Crystallizer copper plate plane correction method based on digital twinning

Abstract

The invention discloses a crystallizer copper plate correction method based on digital twinning, which aims at solving the problem of deformation of a crystallizer copper plate caused by thermal stress and mechanical stress in the using process, realizes real-time monitoring and intelligent correction of the copper plate from a global perspective, and mainly comprises the following steps: constructing a digital twinning model of the crystallizer copper plate; performing multi-physics field coupling simulation by combining the structural characteristics and the use environment of the copper plate; on the basis of real-time data collected by a digital twin system, an improved genetic algorithm is used for online evaluation and pre-judgment of copper plate deformation; generating a correction scheme based on a self-adaptive control strategy, and performing intelligent correction on the copper plate in combination with an execution mechanism; and continuously updating the digital twin model through a feedback mechanism to realize closed-loop optimization of copper plate correction. Through the digital twinning technology, deformation of the crystallizer copper plate is accurately predicted and dynamically corrected, the service life of the copper plate is prolonged, the operation efficiency of the copper plate is improved, and the automation level and the production stability of metallurgical equipment are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of crystallizer copper plate plane correction, and in particular to a crystallizer copper plate plane correction method based on digital twin. Background Art

[0002] As an emerging technology serving fault prediction and health management, digital twins can realize system health management, fault diagnosis and intelligent maintenance. It was first proposed by Professor GRIEVES of the University of Michigan in the product life cycle management course. Later, as a key technology of intelligent manufacturing, digital twins have received widespread attention from scholars and enterprises at home and abroad and have been developed and practiced. Shangguan D proposed a satellite system fault diagnosis and health monitoring method based on digital twins. It uses signal processing and data mining to obtain implicit information about the system's operating status, and provides reasonable fault diagnosis algorithms and maintenance strategies based on data fusion. Tao constructed a five-dimensional digital twin model for complex equipment. The model can effectively perform data interaction to diagnose faults and maintain the design. GrievesM proposed how to synchronize the physical product data with the information contained in the virtual product. By using this information, the digital factory is simulated and how the product is manufactured is predicted to ensure the long-term reliable operation of the product equipment. Zhuang proposed an intelligent production management and control framework based on digital twins to realize the information collection of the physical assembly workshop, the construction of the data twin assembly model, and the model after the construction to provide prediction services and management for the satellite assembly workshop. Benjamin proposed a comprehensive reference model and outlined the application of the reference model in geometric transformation management, providing a conceptual framework to enrich industrial engineering design. Tuegel EJ used the digital twin model to calculate the local damage of the aircraft and the evolution of the material state to predict the structural life of the aircraft and ensure its structural integrity.

[0003] As a key component in mechanical equipment, the flatness of the entire copper plate of the crystallizer has an important impact on the life and performance of complex machinery and mechanisms. Nowadays, the research on the flatness of the copper plate of the crystallizer tends to be based on the experience of workers and masters. The correction data is difficult to quantify, and the experience is difficult to store and accumulate in writing. There is a lack of research on the interaction between virtual space and physical space. Gupta studied the kinematics of the cage and the performance of the copper plate based on the relative position relationship between the components. Based on the model proposed by Gupta, Cao proposed a dynamic model for the vibration of local defects in the plate, and studied the vibration response of the copper plate with single defects, multiple defects and complex faults. AhmadiAM overcomes the limitation of using rolling elements as point masses, and considers the influence of Hertz contact force and damping to regard the shape of the fault defect as a rectangular pit. This method improves the accuracy of predicting low-frequency and high-frequency events. Liu gets rid of the limitations of the traditional rectangular fault defect model and studies the dynamic model of the coupling of time-varying displacement excitation and time-varying contact stiffness excitation considering lubrication traction. Kulkarni PG proposed a dynamic model to predict the vibration behavior of panels under the influence of local defects, simulating the influence of radial load, defect size, position, etc. on the time and frequency domains of panels. However, the above articles are limited to the dynamic analysis of normal panels or faulty panels in virtual space, and no two-way connection between the real space and the virtual space of the whole cycle is established. With the development of sensor technology and various digital equipment, data on the operation stage of parts can be obtained. These data need to be fully explored and utilized to realize the association between the design virtual environment model and the physical industrial manufacturing model. Summary of the invention

[0004] The purpose of the present invention is to provide an intelligent correction method for the plane correction of a crystallizer copper plate based on digital twins to solve the problems raised in the above-mentioned background technology.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] A method for correcting the plane of a crystallizer copper plate based on digital twins comprises the following steps:

[0007] Step S1, point cloud data collection and data filtering and noise reduction processing based on sick laser scanner;

[0008] Step S2, rendering a three-dimensional model based on SolidWorks combined with hydraulic machine tool modeling;

[0009] Step S3, calling the virtual-real interactive connection based on Unity 3D API;

[0010] Step S4, point cloud data processing based on Ransac, ICP and other algorithms;

[0011] Step S5, intelligent application of crystallizer copper plate plane correction based on digital twin.

