A construction method for an intelligent assembly system based on digital twin

Abstract

The present invention discloses a construction method of an intelligent assembly system based on digital twin, belonging to the technical field of intelligent manufacturing, including: aiming at the dynamic data elements of the technical physical workshop in the assembly production site of complex equipment products, using a data acquisition system to collect the dynamic data of the assembly process in the physical workshop in real time, preprocessing and standardizing it, and then storing it in the database; using the digital twin workshop model as the system client, through the interaction with the database, driving the three-dimensional virtual models of assembly parts and the three-dimensional model of the product with unified data services, so as to realize the two-way dynamic mapping between the digital twin workshop and the physical workshop of assembly production; based on the integration of the three-dimensional visualization system, completing the construction of the intelligent workshop digital twin system, thereby realizing functions such as real-time state display, state evolution, data input, and data saving in the process of digital aircraft assembly, and solving the problems of monotonous data presentation, poor interaction and immersion in the existing digital workshop.

Description

Technical Field

[0001] The present invention belongs to the technical field of intelligent manufacturing, and relates to a construction method of an intelligent assembly system based on digital twin. Background Technique

[0002] With the promotion of the integration of information technology and manufacturing industry to the transformation of the manufacturing industry, countries around the world have successively put forward advanced manufacturing development strategies to promote the transformation and upgrading of their own manufacturing industries. For example, the purpose of these measures is to achieve the interconnection and intelligent operation of the physical world and the information world of manufacturing with the help of information technology, and then realize intelligent manufacturing. As one of the key technologies to realize the integration of cyber-physical systems, digital twin technology has been deeply studied in recent years.

[0003] Digital twin technology refers to a process or method that uses digital technology to describe and model a process or method that is consistent with the characteristics, behaviors, and performances of a physical entity. It is an effective way to realize the interaction and integration of the physical space and the information space, and mainly consists of three parts: the physical entity in the real space, the virtual product in the information space, and the data and information that connect the virtual product and the physical product. As the mapping from the physical space to the digital twin model, the construction of the digital twin workshop model, the fusion and interaction of virtual and real data are important bases for realizing the digital twin workshop. Compared with traditional assembly, the product assembly driven by digital twin presents new changes, that is, the process changes from the virtual information assembly process to the virtual-real combination assembly process, the model data changes from the theoretical design model data to the actual measurement model data, the element form changes from a single process element to a multi-dimensional process element, and the assembly process changes from a digital-guided physical assembly process to the co-evolution of physical and virtual assembly processes.

[0004] From the current patent disclosures and literature, some scholars have already conducted research on this. A dynamic linkage control method for the autonomous production process of an intelligent workshop based on digital twin (CN201810731107.1) conducts real-time simulation of the operating state of the intelligent workshop in the digital twin model, calculates the progress deviation in the actual production operation process, and calls the machine tool task scheduling algorithm to predict and determine the machining machine tool number for the next process of the workpiece to be executed, generating the optimal solution of the machine tool task scheduling sequence; A workshop-level intelligent manufacturing system based on digital twin and its configuration method (CN201810339946.9) adopts a system architecture composed of a physical layer, a network layer, and an information layer for the manufacturing system. The configuration method of the intelligent manufacturing system forms an autonomous interaction mechanism of "human-machine-material" in the workshop by establishing digital twins of work-in-progress and manufacturing resources and establishing a mapping relationship with the digital twins, realizing the closed-loop manufacturing logic of "perception - calculation - execution - feedback - decision" in the workshop. The existing inventions' research on digital twin stays at the configuration of the workshop intelligent manufacturing system based on digital twin, unable to reflect the integration into a unified visualization system for real-time scheduling and real-time control, synchronously collaborating the physical workshop and the digital twin workshop, and moreover, the real-time simulation has a long delay time.

[0005] In order to realize the full-process simulation, status monitoring, and assembly result prediction of the assembly production site of complex equipment products such as aircraft, based on the analysis of the structural composition and functional requirements of the equipment product, the present invention proposes a way to construct a digital twin of the equipment product to virtually express the assembly operation process and result prediction. First, it focuses on analyzing the technical route of the digital twin for equipment product assembly, then gives the system architecture for constructing the digital twin of equipment product assembly, and finally gives the implementation approach of the digital twin assembly technology for equipment product assembly, combined with three-dimensional visual interaction design; The research gives an assembly model based on digital twin, breaks through the key technologies, and develops a prototype system to provide technical support for the assembly of equipment products based on digital twin. Therefore, based on the assembly simulation technology of digital twin, an intelligent assembly method for complex equipment products based on digital twin and a dynamic update mechanism for the digital twin assembly model of equipment products are proposed, enabling it to be more widely applied in the fields of intelligent manufacturing and intelligent factories. Summary of the Invention

[0006] In order to overcome the above-mentioned shortcomings of the prior art, the purpose of the present invention is to provide a construction method for an intelligent assembly system based on digital twin, which solves the technical problems of complex technical states, complex and changeable assembly processes and routes at the assembly production site of complex equipment products such as aircraft, and in addition, high requirements for assembly accuracy and quality, prone to assembly errors, low assembly quality and accuracy, low assembly efficiency, long assembly debugging cycle, monotonous data presentation, and poor interaction and immersion sense, and more effectively realizes the two-way mapping of virtual and real data.

[0007] To achieve the above object, the present invention is implemented by the following technical solutions:

[0008] The present invention discloses a construction method of an intelligent assembly system based on digital twin, including:

[0009] 1) For the production elements of the physical assembly workshop, construct a multi-level twin model of all elements of the physical workshop, collect real-time dynamic data of the assembly process in the physical space, perform three-dimensional modeling of multiple components and multiple levels of the aircraft wing, realize the accurate mapping of the digital twin body of the workshop production elements between the physical space and the virtual space, and obtain the digital twin workshop model;

[0010] 2) Take the digital twin workshop model as the system client, construct a data acquisition system to collect multi-source heterogeneous data of the physical workshop, preprocess and standardize the multi-source heterogeneous data and then store it in the database. The digital twin workshop model, as the system client, through interaction with the database, drives the three-dimensional virtual model of the assembly parts and the three-dimensional model of the product with unified data services, so that the digital twin workshop and the physical workshop realize two-way dynamic mapping;

