Industrial digital genealogy system and industrial digital genealogy generation method

Abstract

The invention provides an industrial digital genealogy system and an industrial digital genealogy generation method. The method comprises the following steps: establishing a digital twin product and a digital twin space according to physical interaction information; generating rich geometric representation corresponding to the product, rich feature representation corresponding to the product, environment configuration corresponding to the scene and equipment configuration corresponding to the scene according to the physical interaction information; generating a production object digital genealogy corresponding to the production object according to the digital twin product, the rich geometric representation and the rich feature representation; generating a production space digital genealogy corresponding to the production space according to the digital twin space, the environment configuration and the equipment configuration; according to the digital genealogy, the target digital genealogy is established, a large amount of three-dimensional digital scene data can be generated by combining various random combinations of different parts and assemblies in a production object in a diversified production space, so that an industrial body intelligent world model can learn and train in the digital genealogy, and the adaptability and generalization of the model are improved.

Description

Technical Field

[0001] This application relates to the field of intelligent manufacturing technology, and in particular, to an industrial digital genealogy system and a method for generating an industrial digital genealogy. Background Art

[0002] With the rapid development of large model technology, general large models have achieved remarkable application results in many fields, especially in natural language processing, computer vision and other fields, where extensive breakthroughs have been made. However, general large models are usually disembodied. They exist in the digital domain and cannot directly drive physical devices for embodied perception and intelligent operation. The manufacturing industry has also entered the era of large models. Compared with the disembodied intelligence of general large models, there are a large number of physical production devices that need to be controlled in the manufacturing industry. Therefore, industrial embodied intelligence is required to exert influence in the physical world by controlling physical terminal devices.

[0003] In related technologies, a world model is introduced into the industrial field, and an intelligent agent is trained by constructing a virtual environment simulator, so as to realize industrial embodied control. By learning physical laws such as object dynamics and spatial relationships, the world model provides prior knowledge for industrial embodied intelligence by creating an abstract representation of the external world, thus helping the embodied intelligent agent to perform intelligent perception, planning and decision-making in the physical world.

[0004] However, the existing technology lacks sufficient trainable data and scenarios and cannot be applied to the training of the industrial embodied intelligence world model. Summary of the Invention

[0005] This application provides an industrial digital genealogy system and a method for generating an industrial digital genealogy to solve the technical problem that the existing technology lacks sufficient trainable data and scenarios and cannot be applied to the training of the industrial embodied intelligence world model.

[0006] In a first aspect, this application provides a method for generating an industrial digital genealogy, including: obtaining physical interaction information; establishing a digital twin product and a digital twin space according to the physical interaction information; generating a rich geometric representation corresponding to the product, a rich feature representation corresponding to the product, an environment configuration corresponding to the scenario, and a device configuration corresponding to the scenario according to the physical interaction information; generating a production object digital genealogy corresponding to the production object according to the digital twin product, the rich geometric representation and the rich feature representation; generating a production space digital genealogy corresponding to the production space according to the digital twin space, the environment configuration and the device configuration; and establishing a target digital genealogy according to the production object digital genealogy and the production space digital genealogy, where the target digital genealogy is used for the training of the industrial embodied intelligence world model.

[0007] The present application provides a method for generating an industrial digital family tree. The digital family tree is a new concept proposed in the present application, which provides new possibilities for the construction of an industrial embodied intelligence world model from two perspectives. Specifically, from the perspectives of production objects and production spaces respectively, combining digital twin technology and generation technology, according to the collected physical interaction information, a series of tree-shaped family trees from parts to assemblies and then to products, as well as diverse digital production spaces in the entire life cycle of intelligent manufacturing products, are generated. Furthermore, a production object digital family tree corresponding to the production object and a production space digital family tree corresponding to the production space are constructed. Through the production object digital family tree corresponding to the production object and the production space digital family tree corresponding to the production space, a large amount of three-dimensional digital scene data can be generated by combining diverse random combinations of different parts and assemblies in the production object in diverse production spaces, for the industrial embodied intelligence world model to learn and train in the digital family tree, improving the adaptability and generalization ability of the model.

[0008] Optionally, the physical interaction information includes gene coding information of products or parts, multimodal information, production equipment information, and environmental parameter information; correspondingly, the generating, according to the physical interaction information, a rich geometric representation corresponding to the product, a rich feature representation corresponding to the product, and an environmental configuration corresponding to the scene includes: generating a rich geometric representation corresponding to the product according to the gene coding information, where the gene coding information includes at least one of shape, size, material, type, use, and assembly relationship, and the rich geometric representation includes at least one of boundary representation, modeling sequence, network entity, polyhedron, point cloud, and voxel; generating a rich feature representation corresponding to the product according to the multimodal information, where the multimodal information includes at least one of voice, text, and image; generating a device configuration corresponding to the scene according to the production equipment information; and generating an environmental configuration corresponding to the scene according to the environmental parameter information.

[0009] Here, the present application proposes an information mechanism for the gene (Deoxyribonucleic Acid, DNA) coding of the digital family tree. By designing DNA coding to constrain the phenotypes of the generated parts in the family tree, while ensuring that it meets the requirements of industrial production performance and is interpretable, and then combining multimodal information such as voice, text, or image, a three-dimensional (3D) model that meets engineering requirements can be automatically generated. In terms of the production space, the present application realizes the construction of a digital twin scene by configuring scene elements, and realizes the device configuration and environmental configuration in the scene elements through production equipment information and environmental parameter information. Through the construction of the above series of scene elements and 3D models, sufficient training materials and an accurate simulation environment are provided for the high adaptive ability of industrial intelligent agents, further enriching the trainable data and scenes, and improving the training effect of the industrial embodied intelligence world model.

[0010] Optionally, the physical interaction information includes first target detection information of parts and second target detection information of a scene; correspondingly, establishing a digital twin product and a digital twin space according to the physical interaction information includes: performing approximate matching in a preset parts asset library according to the first target detection information to determine a feature vector corresponding to the first target detection information according to the approximate matching result; performing a matching process of similarity measurement in a predefined parts library according to the feature vector to determine a digital twin product corresponding to the first target detection information according to the matching result of the similarity measurement; and establishing a digital twin space according to the second target detection information.

[0011] Among them, this application extracts data from the physical world and generates digital twins, providing a digital foundation for subsequent family tree generation and agent learning. Specifically, the object detection of each part in the physical world RGB (RGB Color Model) image can be completed through an object detection algorithm, identifying and classifying each part in the image, and using accurate matching methods such as approximate matching and similarity measurement. The part most similar to the identified and detected part is used as a digital entity in the digital twin scene and as a part of the entire scene. Based on the recognition result, the digital twin space, the disadvantages of the data twin product and the digital twin space are determined, making the simulation closer to real production conditions, thereby improving the reliability and accuracy of the simulation reinforcement learning process, further improving the training effect of the industrial embodied intelligence world model, and improving the inference accuracy of the model.

[0012] Optionally, generating a production object digital family tree corresponding to a production object according to the digital twin product, the rich geometric representation, and the rich feature representation includes: obtaining a preset first family tree structure, where the preset first family tree structure is a first tree structure, the first tree structure includes multiple tree nodes at a first level, and the first level includes a product level, an assembly level, and a part level; and performing node filling processing on the first family tree structure according to the digital twin product, the rich geometric representation, and the rich feature representation to obtain a production object digital family tree corresponding to the production object.

[0013] Here, the present application establishes a first family tree structure from the perspective of production objects. From the perspective of production objects, a series of tree-shaped family trees can be generated from parts to assemblies and then to products. Based on this digital family tree structure, 3D parts are not only stored as independent geometric representations but also placed into the family tree structure to reflect the co-generation relationships between parts and their hierarchical connections with downstream assemblies and final products. Through this organizational method, it can clearly express how different requirements give rise to diverse part designs and how a complete product system is formed through assembly relationships, thereby enriching the trainable data and improving the training effect of the industrial embodied intelligence world model.

