the blow molding task information and the matched mold model, and determining the parison mold position number set and the molding mold position number set, it is beneficial to quickly and accurately locate the mold, provide object information for subsequent image acquisition and dimension detection, and help improve detection efficiency and accuracy.

[0034] Step S4000: performing image acquisition and dimension detection based on the parison mold bit number set and the forming mold bit number set by an image acquisition device and a dimension detection device, and constructing a parison mold mesh model set and a forming mold mesh model set;

[0035] Specifically, the image acquisition device and the dimension detection device perform image acquisition and dimension detection based on the obtained parison mold position number set and the molding mold position number set, and construct the parison mold mesh model set and the molding mold mesh model set. The image acquisition device is an image acquisition device used to acquire images of the blow mold and is connected to the server for communication; the dimension detection device is a detection device used to obtain the blow mold dimensional parameters and is connected to the server for communication; the parison mold mesh model set is a set of three-dimensional parison mold mesh models constructed after the parison mold position number set is detected; and the molding mold mesh model set is a set of three-dimensional molding mold mesh models constructed after the molding mold position number set is detected.

[0036] The server sends the parison mold and forming mold location number sets to the image acquisition device and dimension detection device. The image acquisition device and dimension detection device locate the physical mold based on the received location number sets, perform image acquisition and dimension detection on the physical mold, and obtain mold image information and dimension parameters. The server then generates three-dimensional point cloud data of the mold's image and dimensions based on the acquired image information and dimension parameters. The server then aligns the obtained three-dimensional point cloud data, aligning the image point cloud data and dimension point cloud data in the same coordinate system to construct a three-dimensional mesh model, generating the parison mold mesh model set and the forming mold mesh model set.

[0037] By performing image acquisition and dimension detection based on a determined mold position number set, a parison mold mesh model set and a forming mold mesh model set are constructed, providing detection results for subsequent dimension verification and detection analysis, which is conducive to improving detection accuracy and the degree of detection automation.

[0038] Step S5000: performing batch size verification on the parison mold mesh model set and the forming mold mesh model set to obtain size deviation grids;

[0039] Specifically, the server performs batch dimensional verification on the generated parison mold mesh model set and the forming mold mesh model set to obtain dimensional deviation meshes. Batch dimensional verification involves the server automatically using a predefined detection algorithm to detect the dimensional parameters of each 3D mesh model in the received parison mold mesh model set and the forming mold mesh model set, comparing them with the theoretical design parameters to identify mesh areas with dimensional deviations. Dimensional deviation meshes are mesh model areas in the parison mold mesh model set and the forming mold mesh model set that contain dimensional errors, as discovered through dimensional verification.

[0040] First, after the server generates the parison mold mesh model set and the forming mold mesh model set, it automatically applies a dimensional inspection algorithm to each 3D mesh model. The inspection algorithm segments the model into a grid, extracts key dimensional parameters from each mesh within the 3D mesh model, and compares these parameters with the theoretical design parameters for the mold product type. If a parameter is found to be outside the allowable range, the corresponding mesh area is considered to have a dimensional deviation. A dimensional deviation mesh is generated, which contains information about the mold's dimensional issues and their specific locations.

[0041] By performing systematic size deviation detection on the two sets of generated 3D mesh model sets, the mesh areas with size problems are accurately found, and size deviation meshes are generated to provide basic data for subsequent quality inspection and analysis, thereby improving the efficiency and accuracy of quality inspection.

[0042] Step S6000: Acquire mold forming process information, perform blow molding process analysis based on the dimensional deviation grid, and generate blow molding product molding prediction defects;

[0043] Specifically, the server obtains mold forming process information, combines it with the obtained dimensional deviation grid, performs blow molding simulation analysis, and generates a blow molding product molding prediction defect report. Mold forming process information refers to the process parameter information used for blow molding of this model of mold, including process temperature, pressure curve, feed rate, and other parameters; blow molding process analysis refers to the use of blow molding simulation software to simulate and analyze the entire blow molding process based on the obtained mold forming process information and dimensional deviation grid to identify possible product defects; blow molding product molding prediction defects refer to the prediction of quality problems or defects that may occur in the final product through blow molding process analysis.

[0044] First, the server imports the acquired mold forming process information into blow molding simulation software, then builds a 3D model of the mold model within the software environment. Next, the server marks the detected dimensional deviation grids at the corresponding locations on the 3D model, configures the process parameters, and initiates the simulation. The simulation software simulates the actual blow molding process, and any product defects caused by mold dimensional deviations are displayed in the software. The server then analyzes the software results, identifies all potential product defects, and generates predicted defects for the blow molded product.

[0045] By simulating and analyzing the entire blow molding process based on the acquired mold forming process information and dimensional deviation grid, possible product defects are predicted based on specific mold and process information, improving the accuracy and efficiency of inspections, and increasing the detection rate of blow molding product quality inspections.