[0011] 3) Based on the integration of the three-dimensional visualization system, realize the state monitoring and process optimization feedback control of the digital twin body, comprehensively and real-timely and effectively supervise the on-site personnel and operations, and seamlessly integrate the aircraft wing assembly in the real world with the aircraft wing assembly simulation in the virtual environment to complete the construction of the digital twin system of the intelligent workshop;

[0012] Specifically, since the recognition of three-dimensional geometric assembly features is more complex than two-dimensional geometric features, a two-dimensional assembly feature recognition mode is proposed. Project the three-dimensional model wing parts into 6 two-dimensional views, use the contour recognition algorithm to recognize the contours of each projection view, and determine each joint part of the part in combination with the dimensions of the part model to avoid conflicts and collisions when recognizing the features of the same type of joint surface; the assembly features of the three-dimensional model are related to the view direction, rotate the 3D model to ensure that its main direction is the +X direction of the new coordinate system;

[0013] First, calculate the centroid coordinates of the three-dimensional model; establish a new coordinate system with the calculated centroid as the original coordinate point;

[0014]

[0015] Where: C m =(c m1 , c m2 , c m2 ) is the centroid of the three-dimensional model, g i =(g i1 , g i2 , g i3 ) is the centroid of the i-th divided triangle, Si is the area contained in the i-th partition triangle, and n is the number of triangles;

[0016] Secondly, we use the vertex set P of the 3D model to construct a standard set Q using the following transformation;

[0017]

[0018] In the formula: is the conjugate matrix of p j ; is the square of the area of the i-th triangle; p i1 is the column vector of p i ; p i is the i-th matrix; Q i is the standard matrix of p i ;

[0019] Obtain the covariance matrix of the 3D model:

[0020]

[0021] In the formula: U is the correlation matrix of the 3D model;

[0022] Calculate the eigenvalues and obtain their eigenvectors; by normalizing the eigenvectors and arranging them in descending order:

[0023] R = (u1, u2, u3) (17)

[0024] In the formula: u1, u2, and u3 are the eigenvectors of U, and R is the rotation matrix of the 3D model;

[0025] Calculate the new coordinates of the vertices of the 3D model. The coordinates of each vertex and the parameters of the triangles are obtained by traversing through the secondary development of CATIA; a simplified 3D model can be created through a series of Boolean operations and adapted to the initial 3D model to obtain the assembly features;

[0026] p i ' = R · p i (18)

[0027] In the formula: p i ' is the new coordinate of the vertex of the 3D model;

[0028] A new coordinate system is established using the above algorithm; the six view directions of the model are the +X, -X, +Y, -Y, +Z, and -Z directions of the new coordinate system;

[0029] In step 1), for the construction of the multi-component and multi-level twin model of the aircraft wing, the geometric model file is constructed by combining the manual construction using 3D software, the model file in standard format provided by the aircraft manufacturing plant, and the 3D point cloud reconstruction model; through on-site research at the aircraft manufacturing plant, the complex geometric model file obtained by manual construction using 3D software is lightweighted, and the preliminary construction of the geometric model is realized using the professional software CATIA.

[0030] Specifically, the rapid modeling of the CATIA assembly process involves the following two key points: (a) Planning the interaction sequence to determine when the parts move; (b) Motion planning to determine how the parts move; by implementing these two key points, the automation of the assembly simulation is achieved, and the specific implementation is described here:

[0031] 1) Identify the wing box features, and install the front rib, front wall 1, rear wall 1, wing beam 1, sheet metal ribs 1 to 20, front wall 2, rear wall 2, wing beam 2 in sequence, and finally install the stiffening rib 2; perform assembly modeling on the wing box components in sequence.

[0032] 2) The constraint and assembly operation process; the position, coincidence, and distance constraints of the assembled components can be realized. After the installation of the 4 wing box components of the wing is completed, the constraint points can already be seen in CATIA. Next, the update all update button needs to be clicked for update.

[0033] 3) Develop a rapid modeling assembly information system and integrate it into the computer-aided three-dimensional interactive application program CATIA; click the button corresponding to the component, and the picture of the corresponding component will pop up. When the CATIA real-time modeling interface detects the assembly of each component, input the measured dimension value, deviation, measured thickness value, and deviation, and output the assembly dimension information. When all the wing box skeletons of the wing are successfully assembled, the rapid modeling assembly information system synchronously outputs all the assembly tolerance information.

[0034] The specific operation process is as follows:

[0035] The first step: Run the wing box rapid modeling system, input the measured dimension value, deviation, measured thickness value, and deviation of the component, click the button corresponding to the component, and when the CATIA real-time modeling interface detects the assembly of each component, a component will pop up for each assembled component, and the assembly interface can be seen in real time.

[0036] The second step: Assemble the front rib jy-l-001, wing beam zq-01, rear wall zq-02, and front wall zq-03.

[0037] The third step: Click the update button to start the assembly. Use the constraint function to assemble two or more components together. After the update, the mating surfaces 1 and 1 are assembled together, 2 and 2 are assembled together, and 3 and 3 are assembled together.

[0038] Fourth step: Display the assembly dimension information in the rapid modeling dimension information system;

[0039] Fifth step: Then assemble the 5th part, sheet metal rib jy-l-002, the 6th part jy-l-003,..., jy-l-002 to jy-l-019 are all sheet metal ribs;

[0040] Sixth step: Before assembling the 13th part, front wall zq-04..., the 15th part, rear wall zq-07..., the 19th part, wing beam zq-06, the 20th part, front wall zq-05, and finally install the reinforcing rib jy-l-020;

[0041] On the interface of the rapid modeling assembly dimension information system, gather all the part dimension information of jy-l-001 to jy-l-020 and zq-01 to zq-07, and display all the dimension information;

[0042] After the preliminary construction of the geometric model is completed, optimize and render the model. Import the geometric model constructed manually by 3D software and the model after lightweight processing into 3dsMax to build the virtual scene of the digital twin workshop, and modify, render and optimize the model to obtain the geometric model, and then import it into the Posense Manager software and match it with UWB devices to serve as the physical workshop map of the assembly production workshop;

[0043] Develop a function-based computer-aided rapid modeling assembly information management system to identify the constraint positions of component assembly features, and combine with the A* algorithm to complete the assembly path planning;

[0044] Among them, the A* algorithm formula:

[0045] f(n) = g(n) + h(n) (1)

[0046] In the formula: f(n) is the estimated function from the initial point through node n to the target point, g(n) represents the actual distance from the starting point to a certain random point n, and h(n) represents the estimated distance of the best path from n to the target node;

[0047] This formula follows the following characteristics: If it is equal to zero, only g(n) is necessary, and the shortest path from the starting point, where n is any point, becomes a single-source shortest path problem; If h(n) is not greater than the actual distance from n to the target, the optimal solution is obtained;

[0048] The search process of the A* algorithm is defined as follows:

[0049] 1) Define two tables as OPEN and CLOSED; The OPEN table is used to store unvisited vertices; The CLOSED table is used to store visited vertices;

[0050] 2) Assume the graph has n vertices. Select a vertex as the starting point, the target point is a vertex, and the vertices are the remaining locations. Put the starting point into the OPEN list and initialize the CLOSED list.

