Digital twin modeling method, device and equipment and storage medium

Abstract

The invention provides a digital twin modeling method, device and equipment and a storage medium, and relates to the technical field of digital twin modeling, and the method comprises the steps: obtaining initial point cloud data; obtaining preprocessed point cloud data; obtaining first point cloud feature information according to the preprocessed point cloud data; carrying out object classification and segmentation processing to obtain second point cloud feature information; generating an interactive twin model; and acquiring real-time state information, and updating the interactive twinborn model according to the real-time state information. Convolutional feature extraction processing based on point cloud data is realized by generating the point cloud image, global and local features of the point cloud can be efficiently extracted, the modeling efficiency and precision are improved, the high modeling precision is beneficial to guaranteeing the precision of an interactive twinborn model to accurately reflect an actual scene, and the method is suitable for large-scale popularization and application. And meanwhile, the high modeling efficiency is beneficial to realizing synchronous change of the interactive twinborn model and an actual scene, the monitoring and interaction real-time performance of digital twinborn is improved, and the monitoring and management requirements of a machine room can be met.

Description

Technical Field

[0001] The present invention relates to the technical field of digital twin modeling, and in particular to a digital twin modeling method, device, equipment and storage medium. Background Art

[0002] As an important part of network infrastructure, the computer room undertakes the operation and maintenance of key equipment. As the scale and complexity of the computer room continue to increase, the traditional two-dimensional drawing design and manual management methods have gradually shown their limitations and cannot meet the needs of efficient design, fine management and real-time monitoring. By building a three-dimensional model of the computer room, various types of information in the computer room are mapped into the virtual space, which facilitates the monitoring and management of the computer room.

[0003] In the existing technology, 3D modeling technology mainly relies on methods such as geometric modeling, image reconstruction and voxel modeling. However, these methods have defects when dealing with complex scenes in computer rooms: geometric modeling relies on artificially constructed geometric structures and is suitable for regular objects, but it is difficult to ensure the accuracy and integrity of the model when facing the complex environment of a computer room; the image reconstruction method relies on images taken from multiple angles to generate a 3D model, but the reconstruction accuracy is unstable due to factors such as lighting and shooting angle, and it is difficult to apply to complex scenes in a computer room; although the voxel modeling method can handle complex geometric shapes, as the complexity of the scene increases, the computing and storage costs increase significantly, and the processing efficiency is low.

[0004] Therefore, the existing three-dimensional modeling of computer room scenes has problems of insufficient model accuracy and low modeling efficiency, and cannot meet the needs of high-precision monitoring and real-time interaction of computer rooms. Summary of the invention

[0005] The present invention provides a digital twin modeling method, device, equipment and storage medium to solve the problems of insufficient model accuracy and low modeling efficiency in the three-dimensional modeling of computer room scenes in the prior art, and the defect that the requirements for high-precision monitoring and real-time interaction of the computer room cannot be met.

[0006] The present invention provides a digital twin modeling method, comprising: Get initial point cloud data; Preprocessing the initial point cloud data to obtain preprocessed point cloud data; According to the pre-processed point cloud data, a dynamic point cloud image is generated and convolution feature extraction processing is performed according to the dynamic point cloud image to obtain first point cloud feature information; According to the first point cloud feature information, perform object classification and segmentation processing to obtain second point cloud feature information; Generate an interactive twin model according to the second point cloud feature information; Acquire real-time status information, and update the interactive twin model according to the real-time status information.

[0007] According to a digital twin modeling method provided by the present invention, the method of generating a dynamic point cloud map based on the pre-processed point cloud data and performing convolution feature extraction processing on the dynamic point cloud map to obtain first point cloud feature information includes: Determining initial processing point cloud information according to the pre-processed point cloud data; Determining a first preset number of nearest neighbor points for each point according to the processed point cloud information, and constructing the dynamic point cloud map; Calculating point feature information of each point according to the dynamic point cloud image; Perform aggregation processing according to the point feature information to obtain point cloud update feature information; According to the point cloud update feature information, the processed point cloud information is updated to obtain new point cloud update feature information, until the number of times the point cloud update feature information is obtained reaches a preset number, and the point cloud update feature information is globally pooled to obtain the first point cloud feature information.