[0046] Step S7000: performing blow molding process quality inspection based on the predicted molding defects of the blow molded product.

[0047] Specifically, the server performs quality inspections on the actual blow molding process or finished products based on the generated blow molding product molding defect predictions. Blow molding quality inspections refer to quality analysis and inspections of key processes in the actual blow molding process or finished products.

[0048] First, the server uses the generated predicted blow molding defects as input for blow molding quality testing. Then, using video or sensors, the server monitors key blow molding processes in real time, such as profile preheating and product molding. This allows it to specifically detect quality issues contained in the predicted defects and promptly alert users to make operational adjustments or improvements. Furthermore, after product molding, the server can perform offline quality testing on the finished product based on product quality testing standards and the predicted blow molding defects.

[0049] By conducting quality inspections on the actual blow molding process and products based on simulation prediction results, the simulation prediction results are a set of possible defects generated for the specifically selected mold and process information. This set can be used to more accurately detect quality problems in blow molding products, avoiding the subjectivity of manual inspection and achieving the technical effect of improving the precision of blow molding quality inspections, increasing inspection efficiency, and reducing inspection error rates.

[0050] Furthermore, the embodiment of the present application also includes:

[0051] Step S3100: acquiring the number of parallel molds according to the blow molding quantity information and the blow molding time information;

[0052] Step S3200: determining a first idle parison mold position number set and a first idle forming mold position number set according to the parison mold model and the forming mold model;

[0053] Step S3300: traversing the first idle parison mold position number set and the first idle forming mold position number set to perform fault self-diagnosis and troubleshooting, and obtaining a second idle parison mold position number set and a second idle forming mold position number set;

[0054] Step S3400: determining whether the second idle parison mold position number set or the second idle forming mold position number set meets the number of parallel molds;

[0055] Step S3500: If not satisfied, the second idle parison mold bit number set and the second idle forming mold bit number set are set as the parison mold bit number set and the forming mold bit number set;

[0056] Step S3600: If satisfied, randomly select molds that meet the number of parallel molds from the second idle parison mold position number set and the second idle molding mold position number set to obtain the parison mold position number set and the molding mold position number set.

[0057] Specifically, the server first receives the blow molding quantity and blow molding time information uploaded by the user and calculates the number of parallel molds. The number of parallel molds is calculated by dividing the total blow molding time required to complete the total number of blow moldings by the expected blow molding time. This is the number of parallel molds, that is, the number of parison molds and forming molds that need to be used simultaneously.

[0058] At the same time, based on the parison mold model and forming mold model specified by the customer, the mold library is searched to find currently idle molds of that model, obtaining the first idle parison mold bit number set and the first idle forming mold bit number set. The first idle parison mold bit number set refers to the unique identifier set of currently idle parison molds of the given model, based on the status of the mold library at a certain point in time; the first idle forming mold bit number set refers to the unique identifier set of currently idle forming molds of the given model, based on the status of the mold library at a certain point in time.

[0059] The server then performs a fault self-diagnosis procedure on each mold in the first idle parison mold tag set and the first idle forming mold tag set. This procedure performs a series of operational tests on the molds to verify their proper function. Molds that pass these tests are retained in the tag set; molds that fail these tests are removed from the tag set. After the fault self-diagnosis, the second idle parison mold tag set and the second idle forming mold tag set are obtained.

[0060] Next, the server determines whether the number of molds in the second idle parison mold number set and the second idle forming mold number set meet the parallel mold quantity requirement. If not, indicating that the number of selected molds that meet the requirements for the specified blow molding production process is insufficient and all selected models need to be used for this production, the second idle parison mold number set and the second idle forming mold number set are directly set as the final parison mold number set and forming mold number set to be used. If they meet the requirements, indicating that the number of selected molds that meet the requirements for the specified blow molding production process is sufficient and not all need to be put into use, the required number of molds are randomly selected from the second idle parison mold number set and the second idle forming mold number set to form the final parison mold number set and forming mold number set to be used.

[0061] By finding idle molds based on production and mold information and performing fault detection, the quality and efficiency of blow molding production can be improved and the detection error rate can be reduced. The quantity required for production can be determined based on the number and duration of blow molding processes, thereby improving mold utilization and maximizing output.

[0062] Further, such as Figure 2 As shown, the embodiment of the present application also includes:

[0063] Step S4100: using the image acquisition device and the dimension detection device, traverse the parison mold number set and the forming mold number set to perform mold inner wall image acquisition and dimension detection, and construct a parison mold three-dimensional simulation atlas and a forming mold three-dimensional simulation atlas;

[0064] Step S4200: setting the reference grid size and virtual space coordinate system;

[0065] Step S4300: gridding the parison mold 3D simulation atlas and the forming mold 3D simulation atlas according to the reference grid size, positioning and marking the gridding results based on the virtual space coordinate system, and obtaining the parison mold grid model set and the forming mold grid model set.