[0051] 3) Determine if the open list is empty. If it is, it means the search process fails.

[0052] 4) If it is not empty, move the first vertex in the OPEN list to the CLOSED list.

[0053] 5) Judge if it is the target vertex. If it is, it means the search is successful and the algorithm continues to run.

[0054] 6) If it is not, expand the search for the sub-vertices of the vertex, calculate the value, and determine if it exists in the OPEN list or the CLOSED list:

[0055] a) If it does not exist in the OPEN list and the CLOSED list, select to store it in the OPEN list and set a pointer to the parent vertex.

[0056] b) If it exists in the OPEN list, update the value in the OPEN list. That is, select the smaller new value in the OPEN list instead of the larger old value and set a pointer to the parent vertex.

[0057] c) If it already exists in the closed list or they are obstacles, ignore this vertex and return to the first step; sort each vertex in the OPEN list in ascending order according to the value, and then jump to SPP which is the set of all key pose points for forming the path. is the matrix after spatial position transformation;

[0058] Specifically, the essence of the assembly path is composed of multiple connected key pose points; the pose matrix refers to the matrix containing the position information and attitude information of the components in the assembly environment, which is the internal information basis for translating and rotating the components in virtual assembly, that is:

[0059]

[0060] In the formula: SPP is the set of all key pose points for forming the path, is the matrix after spatial position transformation;

[0061]

[0062] In the formula: M1 is geometric transformation of translation, rotation, inversion and scaling; M2 is to generate translation transformation; M3 is to generate projection transformation; M4 is to generate scaling transformation;

[0063]

[0064]

[0065] Where: M zi is the pose matrix with the spatial position of the component unchanged; is the matrix after the spatial position transformation obtained through the CATIA / CAA platform;

[0066] Each part obtains a corresponding different pose matrix due to its different position, which includes the spatial position information of the part.

[0067] As a further solution of the present invention, in step 1), the production factors in the workshop include personnel, equipment, materials and environment. Among them, personnel include the simplified model of workshop operators, personnel position information and key operation information, equipment includes all the equipment participating in the production and processing process in the workshop, materials include production materials and workpieces to be processed, and their twin models include production material and workpiece models, RFID tag information, and processing data in the tag block. The environment includes other static objects in the workshop and workshop walls, and the physical workshop environment is highly restored through the rendering of geometric models.

[0068] As a further solution of the present invention, in step 1), according to the positioning requirements of production factors, RFID positioning tags are attached to materials for identification, regional positioning is performed on them through RFID fixed readers, and the production status information of the materials is obtained. UWB devices are used to accurately position and track the movement trajectories of vehicles, personnel, components, etc. There are many production factors and a complex environment in the assembly production workshop. Accurately positioning personnel through UWB devices can effectively obtain information such as the real-time position, distribution, movement trajectory, and regional residence time of personnel.

[0069] As a further solution of the present invention, in step 2), the multi-source heterogeneous data includes eight categories: production personnel data, instrument and equipment data, tooling data, assembly logistics data, assembly progress data, assembly quality data, actual working hours data, and reverse problem data.

[0070] As a further solution of the present invention, the data required in the digital twin workshop client is divided into real-time drive data, panel display data and historical processing data.

[0071] As a further solution of the present invention, in step 2), the asynchronous data interaction between the digital twin workshop client and the server means that for the instruction issued by the digital twin workshop client, after the server receives the instruction, it connects to the physical workshop to achieve remote control of the physical workshop, so as to realize functions such as real-time status display, status evolution, data input, and data saving of the digital aircraft assembly process; through the interaction with the database, the real-time drive data and the panel display data are stored in the database, and are continuously updated and covered through high-speed data acquisition. The unified data service drives the three-dimensional virtual models of the assembled parts and the three-dimensional model of the product, which is conducive to the reproduction of the historical processing status of the digital twin workshop client, and enables the digital twin workshop and the physical workshop to achieve two-way dynamic mapping.

[0072] As a further solution of the present invention, a joint in-depth study is carried out on the digital aircraft assembly process based on the MBD technology. The MBD technology is combined with the aircraft assembly process design to realize the digital transmission of assembly process information and the parallel cooperation in the assembly process design. Under the MBD manufacturing mode, a three-dimensional assembly process on-site application system for spacecraft is established to complete the construction of 3D-AO. In the preparation stage of assembly, by watching the animation of the assembly process and operating the digital assembly application terminal, the assembly instruction of the process information is described in the form of 3D views and 3D animations, and the structured text data information is supplemented.

[0073] As a further solution of the present invention, an automatic assembly simulation method based on interaction feature pairs is adopted. Random motion planning is carried out in the established C space to find a collision-free path for each part, and the planning result in the C space is converted into the part motion in the simulation environment. The parts interact with each other according to a certain interaction sequence to automatically simulate the assembly process of the product.

[0074] Compared with the prior art, the present invention has the following beneficial effects:

[0075] The present invention discloses a construction method of an intelligent assembly system based on digital twin. Through the multi-element and multi-level twin modeling of the physical workshop and the three-dimensional modeling of multiple components and multiple levels of the aircraft wing, the accurate mapping between the physical space and the virtual space of the digital twin of the production elements in the workshop is realized; through the unified data service to drive the three-dimensional virtual model of the assembly production line and the three-dimensional model of the product, the generation and continuous update of the digital twin instance of the assembled product and the digital twin instance of the assembly space are realized; through the integration of the three-dimensional visualization system, the state monitoring and process optimization feedback control of the digital twin are realized, and the personnel and operations on site are comprehensively and effectively supervised in real time, so that the aircraft wing assembly in the real world is seamlessly integrated with the rapid modeling of the aircraft wing assembly simulation in the virtual environment. In addition, by constructing a three-dimensional visualization system, in-depth research is carried out from new perspectives such as intelligence, modularity, and automation, the application efficiency of the assembly rapid modeling system is comprehensively improved, the scenarios and display modes of digital applications are further expanded, and a foundation is laid for future new application scenarios and new spaces. The assembly process driven by digital twin will be based on the Internet of Things integrating all equipment, realizing the deep integration of the physical world and the information world of the assembly process. Through the intelligent software service platform and tools, the intelligent planning, simulation and optimization of components, equipment and assembly processes are realized, the complex product assembly process is uniformly and efficiently controlled, and the self-organization, self-adaptation and dynamic response of the product assembly system are achieved, so as to achieve the purpose of virtual-real fusion and virtual control of the real. Realizing the real-time, accurate acquisition, effective information extraction and reliable transmission of multi-source heterogeneous data in the complex dynamic entity space is the prerequisite for realizing the digital twin. Therefore, based on the efficient and realistic twin model, the present invention realizes the real-time interaction and fusion of virtual and real data, keeps the virtual workshop highly consistent with the physical workshop, realizes the two-way mapping of real-time data, reduces the complexity of the assembly sequence planning during the assembly process of aircraft components, shortens the assembly cycle of the aircraft, and improves the efficiency and practicality of the assembly planning.

[0076] In summary, the present invention provides a construction method of an intelligent assembly system based on digital twin. In the prior art, the construction of digital twin workshops such as aviation is mostly explored from the perspective of data fusion or three-dimensional monitoring platforms, and the complete construction method of digital twin workshops has not been studied. The present invention proposes a complete and feasible solution for the digital twin workshop from the construction of the twin model, the interaction of virtual and real data to the visualization integration of the digital twin workshop client of virtual simulation, so as to completely twin the rapid modeling of the aircraft wing and automatically simulate the assembly process of the product.

[0077] Through the construction of full-element and multi-level twin models for the physical workshop and aircraft wings, the personnel, equipment, environment, and aircraft wing component models in the digital twin workshop highly restore the physical workshop at multiple levels. Highly realistic simulations are carried out from the appearance rendering to the internal operation rules, achieving the maximum consistency between the physical workshop and the digital twin workshop, and improving the simulation degree of the assembly production site of complex equipment products such as aircraft. In addition, the virtual-real data interaction framework of the present invention is perfect, optimizing the existing digital twin workshop data architecture. BRIEF DESCRIPTION OF THE DRAWINGS

[0078] Figure 1 is the overall framework diagram of the present invention;

[0079] Figure 2 is the data interaction architecture diagram of the present invention;

[0080] Figure 3 is the assembly sequence planning flowchart of the present invention;

[0081] Figure 4 is the deployment diagram of RFID and UWB in the assembly workshop of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0082] In order to enable those skilled in the art to better understand the solution of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without making creative efforts shall fall within the protection scope of the present invention.

[0083] The present invention will be further described in detail below in conjunction with the accompanying drawings:

[0084] Refer to Figure 1 As shown, a construction method of an intelligent assembly system based on digital twin includes three parts: the construction of full-element and multi-component multi-level twin models for the workshop and aircraft wings, the fusion and interaction of virtual-real data, and the integration of a three-dimensional visualization system. The construction of full-element and multi-component multi-level twin models for the workshop and aircraft wings is the physical entity basis of the entire digital twin system; after the construction of full-element and multi-component multi-level twin models for the workshop and aircraft wings, the maximum consistency between the physical workshop and the digital twin workshop is achieved, and the fusion and interaction of virtual-real data are carried out to realize the two-way mapping between the constructed twin models and the physical workshop; finally, the integration of a three-dimensional visualization system is added to make the twin models achieve a more realistic effect, so that the digital twin virtual assembly model has a "real" simulation of the physical assembly process of the product with real-time update and strong sense of presence, thereby completing the construction of the digital twin system of the intelligent workshop.

[0085] The present invention will be further described in detail below in conjunction with specific embodiments:

[0086] 1 Construct a multi-component and multi-level twin model of all elements in the workshop and the aircraft wing

[0087] 1.1 By analyzing and decomposing numerous production elements in the physical workshop, the production elements in the physical workshop can be divided into personnel, equipment, materials, and environment, and the multi-level twin models of all elements in the workshop are respectively modeled.

[0088] Among them, personnel include a simplified model of workshop operators, radio frequency identification, tag information, ultra-wideband technology, personnel location information, and key operation information. According to the positioning requirements of production elements, the present invention attaches RFID positioning tags to materials for identification, uses RFID fixed readers to perform regional positioning on them, and obtains the production status information of materials. UWB devices are used to accurately position and track the movement trajectories of vehicles, personnel, and components. There are numerous production elements and a complex environment in the assembly production workshop. By accurately positioning personnel through UWB devices, information such as the real-time position, distribution, movement trajectory, and regional residence time of personnel can be effectively obtained;

[0089] Equipment includes all equipment participating in the production and processing process in the workshop. Combining the characteristics and requirements of the product production site, Internet of Things technologies such as bar code technology, RFID, and sensors are used to identify assembly resource information, and UWB devices are used to accurately position and track the movement trajectories of vehicles, personnel, components, etc. The information collection points for the perception of the assembly process are designed, and an assembly Internet of Things network is constructed in the assembly workshop to achieve real-time perception of assembly resources;

[0090] Materials include production materials and processed workpieces, and their twin models include production material and processed workpiece models, RFID tag information, and processing data within the tag block;

[0091] The environment includes other static objects in the workshop, workshop walls, etc., and the physical workshop environment is highly restored through the rendering of geometric models.

[0092] 1.2 Construction of a multi-component and multi-level twin model of the aircraft wing.

[0093] Specifically, in this embodiment, the geometric model file can be constructed by combining the manual construction using 3D software, the model file in standard format provided by the aircraft manufacturing factory, and the 3D point cloud reconstruction model.

[0094] Specifically, through on-site research at the aircraft manufacturing factory, the complex geometric model file obtained by manual construction using 3D software is subjected to lightweight processing, and the preliminary construction of the geometric model is realized using CATIA.

[0095] After the preliminary construction of the geometric model is completed, the model is optimized and rendered. The geometric model manually constructed by 3D software and the model after lightweight processing are imported into 3ds Max to build the virtual scene of the digital twin workshop. Then, the model is modified, rendered, and optimized to obtain the geometric model, which is imported into the Posense Manager software and paired with UWB devices to serve as the physical workshop map of the assembly production workshop.