[0014] Optionally, generating a production space digital family tree corresponding to the production space according to the digital twin space, the environment configuration, and the device configuration includes: obtaining a plurality of production scenarios of preset types, where the preset types are determined based on multiple factors among production equipment, production environment, and production personnel; obtaining a second family tree structure corresponding to each production scenario of the preset type; and performing node filling processing on the second family tree structure according to the digital twin space, the environment configuration, and the device configuration to obtain a production space digital family tree corresponding to the production space.

[0015] Among them, the present application establishes a second family tree structure from the perspective of production scenarios. The family tree layer constructs diverse 3D digital scenarios through the methods in the generation layer, providing a rich training environment for subsequent agent learning. Different types of production scenarios are systematically classified, each having different production equipment, production environment, and production personnel. Agents of the industrial embodied intelligence world model can perform simulation learning in these scenarios. Through industrial family tree 3D scene training, the adaptability of intelligent devices such as industrial robots and automated production systems can be improved.

[0016] Optionally, after establishing the target digital family tree according to the production object digital family tree and the production space digital family tree, it further includes: obtaining the original industrial embodied intelligence world model; and training the original industrial embodied intelligence world model according to the target digital family tree to obtain a target industrial embodied intelligence world model.

[0017] Among them, the digital family tree created in this application records the evolution of the family tree of part data from multi-model parts to multi-performance products. Combining its diverse production scenarios, it not only describes the mapping relationship between different part characteristics and different product performances, but also traces the change process of products in the life cycle, ensuring the continuity and consistency in the production process from parts to products. At the same time, it also includes the scenario-level digital space of different tasks in each production and manufacturing stage. Based on the training of the industrial embodied intelligence world model of the digital family tree, it enables the industrial embodied intelligence world model to manipulate physical devices of different models to learn and train in a rich and diverse digital space, further improving the adaptability and generalization of the industrial embodied intelligence world model, so as to achieve interaction and feedback in the industrial physical environment.

[0018] Optionally, the target industrial embodied intelligence world model is used to execute at least one of the applications in the full life cycle of industrial products. The applications in the full life cycle of industrial products include: model simulation design based on the digital family tree; product design based on gene coding constraints; flexible sorting based on the digital family tree; product assembly guidance with reference to gene coding; family tree-driven adaptive scenario understanding; process generation driven by gene coding similarity; part predictive operation and maintenance based on the family tree; operation and maintenance guidance based on gene coding similarity.

[0019] Here, this application can train a target industrial embodied intelligence world model with excellent results based on the rich product and scenario data in the digital atlas, and then realize practical applications such as intelligent design, manufacturing optimization, and predictive maintenance, improving the performance of the intelligent system in a flexible and dynamic environment, and enhancing the automation and intelligence of industrial part intelligent manufacturing in the field of intelligent manufacturing.

[0020] In a second aspect, this application provides a device for generating an industrial digital family tree, including: an acquisition module for acquiring physical interaction information; a twin module for establishing a digital twin product and a digital twin space according to the physical interaction information; a generation module for generating a rich geometric representation corresponding to the product, a rich feature representation corresponding to the product, an environment configuration corresponding to the scenario, and a device configuration corresponding to the scenario according to the physical interaction information; a first family tree module for generating a production object digital family tree corresponding to the production object according to the digital twin product, the rich geometric representation, and the rich feature representation; a second family tree module for generating a production space digital family tree corresponding to the production space according to the digital twin space, the environment configuration, and the device configuration; and an establishment module for establishing a target digital family tree according to the production object digital family tree and the production space digital family tree, where the target digital family tree is used for the training of the industrial embodied intelligence world model.

[0021] Optionally, the physical interaction information includes gene coding information, multimodal information, production equipment information, and environmental parameter information of a product or a part; correspondingly, the generation module is specifically configured to: generate a rich geometric representation corresponding to the product according to the gene coding information, where the gene coding information includes at least one of shape, size, material, type, use, and assembly relationship, and the rich geometric representation includes at least one of boundary representation, modeling sequence, network entity, polyhedron, point cloud, and voxel; generate a rich feature representation corresponding to the product according to the multimodal information, where the multimodal information includes at least one of voice, text, and image; generate a device configuration corresponding to the scenario according to the production equipment information; and generate an environmental configuration corresponding to the scenario according to the environmental parameter information.

[0022] Optionally, the physical interaction information includes first target detection information of a part and second target detection information of a scenario; correspondingly, the twin module is specifically configured to: perform approximate matching in a preset part asset library according to the first target detection information, so as to determine a feature vector corresponding to the first target detection information according to the approximate matching result; perform matching processing of similarity measurement in a predefined part library according to the feature vector, so as to determine a digital twin product corresponding to the first target detection information according to the matching processing result of the similarity measurement; and establish a digital twin space according to the second target detection information.

[0023] Optionally, the first family tree module is specifically configured to: obtain a preset first family tree structure, where the preset first family tree structure is a first tree structure, and the first tree structure includes multiple tree nodes at a first level, and the first level includes a product level, an assembly level, and a part level; perform node filling processing on the first family tree structure according to the digital twin product, the rich geometric representation, and the rich feature representation, so as to obtain a production object digital family tree corresponding to the production object.

[0024] Optionally, the second family tree module is specifically configured to: obtain multiple preset types of production scenarios, where the preset types are determined based on multiple of production equipment, production environment, and production personnel; obtain a second family tree structure corresponding to each of the preset types of production scenarios; perform node filling processing on the second family tree structure according to the digital twin space, the environmental configuration, and the device configuration, so as to obtain a production space digital family tree corresponding to the production space.

[0025] Optionally, after the establishment module is configured to establish a target digital family tree according to the production object digital family tree and the production space digital family tree, the above device further includes a training module, configured to: obtain an original industrial embodied intelligent world model; train the original industrial embodied intelligent world model according to the target digital family tree, so as to obtain a target industrial embodied intelligent world model.

[0026] Optionally, the target industrial embodied intelligence world model is used to execute at least one of the industrial product full - life - cycle applications, and the industrial product full - life - cycle applications include: model simulation design based on digital family trees; product design based on gene - coding constraints; flexible sorting based on digital family trees; product assembly guidance with reference to gene coding; family - tree - driven adaptive scene understanding; process generation driven by gene - coding similarity; part predictive operation and maintenance based on family trees; and operation and maintenance guidance based on gene - coding similarity.

[0027] In a third aspect, the present application provides an industrial digital family - tree system, including: a memory and a processor; the memory stores computer - executable instructions; the processor executes the computer - executable instructions stored in the memory, so that the processor executes the above - mentioned first aspect and / or various possible implementation manners of the first aspect.

[0028] In a fourth aspect, the present application provides a computer - readable storage medium, in which computer - executable instructions are stored, and when the computer - executable instructions are executed by a processor, they are used to implement the above - mentioned first aspect and / or various possible implementation manners of the first aspect.

[0029] In a fifth aspect, the present application provides a computer program product, including a computer program, and when the computer program is executed by a processor, it implements the above - mentioned first aspect and / or various possible implementation manners of the first aspect.

[0030] The industrial digital family - tree system and the method for generating an industrial digital family - tree provided by the present application respectively provide new possibilities for the construction of an industrial embodied intelligence world model from two perspectives. Specifically, from the two perspectives of production objects and production spaces, combining digital twin technology and generation technology, according to the collected physical interaction information, a series of tree - shaped family trees from parts to assemblies and then to products and diverse digital production spaces in the entire life cycle of intelligent manufacturing products are generated. Furthermore, a production - object digital family - tree corresponding to the production object and a production - space digital family - tree corresponding to the production space are constructed. Through the production - object digital family - tree corresponding to the production object and the production - space digital family - tree corresponding to the production space, a large amount of three - dimensional digital scene data can be generated by combining diverse random combinations of different parts and assemblies in the production object in diverse production spaces, for the industrial embodied intelligence world model to learn and train in the digital family - tree, improving the adaptability and generalization of the model. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] The drawings here are incorporated into the specification and form a part of this specification, showing embodiments consistent with the present application, and are used together with the specification to explain the principles of the present application.