[0008] According to a digital twin modeling method provided by the present invention, generating an interactive twin model according to the second point cloud feature information includes: generating triangular mesh model information according to the second point cloud feature information; Optimizing the triangular mesh model information to obtain mesh optimization model information; Rendering processing is performed according to the mesh optimization model information to generate the interactive twin model.

[0009] According to a digital twin modeling method provided by the present invention, generating triangular mesh model information according to the second point cloud feature information includes: Determine a second preset number of nearest neighbor points for each point according to the second point cloud feature information, and obtain point neighborhood information; According to the point neighborhood information, the local plane and normal vector corresponding to each point are calculated to generate a triangular mesh surface; The triangular mesh model is generated according to each of the triangular mesh surfaces.

[0010] According to a digital twin modeling method provided by the present invention, the step of optimizing the triangular mesh model information to obtain mesh optimization model information includes: Performing mesh simplification processing on the triangular mesh model information to obtain first mesh model information; Performing smoothing processing on the first grid model information to obtain second grid model information; Perform detail enhancement processing on the second mesh model information to obtain the mesh optimization model information.

[0011] According to a digital twin modeling method provided by the present invention, after generating triangular mesh model information according to the second point cloud feature information, the method further includes: Get texture information; According to the texture information, texture mapping processing is performed on the triangular mesh model information.

[0012] According to a digital twin modeling method provided by the present invention, the preprocessing of the initial point cloud data to obtain preprocessed point cloud data includes: Performing global registration processing on the initial point cloud data based on a preset registration algorithm to obtain first point cloud data; Performing denoising processing on the first point cloud data to obtain second point cloud data; Down-sampling is performed on the second point cloud data to obtain the pre-processed point cloud data.

[0013] The present invention also provides a digital twin modeling device, comprising: Data acquisition module, used to obtain initial point cloud data; A data preprocessing module, used to preprocess the initial point cloud data to obtain preprocessed point cloud data; A feature extraction module, configured to generate a dynamic point cloud image based on the pre-processed point cloud data and perform convolution feature extraction processing on the dynamic point cloud image to obtain first point cloud feature information; A three-dimensional model construction module, used to perform object classification and segmentation processing according to the first point cloud feature information, obtain second point cloud feature information, and generate an interactive twin model according to the second point cloud feature information; Model optimization module, used to optimize the model; A real-time rendering interaction module, used to generate display information and perform interactive operations according to the interactive twin model; A model updating module, used to obtain real-time status information and update the interactive twin model according to the real-time status information; The data acquisition module, the data preprocessing module, the three-dimensional model construction module, the model optimization module, the real-time rendering interaction module and the model update module cooperate with each other to implement the above-mentioned digital twin modeling method.

[0014] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, a digital twin modeling method as described above is implemented.

[0015] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements a digital twin modeling method as described in any one of the above.

[0016] The digital twin modeling method, device, equipment and storage medium provided by the present invention have at least beneficial effects: obtaining initial point cloud data, which can truly reflect the geometric shape of the object and accurately reflect the details of the object, and provide a data basis for accurate modeling. Preprocessing the initial point cloud data to remove redundant and abnormal data reduces the amount of data processing, which is conducive to improving modeling efficiency. According to the preprocessed point cloud data, convolution feature extraction processing is performed by generating a dynamic point cloud map to obtain the first point cloud feature information, and feature extraction processing based on the point cloud is realized. According to the first point cloud feature information, object classification and segmentation processing are performed to identify the point cloud features of different objects and segment them, and the second point cloud feature information is obtained, which is convenient for subsequent modeling in units of objects to avoid object adhesion in the model. According to the second point cloud feature information, an interactive twin model is generated, and the interactive twin model is updated according to the real-time status information, so that changes in the real scene can be reflected to the interactive twin model in real time, realizing the function of digital twins, and the interactive twin model can achieve the effect of monitoring and interacting with the actual scene based on the three-dimensional model. In this way, by generating a point cloud map, convolutional feature extraction processing based on point cloud data is realized, which can efficiently extract the global and local features of the point cloud, improve the efficiency and accuracy of modeling. High modeling accuracy is conducive to ensuring the accuracy of the interactive twin model to accurately reflect the actual scene. At the same time, high modeling efficiency is conducive to the synchronous changes of the interactive twin model and the actual scene, improving the real-time monitoring and interaction of the digital twin, and meeting the monitoring and management needs of the computer room. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0018] Figure 1 It is a flow chart of a digital twin modeling method provided by the present invention.