[0096] 2 Virtual-real data interaction and fusion method

[0097] As Figure 2 shown, under the local area network, through the efficient data transmission of real-time data between the data acquisition system, database (multi-database collaboration), server, message middleware, and the digital twin workshop model constructed in step (1), the multi-source heterogeneous data in the physical workshop is collected, preprocessed, and standardized. Through the transmission and interaction of data, the digital twin workshop model constructed in step (1) is mapped bidirectionally with the physical workshop, and virtual-real data fusion and interaction are carried out to achieve the virtual-real data fusion and interaction of the digital twin workshop, and further realize the real-time mapping and remote control of the virtual production line and the physical production line.

[0098] Using the CATIA secondary development technology, attaching PYTHON scripts, asynchronously interacting with the server through the script program, writing PYTHON scripts to drive the operation of the digital twin workshop, checking the operation status of the digital twin workshop under given information, and making processing adjustments to the real physical workshop according to the simulation results.

[0099] The secondary development interface method is used to collect data on the real-time situations of AGV, UWB, and aircraft wing assembly in the workshop, construct a data acquisition system, and at the same time store the data in the real-time database and the relational database. At the same time, the server remotely controls the underlying equipment in the workshop and issues user instructions through the communication interface and the secondary development interface.

[0100] 3 Three-dimensional visualization system integration

[0101] The three-dimensional visualization system integration of the digital twin workshop constructed in step (1) includes four parts: scene roaming, monitoring panel display, human-computer interaction, and AR support. The assembly process simulation aims to visually display the assemblability of products through visual methods (such as computer graphics, model technology).

[0102] Specifically, in this embodiment, the three-dimensional visualization system integration of the digital twin workshop constructed in step (1) is implemented based on PYTHON programming.

[0103] For the human-computer interaction, control instructions are sent to the physical workshop by clicking the buttons in the digital twin workshop interface and entering commands.

[0104] The AR support means that the digital twin workshop client supports the application of AR devices. Through the AR glasses, the physical workshop scene can be roamed, the device status can be presented autonomously, and human-computer interaction can be achieved through voice and gesture methods to view the device status and perform remote operations, enhancing the immersion and interactivity.

[0105] The following further elaborates on the design of the 3D visualization system with specific embodiments:

[0106] Step 1: Assembly sequence planning;

[0107] As Figure 3 shown, when planning the assembly sequence, a bottom-up hierarchical plan is adopted. By separating each layer and creating an assembly direction diagram, the complex wing box model is decomposed into smaller numbers and simpler-structured assembly models.

[0108] Step 1.1: Improved A* algorithm in the MBD environment;

[0109] Specifically, the MBD technology is introduced to establish an MBD dataset environment for assembly processes. Global path planning is carried out according to the environmental information to search for the optimal path. In global path search, the most effective method is the A* algorithm. As a typical heuristic priority search algorithm, the A* algorithm effectively solves the point-to-point problem in path planning.

[0110] A* algorithm formula:

[0111] f(n) = g(n) + h(n) (1)

[0112] In the formula: f(n) is the estimated function from the initial point through node n to the target point, g(n) represents the actual distance from the starting point to a certain random point n, and h(n) represents the estimated distance of the best path from n to the target node.

[0113] This formula follows the following characteristics: If it is equal to zero, only g(n) is necessary, and the shortest path from the starting point, where n is any point, becomes a single-source shortest path problem; if h(n) is not greater than the actual distance from n to the target, the optimal solution is obtained. The search process of the A* algorithm is defined as follows:

[0114] 1) Define two tables as OPEN and CLOSED. The OPEN table is used to store unvisited vertices. The CLOSED table is used to store visited vertices;

[0115] 2) Assume that the graph has n vertices, select vertex V s as the starting point, the target point V d is a vertex, and vertex V i is the remaining locations. Put the starting point into the OPEN table and initialize the CLOSED table;

[0116] 3) Determine whether the open list is empty. If it is empty, it indicates that the search process has failed;

[0117] 4) If it is not empty, move the first vertex V in the OPEN list i to the CLOSED list;

[0118] 5) Determine whether V i is the target vertex V d . If it is, it indicates that the search is successful and the algorithm continues to run;

[0119] 6) If it is not, expand the search for the sub-vertices V i of vertex V j (j <= n, j ≠ i), calculate the value of F(V j ), and determine whether V j exists in the OPEN list or the CLOSED list:

[0120] a) If V j does not exist in the OPEN list and the CLOSED list, select to store it in the OPEN list and set a pointer to the parent vertex V i ;

[0121] b) If V j exists in the OPEN list, update the value of F(V j ) in the OPEN list. That is to say, select the smaller new value F(V j ) in the OPEN list, rather than the larger old value F(V j ), and set a pointer to the parent vertex;

[0122] c) If V j already exists in the closed list or they are obstacles, ignore this vertex and return to the first step. Sort each vertex in the OPEN list in ascending order according to the value of F(V j ), and then jump to (3).

[0123] Step 1.2, Model Pose Space Transformation Algorithm;

[0124] Specifically, the essence of the assembly path is composed of multiple key pose points connected. The pose matrix refers to the matrix containing the position information and attitude information of the components in the assembly environment, and it is the internal information basis for translating and rotating the components in virtual assembly, that is:

[0125]

[0126] In the formula: SPP is the set of all key pose points that make up the path, is the matrix after spatial position transformation.

[0127]

[0128] Where: M1 is a geometric transformation of translation, rotation, inversion, and scaling; M2 is a translation transformation; M3 is a projection transformation; M4 is a scaling transformation.

[0129]

[0130] Where: M zi is the pose matrix with the spatial position of the component unchanged; is the matrix after the spatial position transformation obtained through the CATIA / CAA platform.

[0131] Each part obtains a corresponding different pose matrix due to its different position, which includes the spatial position information of the part.

[0132] Step 2: Standardize the assembly model;

[0133] Specifically, since the recognition of three-dimensional geometric assembly features is more complex than that of two-dimensional geometric features, a two-dimensional assembly feature recognition mode is proposed. The wing parts of the three-dimensional model are projected into 6 two-dimensional views, and the contour recognition algorithm is used to recognize the contours of each projection view. The various combination parts of the part are determined by combining the dimensions of the part model to avoid conflicts and collisions when recognizing the features of the same type of joint surface. The assembly features of the three-dimensional model are related to the view direction. Rotate the 3D model to ensure that its main direction is the +X direction of the new coordinate system.