[0032] Figure 1aAn entity comparison diagram of a digital genealogy, digital twin, and digital cousin provided by an embodiment of the present application;

[0033] Figure 1b A concept table comparison diagram of a digital genealogy, digital twin, and digital cousin provided by an embodiment of the present application;

[0034] Figure 2 A schematic diagram of the system architecture for generating an industrial digital genealogy provided by an embodiment of the present application;

[0035] Figure 3 A schematic flowchart of a method for generating an industrial digital genealogy provided by an embodiment of the present application;

[0036] Figure 4 A schematic diagram of the structure of a production object digital genealogy provided by an embodiment of the present application;

[0037] Figure 5 A schematic diagram of the structure of a production space digital genealogy provided by an embodiment of the present application;

[0038] Figure 6 A schematic diagram of the architecture of a digital genealogy system provided by an embodiment of the present application;

[0039] Figure 7 A schematic diagram of the steps of a method for constructing a digital genealogy provided by an embodiment of the present application;

[0040] Figure 8 A schematic diagram of the structure of a device for generating an industrial digital genealogy provided by an embodiment of the present application;

[0041] Figure 9 A schematic diagram of the structure of an industrial digital genealogy system provided by an embodiment of the present application.

[0042] Through the above-mentioned drawings, specific embodiments of the present application have been shown, and there will be more detailed descriptions hereinafter. These drawings and text descriptions are not intended to limit the scope of the concept of the present application in any way, but to illustrate the concept of the present application to those skilled in the art by referring to specific embodiments. Detailed Description of the Embodiments

[0043] Here, the exemplary embodiments will be described in detail, and the examples are shown in the drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present application. On the contrary, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.

[0044] First, the terms involved in the present application will be explained:

[0045] Digital twin, also known as digital double, makes full use of data such as physical models, sensors, and operation history to complete simulation and mapping in the virtual space, thereby reflecting the whole life cycle process of the corresponding entity or system.

[0046] Digital cousin is a digital entity corresponding to a physical entity. Different from digital twins, digital cousins do not fully simulate the physical entities in the real world, but show similar geometric forms and semantic features.

[0047] Digital family tree is a new concept proposed in the embodiments of this application, which has two perspectives of production object and production space at the same time. The digital family tree from the perspective of production object reflects the mapping system of the blood relationship between multiple physical entities in the product family tree. It focuses on the similarity between components and the evolution trend from parts to products; the digital family tree from the perspective of production space reflects the multiple parallel digital worlds parallel to the physical world, and can display the manufacturing processes of different production links through the spatial dimension. Different from digital twins and cousins, the digital family tree pays attention to the pedigree relationship, common characteristics and their change processes among digital entities, rather than the digital mapping of a single entity or its approximate entity.

[0048] World model: By creating an abstract representation of the external world, it helps embodied agents to perform intelligent perception, planning and decision-making in the physical world.

[0049] As a research hotspot in recent years, the world model creates an abstract representation of the external world to help embodied agents perform intelligent perception, planning and decision-making in the physical world. At the same time, by combining embodied intelligence and the world model, the industrial embodied intelligence world model can control physical devices to achieve the deep integration of perception, decision-making and execution in the industrial world. For example, in a flexible manufacturing environment, the diversity and complexity of production tasks require intelligent systems to be able to effectively respond to various changes, such as the diversity of different product types, process flows and production environments. The industrial embodied intelligence world model needs to learn from large-scale data to achieve interaction and feedback in the industrial physical environment, and possess adaptability and generalization to multiple product models and production scenarios.

[0050] However, a major core challenge of the industrial embodied intelligence world model is the lack of sufficient training data and scenarios. Existing general scenarios cannot be directly applied to the training of the industrial embodied intelligence world model, because the diversity, complexity and physical interaction requirements in the industrial flexible production environment all require more diverse and finely labeled data. These data not only need to cover all aspects of product design, manufacturing, assembly, etc., but also need perception and operation data in real scenarios so that the model can learn more accurate operation strategies and feedback mechanisms.

[0051] To solve the above problems, an embodiment of the present application provides an industrial digital family tree system and a method for generating an industrial digital family tree. This method proposes the concept of a digital family tree, providing new possibilities for the construction of an industrial embodied intelligence world model from two perspectives. In a diverse production space, by combining the diverse random combinations of different parts and assemblies in the production object, a large amount of three-dimensional digital scene data is generated for the industrial embodied intelligence world model to learn and train in the digital family tree.

[0052] Optionally, to address the challenge of insufficient data in the construction of an industrial embodied intelligence world model, an embodiment of the present application proposes an innovative concept - the digital family tree, which provides new possibilities for the construction of an industrial embodied intelligence world model from two perspectives. From the perspective of the production object, a series of tree-like family trees can be generated from parts to assemblies and then to products, and a digital family tree DNA gene mechanism is defined. By designing DNA coding to constrain the phenotypes of the generated parts in the family tree, it ensures that they meet the requirements of industrial production performance while being interpretable; from the perspective of the production space, diverse digital production spaces can be generated throughout the entire life cycle of intelligent manufacturing products, including diverse scenarios of different equipment and environmental configurations in warehousing, manufacturing, and operation and maintenance scenarios. Combining the above two perspectives, in a diverse production space, by combining the diverse random combinations of different parts and assemblies in the production object, a large amount of three-dimensional digital scene data can be generated for the industrial embodied intelligence world model to learn and train in the digital family tree, improving the adaptability and generalization of the model.

[0053] Among them, the digital family tree not only describes the mapping relationship between the characteristics of different parts and the performance of different products but also traces the change process of products during their life cycle, ensuring the continuity and consistency during the production process from parts to products; at the same time, it also includes the scene-level digital spaces of different tasks in each production and manufacturing stage, which can be used by the embodied world model to manipulate physical devices of different models to learn and train in rich and diverse digital spaces. Demonstratively, Figure 1a is an entity comparison diagram of a digital family tree, digital twin, and digital cousin provided by an embodiment of the present application; Figure 1b is a conceptual table comparison diagram of a digital family tree, digital twin, and digital cousin provided by an embodiment of the present application. As Figure 1a and Figure 1b show, there are certain similarities between the digital family tree, digital twin, and digital cousin, and the digital family tree has richer information. It can be understood that Figure 1a 、 Figure 1b and Figure 1a 、 Figure 1b The entities in are only schematic and do not affect the protection scope of the embodiments of the present application.

[0054] Optionally, Figure 2Schematic diagram of the architecture of a system for generating an industrial digital family tree provided by an embodiment of the present application. In Figure 2 the above architecture includes at least one of a data acquisition device 201, a processing device 202, and a display device 203.

[0055] It can be understood that the structure schematically shown in the embodiments of the present application does not constitute a specific limitation on the architecture of the system for generating an industrial digital family tree. In other feasible embodiments of the present application, the above architecture may include more or fewer components than those shown in the figure, or combine certain components, or split certain components, or have different component arrangements, which can be specifically determined according to the actual application scenario and will not be limited herein. Figure 2 The components shown can be implemented in hardware, software, or a combination of software and hardware.

[0056] In the specific implementation process, the data acquisition device 201 may include an input / output interface or a communication interface, and the data acquisition device 201 can be connected to the processing device through the input / output interface or the communication interface.

[0057] The processing device 202 can provide new possibilities for the construction of the industrial embodied intelligence world model from two perspectives respectively. In the diverse production space, by combining the diverse random combinations of different parts and assemblies in the production objects, a large amount of three-dimensional digital scene data can be generated for the industrial embodied intelligence world model to learn and train in the digital family tree.

[0058] The display device 203 can also be a touch display screen or the screen of a terminal device, which is used to receive user instructions while displaying the above content to achieve interaction with the user.

