A digital twin integrated system

Abstract

The present invention belongs to the field of data twin technology. The present invention discloses a digital twin integrated system, including: obtaining multi-source heterogeneous data, and performing compression processing, fusion optimization and standard integration processing on the data to obtain a corresponding standardized digital twin set; constructing a corresponding high-precision digital twin scene based on the obtained standardized digital twin data set; performing scene compression and scene optimization processing on the high-precision digital twin scene to obtain a corresponding lightweight digital twin scene package; building a corresponding digital twin platform based on the lightweight digital twin scene package; and performing diversified service publishing and permission management on the digital twin platform to obtain a corresponding standardized digital twin service set; performing real-time monitoring and performance analysis on the digital twin service process based on the standardized digital twin service set, and performing feedback adjustment on the corresponding digital twin platform based on it; the present invention significantly improves the application efficiency and scope of application of digital twin services.

Description

Technical Field

[0001] The present invention relates to the field of digital twin technology, and more specifically, to a digital twin integrated system. Background Art

[0002] With the continuous advancement of smart city construction, digital twin technology, as a key means to achieve deep integration of the physical world and the digital world, is being used more and more widely. Digital twin technology builds digital models of physical entities to achieve real-time monitoring, analysis and prediction of entity status, providing new technical support for urban planning, management and decision-making.

[0003] However, current digital twin technology still faces numerous challenges and limitations in its practical application. First, urban spatial data is multi-source and heterogeneous, with diverse formats and inconsistent standards, making it difficult to effectively integrate. Second, traditional 3D scene construction methods often struggle to balance model accuracy and system performance, resulting in digital twin scenes with insufficient accuracy or excessive system load. Furthermore, existing digital twin systems typically have high requirements for terminal hardware, particularly a strong reliance on GPUs, which limits their deployment and application on low-performance devices.

[0004] In terms of data processing, existing technologies often rely on simple data compression and integration methods, making it difficult to handle massive amounts of heterogeneous data. This leads to data redundancy and inefficient retrieval. They also lack effective data association mechanisms, making it difficult to reflect the complex relationships between urban elements. In terms of scene construction, existing technologies often focus on the construction of static geometric models, lacking effective simulation of dynamic characteristics and temporal changes, making it difficult to truly reflect the characteristics of urban space evolving over time. In terms of system performance, existing digital twin platforms often face problems such as slow loading and delayed response, resulting in a poor user experience, especially when processing large-scale scenes.

[0005] In terms of service release, existing technologies lack unified service standards and interface specifications, making it difficult to support diverse application needs. They also lack effective permission management and security control mechanisms, posing data security risks. In terms of system optimization, existing technologies often rely on static configuration methods, lacking the ability to monitor operating status in real time and dynamically adjust, making it difficult to adapt to changing usage scenarios and load demands.

[0006] In view of this, the present invention proposes a digital twin integrated system to solve the above problems. Summary of the Invention

[0007] In order to overcome the above-mentioned defects of the prior art and achieve the above-mentioned objectives, the present invention provides the following technical solutions:

[0008] A digital twin integrated system, comprising:

[0009] A data acquisition module is used to acquire multi-source heterogeneous data of the target city space, compress and optimize the acquired multi-source heterogeneous data to obtain a corresponding preliminary digital twin database, and perform standard integration processing on the acquired preliminary digital twin database to obtain a corresponding standardized digital twin set; the multi-source heterogeneous data includes spatial data and time series data;

[0010] The scene construction module constructs multi-dimensional spatiotemporal scenes based on the obtained standardized digital twin dataset to obtain the corresponding high-quality visual scenes, and performs attribute mapping and scene verification on them to obtain the corresponding high-precision digital twin scenes;

[0011] The lightweight publishing module builds a data stream distribution solution based on the corresponding high-precision digital twin scene and obtains the corresponding dynamic scene loading framework based on it. At the same time, it builds the corresponding high-performance rendering configuration based on the high-precision digital twin scene, and uses the high-performance rendering configuration and dynamic scene loading framework to package the scene as a whole and configure publishing parameters to obtain the corresponding lightweight digital twin scene package.

[0012] A cross-platform adaptation module is used to perform multi-terminal compatibility processing and GPU-free rendering optimization on the lightweight digital twin scene package to obtain the corresponding digital twin platform;

[0013] The service management module is used to publish diversified services and manage permissions on the digital twin platform to obtain the corresponding standardized digital twin service set;

[0014] Run the monitoring module to perform real-time monitoring and performance analysis of the corresponding digital twin platform based on the standardized digital twin service set, generate the corresponding platform adjustment strategy, and make feedback adjustments to the corresponding digital twin platform based on it.

[0015] Furthermore, the multi-source heterogeneous data includes spatial data and time series data; the spatial data includes BIM model data, GIS map data, point cloud data, and image data; and the time series data includes IoT device data and sensor data.

[0016] Furthermore, the process of obtaining a standardized digital twin dataset includes:

[0017] Perform hierarchical decomposition of the BIM model data and GIS map data in the corresponding spatial data to obtain the hierarchical structure corresponding to the spatial entities in the corresponding urban space, and obtain the subordinate relationships between different spatial entities; and construct the corresponding initial data structure tree based on it;

[0018] Performing data redundancy analysis on the constructed initial data structure tree to obtain a streamlined initial data structure tree, and extracting spatial features of corresponding spatial entities based on the streamlined initial data structure tree; selecting a corresponding spatial index algorithm based on the streamlined structure tree, and constructing an index for spatial entities in the urban space based on the selected spatial index algorithm to obtain corresponding efficient index data;

[0019] Constructing a three-dimensional city model; and simplifying the constructed three-dimensional city model;

[0020] Set different observation distances and angles, and generate LOD models of different levels of detail based on the simplified 3D urban model. Organize the generated LOD models to obtain a corresponding multi-level detail model set. At the same time, establish a transition mapping relationship between LOD models of different levels of detail.

[0021] Perform texture compression and material optimization on the obtained multi-level detail model set to obtain the corresponding lightweight surface model;

[0022] Perform spatiotemporal alignment and semantic annotation on the collected time series data to obtain corresponding multi-source fusion data;

[0023] Construct a mapping rule set; and based on the mapping rule set, perform urban space holographic mapping and data association analysis on the corresponding multi-source fusion data to construct the corresponding urban spatiotemporal association network;

[0024] Using lightweight surface models as the basic framework of urban space and urban spatiotemporal correlation networks as the driving mechanism of urban dynamics, a preliminary digital twin database is constructed by combining pre-established correlation mapping mechanisms between spatial data and time series data.