[0019] Figure 2 It is a schematic diagram of the processing process of a digital twin modeling device provided by the present invention.

[0020] Figure 3 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0021] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0022] Combine the following Figure 1 A digital twin modeling method of the present invention is described, comprising: S100: Acquire initial point cloud data; S200: Preprocessing the initial point cloud data to obtain preprocessed point cloud data; S300: generating a dynamic point cloud image based on the pre-processed point cloud data and performing convolution feature extraction processing on the dynamic point cloud image to obtain first point cloud feature information; S400: performing object classification and segmentation processing according to the first point cloud feature information to obtain second point cloud feature information; S500: Generate an interactive twin model according to the second point cloud feature information; S600: Acquire real-time status information, and update the interactive twin model according to the real-time status information.

[0023] Acquire the initial point cloud data, which can truly reflect the geometric shape of the object and accurately reflect the details of the object, providing a data basis for accurate modeling. Preprocess the initial point cloud data to remove redundant and abnormal data to reduce the amount of data processing, which is conducive to improving modeling efficiency. According to the preprocessed point cloud data, convolution feature extraction is performed by generating a dynamic point cloud map to obtain the first point cloud feature information and implement feature extraction based on the point cloud. According to the first point cloud feature information, object classification and segmentation are performed to identify the point cloud features of different objects and segment them, and the second point cloud feature information is obtained to facilitate subsequent modeling in units of objects and avoid object adhesion in the model. According to the second point cloud feature information, an interactive twin model is generated, and the interactive twin model is updated according to the real-time status information. It can reflect the changes in the real scene to the interactive twin model in real time, realize the function of digital twin, and achieve the effect of monitoring and interacting with the actual scene based on the three-dimensional model through the interactive twin model.

[0024] In this way, by generating a point cloud map, convolutional feature extraction processing based on point cloud data is realized, which can efficiently extract the global and local features of the point cloud, improve the efficiency and accuracy of modeling. High modeling accuracy is conducive to ensuring the accuracy of the interactive twin model to accurately reflect the actual scene. At the same time, high modeling efficiency is conducive to the synchronous changes of the interactive twin model and the actual scene, improving the real-time monitoring and interaction of the digital twin, and meeting the monitoring and management needs of the computer room.

[0025] It should be emphasized that point cloud data cannot be directly processed by convolution feature extraction. In the present invention, based on point cloud data, points are used as nodes of graph data, and the edges between nodes in the graph data are determined by the positional relationship between the points, thereby generating a dynamic point cloud graph, and performing convolution feature extraction in the form of a dynamic point cloud graph, thereby overcoming the problem that point cloud data cannot be processed by convolution feature extraction. The high precision of point cloud data is used to reflect the geometric shape and details of actual objects, so that the model accuracy of the final interactive twin model is higher. At the same time, convolution feature extraction based on dynamic point cloud graphs can efficiently extract the global and local features of point clouds, thereby effectively improving modeling efficiency, especially when processing point cloud data in irregular and sparse scenes.

[0026] The interactive twin model maps various types of information of the physical computer room into the virtual space through three-dimensional modeling, realizing real-time linkage and interaction between the physical and virtual spaces, and using real-time status information to synchronously update the interactive twin model to achieve the effect of digital twins, improve the management efficiency and monitoring accuracy of the computer room, and facilitate operation and maintenance.

[0027] In some embodiments of the present invention, the initial point cloud data may be generated by laser scanning, such as scanning a machine room by a three-dimensional laser radar, and the initial point cloud data may be data in a format such as LAS, PLY, or PCD.