[0012] The digital twin-based crystallizer copper plate correction method provided by the present invention uses augmented reality to achieve multi-perspective collaboration and human-computer interaction, focusing on solving the difficulties in complex product processing collaboration and eliminating communication barriers in the processing process. By establishing a processing technology information integration model based on twin data, the AR processing scene is processed by multi-terminal AR visualization collaboration, so that the online flow of data in the physical space-client-AR-twin space can be achieved through the use of AR technology, and the online supervision, accurate prediction, and optimization of the processing process can be achieved, thereby improving the stability of processing quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0014] Figure 1 A flow chart of a crystallizer copper plate correction method based on digital twin provided by the present invention;

[0015] Figure 2 A processing architecture diagram for implementing the crystallizer copper plate correction method based on digital twin provided by the present invention;

[0016] Figure 3 A schematic diagram of intelligent interaction based on augmented reality technology provided by an embodiment of the present invention;

[0017] Figure 4 A diagram of an AR space anomaly point synchronization process based on a network service provided by an embodiment of the present invention;

[0018] Figure 5 A diagram of a correction process of a mold copper plate based on twin data provided in an embodiment of the present invention;

[0019] Figure 6 A product timing optimization processing diagram based on digital twins provided in an embodiment of the present invention;

[0020] Figure 7 A schematic structural diagram of the crystallizer copper plate plane correction system based on digital twin provided by the present invention. DETAILED DESCRIPTION

[0021] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0022] The purpose of the present invention is to provide a crystallizer copper plate correction method and system based on digital twin, which can adjust and optimize different processing conditions in real time to improve the stability of correction quality.

[0023] An intelligent correction method for mold copper plate plane correction based on digital twin, comprising:

[0024] Based on the point cloud data acquisition and data filtering and noise reduction processing of the sick laser scanner, the processing task constructs a processing technology information integration model; the twin data includes information data and physical data in the assembly process; the processing technology information integration model includes: pre-planned process information set, actual processing process information set, processing process subset, processing process category, processing sliding window information, processing visual information, collaborative processing category and display information;

[0025] 3D model rendering based on SolidWorks combined with hydraulic machine tool modeling;

[0026] Virtual and real interactive connection based on Unity 3D calling API;

[0027] Point cloud data processing based on Ransac, ICP and other algorithms.

[0028] The crystallizer copper plate plane correction intelligent application based on digital twin combines the virtual processing scene with the AR processing scene; the AR processing scene is stored on the client; a virtual processing scene is constructed based on the processing technology information integration model; the virtual processing scene is a mapping scene of the physical processing scene; the virtual processing scene is stored on the server; after obtaining the evolution law of the entire processing cycle from the above steps, the predicted downward pressure data is input into the crystallizer copper plate correction model, and the obtained data is mapped to the corresponding data in the physical space, and then the bearing digital twin model is established, and the crystallizer copper plate correction is realized by using the model.

[0029] Preferably, the constructing of a virtual processing scene based on the processing technology information integration model specifically includes:

[0030] According to the processing technology information integration model combined with the MBD model expression method, the assembly information is added to the assembly three-dimensional model to obtain a virtual assembly scene.

[0031] Preferably, the three-dimensional model rendering based on SolidWorks combined with hydraulic machine tool modeling performs virtual end physical modeling and three-dimensional model rendering of the calibration scene.

[0032] Preferably, the virtual-real interactive connection based on Unity 3D calling API specifically includes: registering an OPC UA server, establishing a TCP / IP network layer communication connection between the physical machine tool and the virtual machine tool through a Socket interface, and using API8070 to establish a connection access between the 8070CNC kernel and the application.