[0025] The metadata of the obtained preliminary digital twin database is normalized to obtain a standardized digital twin dataset.

[0026] Furthermore, the process of constructing a high-quality visual scene includes:

[0027] Based on the standardized digital twin dataset, the overall scene in the target city space is partitioned and its boundaries are divided to obtain the corresponding scene partitioning scheme. Based on the scheme, the scene elements in the corresponding overall scene are classified and their attributes are configured to obtain the corresponding scene element library.

[0028] Based on the streamlined initial data structure tree, the topological relationship between each scene element in the corresponding scene element library is obtained, and the initial scene skeleton is constructed in combination with the obtained scene element library. The constructed initial scene skeleton is then material mapped and illuminated to obtain the corresponding high-quality visual scene.

[0029] Furthermore, the process of acquiring a high-precision digital twin scene includes:

[0030] Configure the physical properties and define the dynamic characteristics of the corresponding high-quality visual scene to obtain the corresponding scene model with physical characteristics; and perform dynamic behavior simulation and state transition rule setting based on the scene model with physical characteristics to obtain the corresponding intelligent interaction scene;

[0031] Perform historical time series data mapping and future state prediction on the obtained intelligent interaction scenarios to obtain the corresponding multi-temporal scenario sequence;

[0032] The obtained multi-phase scene sequence is subjected to spatiotemporal continuity verification and data consistency maintenance to obtain the corresponding high-precision digital twin scene.

[0033] Furthermore, the process of obtaining the lightweight digital twin scene package includes:

[0034] Design on-demand loading strategies and streaming plans for corresponding high-precision digital twin scenarios, obtain corresponding hierarchical loading priorities and progressive transmission strategies, and build corresponding data stream distribution solutions based on them;

[0035] Based on the data stream distribution scheme, the corresponding scene data is hierarchically stored and cached to obtain a corresponding hierarchical data storage structure; based on the hierarchical data storage structure, the viewpoint-related details are progressively loaded and the near-far layer is adaptively controlled to obtain corresponding dynamic loading control parameters; based on the dynamic loading control parameters, the non-visible area data is delayed loaded and automatically unloaded to obtain a corresponding dynamic scene loading framework;

[0036] Perform data extraction on the corresponding scene data to obtain the corresponding scene geometry data and scene texture data, and perform mesh simplification and topology optimization on the obtained scene geometry data to obtain the corresponding lightweight geometry model; at the same time, perform adaptive compression and resolution adjustment on the obtained scene texture data to obtain the corresponding optimized texture set;

[0037] Based on lightweight geometric models and optimized texture sets, rendering pipeline optimization and shader simplification are performed to obtain the corresponding high-performance rendering configuration; and the overall scene packaging and release parameter configuration are performed through high-performance rendering configuration and dynamic scene loading framework to obtain the corresponding lightweight digital twin scene package.

[0038] Furthermore, the process of building a digital twin platform includes:

[0039] Obtain the hardware configuration and performance parameters of various terminal devices, and build corresponding terminal capability classification tables based on them, and obtain corresponding device capabilities based on them;

[0040] Based on the terminal capability classification table, rendering resources are adaptively allocated and prioritized to obtain a corresponding resource scheduling strategy; based on the obtained resource scheduling strategy, software rendering algorithm optimization and CPU computing efficiency improvement are performed to obtain a corresponding efficient soft rendering engine; and based on the efficient soft rendering engine, multi-threaded parallel processing and task allocation optimization are performed to obtain a corresponding parallel rendering framework;

[0041] Based on the parallel rendering framework, scene features are simplified and key information is retained to obtain a corresponding terminal adaptation scene model, and the terminal adaptation scene model is adapted to the multi-platform interface layout and the interaction mode is converted to obtain a corresponding cross-platform interaction interface; the obtained cross-platform interaction interface is program packaged and interface standardized to obtain a corresponding initial twin platform, and the obtained initial twin platform is subjected to environmental compatibility testing and performance optimization to obtain a corresponding digital twin platform.

[0042] Furthermore, the process of building a standardized digital twin service set includes:

[0043] Perform two-dimensional and three-dimensional map service configuration and Internet map service integration on the service interface of the corresponding digital twin platform to obtain a corresponding multi-map service set, and perform service registration and directory management on the multi-map service set to obtain a corresponding service resource directory;

[0044] Based on the service resource directory, user authority level design and access control rule definition are performed to obtain a corresponding service access control matrix, and a corresponding security service access mechanism is constructed based on the service access control matrix; based on the security service access mechanism, service quality monitoring and load balancing configuration are performed on the digital twin platform to obtain a corresponding service quality assurance plan, and service publishing and subscription management are performed through the service quality assurance plan to obtain a corresponding standardized digital twin service set.

[0045] Furthermore, the process of generating corresponding platform adjustment strategies and performing feedback adjustments on the corresponding digital twin platforms based on the strategies includes:

[0046] Monitor the operational data involved in the data twin service process of the corresponding digital twin platform in real time to obtain the corresponding performance data stream and abnormal event set;

[0047] Based on the abnormal event set and combined with the historical operation data of the digital twin platform, performance bottleneck analysis and resource usage evaluation are performed to obtain a corresponding performance bottleneck report;

[0048] Obtain the target user's user interaction behavior and scene loading performance during the corresponding digital twin service process, and generate corresponding user experience evaluation results based on them;

[0049] Based on the user experience evaluation results and performance bottleneck report, a multi-objective optimization problem is constructed and solved, and a corresponding comprehensive optimization plan is constructed based on the solution results; based on the comprehensive optimization plan, parameter adjustment instructions and resource allocation strategy updates are generated to obtain a corresponding platform adjustment plan; based on the platform adjustment plan, the constructed data twin platform is optimized.