[0066] Specifically, the server performs image acquisition and dimension detection on each mold in the parison mold position number set and the molding mold position number set through an image acquisition device and a dimension detection device, constructs a three-dimensional simulation atlas of the parison mold and a three-dimensional simulation atlas of the molding mold, and further converts the two sets of three-dimensional simulation atlases into a parison mold mesh model set and a molding mold mesh model set.

[0067] Among them, the three-dimensional simulation atlas of the preform mold refers to a collection of three-dimensional image data of the inner and outer surfaces of the preform mold obtained by an image acquisition device and a dimension detection device; the three-dimensional simulation atlas of the forming mold refers to a collection of three-dimensional image data of the inner and outer surfaces of the forming mold obtained by an image acquisition device and a dimension detection device; the reference grid size refers to the basic unit volume set by the server for discretizing and gridding the three-dimensional data; the virtual space coordinate system refers to the reference coordinate system constructed by the server in three-dimensional space for locating and identifying a single grid model.

[0068] First, the server controls the image acquisition device and dimension detection device to scan the inner surface of each mold in the parison mold number set and the forming mold number set, obtaining a 3D simulation atlas of the parison mold and the forming mold. Then, the reference grid size and virtual space coordinate system are set, and the two 3D simulation atlases are discretized into a large number of small grids, each with unique spatial coordinates, to form the parison mold mesh model set and the forming mold mesh model set.

[0069] Through automatic scanning and 3D reconstruction of physical molds, a large number of accurate 3D mesh models are obtained, providing basic data for dimensional inspection and quality analysis. This replaces the traditional manual measurement and entry method, greatly improving inspection efficiency, reducing human errors, and improving subsequent inspection accuracy.

[0070] Furthermore, the embodiment of the present application also includes:

[0071] Step S5100: traverse the parison mold number set and the forming mold number set to obtain reference processing positioning information;

[0072] Step S5200: performing positioning pre-control on the parison mold grid model set and the forming mold grid model set to obtain processing pre-control positioning information;

[0073] Step S5300: performing position verification on the processing pre-control positioning information based on the reference processing positioning information to obtain a positioning verification result;

[0074] Step S5400: When the positioning verification result fails, positioning deviation information is generated and the size deviation grid is set.

[0075] Specifically, the server performs batch size verification on the parison mold mesh model set and the molding mold mesh model set to check whether the spatial positioning of each set of mesh models in the mold is accurate. If there is a positioning deviation, a positioning deviation grid will be generated and identified.

[0076] Among them, the benchmark processing positioning information refers to the standard spatial coordinate value set by the server for the feature position in each mold based on the virtual space coordinate system during the mold processing and inspection process; the processing pre-control positioning information refers to the server finding the corresponding feature position of each grid model in the mold through the feature extraction algorithm during the grid model construction process, and obtaining the spatial coordinate values ​​of these positions; the positioning verification result refers to the judgment result obtained by the server by comparing the processing pre-control positioning information with the benchmark processing positioning information, and judging whether the processing pre-control positioning information is within the allowable error range of the benchmark processing positioning information or exceeds the allowable error range; the positioning deviation information refers to the information generated by the server describing the direction, size and position of the positioning deviation when the positioning verification result exceeds the allowable error range.

[0077] First, during the mold design and processing phase, the server sets the baseline machining positioning information for each key feature position within the mold. Then, after the 3D mesh model is constructed, a feature extraction algorithm is used to analyze each mesh model, identify the corresponding feature positions, and obtain the machining pre-control positioning information. Next, the server compares and verifies the machining pre-control positioning information with the baseline machining positioning information to determine whether the pre-control positioning information is within the allowable error range. If it is within the allowable error range, the spatial positioning is accurate; if it is outside the allowable error range, positioning deviation will occur, generating positioning deviation information. The grid cells with positioning deviation will be identified in the 3D mesh model set to form a size deviation grid.

[0078] By automatically testing the positioning accuracy of batch 3D mesh models in space and comparing them with the set benchmark processing positioning information, the detection accuracy is improved, the errors caused by manual measurement are avoided, and accurate positioning information is provided for defect prediction, thereby improving detection accuracy and efficiency.

[0079] Further, such as Figure 3 As shown, the embodiment of the present application also includes:

[0080] Step S5410: When the positioning verification result fails, the image set of the molding die cavity surface acquired by the image acquisition device is retrieved;

[0081] Step S5420: traversing the molding die cavity surface image set and the molding die mesh model set to perform mesh concave-convex positioning and obtain a mesh concave-convex feature set;

[0082] Step S5430: locating abnormal grids according to the grid concave-convex feature set to obtain coarse grid information;

[0083] Step S5440: setting the size deviation grid according to the coarse grid information and the positioning deviation information.

[0084] Specifically, if the positioning verification result fails, the server will further analyze the cause of the positioning deviation to determine whether it is caused by the poor surface condition of the mold inner wall. If so, the relevant dimensional deviation grid will be identified.