[0033] Preferably, the point cloud data processing based on Ransac, ICP and other algorithms specifically includes: filtering, denoising, registering and segmenting the collected data of the correction scene, solving the flatness and verifying the convex point data thereof:

[0034] Preferably, the intelligent application of crystallizer copper plate plane correction based on digital twin performs real-time mapping of the actual processing scene; the real-time mapping of the actual processing scene is performed by the server;

[0035] Determine a reference client, and use the reference client to perform spatial scanning on the mapped actual processing scene to determine spatial anchor point information;

[0036] Determine a root anchor point according to the anchor point information;

[0037] Using the client to scan the feature map on the reference client to obtain the spatial coordinates of the reference client, and then obtaining the coordinate information of the root anchor point through the network service;

[0038] Determine the spatial offset of the coordinate system between the two clients according to the coordinate information of the root anchor point, and synchronize the anchor point information between the two clients based on the spatial offset;

[0039] The client is used to achieve the unification of multi-terminal perspectives based on the synchronous root anchor point and the network service in the asynchronous network communication.

[0040] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0041] The digital twin-based crystallizer copper plate correction method provided by the present invention uses augmented reality to achieve multi-perspective collaboration and human-computer interaction, focusing on solving the difficulties in complex product processing collaboration and eliminating communication barriers in the processing process. By establishing a processing technology information integration model based on twin data, the AR processing scene is processed by multi-terminal AR visualization collaboration, so that the online flow of data in the physical space-client-AR-twin space can be achieved through the use of AR technology, and the online supervision, accurate prediction, and optimization of the processing process can be achieved, thereby improving the stability of processing quality.

[0042] Corresponding to the above-mentioned digital twin crystallizer copper plate plane correction method, the present invention also provides the following implementation system:

[0043] A processing system based on collaborative AR augmented reality technology of digital twins, comprising:

[0044] A processing technology information integration model construction module is used to construct a processing technology information integration model based on twin data combined with processing technology and processing tasks; the twin data includes information data and physical data in the processing process; the processing technology information integration model includes: pre-planned process information set, actual processing technology information set, processing procedure subset, processing procedure category, processing sliding window information, processing visual information, collaborative processing category and enhanced display information;

[0045] A virtual assembly scene construction module is used to construct a virtual processing scene based on the processing technology information integration model; the virtual processing scene is a mapping scene of the physical processing scene; the virtual processing scene is stored on the server;

[0046] An AR processing scene generation module is used to generate an AR processing scene based on the virtual processing scene; the AR processing scene is stored in the client;

[0047] Preferably, the virtual processing scene construction module includes:

[0048] The virtual processing scene construction unit is used to add the processing information to the machine tool correction three-dimensional model to obtain a virtual scene according to the processing technology information integration model combined with the MBD model expression method.

[0049] Preferably, the multi-terminal AR visualization collaborative processing module includes:

[0050] An asynchronous network communication establishing unit, used for establishing asynchronous network communication;

[0051] A multi-terminal AR perspective unifying unit, configured to unify the multi-terminal AR perspectives based on the asynchronous network communication;

[0052] The virtual processing process establishing unit is used to establish a virtual processing process for different processing tasks according to the processing technology information in the AR processing scene based on the unified multi-terminal AR perspective; the virtual processing process includes a processing route and a processing animation.

[0053] Preferably, the multi-terminal AR perspective unification unit includes:

[0054] A real-time mapping subunit is used to perform real-time mapping of the actual processing scene; the real-time mapping of the actual processing scene is performed by the server;

[0055] An outlier information determination subunit is used to determine a reference client, and use the reference client to perform a spatial scan on the mapped actual processing scene to determine the spatial outlier information;

[0056] A root outlier determination subunit, configured to determine a root outlier according to the outlier information;

[0057] A coordinate information acquisition subunit, configured to obtain the spatial coordinates of the reference client by scanning the feature map on the reference client using the client, and then obtain the coordinate information of the root outlier point through a network service;

[0058] An information synchronization subunit, configured to determine a spatial offset of a coordinate system between two clients according to the coordinate information of the root abnormal point, and synchronize abnormal point information between the two clients based on the spatial offset;

[0059] A synchronization root abnormal point determination subunit is used to broadcast the spatial coordinates of the virtual action and the spatial coordinates of the virtual object through a network service when the AR device loads the virtual processing object, so as to obtain a synchronization root abnormal point;

[0060] The multi-terminal AR perspective unification subunit is used to unify the multi-terminal AR perspectives by using the client based on the synchronous root exception point and the network service in the asynchronous network communication.

[0061] Since the technical effect achieved by the crystallizer copper plate correction system based on digital twin provided by the present invention is the same as the technical effect achieved by the crystallizer copper plate correction method based on digital twin provided above, it will not be repeated here.