[0028] It can be understood that real-time status information refers to the current status information of the actual scene. The real-time status information and the initial point cloud data correspond to the same actual scene, such as the same computer room, so that the interactive twin model can be updated according to the real-time status information to reflect the synchronization status of the actual scene, achieve the effect of real-time monitoring, and facilitate monitoring the operating status of each device in the computer room. And by interacting with the interactive twin model, it is possible to obtain information more intuitively and efficiently to improve management efficiency. When an abnormal fault is detected based on the interactive twin model, the location of the fault and related operating status information can be quickly determined, which is convenient for operation and maintenance personnel to quickly locate the fault and handle maintenance operations.

[0029] In some embodiments of the present invention, the real-time status information may be generated by sensors and monitoring devices in actual scenarios, such as environmental sensors, temperature sensors, fitness sensors, power operation monitoring devices, etc. in a computer room.

[0030] In some embodiments of a digital twin modeling method of the present invention, S200 includes: Performing global registration processing on the initial point cloud data based on a preset registration algorithm to obtain first point cloud data; Performing denoising processing on the first point cloud data to obtain second point cloud data; Down-sampling is performed on the second point cloud data to obtain the pre-processed point cloud data.

[0031] The initial point cloud data includes point cloud data of objects at different positions. The initial point cloud data is globally registered through a preset registration algorithm to unify the data into the same coordinate system to form the first point cloud data, which is conducive to ensuring spatial consistency and facilitating subsequent processing. The initial point cloud data is generated by scanning the object. During the scanning process, due to the influence of environmental interference factors, the initial point cloud data includes noise points, abnormal points, isolated points and other data. Through denoising, the outliers, abnormal points and noise points in the first point cloud data are removed to form the second point cloud data, which is conducive to making the second point cloud data more accurately reflect the actual object, and at the same time can reduce the subsequent data processing volume, which is conducive to improving processing efficiency. By downsampling the second point cloud data to obtain the pre-processed point cloud data, the density of the point cloud data is reduced without affecting the accuracy as much as possible, and the amount of data is further reduced to improve processing efficiency, which is conducive to ensuring real-time performance.

[0032] In some embodiments of the present invention, the preset registration algorithm may be an implementation of an algorithm capable of implementing global registration, such as an ICP (Iterative Closest Point) algorithm, to implement global registration processing of initial point cloud data.

[0033] In some embodiments of the present invention, algorithms such as statistical filtering and radius filtering may be used to implement denoising processing on the first point cloud data.

[0034] In some embodiments of the present invention, a speed-up grid downsampling algorithm or the like may be used to implement downsampling processing of the second point cloud data.

[0035] In some embodiments of a digital twin modeling method of the present invention, S300 includes: Determining initial processing point cloud information according to the pre-processed point cloud data; Determining a first preset number of nearest neighbor points for each point according to the processed point cloud information, and constructing the dynamic point cloud map; Calculating point feature information of each point according to the dynamic point cloud image; Perform aggregation processing according to the point feature information to obtain point cloud update feature information; According to the point cloud update feature information, the processed point cloud information is updated to obtain new point cloud update feature information, until the number of times the point cloud update feature information is obtained reaches a preset number, and the point cloud update feature information is globally pooled to obtain the first point cloud feature information.

[0036] Based on the pre-processed point cloud data, the initial processing point cloud information is determined. According to the coordinates of each point in the processing point cloud information, the first preset number of nearest neighbor points are determined, and the coordinates and other information of the point are used as the node features of the graph, and the distance between the point and the nearest neighbor point is used as the edge feature of the graph, so as to form a dynamic point cloud graph, which can reflect the local feature information. According to the dynamic point cloud graph, the point feature information corresponding to each point is calculated, and then the point feature information is aggregated to obtain the point cloud update feature information, which can be understood as updating the point cloud data. According to the point cloud update feature information, the processing point cloud information is updated, that is, the new processing point cloud information is determined, and the dynamic point cloud graph and the point feature information are constructed again, and then the new point cloud update feature information is obtained. Repeat this process to realize the convolution feature extraction process. When the convolution iteration reaches the preset number of times, that is, the number of times the point cloud update feature information is obtained reaches the preset number of times, the features of each point in the last obtained point cloud update feature information are globally pooled to obtain the first point cloud feature information, which can reflect the global feature information in addition to the local feature information of the retained point, so as to facilitate the subsequent classification and segmentation process.