[0050] The technical effects and advantages of the digital twin integrated system of the present invention are as follows:

[0051] 1. Through the efficient fusion and lightweight processing of multi-source heterogeneous data and the high-precision construction of multi-dimensional spatiotemporal scenes, compact management of urban spatial data and realistic simulation of dynamic scenes are achieved, providing accurate data support and visualization foundation for digital twins, improving data processing efficiency and scene realism;

[0052] 2. Through cross-platform compatibility optimization, GPU-free rendering, and standardized service management, the terminal hardware threshold is lowered and the application scenarios are broadened; combined with real-time monitoring and dynamic optimization strategies, the stability, security, and long-term efficient operation of the digital twin platform are guaranteed, promoting the widespread implementation of the technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 Schematic diagram of a digital twin integrated system of the present invention;

[0054] Figure 2 This is a schematic diagram of a digital twin integration method of the present invention. DETAILED DESCRIPTION

[0055] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0056] Example 1

[0057] See also Figure 1 As shown, the digital twin integrated system described in this embodiment includes:

[0058] A data acquisition module is used to acquire multi-source heterogeneous data of the target city space, compress and optimize the acquired multi-source heterogeneous data to obtain a corresponding preliminary digital twin database, and perform standard integration processing on the acquired preliminary digital twin database to obtain a corresponding standardized digital twin set; the multi-source heterogeneous data includes spatial data and time series data;

[0059] The scene construction module constructs multi-dimensional spatiotemporal scenes based on the obtained standardized digital twin dataset to obtain the corresponding high-quality visual scenes, and performs attribute mapping and scene verification on them to obtain the corresponding high-precision digital twin scenes;

[0060] The lightweight publishing module builds a data stream distribution solution based on the corresponding high-precision digital twin scene and obtains the corresponding dynamic scene loading framework based on it. At the same time, it builds the corresponding high-performance rendering configuration based on the high-precision digital twin scene, and uses the high-performance rendering configuration and dynamic scene loading framework to package the scene as a whole and configure publishing parameters to obtain the corresponding lightweight digital twin scene package.

[0061] A cross-platform adaptation module is used to perform multi-terminal compatibility processing and GPU-free rendering optimization on the lightweight digital twin scene package to obtain the corresponding digital twin platform;

[0062] The service management module is used to publish diversified services and manage permissions on the digital twin platform to obtain the corresponding standardized digital twin service set;

[0063] Run the monitoring module to perform real-time monitoring and performance analysis of the corresponding digital twin platform based on the standardized digital twin service set, generate the corresponding platform adjustment strategy, and make feedback adjustments to the corresponding digital twin platform based on it.

[0064] It should be further explained that, in the specific implementation process, the acquisition process of multi-source heterogeneous data includes:

[0065] Setting up a collection unit and deploying the collection unit to pre-set data collection points in the target urban space. The data collection points can be set up at road traffic nodes inside and outside buildings in the target urban space, around public facilities, and pre-set environmental monitoring points, or other locations that can be used to obtain urban space data;

[0066] Based on the acquisition unit, multi-source data is collected from the target urban space to obtain corresponding multi-source heterogeneous data, which is then uploaded to the cloud for storage. The multi-source heterogeneous data includes spatial data such as BIM model data, GIS map data, point cloud data, image data, and time series data such as IoT device data and sensor data.

[0067] It should be further explained that, in the specific implementation process, the acquisition process of the preliminary digital twin database includes:

[0068] Perform hierarchical decomposition of the BIM model data and GIS map data within the corresponding spatial data to obtain the hierarchical structure corresponding to the spatial entities within the corresponding urban space, and obtain the subordinate relationships between different spatial entities; the subordinate relationships include adjacent relationships, connection relationships, and inclusion relationships; for example, the BIM model data can be decomposed into a hierarchical structure such as project, building, floor, room, and component; and the GIS map data can be decomposed into a hierarchical structure such as region, plot, road, and facility;

[0069] Based on the obtained hierarchical structure and subordinate relationships, a corresponding initial data structure tree is constructed, wherein the initial data structure tree is represented as a directed graph, consisting of a number of tree nodes, node edges, and edge attributes. The tree nodes represent spatial entities, the node table represents the subordinate relationships between spatial entities, and the edge attributes include the relationship type, spatial distance, and connection strength between spatial entities;

[0070] Performing data redundancy analysis on the constructed initial data structure tree to obtain a streamlined initial data structure tree, wherein the data redundancy analysis refers to identifying and removing duplicate data, similar attribute data, and low-value information within the constructed initial data structure tree; so as to reduce redundancy and improve access efficiency; for example: for the constructed initial data structure tree, identifying component elements in the spatial entity with a geometric expression repetition rate exceeding 90% and marking them as redundant geometry; identifying edge attributes with a repetition rate exceeding 85% and marking them as redundant attributes; identifying data fields with information entropy below a preset threshold and marking them as low-value information; based on the redundant marking results, optimizing the initial data structure tree, removing duplicate geometric elements, merging similar attribute data, filtering low-value information, and obtaining a streamlined data expression;

[0071] Based on the streamlined initial data structure tree, spatial features of the corresponding spatial entities are extracted. The spatial features include the geometric center coordinates, bounding box boundaries, and geometric shape descriptions of the spatial entities. The corresponding spatial indexing algorithm is selected based on the pre-set query requirements. For example, for dense spatial data in highly urbanized areas, the R-tree indexing algorithm is used; for sparse spatial data in urban peripheries, the quadtree indexing algorithm is used; and for complex three-dimensional scenes, the octree indexing algorithm is used.

[0072] Then, based on the selected spatial indexing algorithm, the spatial entities in the urban space are indexed and constructed to obtain corresponding efficient index data. The efficient index data includes the location information, range information and hierarchical relationship of the spatial entities, so as to achieve efficient organization and rapid retrieval of urban space-related data.

[0073] Based on the collected multi-source heterogeneous data, a three-dimensional urban model corresponding to the target urban space is constructed. The construction process of the corresponding three-dimensional urban model is prior art and will not be elaborated in detail in the present invention.

[0074] The urban 3D model is simplified by using a mesh reduction algorithm and combining the obtained simplification protection priorities;

[0075] Setting different observation distances and angles, and combining the simplified 3D city model to generate LOD models of different levels of detail, including LOD0 (minimal outline), LOD1 (basic blocks), LOD2 (main features), LOD3 (detailed structure) and LOD4 (complete details); wherein the integrity of the geometric information retained by the LOD models of different levels of detail is LOD0 < LOD1 < LOD2 < LOD3 < LOD4;

[0076] Then, a corresponding multi-level detail model set is constructed based on the obtained LOD models of different detail levels; at the same time, a transition mapping relationship between LOD models of different detail levels is established to achieve smooth switching between different LOD models;

[0077] The obtained multi-level detail model set is subjected to texture compression and material optimization to obtain the corresponding lightweight surface model. Texture compression includes three steps: resolution downsampling, image compression, and texture atlas integration. Material optimization clusters and simplifies material parameters to merge known material sets into simplified material sets, reducing the number of material types while maintaining visual expressiveness. For example, materials with similar colors (color difference ΔE < 5) and similar reflective properties are merged into a shared material. For materials used over a large area, atlas technology (such as TextureAtlas) is used to merge multiple small textures into a large texture atlas to reduce rendering state switching.