[0062] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0063] like Figure 1 As shown, a method for correcting the plane of a copper plate of a crystallizer based on digital twins comprises the following steps:

[0064] Step S1, point cloud data collection and data filtering and noise reduction processing based on sick laser scanner;

[0065] Step S2, rendering a three-dimensional model based on SolidWorks combined with hydraulic machine tool modeling;

[0066] Step S3, calling the virtual-real interactive connection based on Unity 3D API;

[0067] Step S4, point cloud data processing based on Ransac, ICP and other algorithms;

[0068] Step S5, intelligent application of crystallizer copper plate plane correction based on digital twin.

[0069] Specifically, the crystallizer copper plate correction method based on digital twin provided by the present invention includes:

[0070] Step 100: Construct a processing technology information integration model based on twin data combined with processing technology and processing tasks. Twin data is an important component of the digital twin architecture. It integrates information data and physical data to achieve consistency between information space and physical space. The existing physical processing object integration model can already manage the processing object model very well. The present invention is mainly designed for the information integration model. The designed processing technology information integration model mainly includes: pre-planned process information set (PAPI), actual processing technology information set (AAPI), processing procedure subset (APS), processing procedure category (CAP), processing sliding window information (ASW), processing visual information (AVI), collaborative processing category (CAC) and enhanced display information (ADI). Among them, the digital twin architecture mainly includes physical processing scenes, virtual processing scenes, twin data, AR processing guidance and time series data analysis and prediction parts.

[0071] AAPI mainly includes the mold copper plate measured data set (PMMD), the processing process measured data set (APMD) and the processing deformation measured data set (ADMD).

[0072] APS is mainly used to store product processing sequences and to achieve high efficiency in processing design and optimization by establishing processing procedure subsets based on actual processing conditions.

[0073] CAP classifies processing categories, mainly into parallel processing procedures and serial processing procedures, and can provide adaptive processing guidance for different process categories.

[0074] ASW determines the processing subset. Due to the different processing complexities, different processing windows will appear during the processing. The window size affects the design and planning of AR processing guidance. The later optimization design correction is carried out by recording the window sequence.

[0075] In order to ensure the unimpeded flow of information in the collaborative processing process, AVI allows AR guidance information to produce positive guidance rather than visual interference, mainly including abnormal point information, device coordinate information, and viewing angle information.

[0076] Among them, ADI is mainly used for display information storage during AR processing guidance.

[0077] The information contained in the processing technology information integration model is stored in the sqlsever database, and a data interface is established for real-time acquisition and dynamic change of information for later digital twin processing. The data connection of the virtual scene is the interactive process of the data flow between the unity virtual scene and the information model. Unity mainly uses the command command to call the database. SQL commands are sent to the database through the server. The command object can be used to add, delete, and modify data and query single data. The dataReader can be used to read the entire row of data, and the dataAdapter can be used to operate the result set. After the server obtains the data, it transmits the data through socket communication and then displays it on the UI interface.

[0078] Step 100: Construct a virtual processing scene based on the processing technology information integration model. The virtual processing scene is a mapping scene of the physical processing scene. The virtual processing scene is stored on the server. In the specific implementation process, this step is mainly based on the processing technology information integration model, combined with the mature MBD model expression method, to add the processing information to the processing three-dimensional model, and simulate the entire processing process in the virtual space. The implementation of step 100 is mainly to ensure the real-time supervision and assembly optimization of the processing process, to ensure the consistency of the established virtual assembly scene and the physical scene, to achieve one-to-one mapping, to reflect the real with the virtual, and to return the virtual with the real.

[0079] The specific implementation process of step 101 is:

[0080] A. Based on the PAPI of the processing technology information integration model, a virtual ideal processing process is created, that is, a virtual reproduction process of the physical processing scene is realized. First, a virtual scene is established based on Unity, and an AR environment is built at the same time.

[0081] B. The virtual scene built based on Unity includes the creation of MBD models and the simulation of processing procedures. The simulation of processing procedures is the reproduction of the processing process. The entire processing process flow is determined based on the existing PAPI, CAP, and ASW information, and the processing sequence is programmed through code. The simulation of the processing process is mainly achieved by creating processing animations, so that the processing guidance process can be realized when it is published to the AR device.