[0059] It should be understood that the above processing device can be implemented by a processor reading and executing instructions in a memory or by a chip circuit.

[0060] In addition, the network architecture and business scenarios described in the embodiments of the present application are for more clearly illustrating the technical solutions of the embodiments of the present application, and do not constitute a limitation on the technical solutions provided by the embodiments of the present application. Those skilled in the art know that with the evolution of the network architecture and the emergence of new business scenarios, the technical solutions provided by the embodiments of the present application are equally applicable to similar technical problems.

[0061] The following will describe the technical solutions of the present application in detail with specific embodiments:

[0062] Optionally, Figure 3 Schematic diagram of the process of a method for generating an industrial digital family tree provided by an embodiment of the present application. The execution subject of the embodiment of the present application can be Figure 2The processing device 202 therein, and the specific execution entity can be determined according to the actual application scenario. For example, Figure 3 As shown, the method includes the following steps:

[0063] S301: Obtain physical interaction information.

[0064] Optionally, based on Figure 2 the data acquisition device 201 therein to obtain physical interaction information.

[0065] Optionally, at the interaction level of the industrial embodied intelligence world model, it includes three core elements: people, terminal devices, and materials, which constitute the physical carrier of actual industrial production and manufacturing activities. At the human level, the physical world involves different roles such as technical workers, designers, and experimenters. By interacting with the terminal devices at the human level, physical interaction information can be obtained. At the terminal device level, the physical world includes industrial automation devices such as robots and machine tools, and physical interaction information can be obtained through the above industrial automation devices. At the material level, the physical world involves raw materials, components, and products, etc. Through data acquisition and modeling, the information at the physical world layer can be mapped to the digital world, thereby providing data sources for digital twins and digital family trees.

[0066] Optionally, the physical interaction information includes the gene coding information of products or parts. The embodiments of this application draw on the concept of genes in the biological world and define the genes of industrial parts from different dimensions, corresponding to different phenotypes of industrial parts. The digital family tree DNA gene mechanism is defined, and by designing DNA coding to constrain the phenotypes of the generated parts in the family tree, it is ensured that while meeting the industrial production performance requirements, it is also interpretable.

[0067] In the concept of the digital pedigree of industrial parts, the basic attributes of parts can be mapped to similar elements in biological DNA. DNA defines the basic characteristics that determine the phenotype of parts, including shape, size, material, type, purpose, fitting relationship, etc. Specifically, these attributes are identified as key elements in the digital pedigree of industrial parts, as follows:

[0068] Shape: The shape configuration of the part is similar to the appearance characteristics of an organism. Different parts have different shapes, such as plate-shaped, T-shaped, block-shaped, etc.

[0069] Size: The size of the part, including length, width, and height, is similar to physical characteristics such as the height and weight of an organism. The sizes of industrial parts are mainly divided into large, medium, and small.

[0070] Material: The material composition of a part, such as metal, plastic, or rubber, corresponds to the concept of skin color in biological DNA and reflects the inherent properties of the part.

[0071] Type: The classification of a part, i.e., the specific category or lineage to which it belongs, such as gears, pipes, or bolts. Similar to species classification in biology, this DNA defines the category and structure of the part.

[0072] Usage: The intended function or role of a part in a system, similar to a human occupation. This attribute determines the contribution of the part to the overall system, such as functions like transmission, fastening, clamping, etc.

[0073] Assembly relationship: This DNA is a unique attribute of a part and has no direct correspondence to organisms. It describes the assembly relationship between a part and other parts in the system, especially how they are combined and interact. The assembly relationship determines the compatibility and performance of the part in the overall assembly, including press fits, threaded connections, clamping fits, etc.

[0074] These elements together constitute the DNA of industrial parts, providing a comprehensive perspective on their characteristics, functions, and evolution in the manufacturing system. Just as biological DNA compresses the genetic information required for the development and adaptation of organisms, the DNA of industrial parts makes it possible to track their historical evolution, modification, and future potential, supporting optimization and customization in modern manufacturing. At the same time, these DNAs can further facilitate the generation of industrial parts under different constraints.

[0075] It can be understood that the DNA encoding of parts here is only schematic. In the actual application process, the DNA encoding of parts can be set according to specific usage situations.

[0076] S302: Establish a digital twin product and a digital twin space according to the physical interaction information.

[0077] Optionally, the physical interaction information includes the first target detection information of the part and the second target detection information of the scene; correspondingly, establishing a digital twin product and a digital twin space according to the physical interaction information includes: performing approximate matching in a preset part asset library according to the first target detection information to determine the feature vector corresponding to the first target detection information according to the approximate matching result; performing a matching process of similarity measurement in a predefined part library according to the feature vector to determine the digital twin product corresponding to the first target detection information according to the matching result of the similarity measurement; establishing a digital twin space according to the second target detection information.

[0078] In a possible implementation, the process of constructing a 3D digital family tree scene usually starts with perception in the physical world, which can typically be achieved through a vision sensor and vision perception algorithms to sense the physical world. For example, object detection algorithms can be used to detect objects in the RGB images of the physical world, identify and classify each part in the image. Once the parts are identified, approximate matching can be performed in the part asset library based on the category of the parts, so as to search for the same or similar parts. This matching process aims to find the parts with the closest DNA match, such as geometric configuration, material properties, or functional characteristics. These features can then be used to create a feature vector as a numerical representation of the basic attributes of the parts. Next, the system searches for parts with similar feature vectors in a part library containing a large number of predefined parts. This can be specifically done using similarity metrics (such as cosine similarity, Euclidean distance, etc.). By sorting the parts according to the similarity scores and selecting the most suitable candidate parts, the part most similar to the detected part in the part asset library is finally identified as the digital entity in the digital twin scene and becomes part of the entire scene.

[0079] Among them, the embodiments of the present application extract data from the physical world and generate digital twins, providing a digital foundation for subsequent family tree generation and agent learning. Specifically, object detection algorithms can be used to detect objects in the RGB images of the physical world, identify and classify each part in the image. An accurate matching method of approximate matching and similarity metric is adopted to use the part most similar to the identified and detected part as the digital entity in the digital twin scene and as part of the entire scene. Based on the recognition results, the digital twin space, the disadvantages of the data twin product, and the digital twin space are determined, making the simulation closer to real production conditions, thereby improving the reliability and accuracy of the simulation reinforcement learning process, further improving the training effect of the industrial embodied intelligence world model, and improving the inference accuracy of the model.

[0080] S303: Generate a rich geometric representation corresponding to the product, a rich feature representation corresponding to the product, an environmental configuration corresponding to the scene, and a device configuration corresponding to the scene according to the physical interaction information.

[0081] Optionally, the physical interaction information includes gene coding information, multi-modal information, production equipment information, and environmental parameter information of the product or part; correspondingly, generating a rich geometric representation corresponding to the product, a rich feature representation corresponding to the product, and an environmental configuration corresponding to the scene according to the physical interaction information includes:

[0082] Generate a rich geometric representation corresponding to the product according to the gene coding information; generate a rich feature representation corresponding to the product according to the multi-modal information. Generate a device configuration corresponding to the scene according to the production equipment information; generate an environmental configuration corresponding to the scene according to the environmental parameter information.

[0083] Among them, the gene coding information includes at least one of shape, size, material, type, use, and assembly relationship.

[0084] Among them, the rich geometric representation includes at least one of boundary representation, modeling sequence, network entity, polyhedron, point cloud, and voxel

[0085] Among them, the multi-modal information includes at least one of speech, text, and image.

[0086] In a possible implementation, the method for intelligent generation of production objects is as follows:

[0087] In the above digital twin modeling process, if no identical or similar parts are found in the part asset library, the system will use generative artificial intelligence methods to generate similar industrial parts. Due to the high requirements for the credibility of the generated content in industrial scenarios, this method should be able to generate new parts based on the given DNA constraints, rather than overusing creativity and distorting the generated parts. According to the definition of the part DNA in the digital family tree, the basic attributes such as shape, size, and material in the DNA can be used to guide the diverse generation of new parts. By specifying these DNA parameters, it can be ensured that the generated parts are both diverse and as compliant with industrial requirements as possible.