[0078] The collected IoT device data, sensor data, and other time series data are subjected to spatiotemporal alignment and semantic annotation to obtain corresponding multi-source fusion data. Spatiotemporal alignment includes two main steps: time standardization and spatial registration. Time standardization refers to unifying data with different time sampling frequencies into the same time reference system, such as converting hourly, daily, and monthly data into a unified time granularity. Spatial registration converts spatial data in different coordinate systems into a unified geographic coordinate system to ensure consistency of spatial location. Semantic annotation adds standardized descriptive information to time series data based on urban spatial entities, including data type, unit, numerical range, and semantic association.

[0079] Furthermore, a mapping rule set between the physical space and the digital space is constructed. The mapping rule set includes geometric mapping rules, attribute mapping rules, and behavioral mapping rules. The geometric mapping rules are used to define how physical entities are represented in the digital space; the attribute mapping rules are used to define how physical attributes are converted into digital attributes; and the behavioral mapping rules are used to define how physical behaviors are simulated into digital behaviors. It should be further noted that the construction process of the mapping rule set is prior art and will not be elaborated in detail in this invention.

[0080] Based on the mapping rule set, the obtained multi-source fusion data is subjected to urban space holographic mapping and data association analysis to construct a corresponding urban spatiotemporal association network. The urban spatiotemporal association network consists of network nodes and network edges. The network nodes represent urban events or states, and the network edges represent the influence relationship between events or states. Among them, urban space holographic mapping refers to mapping the urban space from physical space to digital space based on the mapping rule set; data association analysis refers to identifying the association relationship between different data through association rule mining;

[0081] Using lightweight surface models as the basic framework of urban space and urban spatiotemporal correlation networks as the driving mechanism of urban dynamics, a preliminary digital twin database is constructed by combining pre-established correlation mapping mechanisms between spatial data and time series data.

[0082] It should be further explained that, in the specific implementation process, the association mapping mechanism refers to a mapping relationship matrix used to describe the bidirectional association relationship between spatial data and time series data; for example: for each spatial entity, its corresponding time series data stream is associated; for each time series data point, its affected spatial range is associated; and a many-to-many mapping relationship matrix is ​​established based on it.

[0083] It should be further explained that, in the specific implementation process, the acquisition process of the standardized digital twin set includes:

[0084] The metadata of the preliminary digital twin database obtained is normalized to obtain a standardized digital twin dataset; metadata normalization defines a unified data description framework, which includes basic information such as data source, acquisition time, accuracy level, and update frequency.

[0085] It should be further explained that, in the specific implementation process, the construction process of high-quality visual scenes includes:

[0086] Based on the standardized digital twin dataset, the overall scene in the target city space is partitioned and divided into boundaries to obtain a corresponding scene partitioning scheme, which includes main partition information, sub-partition information, partition boundary descriptions, and inter-partition associations. In the scene partitioning process, indicators such as data density, functional relevance, and load balancing need to be considered. The scene partitioning refers to dividing the overall scene corresponding to the target city space into multiple sub-areas by using a spatial partitioning algorithm. The spatial partitioning algorithm used in the present invention is a grid partitioning algorithm.

[0087] Based on the scene partitioning scheme, scene elements within the corresponding overall scene are classified and attributed to obtain a corresponding scene element library. Element classification refers to categorizing various types of scene elements within the sub-area and configuring corresponding attribute parameters for the classified scene elements, forming a rich scene element library. For example, from the time dimension, scene element types can be divided into static elements (such as buildings, roads, terrain, municipal facilities, vegetation, and fixed facilities) and dynamic elements (such as vehicles, pedestrians, environmental factors, temporary facilities, and urban events). Attribute configuration refers to configuring corresponding attribute parameters for the classified scene elements. For example, the attribute parameters required to be configured for building elements within the static elements include building height, number of floors, building age, structural type, usage function, energy consumption characteristics, etc.; the attribute parameters required to be configured for vehicle elements within the dynamic elements include: vehicle model, speed range, turning rules, dwell time, etc.

[0088] Furthermore, the topological relationships between the various scene elements in the corresponding scene element library are obtained based on the streamlined initial data structure tree. The topological relationships include spatial connection relationships (such as adjacent, intersecting, and containing), functional connection relationships (such as power supply, water supply, and transportation), and logical connection relationships (such as subordination, control, and influence).

[0089] Then, an initial scene skeleton is constructed based on the obtained topological relationships and scene element library. The initial scene skeleton is represented by an attribute graph model. In the attribute graph model, nodes represent scene elements, and edges represent relationships between scene elements. Node attributes include element ID, type, geometric features, and attribute set; edge attributes include relationship type, directionality, etc.

[0090] After the initial scene skeleton is constructed, the constructed initial scene skeleton is subjected to material mapping and lighting optimization to obtain a corresponding high-quality visual scene; wherein, the material mapping refers to accurately simulating the optical characteristics of various materials by applying physical-based rendering technology, so that the material performance simulated in the corresponding initial scene skeleton is more consistent with the actual material performance; the lighting optimization is to enhance the light and shadow effects of the initial scene skeleton by applying global illumination, ambient occlusion and other technologies.

[0091] It should be further explained that, in the specific implementation process, the acquisition process of high-precision digital twin scenes includes:

[0092] Physical attribute configuration and dynamic characteristic definition are performed on high-quality visual scenes to obtain a scene model with physical characteristics. Physical attribute configuration sets physical parameters for high-quality visual scenes based on a pre-built physical attribute database. These physical parameters include mass, density, elasticity, friction coefficient, etc. The physical attribute database contains reference data sets such as material density tables, friction coefficient tables, and elasticity coefficient tables, which are usually prepared in advance based on industry literature. Dynamic characteristic definition is used to set behavior rules for scene elements that change over time; for example, the change pattern of traffic flow over time.

[0093] Based on a scene model with physical characteristics, dynamic behavior simulation and state transition rule setting are performed to obtain the corresponding intelligent interaction scene. The dynamic behavior simulation is based on the physical engine for motion calculation and collision detection. The state transition rule is used to define the state changes of scene elements under different conditions, such as the red, yellow, and green state transitions of traffic lights.