[0082] Among them, the main process of establishing a virtual processing scene is:

[0083] a. Import of processing models: UG NX10.0 can establish digital models. If the format of the three-dimensional model in the physical processing object integrated model is not supported by Unity, it can be rebuilt using UG NX10.0. At the same time, the model needs to be simplified to a certain extent, and its unnecessary model features need to be removed to reduce the amount of calculation in the model rendering process to improve efficiency. The model generated by UGNX10.0 can import all its material maps into Unity without information loss and format conversion. Unity can use the sub-object relationship of the UG NX10.0 model to add rigid body collision and realize modular control of the model. In Unity's three-dimensional scene, X is to the right, Y is up, and Z is forward. Models imported from outside also need to be established according to this rule to avoid spatial coordinate problems in the assembly process affecting the control of the script. When the model coordinate system is not unified, an empty object of the standard coordinate system can be established as a sub-object to meet the normal control in the later stage. Unity cannot import DXF models. The assembly model uses an existing model. By using Blender to convert the model format, it is converted into FBX format that can be used directly in Unity. Since the model size is too large and the model is an integral model, Blender is used to reduce the size and obtain the model of the independent plate, so as to obtain each processing component for the convenience of later processing and display.

[0084] b. Creation of MBD model: Process the imported model and convert it into an MBD model. The MBD model mainly adds basic information such as dimensions and tolerances to the three-dimensional entity of the processing object. This can be achieved using Unity. By establishing a canvas in the world coordinate system in the scene, and then adding sub-objects to the canvas, including images, texts, etc., the database is bound to obtain real-time on-site data for update. In order to ensure the unity of model information and the spatial coordinates of the model, a prefab is created on the canvas and used as a sub-object of the model. In this way, the assembly model is no longer in a fixed state, but changes in real time, and can provide real-time feedback to the physical scene.

[0085] c. Coordinate unification of virtual space and physical space: The processing process is mainly a real-time response to the physical processing process. By establishing a processing origin, it is used to determine the spatial relationship with the physical scene. ARcamera is used to achieve markerless spatial tracking and positioning. In the actual AR processing guidance process, the processing operator adjusts the spatial coordinates of the processing origin to achieve dynamic presentation of the virtual scene position.

[0086] d. Layout and modularization of the processing process: Based on the determined processing origin, the acquisition of PAPI information can determine the entire processing planning process, and the processing sequence can be determined according to APS. Each processing sequence can obtain the relative coordinates of the processing relative to the processing origin, so only modular design is required for each processing sequence. Each processing sequence includes MBD model creation, prefabricated body production, collider addition, shader addition, trigger addition, Animation creation, and control script creation. The modular assembly guidance design is used for the adaptive adjustment of later serial and parallel processing procedures and AR guidance.

[0087] e. Processing adaptability adjustment function: Based on the determined processing sequence and assembly module, the logic control of the processing sequence is created by using the tree structure through C# script control, and the processing module is virtually presented by sequential traversal of the tree structure. The serial and parallel processes of the processing process are represented by the depth and breadth of the tree, and the adjustment of the processing sequence of the actual assembly process is achieved by moving the leaf nodes. The processing information of each node is obtained through query, and the processing process is optimized by changing the node information.

[0088] Step 102: Generate an AR processing scene based on the virtual processing scene. The AR processing scene is stored on the client.

[0089] The AR processing scene is generated based on the virtual scene. By adding human-computer interaction methods to each object or process, the virtual-reality interaction process is realized (the realization of the interaction process is described in the collaborative processing process). The basic environment configuration is performed through Unity to add the system's spatial perception ability. Then, the generated application is published and installed on the Android terminal or HoloLens to realize the presentation of the AR processing scene.

[0090] The difference between virtual scenes and AR processing scenes is that virtual scenes are mainly placed on the server side, and are mainly used for real-time supervision of physical scenes and optimization of processing technology. AR processing scenes are mainly placed on the client side and published on AR devices (such as head-mounted display devices (hololens2) or tablets that support AR development, etc.), and are mainly used for processing guidance, that is, to superimpose the virtual scene on the physical scene, and observe the virtual projection of the processing animation while observing the processing object, so as to obtain more processing information. When the physical processing scene fails or there is a processing error, the server side responds accordingly by obtaining the AAPI information of the information integration model, such as adjusting the processing position. This processing information is then transmitted to the AR assembly scene through network communication to update the AR processing guidance and realize online optimization of processing.

[0091] like Figure 3As shown in the figure, different scenarios and working conditions guided by augmented reality can be used to assist processing based on the processing process. Augmented reality technology is used to improve the virtual-real interaction of the digital twin processing process, and network services are used to transmit the operator's information in real time to optimize the processing technology in real time, thereby reducing the cognitive burden of employees during the processing and improving processing efficiency.

[0092] Step 104: Establish asynchronous network communication and perform multi-terminal AR visualization collaborative processing on the AR processing scene.