[0134] First, calculate the centroid coordinates of the three-dimensional model. Establish a new coordinate system with the calculated centroid as the original coordinate point.

[0135]

[0136] Where: C m =(c m1 , c m2 , c m2 ) is the centroid of the three-dimensional model, g i =(g i1 , g i2 , g i3 ) is the centroid of the i-th divided triangle, S i is the area contained in the i-th divided triangle, and n is the number of triangles.

[0137] Secondly, we use the vertex set P of the three-dimensional model to construct a standard set Q using the following transformation.

[0138]

[0139] Where: is pj Conjugate matrix of; is the square of the area of the i-th triangle; p i1 is p i column vector of; p i is the i-th matrix; Q i is p i standard matrix of.

[0140] Obtain the covariance matrix of the 3D model:

[0141]

[0142] In the formula: U is the correlation matrix of the 3D model.

[0143] Calculate the eigenvalues obtained, and obtain its eigenvectors. By normalizing the eigenvectors and arranging them in descending order:

[0144] R = (u1, u2, u3) (17)

[0145] In the formula: u1, u2, and u3 are the eigenvectors of U, and R is the rotation matrix of the 3D model.

[0146] Calculate the new coordinates of the vertices of the 3D model. The coordinates of each vertex and the parameters of the triangle are obtained through traversal of CATIA secondary development. A simplified 3D model can be created through a series of Boolean operations, and it is adapted to the initial 3D model to obtain assembly features.

[0147] p i ′ = R · p i (18)

[0148] In the formula: p i ′ is the new coordinate of the vertex of the 3D model.

[0149] A new coordinate system is established using the above algorithm. The six view directions of the model are the +X, -X, +Y, -Y, +Z, and -Z directions of the new coordinate system.

[0150] Step 3: Automatically plan the path for assembly operation;

[0151] Specifically, rapid modeling in the CATIA assembly process involves the following two key points: (a) Planning the interaction sequence to determine when the parts move; (b) Motion planning to determine how the parts move. By implementing these two key points, the automation of assembly simulation is achieved, and the specific implementation is described here:

[0152] 1) Identify the wing box features and install the front rib, front wall 1, rear wall 1, wing beam 1, sheet metal ribs 1 to 20, front wall 2, rear wall 2, wing beam 2, and finally install the strengthening rib 2 in sequence. Perform assembly modeling on the wing box components in sequence;

[0153] 2) Constraints and assembly operation process. It can achieve position, mating, and distance constraints for assembled components. After the installation of 4 wing box components is completed, the constraint points can already be seen in CATIA. Next, it is necessary to click the update all update button for updating;

[0154] 3) Develop a rapid modeling assembly information system and integrate it into the computer-aided three-dimensional interactive application program (CATIA). Click the button corresponding to the component, and the picture of the corresponding component will pop up. When the CATIA real-time modeling interface detects the assembly of each component, input the measured size value, deviation, measured thickness value, and deviation, and output the assembly size information. When all the wing box skeletons are successfully assembled, the rapid modeling assembly information system synchronously outputs all the assembly tolerance information.

[0155] The specific operation process is as follows:

[0156] (a) Run the wing box rapid modeling system, input the measured size value, deviation, measured thickness value, and deviation of the component, click the button corresponding to the component. When the CATIA real-time modeling interface detects the assembly of each component, a component will pop up for each assembled component, and the assembly interface can be seen in real time;

[0157] (b) Assemble the front rib jy-l-001, wing beam zq-01, rear wall zq-02, and front wall zq-03;

[0158] (c) Click the update button to start the assembly. Use the constraint function to assemble two or more components together. After updating, the mating surfaces 1 and 1 are assembled together, 2 and 2 are assembled together, and 3 and 3 are assembled together;

[0159] (d) Display the assembly size information in the rapid modeling dimension information system;

[0160] (e) Then assemble the 5th component, sheet metal rib jy-l-002, the 6th component jy-l-003,..., and jy-l-002 to jy-l-019 are all sheet metal ribs;

[0161] (f) Assemble the 13th component, front wall zq-04..., the 15th component, rear wall zq-07..., the 19th component, wing beam zq-06, the 20th component, front wall zq-05, and finally install the reinforcing rib jy-l-020.

[0162] In the interface of the rapid modeling assembly dimension information system, gather all the component dimension information of jy-l-001 to jy-l-020 and zq-01 to zq-07, and display all the dimension information.

[0163] In summary, through the construction of a digital twin workshop and the integration of a 3D visualization system, the integration of functions such as real-time status display, historical processing status reproduction, wing assembly motion simulation, remote control, and processing situation prediction is achieved. Among them, in the MBD environment, combined with the automatic recognition process of assembly features, the assembly model is standardized, and the contour recognition algorithm is used for feature recognition. The A* algorithm is adopted to obtain the optimal assembly path. When the relationships among geometric features, assembly features, and constraints are correctly obtained, a fast modeling assembly information display system is developed with the improved A* algorithm - pose transformation algorithm as the core. The required functions are integrated into the interface of the fast modeling assembly information display system for interactive fast modeling and the display of more assembly information. Experimental verification is carried out on the assembly process. Using the improved A* algorithm - pose transformation algorithm, the assembly cycle is shortened by 37.398%; the staged assembly efficiency in the reverse modeling stage is increased by 56.370%; the number of errors using the improved A* algorithm - pose transformation algorithm is 71, a reduction of 32.381%. Compared with the pose matrix algorithm, the efficiency and accuracy of the wing box assembly are improved, thus verifying the effectiveness of this method.

[0164] Therefore, the present invention provides a construction method for an intelligent assembly system based on digital twins, specifically including the multi-level modeling of all elements of the physical workshop, the fusion and interaction of virtual and real data, and the integration of a 3D visualization interaction system. First, multiple software are comprehensively used to model all production elements of the physical workshop, namely personnel, equipment, environment, and aircraft wing parts at multiple levels, so that the virtual workshop maximally restores the real physical workshop to obtain a digital twin workshop model. Secondly, a data acquisition system, a database, a database server, a message middleware, and the obtained digital twin workshop model are used as the client of 3D virtual simulation to collect, preprocess, and standardize multi-source heterogeneous data, and through the high-speed transmission and interaction of data, the real-time mapping and remote control of the virtual production line and the physical production line are realized. Finally, the 3D visualization system integration of the digital twin workshop is realized through four methods: scene roaming, monitoring panel display, human-computer interaction, and AR support, enabling the deep fusion and interaction of the physical space and the virtual space, so as to realize functions such as real-time status display, historical processing status reproduction, wing assembly motion simulation, and remote control.