[0037] In this way, by repeatedly constructing a dynamic point cloud map and feature calculation based on the processed point cloud information, convolution feature extraction is achieved based on the dynamic point cloud map, which integrates the local neighborhood point information of the point, captures local geometric structure and topological relationship between points and other features, and performs a preset number of rounds of processing to gradually extract higher-level and more abstract features. In this process, the dynamic point cloud map is dynamically updated as the point cloud update feature information changes, so as to capture new local spatial features, which is conducive to making the first point cloud feature information more accurately reflect the characteristics of the object.

[0038] In some embodiments of the present invention, the point cloud update feature information may be calculated by the following expression:

[0039] in, is the point feature information of the i-th point when the point cloud is updated for the l-th time. is the set of nearest neighbor points of the i-th point, ( ) is a nonlinear function of feature combination (which can be a multi-layer perceptron MLP), and Aggregate is aggregation processing (such as summing, finding the maximum value, etc.).

[0040] In some embodiments of the present invention, multiple EdgeConv layers can be stacked to generate a dynamic point cloud map and perform convolution feature extraction processing based on the dynamic point cloud map, wherein the number of EdgeConv layers is equal to a preset number of times, that is, equal to the number of times point cloud update feature information is obtained, the preprocessed point cloud data is used as the input data of the first EdgeConv layer, the output data of the previous EdgeConv layer is used as the input data of the next EdgeConv layer, and the output data of the last layer of EdgeConv is the first point cloud feature information.

[0041] In some embodiments of the present invention, a DGCNN processing model can be used to implement a process of obtaining first point cloud feature information based on preprocessed point cloud data, generating a dynamic point cloud map, and performing convolution feature extraction processing based on the dynamic point cloud map. Multiple EdgeConv layers can be stacked in the DGCNN processing model.

[0042] In some embodiments of the present invention, constructing a dynamic point cloud map can be to find K (a first preset number) nearest neighboring points for each point based on the Euclidean distance or other metrics, thereby constructing a local neighborhood for each point. This process regards the point cloud as a graph structure, where the point is a node in the graph, and the edges between the point and the adjacent points in the local neighborhood constitute the edges of the graph. Since this process is based on the dynamic determination between the points, each time the processed point cloud information is updated according to the point cloud update feature information, a new adjacency relationship will be constructed, thereby dynamically forming a different graph, that is, a new dynamic point cloud map.

[0043] Since the initial point cloud data is obtained by scanning, the point cloud data between different objects are continuous. Therefore, the first point cloud feature information is subjected to object classification and segmentation processing to identify different objects and segment the point cloud features of different objects to facilitate subsequent modeling processing.

[0044] In some embodiments of a digital twin modeling method of the present invention, S500 includes: generating triangular mesh model information according to the second point cloud feature information; Optimizing the triangular mesh model information to obtain mesh optimization model information; Rendering processing is performed according to the mesh optimization model information to generate the interactive twin model.

[0045] According to the second point cloud feature information after object classification and segmentation, surface reconstruction is performed based on the point cloud features of different objects to build and integrate the object model to generate triangular mesh model information. By optimizing the triangular mesh model information to appropriately optimize the number of polygons, reduce the redundant complexity of the model while ensuring the accuracy of the model as much as possible, and obtain the mesh optimization model information, it is helpful to avoid the final model being too large and ensure processing efficiency. Rendering is performed based on the mesh optimization model information to obtain an interactive twin model for easy display and interactive operations.

[0046] In some embodiments of a digital twin modeling method of the present invention, generating triangular mesh model information according to the second point cloud feature information includes: Determine a second preset number of nearest neighbor points for each point according to the second point cloud feature information, and obtain point neighborhood information; According to the point neighborhood information, the local plane and normal vector corresponding to each point are calculated to generate a triangular mesh surface; The triangular mesh model is generated according to each of the triangular mesh surfaces.