[0088] During the generation process, specific designs can be made to the parts through DNA editing methods to achieve the effect of personalized customization. The parts edited by DNA will retain other unmodified DNA attributes, enabling the newly generated parts to inherit most of the characteristics of the original parts while also having some customized characteristics that are different from them. Through this intelligent generation method, new parts that can be interpreted from an industrial perspective can be efficiently generated in batches, thus expanding the part asset library of the digital family tree and empowering the construction of the digital family tree 3D scene.

[0089] In a possible implementation, the method for intelligent generation of production space is as follows:

[0090] Once the parts are matched or generated, they can be imported into the digital family tree 3D scene. As a digital world corresponding to the physical world and with greater scalability, multiple parts under the family tree can be placed in it and arranged in various combinations. The digital family tree 3D scene can not only reflect the current state and configuration of the physical world but also record and deduce its changes over time, including the addition of new parts and the changes of existing parts during the manufacturing process, which will enable the embodied intelligent world model to have better generalization and adaptation capabilities.

[0091] After matching or generating parts, these parts will be imported into the digital family tree 3D scene and placed in the 3D scene space while satisfying physical constraints. Subsequently, domain randomization parameters can also be set in the digital family tree 3D scene. For example, by changing environmental parameters to enhance the diversity of synthetic data, the robustness of the embodied intelligent world model in different environments can be improved. By creating diverse and realistic scenes, various conditions and challenges that may be encountered in the real manufacturing environment can be simulated. Specifically, the domain randomization parameters in the digital family tree scene are mainly related to the environment, such as lighting conditions, object spatial relationships, object poses, and backgrounds. For example, the lighting intensity and angle can be randomly adjusted to simulate changes in the lighting conditions in a factory workshop; the spatial relationships and poses between parts in the scene can also be randomized, including changing the positions, rotations, and relative relationships in space of the parts; at the same time, the background of the object working space can be configured. For example, the material of the workbench can be changed to different backgrounds such as metal and plastic, and different scales of wear can also be added. The scenes configured with domain randomization can generate more diverse synthetic data, which can be used to train the embodied intelligent world model. Domain randomization helps improve the generalization ability of the model, ensuring that they can perform well in different scenes and adapt to unpredictable conditions in real applications. Therefore, the digital family tree framework becomes a powerful tool for creating multifunctional training environments, capable of simulating complex manufacturing processes and enhancing the performance of intelligent systems in flexible and dynamic environments.

[0092] Here, the embodiment of the present application proposes an information mechanism for DNA encoding of the digital family tree. By designing DNA encoding to constrain the phenotypes of the generated parts in the family tree, while ensuring that they meet the industrial production performance requirements and are interpretable, and then combining multi-modal information such as voice, text, or images, a 3D model that meets the engineering requirements can be automatically generated. In terms of the production space, the embodiment of the present application realizes the construction of the digital twin scene by configuring scene elements, realizes the equipment configuration and environmental configuration in the scene elements through production equipment information and environmental parameter information. Through the construction of the above series of scene elements and 3D models, sufficient training materials and accurate simulation environments are provided for the high adaptive ability of industrial agents, further enriching the trainable data and scenes, and improving the training effect of the industrial embodied intelligent world model.

[0093] S304: Generate a production object digital family tree corresponding to the production object according to the digital twin product, rich geometric representation, and rich feature representation.

[0094] Optionally, the generating a production object digital family tree corresponding to the production object according to the digital twin product, rich geometric representation, and rich feature representation includes:

[0095] Obtain a preset first family tree structure, where the preset first family tree structure is a first tree structure, and the first tree structure includes multiple tree nodes at the first level, and the first level includes a product level, an assembly level, and a part level; according to the digital twin product, rich geometric representation, and rich feature representation, perform node filling processing on the first family tree structure to obtain the production object digital family tree corresponding to the production object.

[0096] Exemplarily, Figure 4 is a schematic diagram of the structure of a production object digital family tree provided by an embodiment of the present application. As Figure 4 shown, it shows a digital family tree from the perspective of a production object. In terms of the production object, the previously generated 3D parts are not only stored as independent geometric representations, but also placed in the family tree structure to reflect the co-generation relationship between the parts, as well as their hierarchical connections with downstream assemblies and final products. Through this organizational method, it is possible to clearly express how different requirements give rise to diverse part designs and how a complete product system is formed through assembly relationships.

[0097] Here, an embodiment of the present application establishes a first family tree structure from the perspective of a production object. From the perspective of a production object, a series of tree-shaped family trees can be generated from parts to assemblies and then to products. Based on this digital family tree structure, 3D parts are not only stored as independent geometric representations, but also placed in the family tree structure to reflect the co-generation relationship between the parts, as well as their hierarchical connections with downstream assemblies and final products. Through this organizational method, it is possible to clearly express how different requirements give rise to diverse part designs and how a complete product system is formed through assembly relationships, thereby enriching the trainable data and improving the training effect of the industrial embodied intelligence world model.

[0098] S305: Generate a production space digital family tree corresponding to the production space according to the digital twin space, environmental configuration, and device configuration.

[0099] Optionally, the generating a production space digital family tree corresponding to the production space according to the digital twin space, environmental configuration, and device configuration includes: obtaining multiple production scenarios of preset types, where the preset types are determined based on multiple of production equipment, production environment, and production personnel; obtaining a second family tree structure corresponding to each production scenario of the preset type; and performing node filling processing on the second family tree structure according to the digital twin space, environmental configuration, and device configuration to obtain the production space digital family tree corresponding to the production space.

[0100] Exemplarily, Figure 5 is a schematic diagram of the structure of a production space digital family tree provided by an embodiment of the present application. As Figure 5As shown, a digital family tree from the perspective of production space is presented. In terms of production space, diverse 3D digital scenes are constructed to provide a rich training environment for subsequent agent learning. Different types of production scenes are systematically classified, each with different production equipment, production environments, and production personnel. Agents can conduct simulation learning in these scenes. Through training in the 3D industrial family tree scenes, the adaptability of intelligent devices such as industrial robots and automated production systems can be improved. It can be understood that Figure 4 、 Figure 5 and those in the embodiments of this application Figure 6 are merely illustrative and do not affect the protection scope of the embodiments of this application.

[0101] Among them, the embodiments of this application establish a second family tree structure from the perspective of production scenes. The family tree layer constructs diverse 3D digital scenes through the methods in the generation layer to provide a rich training environment for subsequent agent learning. Different types of production scenes are systematically classified, each with different production equipment, production environments, and production personnel. Agents of the industrial embodied intelligent world model can conduct simulation learning in these scenes. Through training in the 3D industrial family tree scenes, the adaptability of intelligent devices such as industrial robots and automated production systems can be improved.

[0102] S306: Establish a target digital family tree according to the production object digital family tree and the production space digital family tree.

[0103] Among them, the target digital family tree is used for the training of the industrial embodied intelligent world model.

[0104] The embodiments of this application provide a method for generating an industrial digital family tree. The digital family tree is a new concept proposed in the embodiments of this application, which provides new possibilities for the construction of the industrial embodied intelligent world model from two perspectives. Specifically, from the two perspectives of production objects and production space, combining digital twin technology and generation technology, according to the collected physical interaction information, a series of tree-shaped family trees from parts to assemblies and then to products and diverse digital production spaces in the whole life cycle of intelligent manufacturing products are generated. Furthermore, a production object digital family tree corresponding to the production object and a production space digital family tree corresponding to the production space are constructed. Through the production object digital family tree corresponding to the production object and the production space digital family tree corresponding to the production space, a large amount of three-dimensional digital scene data can be generated by combining diverse random combinations of different parts and assemblies in the production object in a diverse production space, for the industrial embodied intelligent world model to learn and train in the digital family tree, improving the adaptability and generalization of the model.