[0085] Among them, the molding mold cavity surface image set refers to a collection of multiple image data of the molding mold inner wall surface collected by the server through the image acquisition device; the grid concave-convex feature set refers to a set of characteristic parameters describing the concave-convex state of the grid model in space extracted by the server by comparing the molding mold cavity surface image set and the molding mold grid model set to determine whether the molding mold inner wall surface corresponding to each grid model is flat; abnormal grid positioning refers to the positioning of the grid with surface roughness or defects in the grid model when the concave-convex state of the grid model corresponding to the molding mold inner wall surface is determined to be beyond the normal range; rough grid information is the spatial distribution information describing the abnormality of the grid model formed based on the located abnormal grid.

[0086] If the positioning verification result fails, the server first retrieves the inner wall surface image of the forming mold captured by the image acquisition device to form a forming mold cavity surface image set. Then, the forming mold cavity surface image set and the forming mold mesh model set are matched and compared to determine whether the inner wall surface area corresponding to each mesh model in the image is flat and smooth, and the mesh concave-convex feature set is extracted. If the inner wall surface area corresponding to some mesh models is found to be too rough or has obvious depressions or protrusions, the inner wall state of the spatial area where these mesh models are located will be determined to be abnormal, and rough mesh information will be generated. Finally, the rough mesh information and positioning deviation information are combined to identify the mesh models corresponding to the spatial range described by the two sets of information from the forming mold mesh model set to form a size deviation mesh.

[0087] By analyzing and evaluating the inner wall condition of the molding die, if the positioning deviation is caused by the inner wall surface condition, the relevant spatial range can be accurately extracted and the corresponding dimensional deviation grid can be identified, providing practical detection and positioning information, avoiding subjective errors caused by manual judgment, and improving the accuracy and efficiency of detection.

[0088] Furthermore, the embodiment of the present application also includes:

[0089] Step S5431: Acquire a cavity surface conceptual mesh model, wherein the cavity surface conceptual mesh model is a desired cavity surface model;

[0090] Step S5432: extracting the corner point change mesh area based on the cavity surface conceptual mesh model;

[0091] Step S5433: Acquire a first adjacent grid area and a second adjacent grid area of ​​the corner point change grid area;

[0092] Step S5434: Obtain a first adjacent tangent line set between the first adjacent mesh area and the corner point change mesh area;

[0093] Step S5435: Obtain a second adjacent tangent line set between the second adjacent mesh area and the corner point change mesh area;

[0094] Step S5436: Calculating a set of tangent angles between the first adjacent tangent set and the second adjacent tangent set;

[0095] Step S5437: When the deviation mean of the tangent angle set is less than or equal to the deviation threshold, and the tangent angle mean meets the preset angle range, dividing the first adjacent grid area and the second adjacent grid area into different areas;

[0096] Step S5438: Repeat the division to obtain the grid area division result;

[0097] Step S5439: traverse the grid area division results to locate abnormal grids and obtain the coarse grid information.

[0098] Specifically, by analyzing the grid concave-convex feature set, regional division and abnormal grid positioning technology are used to determine which spatial ranges on the inner wall surface of the molding die are abnormally rough, and rough grid information describing these abnormal spatial ranges is generated.

[0099] First, the server generates a three-dimensional mesh model of the ideal inner wall surface of the forming mold as a conceptual mesh model of the cavity surface. The conceptual mesh model of the cavity surface refers to the ideal three-dimensional mesh model of the inner wall surface of the forming mold generated by the server based on the theoretical inner wall contour design of the forming mold, which describes the desired three-dimensional shape of the inner wall surface of the forming mold. Then, based on the conceptual mesh model of the cavity surface, the area where the plane changes is extracted as the corner point change mesh area. Next, the two mesh areas adjacent to the corner point change mesh area are analyzed and determined to be the first adjacent mesh area and the second adjacent mesh area. For example, two adjacent corner points of a certain mold are connected to generate a verification line. The deviation of the verification line from other consecutive corner points is determined. If the deviation of the distance between the verification line and the other consecutive corner points is less than the preset deviation, the line containing the corner point is regarded as the critical area of ​​the mold shape change, and the corner point change mesh area is obtained. The two areas adjacent to the corner point change mesh area are respectively the first adjacent mesh area and the second adjacent mesh area.

[0100] Then calculate the tangent formed by the contact boundary of the first adjacent grid area and the second adjacent grid area to obtain the first adjacent tangent set and the second adjacent tangent set. Calculate the angle between the intersection points of the two sets of tangents to obtain the tangent angle set, which describes the relative inclination of the two sets of tangents. If the statistical parameters of the tangent angle set are within the preset range, it means that the two tangent sets should belong to the same straight line, and the first adjacent grid area and the second adjacent grid area are divided; otherwise, no division is required. Repeat this process until all sub-areas are verified, and divide the inner wall of the mold according to the boundary of the corner point change, so as to obtain multiple areas, each area is relatively in the same plane, and obtain the grid area division result.