[0165] The above content is only to illustrate the technical idea of the present invention and cannot be used to limit the protection scope of the present invention. Any modification made on the basis of the technical solution according to the technical idea proposed by the present invention falls within the protection scope of the claims of the present invention.

Claims

1. A construction method of an intelligent assembly system based on digital twin, characterized in that, Including: 1) For the production elements of the physical assembly workshop, construct a multi-level twin model of all elements in the physical workshop, collect real-time dynamic data of the assembly process in the physical space, perform three-dimensional modeling of multiple components and levels of the aircraft wing, and achieve an accurate mapping between the digital twin body's physical space and virtual space in the workshop production elements, obtaining a digital twin workshop model; 2) Use the digital twin workshop model as the system client, construct a data acquisition system to collect multi-source heterogeneous data from the physical workshop, preprocess and standardize the multi-source heterogeneous data and then store it in the database; The digital twin workshop model, as the system client, through interaction with the database, drives the three-dimensional virtual models of the assembled parts and the product three-dimensional model with unified data services, enabling two-way dynamic mapping between the digital twin workshop and the physical workshop; 3) Based on the integration of the three-dimensional visualization system, realize the state monitoring and process optimization feedback control of the digital twin body, comprehensively and real-timely and effectively supervise the on-site personnel and operations, and seamlessly integrate the actual aircraft wing assembly with the aircraft wing assembly simulation in the virtual environment for rapid modeling, completing the construction of the intelligent workshop digital twin system; Specifically, since the recognition of three-dimensional geometric assembly features is more complex than two-dimensional geometric features, a two-dimensional assembly feature recognition mode is proposed. Project the three-dimensional model of the wing part into 6 two-dimensional views, use the contour recognition algorithm to recognize the contours of each projection view, and determine each joint part of the part in combination with the various dimensions of the part model to avoid conflicts and collisions when recognizing the features of the same type of joint surface; the assembly features of the three-dimensional model are related to the view direction. Rotate the 3D model to ensure that its main direction is the +X direction of the new coordinate system; First, calculate the centroid coordinates of the three-dimensional model; establish a new coordinate system with the calculated centroid as the original coordinate point; Where: C m =(c m1 , c m2 , c m2 ) is the centroid of the 3D model, g i =(g i1 , g i2 , g i3 ) is the centroid of the i-th divided triangle, S i is the area contained in the i-th divided triangle, and n is the number of triangles; Secondly, we use the vertex set P of the three-dimensional model to construct a standard set Q, using the following transformation; p i = (p i1 , p i2 , p i3 , ……) (14) Q i = (q i1 , q i2 , q i3 , ……) (15) In the formula: is the conjugate matrix of p j ; is the square of the area of the i-th triangle; p i1 is the column vector of p i ; p i is the i-th matrix; Q i is the standard matrix of p i ; Obtain the covariance matrix of the three-dimensional model: In the formula: U is the correlation matrix of the 3D model; Calculate the eigenvalues and obtain their eigenvectors; arrange the eigenvectors in descending order through normalization: R=(u1,u2,u3) (17) In the formula: u1, u2 and u3 are the eigenvectors of U, and R is the rotation matrix of the 3D model; Calculate the new coordinates of the vertices of the three-dimensional model. The coordinates of each vertex and the parameters of the triangle are obtained through traversal of the secondary development of CATIA; a simplified 3D model can be created through a series of Boolean operations and adapted to the initial 3D model to obtain assembly features; p i p' = R·p i (18) Where: p i ' is the new coordinate of the vertex of the 3D model; A new coordinate system is established using the above algorithm; the six view directions of the model are the +X, -X, +Y, -Y, +Z and -Z directions of the new coordinate system; In step 1), for the construction of the multi-component and multi-level twin model of the aircraft wing, the geometric model file is constructed by combining the methods of manual construction by three-dimensional software, providing model files in standard format by the aircraft manufacturing factory, and three-dimensional point cloud reconstruction model; Conduct on-site research at the aircraft manufacturing factory, perform lightweight processing on the complex geometric model file obtained through manual construction by three-dimensional software, and use professional CATIA software to realize the preliminary construction of the geometric model; Specifically, the rapid modeling of the CATIA assembly process involves the following two key points: (a) Interaction order planning determines when the parts move; (b) Motion planning determines how the parts move; By implementing these two key points, the automation of the assembly simulation is achieved, and the specific implementation is described here: 1) Identify the wing box features and install the front rib, front wall 1, rear wall 1, spar 1, sheet metal ribs 1 to 20, front wall 2, rear wall 2, spar 2, and finally install the stiffening rib 2 in sequence; Perform assembly modeling on the wing box components in sequence. 2) Constrain and run the assembly process; It is possible to achieve position, coincidence, and distance constraints for the assembly components. After the installation of the 4 wing box components of the wing is completed, the constraint points can already be seen in CATIA. Next, it is necessary to click the update all update button for updating. 3) Develop a rapid modeling assembly information system and integrate it into the computer-aided three-dimensional interactive application program CATIA; Click the button corresponding to the component, and the picture of the corresponding component will pop up. When the CATIA real-time modeling interface detects each component of the assembly, input the measured dimension value, deviation, measured thickness value, and deviation, and output the assembly dimension information. When all the wing box skeletons are successfully assembled, the rapid modeling assembly information system synchronously outputs all the assembly tolerance information. The specific operation process is as follows: The first step: Run the wing box rapid modeling system, input the measured dimension value, deviation, measured thickness value, and deviation of the component, click the button corresponding to the component. When the CATIA real-time modeling interface detects each component of the assembly, a component will pop up for each assembled component, and the assembly interface can be seen in real time. The second step: Assemble the front rib jy-l-001, spar zq-01, rear wall zq-02, and front wall zq-03. The third step: Click the update button to start the assembly. Use the constraint function to assemble two or more components together. After the update, mating surface 1 and 1 are assembled together, 2 and 2 are assembled together, and 3 and 3 are assembled together. The fourth step: Display the assembly dimension information in the rapid modeling dimension information system. The fifth step: Then assemble the 5th component, sheet metal rib jy-l-002, the 6th component jy-l-003,..., jy-l-002 to jy-l-019 are all sheet metal ribs. The sixth step: Assemble the 13th component, front wall zq-04..., the 15th component, rear wall zq-07..., the 19th component, spar zq-06, the 20th component, front wall zq-05, and finally install the stiffening rib jy-l-020. In the interface of the rapid modeling assembly dimension information system, gather all the component dimension information of jy-l-001 to jy-l-020 and zq-01 to zq-07, and display all the dimension information. After the preliminary construction of the geometric model is completed, the model is optimized and rendered. The geometric model constructed manually by 3D software and the model after lightweight processing are imported into 3dsMax to build the virtual scene of the digital twin workshop. Then, the model is modified, rendered, and optimized to obtain the geometric model, which is imported into the Posense Manager software and paired with UWB devices to serve as the physical workshop map of the assembly production workshop; Develop a function-based computer-aided rapid modeling assembly information management system to identify the constrained positions of component assembly features and complete the assembly path planning in combination with the A* algorithm; Among them, the A* algorithm formula: f(n) = g(n) + h(n) (1) In the formula: f(n) is the estimated function from the initial point through node n to the target point, g(n) represents the actual distance from the starting point to a random point n, and h(n) represents the estimated distance of the best path from n to the target node; This formula follows the following characteristics: If it is equal to zero, only g(n) is necessary, and the shortest path from the starting point, where n is any point, becomes a single-source shortest path problem; if h(n) is not greater than the actual distance from n to the target, the optimal solution is obtained; The search process of the A* algorithm is defined as follows: 1) Define two tables as OPEN and CLOSED; the OPEN table is used to store unvisited vertices; the CLOSED table is used to store visited vertices; 2) Assume that the graph has n vertices. Select a vertex as the starting point, the target point is a vertex, and the vertices are the remaining locations. Put the starting point into the OPEN table and initialize the CLOSED table; 3) Determine whether the open table is empty. If it is empty, it means the search process fails; 4) If it is not empty, move the first vertex in the OPEN table to the CLOSED table; 5) Determine whether it is the target vertex. If it is, it means the search is successful and the algorithm continues to run; 6) If it is not, expand the search for the sub-vertices of the vertex, calculate the value, and determine whether it exists in the OPEN table or the CLOSED table: a) If it does not exist in the OPEN table and the CLOSED table, select to store it in the OPEN table and set a pointer to the parent vertex; b) If it exists in the OPEN table, update the value in the OPEN table. That is, select the smaller new value in the OPEN table instead of the larger old value and set a pointer to the parent vertex; c) If they already exist in the closed list or they are obstacles, ignore this vertex and return to the first step; sort each vertex in the OPEN list in ascending order according to the value, and then jump to SPP which is the set of all key pose points that make up the path. is the matrix after the spatial position transformation; Specifically, the essence of the assembly path is composed of multiple key pose points connected. The pose matrix refers to the matrix containing the position information and attitude information of components in the assembly environment and is the internal information basis for translating and rotating components in virtual assembly, that is: where: SPP is the set of all key pose points that make up the path, is the matrix after spatial position transformation; M4 = [a 44 (7) In the formula: M1 is the geometric transformation of translation, rotation, inversion, and scaling; M2 is the generation of translation transformation; M3 is the generation of projection transformation; M4 is the generation of scaling transformation; Where: M zi is the pose matrix with the spatial position of the component unchanged; is the matrix after the spatial position transformation obtained through the CATIA / CAA platform; Each part obtains a corresponding different pose matrix due to its different positions, which includes the spatial position information of the part.