[0094] The obtained intelligent interaction scenarios are mapped to historical time series data and predicted for future states to obtain corresponding multi-temporal scenario sequences. Historical time series data mapping refers to associating historical multi-source heterogeneous data with scenario elements. For example, historical traffic flow data can be mapped to road networks, or historical energy consumption data can be mapped to buildings. Future state prediction uses a time series model to predict the possible future evolution of the scenario. For example, a long short-term memory network (LSTM) can be used to predict traffic flow:

[0095] The obtained multi-temporal scene sequence is subjected to spatiotemporal continuity verification and data consistency maintenance to obtain the corresponding high-precision digital twin scene; spatiotemporal continuity verification is used to ensure smooth transition of the scene in the time dimension; data consistency maintenance is used to handle data conflicts and differences caused by different data sources;

[0096] In this embodiment, the multi-dimensional spatiotemporal scene construction process adopts a layered construction strategy, first building the basic spatial framework, then adding dynamic behavior and temporal changes, and finally performing overall optimization and verification to ensure the accuracy and realism of the scene.

[0097] It should be further explained that, in the specific implementation process, the acquisition process of the lightweight digital twin scene package includes:

[0098] Design on-demand loading strategies and stream transmission planning based on high-precision digital twin scenarios. The on-demand loading strategy determines hierarchical loading priorities based on user visual needs. The process of obtaining hierarchical loading priorities includes:

[0099] Obtaining the user's historical viewpoint information, including parameters such as the user's viewpoint position, viewpoint direction, viewpoint angle, and movement speed, and constructing a corresponding viewpoint heat map based on the historical viewpoint information. The viewpoint heat map can be used to reflect the distribution of the user's viewpoint movement trajectory;

[0100] Then, based on the viewpoint distribution map, the viewpoint area of ​​interest to the corresponding user is obtained, and based on this, the initial loading priority of the scene data in the corresponding high-precision digital twin scene is determined; for example, the current visible area is set as the first priority, and the expected visible area is set as the second priority; the expected visible area refers to the visible area that is predicted to enter the user's field of view based on the viewpoint movement trajectory and combined with the current visible area;

[0101] Then, the semantic relationships of each scene element in the corresponding high-precision digital twin scene are obtained and identified, and based on this, the corresponding initial loading priority is further subdivided to obtain the corresponding hierarchical loading priority. For example, key facilities with interactive functions (such as traffic control equipment and important public buildings) are assigned the second highest priority to ensure that these elements are loaded in time; purely decorative elements (such as decorative trees and landscape ornaments) are assigned the second lowest priority and can be loaded later.

[0102] The streaming transmission planning refers to the packaging of scene data in high-precision digital twin scenes into fixed-size blocks based on a pre-planned streaming transmission mechanism, and the design of a progressive data transmission strategy based on network conditions and user experience requirements. Each data block contains a metadata header and actual scene data content; the metadata header is used to describe the block content, dependencies, and display requirements.

[0103] Furthermore, a corresponding data stream distribution scheme is constructed based on the obtained hierarchical loading priority and progressive transmission strategy;

[0104] Based on the data stream distribution scheme, the corresponding scene data is hierarchically stored and cached to obtain a corresponding hierarchical data storage structure; wherein the analytical data storage structure adopts a pyramid structure; the hierarchical storage distributes the scene data to different levels of storage media, for example: obtaining the data importance of the corresponding scene data, and obtaining the user's access frequency to different scene data based on historical visual information; then, the corresponding scene data is divided into a storage structure based on the data importance and access frequency, and hierarchical storage is performed based on the storage structure, the storage structure including four levels: core data, high-frequency data, regular data, and low-frequency data; for example, the scene skeleton can be divided into core data, and a solid-state storage strategy can be adopted to ensure fast startup;

[0105] The cache optimization selects the corresponding cache method based on the storage structure of the corresponding scene data and user behavior. For example, for low-frequency data, the low-frequency detail data will only be loaded after the user stays in a specific area for longer than the expected time.

[0106] Based on the hierarchical data storage structure, progressive loading of viewpoint-related details and adaptive control of near and far levels are performed to obtain corresponding dynamic loading control parameters, and according to the dynamic loading control parameters, delayed loading and automatic unloading of non-visual area data are performed to obtain a corresponding dynamic scene loading framework; wherein, the dynamic loading control parameters include LOD switching distance and memory usage upper limit, etc.; the dynamic scene loading framework is used to realize real-time memory management and resource scheduling to ensure stable operation performance.

[0107] Among them, viewpoint-dependent progressive loading of details refers to dynamically adjusting the detail level of the loaded content based on the user's viewpoint position and movement trend; for example, according to a pre-set level selection function, the appropriate level of detail is selected in combination with the actual viewpoint distance; the mathematical formula of the level selection function is: Where, represents the model precision factor, Indicates the benchmark accuracy, usually 1.0; and Represent the reference distance and the actual observation distance respectively; Represents the visibility factor, which is related to the importance of scene elements in the field of view. The higher the model precision factor, the higher the level of detail selected.

[0108] The near-far layer adaptive control dynamically adjusts the LOD model accuracy according to the visual distance; for example, the required LOD model accuracy is obtained based on a pre-set adaptive control function; wherein the mathematical expression formula of the adaptive control function is: Where, LOD i is the level of detail of the selected LOD model; i=0, 1, 2, 3, 4; Indicates the viewpoint angle; Indicates the radius of the circumscribed circle of the corresponding spatial entity; dynamically switches LOD models of different detail levels based on progressive loading of viewpoint-related details and adaptive control of near and far levels;

[0109] Delayed loading of non-visible areas uses a multi-level priority queue to sort loading tasks according to the visibility and importance of spatial entities; automatic unloading monitors memory usage and unloads non-essential data in reverse order of priority when memory pressure increases; for example, the frustum culling technology is used to identify the current non-visible area; for occluded objects, the occlusion culling technology is applied to avoid loading invisible parts; for data in areas that have not been visible for a long time, the unloading mechanism is triggered to release memory; low-precision proxy data of non-visible areas is retained to facilitate rapid display restoration.

[0110] Data extraction is performed on the obtained scene data to obtain corresponding scene geometry data and scene texture data. The obtained scene geometry data is then mesh simplified and topologically optimized to obtain a corresponding lightweight geometric model. Mesh simplification refers to reducing the model complexity by reducing the number of vertices and facets. For example, the edge collapse algorithm is used to iteratively calculate the collapse cost and select the edge with the least impact to collapse until the target simplification rate is reached. Topology optimization adjusts the connection relationship of the mesh to improve rendering efficiency.