[0101] After obtaining the final mesh region division results, the server sets a window size within each subregion and, with each mesh as the center, determines the difference in concavity and convexity between the other meshes within the window and the central mesh model. If the difference exceeds the predetermined range, the inner wall surface corresponding to the mesh model is determined to be abnormally rough, and its spatial information is extracted to form rough mesh information.

[0102] By adopting regional division and tangent angle judgment to carry out refined detection and evaluation of the inner wall surface state of the molding mold, the precise spatial range of abnormally rough inner wall surface is accurately located, high-precision rough grid information is obtained, and reliable detection and positioning information is provided, laying a foundation for improving the detection accuracy of blow molding processing quality.

[0103] Furthermore, the embodiment of the present application also includes:

[0104] Step S6100: collecting mold forming process record data, dimension deviation grid identification data, and blow molded product forming defect record data;

[0105] Step S6200: using the blow molded product molding defect record data as supervision data, the mold molding process record data and the dimensional deviation grid identification data as input data, and training a first pre-learner based on a BP neural network;

[0106] Step S6300: using the blow molded product molding defect record data as supervision data, the mold molding process record data and the dimensional deviation grid identification data as input data, and training a second pre-learner based on random forest;

[0107] Step S6400: using the first output data of the first pre-learner and the second output data of the second pre-learner as input data and the blow molded product molding defect record data as supervision data to train a feature fusion learner;

[0108] Step S6500: merging the first output layer of the first pre-learner and the first output layer of the second pre-learner with the input layer of the feature fusion learner to obtain a blow molding process analysis model;

[0109] Step S6600: According to the blow molding process analysis model, the mold forming process information is obtained, and blow molding process analysis is performed in combination with the dimensional deviation grid to generate the blow molding product molding prediction defect.

[0110] Specifically, a machine learning approach was used to build a blow molding analysis model by training on historical mold forming process information, dimensional deviation grid data, and blow molded product molding defect records. Current mold forming process information and dimensional deviation grid data were then input into the analysis model to predict potential molding defects during the current blow molding process, providing decision support for quality control.

[0111] First, historical mold forming process information, dimensional deviation grid data, and blow molding product molding defect record data are collected. The mold forming process record data refers to mold-related process parameter information during the blow molding process, such as exhaust settings, molding wall temperature, and pressure curve data, which describes the mold's working state during the molding process. The dimensional deviation grid identification data describes the spatial distribution of the dimensional deviation grid. The blow molding product molding defect record data describes various molding defects generated during the historical blow molding process, such as extrusion marks and flash.

[0112] Then, two machine learning algorithms, BP neural network and random forest, were used to train the first and second pre-learners, respectively. The first pre-learner is the first machine learning model trained on the server using the BP neural network algorithm. This model takes as input the mold forming process record data and dimensional deviation grid identification data, and outputs first output data, which is used to describe the molding defects that may occur under the molding conditions. The second pre-learner is the second machine learning model trained on the server using the random forest algorithm. This model takes as input the same pre-learner as the first pre-learner and outputs second output data, which is used to describe the molding defects that may occur under the molding conditions.

[0113] Next, the third machine learning algorithm is used to input the output data of the first and second pre-learners, and the blow molding product molding defect record data is used as the supervision data to train the feature fusion learner. The feature fusion learner is a machine learning model that can combine the outputs of the two models and more accurately predict molding defects by using the third machine learning algorithm on the server side. The input and output layers of the three learners are then connected to construct a blow molding process analysis model. During the analysis phase, the server side inputs the current mold molding process information and dimensional deviation grid data into the blow molding process analysis model. The model integrates the prediction results of the three learners, outputs various possible molding defects and their probabilities under the current molding conditions, and generates predicted molding defects for blow molding products.

[0114] By predicting and evaluating the blow molding process quality, and analyzing the possible molding defects based on the current working status and existing problems of the mold, we can provide important decision-making basis for quality control, improve the precision of blow molding process quality inspection, improve detection efficiency, and reduce detection error rate.

[0115] In summary, the blow molding process quality inspection method based on the blow molding mold provided in the embodiments of the present application has the following technical effects:

[0116] Receive blow molding task information uploaded by the user end, the blow molding task information includes blow molding product model, blow molding quantity information and blow molding time information, and provide product and process information basis for quality inspection; match the parison mold model and the molding mold model according to the blow molding product model, and provide mold position information for image acquisition and size detection; according to the blow molding quantity information and blow molding time information, combine the parison mold model and the molding mold model to perform mold allocation, obtain the parison mold position number set and the molding mold position number set, perform mold allocation according to the output and process information, obtain the mold position information that needs to be detected, and provide an operational basis for image acquisition and size detection; through the image acquisition device and the size detection device, based on the parison mold position number The image acquisition and dimension detection of the set and the molding mold position number set are carried out, and the parison mold grid model set and the molding mold grid model set are constructed to provide data support for quality prediction and detection; batch dimension verification is carried out on the parison mold grid model set and the molding mold grid model set, and the dimension deviation grid is obtained. The dimension verification of the mold model is carried out to detect the mold dimension accuracy to provide a basis for quality prediction; the mold molding process information is obtained, and the blow molding processing analysis is carried out in combination with the dimension deviation grid to generate blow molding product molding prediction defects, realize the predictive quality detection, and improve the detection efficiency; the blow molding processing quality detection is carried out based on the blow molding product molding prediction defects, realize automated detection, reduce manual detection errors, and improve detection accuracy.