[0047] According to the second point cloud feature information, based on the second preset number of nearest neighboring points around the point, the point neighborhood information is obtained, which can reflect the local feature information around each store, and then according to the point domain information, based on the position coordinates and features of each point, the local plane and normal vector are calculated to generate a continuous triangular mesh surface to achieve modeling of the object. The triangular mesh surfaces of each object are integrated to form a triangular mesh model to achieve modeling of the actual scene. In this way, the triangular mesh surface is constructed from the local part, and the continuous triangular mesh surface achieves the modeling of the object, and finally forms a triangular mesh model, which is conducive to ensuring the accuracy of local details, and the form of the triangular mesh surface can also accurately reflect the details of the modeling, so that the triangular mesh model has the advantage of high precision and meets the modeling requirements.

[0048] In some embodiments of the present invention, the triangular mesh surface is generated according to the point neighborhood information, which may be achieved by using a surface reconstruction algorithm, such as a Poisson surface reconstruction algorithm.

[0049] In some embodiments of a digital twin modeling method of the present invention, the step of optimizing the triangular mesh model information to obtain mesh optimization model information includes: Performing mesh simplification processing on the triangular mesh model information to obtain first mesh model information; Performing smoothing processing on the first grid model information to obtain second grid model information; Perform detail enhancement processing on the second mesh model information to obtain the mesh optimization model information.

[0050] By performing mesh simplification on the triangular mesh model to reduce the number of redundant polygons and obtain the first mesh model information, the complexity of the mesh is reduced while ensuring the details of the model, which is conducive to reducing the amount of data and complexity of subsequent model processing and improving processing efficiency. By performing smoothing on the first mesh model information to adjust the vertex positions to make the mesh surface smoother and obtain the second mesh model information, it is conducive to eliminating irregular planes on the model surface caused by interference and improving the accuracy of the model. By performing detail enhancement on the second mesh model information to enhance the details of important objects and obtain a mesh optimization model, the key details are highlighted while keeping the model simple. In this way, the model is optimized through mesh simplification, smoothing, and detail enhancement, which is conducive to improving the quality of the final interactive twin model.

[0051] In some embodiments of the present invention, when performing detail enhancement processing, classification information in the object classification and segmentation process may be combined, that is, the types of objects may be identified to perform detail enhancement processing on important and critical objects.

[0052] In some embodiments of the present invention, a Quadric Edge Collapse Decimation algorithm or the like may be used for mesh simplification processing, and a Laplacian smoothing algorithm or the like may be used for smoothing processing.

[0053] In some embodiments of a digital twin modeling method of the present invention, after generating triangular mesh model information according to the second point cloud feature information, the method further includes: Get texture information; According to the texture information, texture mapping processing is performed on the triangular mesh model information.

[0054] When obtaining the initial point cloud data, you can obtain the color point cloud or the corresponding image data, and then obtain the corresponding texture information, and map the texture to the triangular mesh model, which is beneficial to improve the quality of the model and the visual realism of the final interactive twin model.

[0055] A digital twin modeling device provided by the present invention is described below. The digital twin modeling device described below and the digital twin modeling method described above can refer to each other.

[0056] refer to Figure 2 The present invention also provides a digital twin modeling device, comprising: Data acquisition module, used to obtain initial point cloud data; A data preprocessing module, used to preprocess the initial point cloud data to obtain preprocessed point cloud data; A feature extraction module, configured to generate a dynamic point cloud image based on the pre-processed point cloud data and perform convolution feature extraction processing on the dynamic point cloud image to obtain first point cloud feature information; A three-dimensional model construction module, used to perform object classification and segmentation processing according to the first point cloud feature information, obtain second point cloud feature information, and generate an interactive twin model according to the second point cloud feature information; Model optimization module, used to optimize the model; A real-time rendering interaction module, used to generate display information and perform interactive operations according to the interactive twin model; A model updating module, used to obtain real-time status information and update the interactive twin model according to the real-time status information; The data acquisition module, the data preprocessing module, the three-dimensional model construction module, the model optimization module, the real-time rendering interaction module and the model update module cooperate with each other to implement the above-mentioned digital twin modeling method.

[0057] The data acquisition module obtains the initial point cloud data, which can truly reflect the geometric shape of the object and accurately reflect the details of the object, providing a data basis for accurate modeling.