[0105] Optionally, after establishing the target digital family tree according to the production object digital family tree and the production space digital family tree, it further includes: obtaining the original industrial embodied intelligent world model; training the original industrial embodied intelligent world model according to the target digital family tree to obtain the target industrial embodied intelligent world model.

[0106] Among them, the digital family tree created in the embodiments of this application records the evolution of the family tree of part data from multi-model parts to multi-performance products. Combining its various production scenarios, it not only describes the mapping relationship between different part characteristics and different product performances, but also traces the change process of products in the life cycle, ensuring the continuity and consistency in the production process from parts to products. At the same time, it also includes the scene-level digital space of different tasks in each production and manufacturing stage. The training of the industrial embodied intelligent world model based on the digital family tree allows the industrial embodied intelligent world model to manipulate different models of physical devices to learn and train in a rich and diverse digital space, further improving the adaptability and generalization of the industrial embodied intelligent world model, thereby realizing interaction and feedback in the industrial physical environment.

[0107] Optionally, the target industrial embodied intelligent world model is used to execute at least one of the applications in the full life cycle of industrial products. The applications in the full life cycle of industrial products include: model simulation design based on the digital family tree; product design based on gene coding constraints; flexible sorting based on the digital family tree; product assembly guidance with reference to gene coding; spectrum-driven adaptive scene understanding; process generation driven by gene coding similarity; predictive operation and maintenance of parts based on the spectrum; operation and maintenance guidance based on gene coding similarity.

[0108] Here, the embodiments of this application can train the target industrial embodied intelligent world model with excellent effects based on the rich product and scene data in the digital map, and then realize practical applications such as intelligent design, manufacturing optimization, and predictive maintenance, improving the performance of the intelligent system in a flexible and dynamic environment, and enhancing the automation and intelligence of industrial part intelligent manufacturing in the field of intelligent manufacturing.

[0109] Optionally, Figure 6 is a schematic diagram of the architecture of a digital family tree system provided by the embodiments of this application. This digital family tree system can be the industrial digital family tree system in the embodiments of this application, or be implemented relying on the industrial digital family tree system. As Figure 6 shown, the architecture of the digital family tree includes a physical world layer, a digital family tree layer, an intelligent interaction layer, and a typical application layer. Among them, the digital family tree layer includes 3 sub-layers: a twin layer, a generation layer, and a digital family tree.

[0110] Among them, the physical world layer is the real foundation of the digital family tree. It covers three core elements: people, terminal devices, and materials, constituting the physical carrier of actual industrial production and manufacturing activities. The data and structure at this layer are not only the direct objects of industrial agent interaction but also the real sources of digital twins and agent learning, providing rich physical interaction information for industrial agents. At the human level, the physical world involves different roles such as technical workers, designers, and experimenters; at the terminal device level, the physical world includes industrial automation devices such as robots and machine tools; at the material level, the physical world involves raw materials, components, and products, etc. Through data collection and modeling, the information in the physical world layer can be mapped to the digital world, and then digital twins and digital family trees are formed.

[0111] The twin layer is a prerequisite step in constructing the digital family tree, responsible for extracting data from the physical world and generating digital twins, providing a digital foundation for subsequent family tree generation and agent learning. At this layer, industrial systems map the production elements in the physical world to the digital space through methods such as 3D modeling and data collection. In terms of production objects, the twin layer focuses on the digital transformation of specific products. Through means such as 3D modeling, structure scanning, and material analysis, physical products are transformed into 3D models with detailed feature data. These data include feature information such as the geometric structure, assembly method, and process parameters of the product, providing a basis for subsequent agent training. In terms of the production space, the twin layer digitally maps elements such as the production environment, manufacturing process, and human interaction, covering manufacturing scenarios such as automated production lines, intelligent workshops, and robot workstations, and combines sensor data to enable it to dynamically reflect the operating status of the actual industrial environment. The environmental twin data integrates parameters such as workshop temperature, humidity, light, and energy consumption, making the simulation closer to real production conditions, thereby improving the reliability and accuracy of the simulation reinforcement learning process.

[0112] The generation layer of the digital family tree is the core link in the intelligent modeling of industrial large models, including two aspects: product family tree generation and scenario family tree generation. Through multimodal data fusion and deep learning, this layer constructs a digital family tree with rich features and geometric information. In terms of production objects, based on the input "DNA data" (i.e., the gene coding of products or components, including structural parameters, functional characteristics, manufacturing processes, etc.), the system automatically generates a 3D model that meets engineering requirements by combining multimodal information such as text, images, and voices. This model can adopt different geometric representation forms such as Boundary Representation (B-rep), modeling sequences, meshes, polyhedra, point clouds, and voxels. In terms of production space, the generation layer realizes the construction of digital twin scenarios by configuring scenario elements. By selecting different environmental parameters (such as temperature, humidity, light, etc.) and different production equipment, the generation layer completes the configuration of the entire production scenario according to the given input parameters, so as to display the simulation production performance under various working conditions. By constructing a series of family tree scenarios, it provides sufficient training materials and accurate simulation environments for the high adaptability of industrial agents.

[0113] The digital family tree sub-layer, based on the generation of 3D models, further constructs the family tree structures of products and scenarios, establishing product family trees and scenario family trees. In terms of production objects, the previously generated 3D parts are not only stored as independent geometric representations but also placed into the family tree structure to reflect the co-generation relationships between parts and their hierarchical connections with downstream assemblies and final products. Through this organizational method, it can clearly express how different requirements give rise to diverse part designs and how they form a complete product system through assembly relationships. In terms of production space, the digital family tree sub-layer constructs diverse 3D digital scenarios through the methods in the generation layer, providing a rich training environment for subsequent agent learning. Different types of production scenarios are systematically classified, each with different production equipment, production environments, and production personnel, and agents can conduct simulation learning in these scenarios. Through training in industrial family tree 3D scenarios, the adaptability of intelligent devices such as industrial robots and automated production systems is improved.

[0114] The intelligent interaction layer is a key step from agent learning to actual industrial applications. During the interaction process, the industrial embodied intelligence model uses the learned production object family tree and production space family tree data for learning. The trained agent can interact efficiently with humans, terminal devices, and materials, enabling the digital family tree to play a role in the real production environment. In the interaction layer, human operators communicate with the industrial embodied intelligence model through natural language, instructions, and other interaction methods, and make adjustments when necessary. As the intelligent core, the industrial embodied intelligence model performs intelligent operations on terminal devices to achieve intelligent production. During the task execution process, the industrial embodied intelligence model continuously conducts overall management of production materials.

[0115] The typical application layer is the practical link where the industrial large model is finally implemented. Relying on the foregoing layers, it realizes practical applications such as intelligent design, manufacturing optimization, and predictive maintenance. In the design stage, based on the historical product data, process parameters, and production experience of similar products within the family tree, an efficient and reliable design plan is automatically generated to achieve simulation design driven by the digital family tree, reducing the design time. In the manufacturing stage, the application layer realizes the simulation of the production process based on the family tree scenario, optimizes the production process in the virtual digital family tree scenario, and improves the generalization ability of the model. At the same time, the intelligent agent can generate processing plans or operation methods for parts based on the similarity between the DNAs related to part manufacturing, enabling the large model to exhibit excellent flexibility in a flexible manufacturing environment. In the operation and maintenance stage, predictive operation and maintenance of parts can be carried out using the part family tree. By analyzing the DNA information such as the material and shape of the parts and combining with the operation status data of the parts, potential faults can be detected in advance, improving the reliability of the equipment.