[0117] Example 2

[0118] Based on the same inventive concept as the blow molding quality detection method based on the blow molding mold in the above embodiment, Figure 4 As shown, the embodiment of the present application provides a blow molding process quality detection system based on a blow mold, which includes a server, an image acquisition device, and a size detection device in communication connection, including:

[0119] The task information receiving module 11 is used to receive the blow molding task information uploaded by the user terminal, wherein the blow molding task information includes the blow molding product model, blow molding quantity information and blow molding time information;

[0120] A mold model matching module 12 is used to match the parison mold model and the molding mold model according to the blow molded product model;

[0121] A mold number acquisition module 13 is configured to allocate molds based on the blow molding quantity information and the blow molding time information, in combination with the parison mold model and the forming mold model, and acquire a parison mold number set and a forming mold number set;

[0122] A mold mesh construction module 14 is configured to perform image acquisition and size detection based on the parison mold bit number set and the forming mold bit number set through an image acquisition device and a size detection device, and to construct a parison mold mesh model set and a forming mold mesh model set;

[0123] A batch size verification module 15 is used to perform batch size verification on the parison mold grid model set and the forming mold grid model set to obtain size deviation grids;

[0124] a blow molding process analysis module 16 for acquiring mold forming process information, performing blow molding process analysis in combination with the dimensional deviation grid, and generating blow molding product molding defect predictions;

[0125] The processing quality detection module 17 performs blow molding processing quality detection based on the blow molding product molding prediction defects.

[0126] Furthermore, the mold number obtaining module 13 includes the following execution steps:

[0127] Acquire the number of parallel molds according to the blow molding quantity information and the blow molding time information;

[0128] Determining a first idle parison mold position number set and a first idle forming mold position number set according to the parison mold model and the forming mold model;

[0129] Traversing the first idle preform mold position number set and the first idle forming mold position number set to perform fault self-diagnosis and troubleshooting, and obtaining a second idle preform mold position number set and a second idle forming mold position number set;

[0130] Based on whether the second idle parison mold bit number set or the second idle forming mold bit number set meets the number of parallel molds;

[0131] If not, the second idle parison mold bit number set and the second idle forming mold bit number set are set as the parison mold bit number set and the forming mold bit number set;

[0132] If so, randomly select molds that meet the number of parallel molds from the second idle parison mold bit number set and the second idle forming mold bit number set to obtain the parison mold bit number set and the forming mold bit number set.

[0133] Furthermore, the mold grid construction module 14 includes the following execution steps:

[0134] By using the image acquisition device and the size detection device, the parison mold position number set and the forming mold position number set are traversed to perform mold inner wall image acquisition and size detection, and a parison mold three-dimensional simulation atlas and a forming mold three-dimensional simulation atlas are constructed;

[0135] Set the base grid size and virtual space coordinate system;

[0136] The parison mold three-dimensional simulation atlas and the forming mold three-dimensional simulation atlas are gridded according to the reference grid size, and the gridded results are positioned and marked based on the virtual space coordinate system to obtain the parison mold grid model set and the forming mold grid model set.

[0137] Furthermore, the batch size verification module 15 includes the following execution steps:

[0138] Traversing the parison mold position number set and the forming mold position number set to obtain reference processing positioning information;

[0139] Performing positioning pre-control on the parison mold grid model set and the forming mold grid model set to obtain processing pre-control positioning information;

[0140] Performing position verification of the machining pre-control positioning information based on the reference machining positioning information to obtain a positioning verification result;

[0141] When the positioning verification result fails, positioning deviation information is generated and the size deviation grid is set.

[0142] Furthermore, the batch size verification module 15 further includes the following execution steps:

[0143] When the positioning verification result fails, retrieving a set of images of the forming mold cavity surface acquired by the image acquisition device;

[0144] Traversing the molding die cavity surface image set and the molding die mesh model set to perform mesh concave-convex positioning and obtain a mesh concave-convex feature set;

[0145] Positioning abnormal grids according to the grid concave-convex feature set to obtain coarse grid information;

[0146] The size deviation grid is set according to the coarse grid information and the positioning deviation information.