[0058] The data preprocessing module preprocesses the initial point cloud data to remove redundant and abnormal data and reduce the amount of data processing, which is conducive to improving modeling efficiency.

[0059] The feature extraction module performs convolution feature extraction processing based on the pre-processed point cloud data by generating a dynamic point cloud map, obtains the first point cloud feature information, and realizes feature extraction processing based on the point cloud.

[0060] The 3D model building module performs object classification and segmentation processing based on the first point cloud feature information to identify the point cloud features of different objects and segment them, and obtain the second point cloud feature information to facilitate subsequent modeling based on objects and avoid objects sticking together in the model. Based on the second point cloud feature information, an interactive twin model is generated.

[0061] The model optimization module performs model optimization processing in the process of generating the interactive twin model.

[0062] The real-time rendering interaction module performs rendering processing according to the interactive twin model, generates corresponding model display information, and allows interactive operations on the model display information.

[0063] The model update module updates the interactive twin model according to the real-time status information, and can reflect the changes in the real scene to the interactive twin model in real time, realizing the function of the digital twin. Through the interactive twin model, it is possible to monitor and interact with the actual scene based on the three-dimensional model.

[0064] In this way, by generating a point cloud map, convolutional feature extraction processing based on point cloud data is realized, which can efficiently extract the global and local features of the point cloud, improve the efficiency and accuracy of modeling. High modeling accuracy is conducive to ensuring the accuracy of the interactive twin model to accurately reflect the actual scene. At the same time, high modeling efficiency is conducive to the synchronous changes of the interactive twin model and the actual scene, improving the real-time monitoring and interaction of the digital twin, and meeting the monitoring and management needs of the computer room.

[0065] In some embodiments of the present invention, the real-time rendering interaction module can perform rendering processing according to the interactive twin model to realize functions such as virtual reality (VR) display and augmented reality (AR) display, allowing users to perform interactive operations in a virtual environment in a more three-dimensional and realistic manner.

[0066] In some embodiments of the present invention, the model update module can obtain the detection data of the sensor and automatically update the parameters in the interactive twin model and the associated operating status information.

[0067] In some embodiments of a digital twin modeling device of the present invention, it also includes: a data storage and management module, which is used to store data information such as point cloud data, model data, feature attribute information, and historical records, and at the same time manage the user's access rights and operation rights to ensure the security and confidentiality of the data.

[0068] In some embodiments of the present invention, the feature volume extraction module may include a DGCNN processing model, in which multiple EdgeConv layers are stacked. The DGCNN processing model is obtained through a pre-model training process. During the model training process, data enhancement processing, such as random rotation, scaling, flipping, etc., is used to improve the generalization ability of the DGCNN processing model. At the same time, optimization algorithms, such as Adam and SGD algorithms, can be used in combination with regularization processing, such as Dropout, Batch, Normalization, etc., to prevent model overfitting and optimize the model. Then, by evaluating the precision and accuracy of the output results of the model, the performance of the trained model is verified using methods such as cross-validation to determine whether the model meets the usage requirements.

[0069] Figure 3 An example of a physical structure diagram of an electronic device is shown in FIG. Figure 3As shown, the electronic device may include: a processor 810, a communication interface 820, a memory 830 and a communication bus 840, wherein the processor 810, the communication interface 820 and the memory 830 communicate with each other through the communication bus 840. The processor 810 may call the logic instructions in the memory 830 to execute the above-mentioned digital twin modeling method.

[0070] In addition, the logic instructions in the above-mentioned memory 830 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art or the part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0071] On the other hand, the present invention also provides a computer program product, which includes a computer program. The computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute a digital twin modeling method provided by the above methods.

[0072] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to execute a digital twin modeling method provided by the above-mentioned methods.

[0073] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0074] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0075] In the description of the present invention, it should be understood that the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of the features. In the description of the present invention, the meaning of "plurality" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0076] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, unless they are contradictory.

[0077] All actions to obtain signals, information or data in this application are carried out in compliance with the relevant data protection laws and policies of the country where they are located, and with the authorization given by the owner of the corresponding device.