[0093] The basic AR-assisted processing process is realized by creating virtual scenes and AR scenes. The present invention needs to further operate the processing information model to realize AR visualization collaboration. Based on this, step 103 specifically includes:

[0094] A. Establish asynchronous network communication. Figure 4 As shown, the purpose of establishing asynchronous network communication is to achieve the collaborative work of a server and multiple clients. The process of creating a server in the present invention is as follows:

[0095] a. Create a server socket.

[0096] b. Receive client connection StartAccept(SocketAsyncEventArgs e);

[0097] c. Asynchronous connection AcceptAsync;

[0098] d. I / O operation judgment: determine whether it is in a suspended state.

[0099] e. If not suspended, process the client connection ProcessAccept(AsyncEventArgs e)

[0100] f. Create a Usertoken class to process the client sending and receiving data (ReceiveAsync performs asynchronous reception / SendAsync performs asynchronous sending)

[0101] g. Create a message processing center ABSHandlerCenter to process client and server business. i. Based on the implementation of asynchronous network communication, the information of the virtual scene and AR guidance scene of the entire system can flow bidirectionally, thereby realizing the collaboration of multiple AR clients.

[0102] In order to improve the accuracy of information collaboration and enhance the immersion and auxiliary functions of AR processing guidance, the present invention needs to determine the processing information transmission rules of multiple clients during the information collaboration process, and add human-computer interaction means. Figure 6As shown in the figure, for the processing categories of the actual processing process, single-station processing collaboration and multi-station processing collaboration design are mainly carried out. Complex products are processed at different stations based on an assembly reference component, mainly in a multi-station collaborative processing environment. During multi-station AR collaborative processing based on digital twins, multiple clients need to be independent and collaborative.

[0103] Independence means that different processing guidance can be achieved for different workstations, and the client can achieve independent processing guidance work. Considering that the range of action of employees in the single-station processing process is relatively larger, in order to ensure the movement of employees in the processing process, or to achieve the loss of processing UI caused by irregular movements caused by processing actions, by giving it a follow-up attribute, that is, determining a relative spatial coordinate, obtaining the spatial position of the AR device in real time through the sensor, and determining the real-time display position of the UI through calculation, the UI spatial position is relatively determined. At the same time, in order to achieve the optimal viewpoint of the processing process, an assembly animation is created based on the local coordinate system, and the MBD model and sub-object Bounding Box are magnified and rotated to achieve comprehensive acquisition of process information.

[0104] To achieve collaboration, the client can observe the processing conditions of other stations, and can provide feedback and data exchange on the processing operations of other stations. Through collaborative processing, the system can be managed and optimized according to the processing of each station, that is, the positioning reference and error transmission of multiple stations are considered, and interface parts are established as connectors for multi-station processing in complex processing processes to reduce the impact of cumulative errors. The design size of the interface part is not a fixed value, but a size distribution. Through the establishment of a set of interface parts, based on the information collection and analysis of different stations in the actual processing process, the multi-station processing parameters are coordinated and optimized, and the selection of interface parts in the dynamic processing process is realized. The information is transmitted to the AR equipment of each station in real time through network communication, and the AR-guided scene processing information is optimized to ensure that the overall processing meets the process specifications.

[0105] The following is based on Figure 2 The digital twin architecture shown in the figure illustrates the specific implementation process of the crystallizer copper plate correction method based on digital twin provided by the present invention. In actual application, the crystallizer copper plate correction method based on digital twin provided by the present invention can also be applied to other architectures.

[0106] The overall idea of ​​implementing the digital twin-based crystallizer copper plate correction method provided by the present invention is: by constructing a processing design model of a complex product in a virtual assembly space, and pre-designing and planning the ideal processing technology based on the theoretical digital model, a complete virtual processing scene is generated. The entire processing process is divided into many processing steps. In the preparation stage, the process objects, process resources, etc. are analyzed, and processing decisions are made to consider whether the assembly conditions are met. In the physical processing scenario, augmented reality technology is used to implement the AR processing guidance process. In the process of human-computer interaction with AR, the actual data collaborative processing model is used to synchronize the processing state of the physical model with the virtual model in real time. As the processing process proceeds, the actual processing time series data and the twin data of virtual processing are combined to perform time series data analysis and prediction, and processing supervision and processing optimization are performed according to the data analysis results. The analysis results and optimization schemes are visualized in real time using the AR system to provide the best processing operation.