[0147] Furthermore, the batch size verification module 15 further includes the following execution steps:

[0148] Acquire a cavity surface conceptual mesh model, wherein the cavity surface conceptual mesh model is a desired cavity surface model;

[0149] Extracting a corner point change mesh area based on the cavity surface conceptual mesh model;

[0150] Acquire a first adjacent grid area and a second adjacent grid area of ​​the corner point change grid area;

[0151] Acquire a first adjacent tangent line set between the first adjacent grid area and the corner point change grid area;

[0152] Acquire a second adjacent tangent line set between the second adjacent grid area and the corner point change grid area;

[0153] Calculating a tangent angle set between the first adjacent tangent line set and the second adjacent tangent line set;

[0154] When the deviation mean of the tangent angle set is less than or equal to a deviation threshold, and the tangent angle mean meets a preset angle range, dividing the first adjacent grid area and the second adjacent grid area into different areas;

[0155] Repeat the division to obtain the grid area division results;

[0156] Traverse the grid area division result to locate abnormal grids and obtain the coarse grid information.

[0157] Furthermore, the blow molding process analysis module 16 includes the following execution steps:

[0158] Collect mold forming process record data, size deviation grid identification data and blow molding product molding defect record data;

[0159] Using the blow molded product molding defect record data as supervision data, the mold molding process record data and the dimensional deviation grid identification data as input data, and training a first pre-learner based on a BP neural network;

[0160] Using the blow molded product molding defect record data as supervision data, and the mold molding process record data and the dimensional deviation grid identification data as input data, a second pre-learner is trained based on random forest;

[0161] Training a feature fusion learner using the first output data of the first pre-learner and the second output data of the second pre-learner as input data and the blow molded product molding defect record data as supervision data;

[0162] Merging the first output layer of the first pre-learner and the first output layer of the second pre-learner with the input layer of the feature fusion learner to obtain a blow molding process analysis model;

[0163] According to the blow molding process analysis model, the mold forming process information is obtained, and blow molding process analysis is performed in combination with the dimensional deviation grid to generate the blow molding product molding defect prediction.

[0164] Any step of the method described above can be stored as a computer instruction or program in an unlimited computer memory, and can be called and recognized by an unlimited computer processor to implement any method in the embodiments of the present application, without any unnecessary restrictions.

[0165] Furthermore, the terms "first" or "second" as described above may not only represent an order relationship but may also represent a specific concept and / or refer to the selection of multiple elements individually or collectively. Obviously, those skilled in the art may make various modifications and variations to this application without departing from the scope of this application. Thus, if such modifications and variations fall within the scope of this application and its equivalents, this application is intended to include such modifications and variations.

Claims

1. A blow molding quality inspection method based on a blow mold, characterized in that: The system is applied to a blow molding process quality detection system based on a blow mold. The system includes a server communicating with an image acquisition device and a size detection device, including: Receive blow molding task information uploaded by the user, wherein the blow molding task information includes blow molding product model, blow molding quantity information, and blow molding time information; According to the blow molding product model, matching the parison mold model and the forming mold model; According to the blow molding quantity information and the blow molding time information, the mold allocation is performed in combination with the parison mold model and the forming mold model to obtain a parison mold bit number set and a forming mold bit number set; Performing image acquisition and size detection based on the parison mold bit number set and the forming mold bit number set by an image acquisition device and a size detection device to construct a parison mold grid model set and a forming mold grid model set; Performing batch size verification on the parison mold grid model set and the forming mold grid model set to obtain size deviation grids; Acquire mold forming process information, perform blow molding process analysis based on the dimensional deviation grid, and generate blow molding product molding prediction defects; Blow molding process quality detection is performed based on the predicted defects in blow molding of the blow molded product.

2. The method according to claim 1, wherein According to the blow molding quantity information and the blow molding time information, mold allocation is performed in combination with the parison mold model and the forming mold model to obtain the parison mold bit number set and the forming mold bit number set, including: Acquire the number of parallel molds according to the blow molding quantity information and the blow molding time information; Determining a first idle parison mold position number set and a first idle forming mold position number set according to the parison mold model and the forming mold model; Traversing the first idle preform mold position number set and the first idle forming mold position number set to perform fault self-diagnosis and troubleshooting, and obtaining a second idle preform mold position number set and a second idle forming mold position number set; The server determines whether the number of molds in the second idle parison mold position number set and the second idle forming mold position number set meets the requirement of the number of parallel molds; If not, the second idle parison mold bit number set and the second idle forming mold bit number set are set as the parison mold bit number set and the forming mold bit number set; If so, randomly select molds that meet the number of parallel molds from the second idle parison mold bit number set and the second idle forming mold bit number set to obtain the parison mold bit number set and the forming mold bit number set.