[0078] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A digital twin modeling method, characterized in that: include: Obtain initial point cloud data; Preprocessing the initial point cloud data to obtain preprocessed point cloud data; According to the pre-processed point cloud data, a dynamic point cloud image is generated and convolution feature extraction processing is performed according to the dynamic point cloud image to obtain first point cloud feature information; According to the first point cloud feature information, perform object classification and segmentation processing to obtain second point cloud feature information; Generate an interactive twin model according to the second point cloud feature information; Acquire real-time status information, and update the interactive twin model according to the real-time status information.

2. A digital twin modeling method according to claim 1, characterized in that: The step of generating a dynamic point cloud image based on the pre-processed point cloud data and performing convolution feature extraction processing on the dynamic point cloud image to obtain first point cloud feature information includes: Determining initial processing point cloud information according to the pre-processed point cloud data; Determining a first preset number of nearest neighbor points for each point according to the processed point cloud information, and constructing the dynamic point cloud map; Calculating point feature information of each point according to the dynamic point cloud image; Perform aggregation processing according to the point feature information to obtain point cloud update feature information; According to the point cloud update feature information, the processed point cloud information is updated to obtain new point cloud update feature information, until the number of times the point cloud update feature information is obtained reaches a preset number, and the point cloud update feature information is globally pooled to obtain the first point cloud feature information.

3. A digital twin modeling method according to claim 1, characterized in that: The step of generating an interactive twin model according to the second point cloud feature information includes: generating triangular mesh model information according to the second point cloud feature information; Optimizing the triangular mesh model information to obtain mesh optimization model information; Rendering processing is performed according to the mesh optimization model information to generate the interactive twin model.

4. A digital twin modeling method according to claim 3, characterized in that: The step of generating triangular mesh model information according to the second point cloud feature information includes: Determine a second preset number of nearest neighbor points for each point according to the second point cloud feature information, and obtain point neighborhood information; According to the point neighborhood information, the local plane and normal vector corresponding to each point are calculated to generate a triangular mesh surface; The triangular mesh model is generated according to each of the triangular mesh surfaces.

5. A digital twin modeling method according to claim 3, characterized in that: The step of optimizing the triangular mesh model information to obtain mesh optimization model information includes: Performing mesh simplification processing on the triangular mesh model information to obtain first mesh model information; Performing smoothing processing on the first grid model information to obtain second grid model information; Perform detail enhancement processing on the second mesh model information to obtain the mesh optimization model information.

6. A digital twin modeling method according to claim 3, characterized in that: After generating the triangular mesh model information according to the second point cloud feature information, the method further includes: Get texture information; According to the texture information, texture mapping processing is performed on the triangular mesh model information.

7. A digital twin modeling method according to claim 1, characterized in that: The preprocessing of the initial point cloud data to obtain preprocessed point cloud data includes: Performing global registration processing on the initial point cloud data based on a preset registration algorithm to obtain first point cloud data; Performing denoising processing on the first point cloud data to obtain second point cloud data; Down-sampling is performed on the second point cloud data to obtain the pre-processed point cloud data.

8. A digital twin modeling device, characterized in that: include: Data acquisition module, used to obtain initial point cloud data; A data preprocessing module, used to preprocess the initial point cloud data to obtain preprocessed point cloud data; A feature extraction module, configured to generate a dynamic point cloud image based on the pre-processed point cloud data and perform convolution feature extraction processing on the dynamic point cloud image to obtain first point cloud feature information; A three-dimensional model construction module, used to perform object classification and segmentation processing according to the first point cloud feature information, obtain second point cloud feature information, and generate an interactive twin model according to the second point cloud feature information; Model optimization module, used to optimize the model; A real-time rendering interaction module, used to generate display information and perform interactive operations according to the interactive twin model; A model updating module, used to obtain real-time status information and update the interactive twin model according to the real-time status information; The data acquisition module, the data preprocessing module, the three-dimensional model construction module, the model optimization module, the real-time rendering interaction module and the model update module cooperate with each other to implement a digital twin modeling method as described in any one of claims 1 to 7.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, it implements a digital twin modeling method as described in any one of claims 1 to 7.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, a digital twin modeling method as described in any one of claims 1 to 7 is implemented.

Images