[0111] At the same time, the obtained scene texture data is adaptively compressed and the resolution is adjusted to obtain the corresponding optimized texture set; adaptive compression selects different compression algorithms and compression rates according to the characteristics of the texture content; for example, a low compression rate is used to preserve details in high-frequency detail areas, and a high compression rate is used to reduce the amount of data in flat areas; resolution adjustment assigns appropriate texture resolutions to different objects based on their importance and visual impact. For example, high-resolution textures are used for important buildings, and low-resolution textures are used for ordinary vegetation, thus reducing memory usage while ensuring visual effects;

[0112] Rendering pipeline optimization and shader simplification are performed based on lightweight geometry models and optimized texture sets to obtain corresponding high-performance rendering configurations. Rendering pipeline optimization includes batch merging, draw call reduction, and state sorting to reduce API call overhead. Shader simplification reduces shader complexity through algorithm optimization and precision adjustment. For example, for distant objects, a simplified lighting model and low-precision calculations are used; for nearby important objects, a complete physical lighting model and high-precision calculations are used.

[0113] The high-performance rendering configuration and dynamic scene loading framework are used to package the entire scene and configure publishing parameters to obtain the corresponding lightweight digital twin scene package. The overall scene packaging unifies and compresses the dynamic scene loading framework, lightweight geometric models, optimized texture sets, and high-performance rendering configuration. The publishing parameter configuration sets the scene loading strategy, rendering options, and interaction parameters to ensure the best performance experience on different devices. For example, for high-performance PC devices, high-quality rendering options are configured; for mobile devices, power-saving optimization options are configured. The resulting lightweight digital twin scene package has high compression ratio, low memory usage, and fast loading, making it suitable for deployment on various terminals.

[0114] It should be further explained that, in the specific implementation process, the construction process of the digital twin platform includes:

[0115] Obtain the hardware configuration and performance parameters of various terminal devices and construct a corresponding terminal capability grading table based on them. The terminal capability grading table can be used to categorize terminal devices according to indicators such as CPU performance, memory capacity, and network bandwidth to obtain corresponding device capabilities, including device scores and device levels. The terminal devices include but are not limited to mobile terminals and computer terminals;

[0116] Based on the terminal capability classification table, rendering resources are adaptively allocated and prioritized for the corresponding lightweight digital twin scene package to obtain a corresponding resource scheduling strategy; the resource scheduling strategy refers to dynamically adjusting the scene complexity and loaded content according to the device level; wherein, adaptive allocation of rendering resources refers to dynamically adjusting the rendering resolution, texture quality, geometric accuracy and special effects level of the corresponding terminal device according to the device capabilities; for example, for high-scoring devices, 4K rendering resolution and high-quality textures are allocated; for low-scoring devices, 720p rendering resolution and compressed textures are used; the priority scheduling is to reasonably allocate computing resources according to the importance of scene elements and device capabilities; for example, on low-performance devices, priority is given to ensuring the rendering quality of scene elements under the main vision, and reducing the level of detail of secondary scene elements;

[0117] Based on resource scheduling strategies, software rendering algorithms are optimized and CPU computing efficiency is improved for high-performance rendering configurations within the corresponding lightweight digital twin scenes to obtain a corresponding high-efficiency soft rendering engine. This high-efficiency soft rendering engine can be used to implement functions such as rasterization, texture mapping, and basic lighting. Software rendering algorithm optimization refers to reducing the amount of computation by reducing computational precision, while CPU computing efficiency improvement is achieved by improving the CPU rendering performance of terminal devices through SIMD instruction set optimization and other methods.

[0118] The efficient soft rendering engine is subjected to multi-threaded parallel processing and task allocation optimization to obtain a corresponding parallel rendering framework. The parallel rendering framework fully utilizes the multi-core CPU resources of the terminal device by adopting task partitioning and thread pooling technology. Multi-threaded parallel processing is used to decompose the rendering tasks generated by the efficient soft rendering engine into independent work units and assign them to multiple CPU cores for parallel execution. For example, a tile partitioning strategy is adopted to divide the screen into multiple tiles, with each thread responsible for rendering one or more tiles. Task allocation optimization improves multi-core utilization and achieves dynamic load balancing through work stealing algorithms and load balancing technology.

[0119] Based on the parallel rendering framework, the lightweight digital twin scene is simplified in scene features and key information is retained to obtain a corresponding terminal-adapted scene model, and the terminal-adapted scene model is adapted to the multi-platform interface layout and the interaction mode is converted to obtain a corresponding cross-platform interaction interface; wherein, the terminal-adapted scene model is used to dynamically adjust the geometric complexity and texture details according to the device capabilities; the cross-platform interaction interface supports multiple input methods such as touch, mouse and keyboard, and gestures to achieve consistency and smoothness of operation; it should be further explained that scene feature simplification refers to removing visual redundancy and reducing scene complexity by identifying and extracting key features of the scene; for example, simplifying a building into a representation of an outline plus key feature points retains recognizability while significantly reducing the amount of data; key information retention is used to ensure that the functionality and interactivity of the scene are not affected; for example, even if the visual representation is simplified, the attribute data and interaction functions of the building are still retained;

[0120] The obtained cross-platform interactive interface is packaged and interface standardized to obtain the corresponding initial twin platform, and the obtained initial twin platform is tested for environmental compatibility and performance optimized to obtain the corresponding digital twin platform; the program packaging is used to integrate the lightweight digital twin scene package, efficient rendering engine and cross-platform interactive interface into the initial twin platform, and the interface standardization is used to perform compatibility processing on the service interface of the initial twin platform so that it can adapt to different terminal devices and service platforms (that is, the service interface after interface standardization can support multiple API interface protocols to cope with the API interface differences of different terminal devices or service platforms); performance optimization refers to optimizing the compatibility issues tested during the environmental compatibility test to ensure that it can achieve the expected performance under different terminal devices and service platforms. The service platform includes GIS software platform and BIM software platform, etc.

[0121] It should be further explained that, in the specific implementation process, the construction process of the standardized digital twin service set includes:

[0122] Perform two- and three-dimensional map service configuration and Internet map service integration on the service interface of the corresponding digital twin platform to obtain a corresponding multi-map service set, and perform service registration and directory management on the multi-map service set to obtain a corresponding service resource directory; wherein, the multi-map service set provides diversified map services such as vectors, images, terrains and three-dimensional models; the service resource directory adopts classification organization and metadata description to support service discovery and combined calling.