3. The method according to claim 1, wherein By using an image acquisition device and a dimension detection device, image acquisition and dimension detection are performed based on the parison mold bit number set and the forming mold bit number set, and a parison mold grid model set and a forming mold grid model set are constructed, including: By using the image acquisition device and the size detection device, the parison mold position number set and the forming mold position number set are traversed to perform mold inner wall image acquisition and size detection, and a parison mold three-dimensional simulation atlas and a forming mold three-dimensional simulation atlas are constructed; Set the base grid size and virtual space coordinate system; The parison mold three-dimensional simulation atlas and the forming mold three-dimensional simulation atlas are gridded according to the reference grid size, and the gridded results are positioned and marked based on the virtual space coordinate system to obtain the parison mold grid model set and the forming mold grid model set.

4. The method according to claim 1, wherein Batch dimension verification is performed on the parison mold mesh model set and the forming mold mesh model set to obtain dimension deviation grids, including: Traversing the parison mold position number set and the forming mold position number set to obtain reference processing positioning information; Performing positioning pre-control on the parison mold grid model set and the forming mold grid model set to obtain processing pre-control positioning information; Performing position verification of the machining pre-control positioning information based on the reference machining positioning information to obtain a positioning verification result; When the positioning verification result fails, positioning deviation information is generated and the size deviation grid is set.

5. The method according to claim 4, wherein When the positioning verification result fails, generating positioning deviation information and setting the size deviation grid include: When the positioning verification result fails, retrieving a set of images of the forming mold cavity surface acquired by the image acquisition device; Traversing the molding die cavity surface image set and the molding die mesh model set to perform mesh concave-convex positioning and obtain a mesh concave-convex feature set; Positioning abnormal grids according to the grid concave-convex feature set to obtain coarse grid information; The size deviation grid is set according to the coarse grid information and the positioning deviation information.

6. The method according to claim 5, wherein Abnormal grid positioning is performed according to the grid concave-convex feature set to obtain coarse grid information, including: Acquire a cavity surface conceptual mesh model, wherein the cavity surface conceptual mesh model is a desired cavity surface model; Extracting a corner point change mesh area based on the cavity surface conceptual mesh model; Acquire a first adjacent grid area and a second adjacent grid area of ​​the corner point change grid area; Acquire a first adjacent tangent line set between the first adjacent grid area and the corner point change grid area; Acquire a second adjacent tangent line set between the second adjacent grid area and the corner point change grid area; Calculating a tangent angle set between the first adjacent tangent line set and the second adjacent tangent line set; When the deviation mean of the tangent angle set is less than or equal to a deviation threshold, and the tangent angle mean meets a preset angle range, dividing the first adjacent grid area and the second adjacent grid area into different areas; Repeat the division to obtain the grid area division results; Traverse the grid area division result to locate abnormal grids and obtain the coarse grid information.

7. The method according to claim 5, wherein Obtain mold forming process information, combine it with the dimensional deviation grid to perform blow molding process analysis, and generate blow molding product molding prediction defects, including: Collect mold forming process record data, size deviation grid identification data and blow molding product molding defect record data; Using the blow molded product molding defect record data as supervision data, the mold molding process record data and the dimensional deviation grid identification data as input data, and training a first pre-learner based on a BP neural network; Using the blow molded product molding defect record data as supervision data, and the mold molding process record data and the dimensional deviation grid identification data as input data, a second pre-learner is trained based on random forest; Training a feature fusion learner using the first output data of the first pre-learner and the second output data of the second pre-learner as input data and the blow molded product molding defect record data as supervision data; Merging the first output layer of the first pre-learner and the first output layer of the second pre-learner with the input layer of the feature fusion learner to obtain a blow molding process analysis model; According to the blow molding process analysis model, the mold forming process information is obtained, and blow molding process analysis is performed in combination with the dimensional deviation grid to generate the blow molding product molding defect prediction.

8. The blow molding processing quality detection system based on the blow molding mold is characterized by: The system includes a server communicating with an image acquisition device and a size detection device, including: A task information receiving module, which is used to receive blow molding task information uploaded by the user end, wherein the blow molding task information includes blow molding product model, blow molding quantity information and blow molding time information; A mold model matching module, which is used to match the parison mold model and the molding mold model according to the blow molded product model; A mold number acquisition module is used to allocate molds based on the blow molding quantity information and the blow molding duration information, in combination with the parison mold model and the forming mold model, to acquire a parison mold number set and a forming mold number set; A mold grid construction module, the mold grid construction module is used to perform image acquisition and size detection based on the parison mold bit number set and the forming mold bit number set through an image acquisition device and a size detection device, and to construct a parison mold grid model set and a forming mold grid model set; A batch size verification module, the batch size verification module is used to perform batch size verification on the parison mold grid model set and the forming mold grid model set to obtain a size deviation grid; a blow molding process analysis module, the blow molding process analysis module being used to obtain mold molding process information, perform blow molding process analysis in combination with the dimensional deviation grid, and generate blow molding product molding defect predictions; A processing quality detection module is used to detect the blow molding process quality based on the predicted defects of the blow molding product.

Images