[0107] Based on the above ideas, the implementation process of the present invention is:

[0108] 1) Use the processing technology information integration model to create a virtual processing scene and achieve one-to-one mapping with the physical processing scene:

[0109] According to the processing technology information integration model combined with the mature MBD model expression method, the processing information is added to the processing 3D model, and the entire processing process is simulated in the virtual space. Based on the theoretical model, the processing technology design and planning are realized, and reasonable and effective processing technology parameters for the physical processing space are determined. The network communication technology is used to obtain the multi-force processing measured data in the participating processing process that is perceived and collected in real time from the information model, so as to realize the dynamic construction and iterative update of the processing design model. Figure 3 shown.

[0110] 2) AR visualization collaboration based on network services:

[0111] Use network communication technology to realize the information flow between the main scene and the sub-scenes, that is, different workstations. Based on ARFoundation, realize the compatibility of AR development environment of multiple devices, determine the spatial abnormal point information based on spatial scanning, and realize the abnormal point synchronization of multiple devices through image recognition and spatial coordinate transformation. Finally, realize the fusion of virtual and real through 3D registration technology, and enhance the visualization expression of virtual processing scene.

[0112] 3) Augmented reality processing collaboration based on twin data:

[0113] For complex processing, the assembly personnel may work at different positions, with different processing focuses and observation angles. Therefore, single-view processing is problematic. To address this problem, the unified model perspective of multiple clients can realize multi-person collaborative operation of processing scenes, that is, the processing process requires real-time update of processing information on each client, which requires analysis of processing categories, processing relationships and collaboration, and processing path planning. Here, the processing needs of complex products are met through the augmented reality processing design of single and multi-stations.

[0114] 4) Time series data analysis and prediction:

[0115] The processing process obtains the actual processing information (AAPI) through ADMD, PMMD, and APMD, and feeds back to the pre-planned process information PAPI through the mapping mechanism for comparative analysis. Then, the processing procedure category (CAP) is distinguished, and the pre-planned process information set PAPI is analyzed and matched through the processing sliding window information (ASW). The matching degree analysis determines whether the assembly process subset (APS) is optimized. When optimization is required, parallel processing performs adaptive processing optimization adjustment. For serial processing procedures, the main optimization is processing coordination. According to the determined processing category, the division of labor information of processing is determined, and the difference optimization of CAC is performed according to the actual processing visual information (AVI). Guidance and error warning for employee processing are mainly achieved through ADI. The comparative analysis of actual processing data and assembly process design information is mainly achieved through algorithms, and abnormal information generated during the processing is discovered and augmented reality visualization warnings are performed.

[0116]

[0117] Where D represents the distance from the point outside the reference plane in the point cloud data to the fitting plane, m and n represent any two points, A and B represent the coefficients of the reference plane equation, Δt represents the flatness error, and δ represents the variance of the vertical distance from the position of the point cloud point outside the reference plane to the reference plane.

[0118] The processing technology information integration model construction module 1 is used to construct a processing technology information integration model based on twin data combined with processing technology and assembly tasks. Twin data includes information data and physical data in the processing process. The processing technology information integration model includes: pre-planned process information set, actual processing technology information set, processing process subset, processing process category, processing sliding window information, processing visual information, collaborative processing category and enhanced display information.

[0119] The virtual processing scene construction module 2 is used to construct a virtual processing scene based on the processing technology information integration model. The virtual processing scene is a mapping scene of the physical processing scene. The virtual processing scene is stored on the server.

[0120] The AR processing scene generation module 3 is used to generate an AR assembly scene based on the virtual processing scene. The AR processing scene is stored in the client.

[0121] Specifically, the virtual processing scene construction module used above may further include: a virtual processing scene construction unit.

[0122] The virtual processing scene construction unit is used to add processing information to the assembly three-dimensional model to obtain a virtual processing scene according to the processing technology information integration model combined with the MBD model expression method.

[0123] Furthermore, in order to achieve accurate prediction of the processing process, the multi-terminal AR visualization collaborative processing module adopted above includes: an asynchronous network communication establishment unit, an AR perspective unification unit and a virtual processing process establishment unit.

[0124] The asynchronous network communication establishing unit is used to establish asynchronous network communication.

[0125] The virtual processing process establishment unit is used to establish a virtual assembly process for different processing tasks in the AR processing scene based on the unified multi-terminal AR perspective and the processing process information. The virtual processing process includes processing routes and processing animations.

[0126] The coordinate information acquisition subunit is used to use the client to scan the feature map on the reference client to obtain the spatial coordinates of the reference client, and then obtain the coordinate information of the root abnormal point through the network service.