[0123] It should be further explained that the configuration of 2D and 3D map services includes the configuration and release of basemap services, tile services, and scene services. For example, the release of multi-level tile map services that comply with the WMTS standard supports efficient map browsing. Internet map service integration connects to mainstream online map services such as Amap, Baidu Map, and Tiandi Map, expanding data coverage.

[0124] Service registration refers to recording the published service information into the service registry by using existing service registration centers such as Eureka or Consul. The service information includes the service name, access address, interface description and version information; directory management organizes the hierarchical relationships and dependencies between services to form a service resource directory tree, which facilitates browsing and retrieval of services.

[0125] Based on the service resource directory, user authority levels are designed and access control rules are defined, and a corresponding service access control matrix is ​​constructed based on it. According to the service access control matrix, service call authentication and authorization processes are implemented to obtain a corresponding secure service access mechanism; the service access control matrix is ​​used to define the authority relationship between user roles and resource operations; the secure service access mechanism refers to user access verification through token authentication and encrypted communication to ensure the security of data and services.

[0126] Based on the secure service access mechanism, the digital twin platform is subjected to service quality monitoring and load balancing configuration to obtain a corresponding service quality assurance plan. The service quality assurance plan is then used to manage service publishing and subscription for the corresponding digital twin platform, thereby constructing a corresponding standardized digital twin service set. The service quality assurance plan includes performance monitoring, concurrency control, and fault recovery strategies. The standardized digital twin service set provides a unified service catalog, registration, discovery, and invocation mechanism to achieve flexible combination and convenient access to digital twin capabilities.

[0127] Furthermore, the digital twin platform can provide digital twin services to target users based on the constructed standardized digital twin service set; and enable the digital twin platform to have reliable digital twin service capabilities.

[0128] It should be further explained that, during the specific implementation process, the process of generating the corresponding platform adjustment strategy and making feedback adjustments to the corresponding digital twin platform based on it includes:

[0129] Monitor the operational data involved in the data twin service process of the corresponding digital twin platform in real time to obtain the corresponding performance data stream, which includes key indicators such as CPU occupancy, memory usage, response time, and throughput;

[0130] Based on the performance data stream, anomaly detection and threshold alarm are performed on the digital twin platform to obtain a corresponding abnormal event set; the abnormal event set includes performance fluctuations and abnormal behaviors in the digital twin platform that exceed preset indicator thresholds;

[0131] Furthermore, based on the abnormal event set and combined with the historical operation data of the digital twin platform, performance bottleneck analysis and resource usage evaluation are performed to obtain a corresponding performance bottleneck report; the performance bottleneck analysis is used to identify performance constraints in the process of obtaining digital twin services. For example, based on the abnormal event set and historical operation data, performance constraints are analyzed (such as by analyzing the usage of resources such as CPU, memory, network and disk to identify the resources that first reach the bottleneck); resource usage evaluation is used to analyze the efficiency of resource usage and the rationality of resource allocation;

[0132] Obtain the target user's user interaction behavior during the corresponding digital twin service process, including user behavior data such as user operation frequency, dwell time, and interaction path; and simultaneously obtain the scene loading performance of the digital twin platform during the corresponding digital twin service process, including indicators such as scene initialization time, frame rate stability, and interaction response delay;

[0133] Generate corresponding user experience evaluation results based on user interaction behavior and scene loading performance; user experience evaluation results can be used to quantify operation smoothness, loading waiting time and interactive responsiveness;

[0134] Furthermore, a multi-objective optimization problem is constructed based on the user experience evaluation results and the performance bottleneck report. The multi-objective optimization problem includes multiple objectives such as balancing load performance, data accuracy, and user experience. The constructed multi-objective optimization problem is solved, and a corresponding comprehensive optimization solution is constructed based on the solution results. The particle swarm optimization algorithm is used in the solution process of the multi-objective optimization problem. The particle swarm optimization algorithm is a prior art and will not be elaborated in detail in the present invention.

[0135] Furthermore, based on the comprehensive optimization plan, parameter adjustment instructions and resource allocation strategy updates are generated to obtain the corresponding platform adjustment plan; the parameter adjustment instructions are used to guide the adjustment process of each parameter in the data twin platform; the resource allocation strategy update is used to optimize the allocation method of computing and storage resources;

[0136] Furthermore, based on the platform adjustment plan, the corresponding compression processing, fusion optimization and corresponding hierarchical loading priority process are dynamically adjusted to achieve continuous optimization of the data twin platform.

[0137] The digital twin integration system described in the embodiment of the present application realizes the compact expression and unified management of massive map data by efficiently compressing and fusion-optimizing multi-source heterogeneous data; creates a digital twin scene with high precision and the ability to evolve over time through multi-dimensional spatiotemporal scene construction and dynamic updating; greatly reduces the amount of data transmission through extreme compression and optimization processing, and realizes efficient and smooth scene loading; greatly reduces the terminal hardware requirements through multi-terminal compatibility processing and GPU-free rendering optimization; meets the diverse application needs through diversified service publishing and authority management; and realizes the continuous optimization of the digital twin platform performance through real-time monitoring and performance analysis, ensuring long-term stable and efficient operation. This system significantly improves the efficiency and accuracy of digital twin technology, greatly broadens the boundaries of application scenarios, and opens up new paths for the widespread application of digital twin technology.

[0138] Example 2

[0139] See also Figure 2 As shown, for the parts not described in detail in this embodiment, please refer to the description of Example 1. A digital twin integration method is provided, including:

[0140] Step 1: Obtain multi-source heterogeneous data of the target city space, compress and optimize the obtained multi-source heterogeneous data to obtain a corresponding preliminary digital twin database, and perform standard integration processing on the obtained preliminary digital twin database to obtain a corresponding standardized digital twin set; the multi-source heterogeneous data includes spatial data and time series data;

[0141] Step 2: Based on the obtained standardized digital twin dataset, a multi-dimensional spatiotemporal scene is constructed to obtain the corresponding high-quality visual scene, and attribute mapping and scene verification are performed to obtain the corresponding high-precision digital twin scene;

[0142] Step 3: Build a data stream distribution solution based on the corresponding high-precision digital twin scene, and obtain the corresponding dynamic scene loading framework based on it;

[0143] Step 4: Build a corresponding high-performance rendering configuration based on the high-precision digital twin scene, and use the high-performance rendering configuration and dynamic scene loading framework to package the entire scene and configure the release parameters to obtain the corresponding lightweight digital twin scene package;

[0144] Step 5: Perform multi-terminal compatibility processing and GPU-free rendering optimization on the lightweight digital twin scene package to obtain the corresponding digital twin platform;

[0145] Step 6: Publish diversified services and manage permissions on the digital twin platform to obtain the corresponding standardized digital twin service set;

[0146] Step 7: Based on the standardized digital twin service set, conduct real-time monitoring and performance analysis of the corresponding digital twin platform, generate the corresponding platform adjustment strategy, and make feedback adjustments to the corresponding digital twin platform based on it.