[0116] In a possible implementation manner, Figure 7 is a schematic diagram of the steps of a method for constructing a digital family tree provided by an embodiment of the present application. Combining Figure 7 , the embodiments of the present application can achieve the following functions:

[0117] It provides new possibilities for the construction of the industrial embodied intelligence world model from two perspectives respectively. From the perspective of production objects, a series of tree-shaped family trees from parts to assemblies and then to products can be generated, and a digital family tree DNA gene mechanism is defined. By designing DNA coding to constrain the phenotypes of the generated parts in the family tree, it ensures that they meet the industrial production performance requirements while being interpretable; from the perspective of production space, diverse digital production spaces in the entire life cycle of intelligent manufacturing products can be generated, including diverse scenarios of different equipment and environmental configurations in warehousing, manufacturing, and operation and maintenance scenarios. Combining the above two perspectives, a large amount of three-dimensional digital scene data can be generated by combining diverse random combinations of different parts and assemblies in production objects in diverse production spaces for the industrial embodied intelligence world model to learn and train in the digital family tree, improving the adaptability and generalization ability of the model.

[0118] The digital family tree not only describes the mapping relationship between the characteristics of different parts and the performance of different products, but also traces the changes in the product during its life cycle, ensuring the continuity and consistency in the production process from parts to products. It also includes a scenario-level digital space for different tasks in each production and manufacturing stage, where the embodied world model can manipulate physical devices of different models to learn and train in a rich and diverse digital space.

[0119] Figure 8 The following is a schematic structural diagram of a generating device for an industrial digital family tree provided by an embodiment of the present application. As Figure 8 shown, the device of the embodiment of the present application includes: an acquisition module 801, a twin module 802, a generation module 803, a first family tree module 804, a second family tree module 805, and an establishment module 806. The generating device of the industrial digital family tree here can be the above-mentioned processing device itself, or a chip or integrated circuit that implements the functions of the processing device. It should be noted here that the division of the acquisition module 801, the twin module 802, the generation module 803, the first family tree module 804, the second family tree module 805, and the establishment module 806 is only a division of logical functions, and physically, the two can be integrated or independent.

[0120] Among them, the acquisition module is used to acquire physical interaction information; the twin module is used to establish a digital twin product and a digital twin space according to the physical interaction information; the generation module is used to generate a rich geometric representation corresponding to the product, a rich feature representation corresponding to the product, an environment configuration corresponding to the scenario, and a device configuration corresponding to the scenario according to the physical interaction information; the first family tree module is used to generate a production object digital family tree corresponding to the production object according to the digital twin product, the rich geometric representation, and the rich feature representation; the second family tree module is used to generate a production space digital family tree corresponding to the production space according to the digital twin space, the environment configuration, and the device configuration; the establishment module is used to establish a target digital family tree according to the production object digital family tree and the production space digital family tree, where the target digital family tree is used for the training of the industrial embodied intelligent world model.

[0121] Optionally, the physical interaction information includes gene coding information, multi-modal information, production equipment information, and environmental parameter information of a product or a part; correspondingly, the generation module is specifically configured to: generate a rich geometric representation corresponding to the product according to the gene coding information, where the gene coding information includes at least one of shape, size, material, type, use, and assembly relationship, and the rich geometric representation includes at least one of boundary representation, modeling sequence, network entity, polyhedron, point cloud, and voxel; generate a rich feature representation corresponding to the product according to the multi-modal information, where the multi-modal information includes at least one of voice, text, and image; generate a device configuration corresponding to the scenario according to the production equipment information; and generate an environmental configuration corresponding to the scenario according to the environmental parameter information.

[0122] Optionally, the physical interaction information includes first target detection information of a part and second target detection information of a scenario; correspondingly, the digital twin module is specifically configured to: perform approximate matching in a preset part asset library according to the first target detection information, so as to determine a feature vector corresponding to the first target detection information according to the approximate matching result; perform a matching process of similarity measurement in a predefined part library according to the feature vector, so as to determine a digital twin product corresponding to the first target detection information according to the matching result of the similarity measurement; and establish a digital twin space according to the second target detection information.

[0123] Optionally, the first family tree module is specifically configured to: obtain a preset first family tree structure, where the preset first family tree structure is a first tree structure, and the first tree structure includes multiple tree nodes at a first level, and the first level includes product level, assembly level, and part level; perform node filling processing on the first family tree structure according to the digital twin product, the rich geometric representation, and the rich feature representation, so as to obtain a production object digital family tree corresponding to the production object.

[0124] Optionally, the second family tree module is specifically configured to: obtain multiple preset types of production scenarios, where the preset types are determined based on multiple of production equipment, production environment, and production personnel; obtain a second family tree structure corresponding to each preset type of production scenario; perform node filling processing on the second family tree structure according to the digital twin space, the environmental configuration, and the device configuration, so as to obtain a production space digital family tree corresponding to the production space.

[0125] Optionally, after the establishment module is used to establish a target digital family tree according to the production object digital family tree and the production space digital family tree, the above device further includes a training module, configured to: obtain an original industrial embodied intelligent world model; train the original industrial embodied intelligent world model according to the target digital family tree, so as to obtain a target industrial embodied intelligent world model.

[0126] Optionally, the target industrial embodied intelligent world model is used to execute at least one of the applications in the entire life cycle of industrial products. The applications in the entire life cycle of industrial products include: model simulation design based on digital family trees; product design based on gene coding constraints; flexible sorting based on digital family trees; product assembly guidance with reference to gene coding; family tree-driven adaptive scene understanding; process generation driven by gene coding similarity; part predictive operation and maintenance based on family trees; operation and maintenance guidance based on gene coding similarity.

[0127] Reference Figure 9 , which shows a schematic structural diagram of an industrial digital family tree system 900 suitable for implementing the embodiments of the present disclosure. The industrial digital family tree system 900 can be a terminal device or a server. Among them, the terminal device can include, but is not limited to, mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, personal digital assistants (Personal Digital Assistant, abbreviated as PDA), tablet computers (Portable Android Device, abbreviated as PAD), portable multimedia players (Portable Media Player, abbreviated as PMP), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 9 The industrial digital family tree system shown is only an example and should not impose any limitations on the functions and usage scope of the embodiments of the present disclosure.

[0128] As Figure 9 shown, the industrial digital family tree system 900 can include a processing device (such as a central processing unit, a graphics processing unit, etc.) 901, which can execute various appropriate actions and processes according to the program stored in the read-only memory (Read Only Memory, abbreviated as ROM) 902 or the program loaded from the storage device 908 into the random access memory (Random Access Memory, abbreviated as RAM) 903. In the RAM 903, various programs and data required for the operation of the industrial digital family tree system 900 are also stored. The processing device 901, the ROM 902, and the RAM 903 are connected to each other through a bus 904. The input / output (I / O) interface 905 is also connected to the bus 904.

[0129] Typically, the following devices can be connected to the I / O interface 905: input devices 906 including, for example, a touch screen, a touchpad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; output devices 907 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; storage devices 908 including, for example, magnetic tapes, hard disks, etc.; and communication devices 909. The communication devices 909 can allow the industrial digital pedigree system 900 to communicate with other devices wirelessly or wiredly to exchange data. Although Figure 9 the industrial digital pedigree system 900 with various devices is shown, it should be understood that it is not required to implement or have all the shown devices. More or fewer devices can be alternatively implemented or had.

[0130] Specifically, according to an embodiment of the present disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, an embodiment of the present disclosure includes a computer program product that includes a computer program carried on a computer-readable medium, and the computer program contains program codes for performing the methods shown in the flowcharts. In such an embodiment, the computer program can be downloaded and installed from a network through the communication device 909, or installed from the storage device 908, or installed from the ROM 902. When the computer program is executed by the processing device 901, the above functions defined in the methods of the embodiments of the present disclosure are performed.

[0131] It should be noted that the above-mentioned computer-readable medium in the present disclosure may be a computer-readable signal medium, a computer-readable storage medium, or any combination of the two. A computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples of the computer-readable storage medium may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present disclosure, a computer-readable storage medium may be any tangible medium that contains or stores a program, and this program can be used by or in combination with an instruction execution system, apparatus, or device. In the present disclosure, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries computer-readable program code. Such a propagated data signal may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium, and this computer-readable signal medium can send, propagate, or transmit a program for use by or in combination with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted by any appropriate medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0132] The above-mentioned computer-readable medium may be included in the above industrial digital family tree system; or it may exist independently and not be assembled into the industrial digital family tree system.