2. The construction method of an intelligent assembly system based on digital twin according to claim 1, characterized in that In step 1), the production elements in the workshop include personnel, equipment, materials, and environment. Among them, personnel include the simplified model of workshop operators, personnel location information, and key operation information; equipment includes all the equipment involved in the production and processing process in the workshop; materials include production materials and workpieces to be processed, and their twin models include production material and workpiece models, RFID tag information, and processing data within the tag block; environment includes other static objects in the workshop and workshop walls, and the highly realistic physical workshop environment is achieved through the rendering of geometric models.

3. The construction method of an intelligent assembly system based on digital twin according to claim 1, characterized in that, In step 1), according to the positioning requirements of production elements, RFID positioning tags are attached to materials for identification, and regional positioning is carried out on them through RFID fixed readers and writers, and the production status information of materials is obtained. UWB devices are used to accurately position and track the movement trajectories of carriers, personnel, and components. There are numerous production elements and a complex environment in the assembly production workshop. Accurately positioning personnel through UWB devices can effectively obtain the real-time position, distribution, movement trajectory, and regional residence time information of personnel.

4. A construction method of an intelligent assembly system based on digital twin according to claim 1, characterized in that, In step 2), the multi-source heterogeneous data includes eight categories: production personnel data, instrument and equipment data, tooling data, assembly logistics data, assembly progress data, assembly quality data, actual working hours data, and reverse problem data.

5. The construction method of an intelligent assembly system based on digital twin according to claim 1, characterized in that The data required in the digital twin workshop client is divided into real-time drive data, panel display data, and historical processing data.

6. A construction method of an intelligent assembly system based on digital twin according to claim 1, characterized in that, In step 2), the asynchronous data interaction between the digital twin workshop client and the server means that for the instruction issued by the digital twin workshop client, after the server receives the instruction, it connects to the physical workshop to achieve remote control of the physical workshop, thereby realizing functions such as real-time status display, status evolution, data input, and data saving during the digital aircraft assembly process; through the interaction with the database, the real-time drive data and panel display data are stored in the database and continuously updated and overwritten through high-speed data acquisition. Driving the three-dimensional virtual models of assembly parts and the three-dimensional model of the product with unified data services is conducive to the reproduction of the historical processing status of the digital twin workshop client, enabling the digital twin workshop and the physical workshop to achieve two-way dynamic mapping.

7. A method for constructing an intelligent assembly system based on digital twin according to claim 1, characterized in that Based on the MBD technology, a joint in-depth study is carried out on the digital aircraft assembly process. The MBD technology is combined with the design of the aircraft assembly process to achieve digital transmission of assembly process information and parallel collaboration in the assembly process design. Under the MBD manufacturing mode, a three-dimensional assembly process on-site application system for spacecraft is established to complete the construction of 3D-AO. In the preparation stage of assembly, by watching the animation of the assembly process and operating the digital assembly application terminal, the assembly instruction of process information is described in the form of 3D views and 3D animations to supplement the structured text data information.

Images