[0127] The information synchronization subunit is used to determine the spatial offset of the coordinate system between the two clients according to the coordinate information of the abnormal point, and realize the synchronization of the abnormal point information between the two clients based on the spatial offset.

[0128] The synchronization abnormal point determination subunit is used to broadcast the spatial coordinates of the virtual action and the spatial coordinates of the virtual object through the network service when the AR device loads the virtual processing object, so as to obtain the synchronization root abnormal point.

[0129] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0130] The principles and implementation methods of the present invention are described in this article using specific examples. The description of the above embodiments is only used to help understand the method and core idea of ​​the present invention. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. A crystallizer copper plate plane correction method based on digital twin, characterized in that The following steps are involved: Step S1, point cloud data collection and data filtering and noise reduction processing based on sick laser scanner; Step S2, rendering a three-dimensional model based on SolidWorks combined with hydraulic machine tool modeling; Step S3, calling the virtual-real interactive connection based on Unity 3D API; Step S4, point cloud data processing based on Ransac, ICP and other algorithms; Step S5, intelligent application of crystallizer copper plate plane correction based on digital twin.

2. The method for plane correction of a mold copper plate based on digital twinning according to claim 1, characterized in that: The point cloud data acquisition and data filtering and noise reduction processing based on the sick laser scanner are specifically It includes: It is characterized in that: according to the correction process information of the crystallizer copper plate, the system is equipped with a laser scanner driven by a servo motor to collect point cloud data of the crystallizer copper plate for standby use.

3. The method for plane correction of a mold copper plate based on digital twinning according to claim 2, characterized in that: The three-dimensional model rendering based on SolidWorks combined with hydraulic machine tool modeling, virtual end physical modeling and three-dimensional model rendering of the calibration scene, specifically includes: After collecting the real-world physical data of the hydraulic machine tool, crystallizer copper plate, and laser scanner gantry, the physical model of the crystallizer copper plate processing scene was established by taking into account the mechanical structure accuracy error, laser scanner data acquisition error, and algorithm error; the three-dimensional model was drawn in UGNX software according to the size ratio and exported to the STL format file, which was imported into the 3dsMax software for rendering, and then the rendered FBX file was exported to Unity3D.

4. The method for plane correction of a crystallizer copper plate based on digital twinning according to claim 3 is characterized in that: The virtual-real interactive connection based on Unity 3D API call is used to build real-time data interaction between the physical entity and the virtual model of the calibration scene processing environment, specifically including: Register the OPC UA server to establish a TCP / IP network layer communication connection between the physical machine tool and the virtual machine tool through the Socket interface, and use API8070 to establish a connection access between the 8070CNC kernel and the application.

5. The method for plane correction of a mold copper plate based on digital twinning according to claim 4, characterized in that: The point cloud data processing based on Ransac, ICP and other algorithms is to collect data from the correction scene for filtering, noise reduction, registration, segmentation, flatness solution and convex point data verification: The point cloud data in step S1 and the point cloud data generated by the virtual model in step S2 are matched at the unit level through data connection to reveal the relationship between the flatness value and the pressing condition of the hydraulic machine tool, and the relationship between the pressing amount of the press and the comprehensive equivalent arc, and the digital twin model of the bearing is obtained; as the deformation caused by the error, when the copper plate is in the healthy stage, the flatness error is 0, and when the copper plate is in the fault stage, the fault model is established to maximize the actual situation of the copper plate fault, and its flatness is specifically calculated as follows: Where D represents the distance from the point outside the reference plane in the point cloud data to the fitting plane, m and n represent any two points, A and B represent the coefficients of the reference plane equation, Δt represents the flatness error, and δ represents the variance of the vertical distance from the position of the point cloud point outside the reference plane to the reference plane.

6. The method for plane correction of a crystallizer copper plate based on digital twinning according to claim 5, characterized in that: The intelligent application of crystallizer copper plate plane correction based on digital twin performs plane correction on the crystallizer copper plate of the correction scene workpiece: In the above step S4, the curvature and flatness of the copper plate matched in step S3 and the equivalent pressing amount of the hydraulic machine tool are used to obtain the pressing law of the whole processing cycle through a neural network; in the above step S5, after obtaining the evolution law of the whole processing cycle from step S4, the predicted pressing data is input into the crystallizer copper plate correction model, and the obtained data is mapped to the corresponding data in the physical space, and then the digital twin model of the bearing is established, and the crystallizer copper plate correction is realized by using the model.

Images