[0133] The above-mentioned computer-readable medium carries one or more programs, and when the one or more programs are executed by the industrial digital family tree system, the industrial digital family tree system is caused to execute the method shown in the above embodiments.

[0134] Computer program code for performing the operations of the present disclosure may be written in one or more programming languages or combinations thereof. The programming languages include object-oriented programming languages such as Java, Smalltalk, C++, and also include conventional procedural programming languages such as the "C" language or similar programming languages. The program code may execute entirely on the user's computer, partially on the user's computer, execute as a stand-alone software package, execute partially on the user's computer and partially on a remote computer, or execute entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or alternatively, may be connected to an external computer (e.g., through the Internet using an Internet service provider).

[0135] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagram may represent a module, a segment of a program, or a part of code that contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than marked in the accompanying drawings. For example, two consecutive blocks shown may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, and combinations of blocks in the block diagram and / or flowchart, may be implemented by a dedicated hardware-based system for performing the specified functions or operations, or may be implemented by a combination of dedicated hardware and computer instructions.

[0136] The units involved in the embodiments described in the present disclosure may be implemented in software or in hardware. Among them, the name of the unit does not constitute a limitation on the unit itself in some cases. For example, the first acquisition unit may also be described as "the unit for acquiring at least two Internet protocol addresses".

[0137] The functions described above herein may be performed at least in part by one or more hardware logic components. For example, by way of non-limitation, exemplary types of hardware logic components that may be used include: Field Programmable Gate Arrays (FPGA), Application Specific Integrated Circuits (ASIC), Application Specific Standard Products (ASSP), Systems on Chip (SOC), Complex Programmable Logic Devices (CPLD), and so on.

[0138] In the context of the present disclosure, a machine-readable medium can be a tangible medium that can contain or store a program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. The machine-readable medium can include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of the machine-readable storage medium would include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0139] The industrial digital family tree system of the embodiments of the present application can be used to execute the technical solutions in the foregoing method embodiments of the present application. The implementation principles and technical effects are similar and will not be elaborated herein.

[0140] The embodiments of the present application also provide a computer-readable storage medium storing computer-executable instructions that, when executed by a processor, are used to implement the method for generating an industrial digital family tree according to any one of the foregoing.

[0141] The embodiments of the present application also provide a computer program product including a computer program that, when executed by a processor, is used to implement the method for generating an industrial digital family tree according to any one of the foregoing.

[0142] In several embodiments provided by the present application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative. For example, the division of units is only a logical function division, and there can be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be indirect couplings or communication connections through some interfaces, apparatuses, or units, and can be in electrical, mechanical, or other forms.

[0143] In addition, in each embodiment of the present application, the functional units can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit. The above-mentioned integrated units can be implemented in the form of hardware or in the form of software functional units.

[0144] Those skilled in the art will readily conceive of other embodiments of the present application after considering the specification and practicing the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the present application, which follow the general principles of the present application and include well-known common general knowledge or conventional technical means in the technical field not disclosed in the present application. The specification and examples are only regarded as exemplary, and the true scope and spirit of the present application are pointed out by the following claims.

[0145] It should be understood that the present application is not limited to the exact structures described above and shown in the drawings, and various modifications and changes can be made without departing from its scope. The scope of the present application is only limited by the appended claims.

Claims

1. A method for generating an industrial digital family tree, characterized in that: include: Obtain physical interaction information; Establishing a digital twin product and a digital twin space according to the physical interaction information; Generate, according to the physical interaction information, a rich geometric representation corresponding to the product, a rich feature representation corresponding to the product, an environment configuration corresponding to the scene, and a device configuration corresponding to the scene; Generate a production object digital family tree corresponding to the production object according to the digital twin product, the rich geometric representation and the rich feature representation; Generate a production space digital genealogy corresponding to the production space according to the digital twin space, the environment configuration and the equipment configuration; A target digital family tree is established according to the production object digital family tree and the production space digital family tree, wherein the target digital family tree is used for training the industrial embodied intelligent world model.

2. The method according to claim 1, characterized in that The physical interaction information includes genetic coding information, multimodal information, production equipment information and environmental parameter information of the product or part; Accordingly, generating a rich geometric representation corresponding to the product, a rich feature representation corresponding to the product, and an environment configuration corresponding to the scene according to the physical interaction information includes: Generate a rich geometric representation corresponding to the product according to the genetic encoding information, wherein the genetic encoding information includes at least one of shape, size, material, type, purpose, and assembly relationship, and the rich geometric representation includes at least one of boundary representation, modeling sequence, network entity, polygon, point cloud, and voxel; Generating a rich feature representation corresponding to the product according to the multimodal information, wherein the multimodal information includes at least one of voice, text, and image; Generate equipment configuration corresponding to the scenario according to the production equipment information; Generate an environment configuration corresponding to the scene based on the environment parameter information.

3. The method according to claim 1, characterized in that The physical interaction information includes first target detection information of the part and second target detection information of the scene; Accordingly, the digital twin product and digital twin space are established according to the physical interaction information, including: According to the first target detection information, an approximate match is performed in a preset part asset library to determine a feature vector corresponding to the first target detection information according to the approximate matching result; According to the feature vector, a matching process of similarity measurement is performed in a predefined parts library to determine the digital twin product corresponding to the first target detection information according to the matching process result of the similarity measurement; A digital twin space is established based on the second target detection information.

4. The method according to any one of claims 1 to 3, characterized in that: The step of generating a production object digital family tree corresponding to the production object according to the digital twin product, the rich geometric representation, and the rich feature representation includes: Obtaining a preset first family tree structure, wherein the preset first family tree structure is a first tree structure, the first tree structure includes a plurality of first-level tree nodes, and the first level includes a product level, an assembly level, and a part level; According to the digital twin product, the rich geometric representation and the rich feature representation, node filling processing is performed on the first family tree structure to obtain a production object digital family tree corresponding to the production object.

5. The method according to claim 4, characterized in that The generating a production space digital genealogy corresponding to the production space according to the digital twin space, the environment configuration and the equipment configuration includes: Acquire a plurality of preset types of production scenarios, wherein the preset types are determined based on a plurality of production equipment, production environment, and production personnel; Obtaining a second family tree structure corresponding to each of the preset types of production scenarios; According to the digital twin space, the environment configuration and the equipment configuration, node filling processing is performed on the second family tree structure to obtain a production space digital family tree corresponding to the production space.

6. The method according to claim 5, characterized in that After establishing a target digital family tree according to the production object digital family tree and the production space digital family tree, the method further includes: Get the original industrial embodied intelligence world model; According to the target digital family tree, the original industrial embodied intelligent world model is trained to obtain a target industrial embodied intelligent world model.

7. The method according to claim 6, characterized in that The target industrial embodied intelligent world model is used to execute at least one of the industrial product full life cycle applications, and the industrial product full life cycle applications include: Model simulation design based on digital genealogy; Product design based on genetically encoded constraints; Flexible sorting based on digital genealogy; Product assembly instructions for reference gene encoding; Genealogy-driven adaptive scene understanding; Genetically encoded similarity-driven process generation; Predictive maintenance of parts based on genealogy; Operation and maintenance guidance based on genetic coding similarity.

8. An industrial digital genealogy system, characterized in that: include: Memory, processor; The memory stores computer-executable instructions; The processor executes the computer-executable instructions stored in the memory, so that the processor performs the method according to any one of claims 1 to 7.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, which are used to implement the method according to any one of claims 1 to 7 when executed by a processor.

10. A computer program product, characterized in that The invention comprises a computer program, which implements the method according to any one of claims 1 to 7 when being executed by a processor.

Images