[0147] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art will be able to modify the technical solutions described in the foregoing embodiments or to substitute equivalents for some of the technical features. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

[0148] It should be noted that, in this document, the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, article, or apparatus comprising a series of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or apparatus comprising the element.

[0149] In the description of the present invention, it should be understood that the terms "first", "second", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0150] In the description of the present invention, unless otherwise specified, "plurality" means two or more.

[0151] In the description of the present invention, “several” means one or more, and “a large number” means two or more.

[0152] Throughout this specification, reference to terms such as "one embodiment," "some embodiments," "examples," "specific examples," or "some examples" means that a specific feature, structure, material, or characteristic described in conjunction with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

[0153] The formulas in this manual are all dimensionless and calculated using numerical values. The formulas are obtained by collecting a large amount of data and performing software simulation to obtain the corresponding recent real situation. The preset parameters and thresholds in the formulas are set by technicians in this field according to actual conditions.

[0154] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to the embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the claims and their equivalents.

Claims

1. A digital twin integrated system, characterized in that: include: A data acquisition module is used to acquire multi-source heterogeneous data of the target city space, compress and optimize the acquired multi-source heterogeneous data to obtain a corresponding preliminary digital twin database, and perform standard integration processing on the acquired preliminary digital twin database to obtain a corresponding standardized digital twin set; the multi-source heterogeneous data includes spatial data and time series data; The scene construction module constructs multi-dimensional spatiotemporal scenes based on the obtained standardized digital twin dataset to obtain the corresponding high-quality visual scenes, and performs attribute mapping and scene verification on them to obtain the corresponding high-precision digital twin scenes; The lightweight publishing module builds a data stream distribution solution based on the corresponding high-precision digital twin scene and obtains the corresponding dynamic scene loading framework based on it, including: Based on the corresponding high-precision digital twin scenario, on-demand loading strategy design and streaming transmission planning are carried out to obtain the corresponding hierarchical loading priority and progressive transmission strategy, and based on this, the corresponding data stream distribution solution is built; Based on the data stream distribution scheme, the corresponding scene data is hierarchically stored and cached to obtain a corresponding hierarchical data storage structure; and based on the hierarchical data storage structure, the scene data is progressively loaded with viewpoint-related details and adaptively controlled in terms of near and far levels to obtain corresponding dynamic loading control parameters, including: Viewpoint-dependent progressive loading of details refers to dynamically adjusting the level of detail of the loaded content based on the user's viewpoint position and movement trend. The appropriate level of detail is selected based on a pre-set level selection function and the actual viewpoint distance. The mathematical formula of the level selection function is: Where, represents the model precision factor, Indicates benchmark accuracy; and Represent the reference distance and the actual observation distance respectively; represents the visibility factor; The near-far layer adaptive control dynamically adjusts the LOD model accuracy according to the visual distance; the required LOD model accuracy is obtained based on the pre-set adaptive control function; the mathematical expression of the adaptive control function is: Where, is the level of detail of the selected LOD model; i=0, 1, 2, 3, 4; Indicates the viewpoint angle; Indicates the radius of the circumscribed circle of the corresponding spatial entity; dynamically switches LOD models of different detail levels based on progressive loading of viewpoint-related details and adaptive control of near and far levels; Perform delayed loading and automatic unloading of non-visual area data according to the dynamic loading control parameters to obtain a corresponding dynamic scene loading framework; At the same time, a corresponding high-performance rendering configuration is built based on the high-precision digital twin scene. The scene is packaged as a whole and the release parameters are configured through the high-performance rendering configuration and dynamic scene loading framework to obtain the corresponding lightweight digital twin scene package. A cross-platform adaptation module is used to perform multi-terminal compatibility processing and GPU-free rendering optimization on the lightweight digital twin scene package to obtain the corresponding digital twin platform; The service management module is used to publish diversified services and manage permissions on the digital twin platform to obtain the corresponding standardized digital twin service set; Run the monitoring module to perform real-time monitoring and performance analysis of the corresponding digital twin platform based on the standardized digital twin service set, generate the corresponding platform adjustment strategy, and make feedback adjustments to the corresponding digital twin platform based on it.

2. A digital twin integrated system according to claim 1, characterized in that: The multi-source heterogeneous data includes spatial data and time series data; the spatial data includes BIM model data, GIS map data, point cloud data, and image data; the time series data includes IoT device data and sensor data.

3. A digital twin integrated system according to claim 2, characterized in that: The process of obtaining a standardized digital twin dataset includes: Perform hierarchical decomposition of the BIM model data and GIS map data in the corresponding spatial data to obtain the hierarchical structure corresponding to the spatial entities in the corresponding urban space, and obtain the subordinate relationships between different spatial entities; and construct the corresponding initial data structure tree based on it; Performing data redundancy analysis on the constructed initial data structure tree to obtain a streamlined initial data structure tree, and extracting spatial features of corresponding spatial entities based on the streamlined initial data structure tree; selecting a corresponding spatial index algorithm based on the streamlined structure tree, and constructing an index for spatial entities in the urban space based on the selected spatial index algorithm to obtain corresponding efficient index data; Constructing a three-dimensional city model; and simplifying the constructed three-dimensional city model; Set different observation distances and angles, and generate LOD models of different levels of detail based on the simplified 3D urban model. Organize the generated LOD models to obtain a corresponding multi-level detail model set. At the same time, establish a transition mapping relationship between LOD models of different levels of detail. Perform texture compression and material optimization on the obtained multi-level detail model set to obtain the corresponding lightweight surface model; Perform spatiotemporal alignment and semantic annotation on the collected time series data to obtain corresponding multi-source fusion data; Construct a mapping rule set; and based on the mapping rule set, perform urban space holographic mapping and data association analysis on the corresponding multi-source fusion data to construct the corresponding urban spatiotemporal association network; Using lightweight surface models as the basic framework of urban space and urban spatiotemporal correlation networks as the driving mechanism of urban dynamics, a preliminary digital twin database is constructed by combining pre-established correlation mapping mechanisms between spatial data and time series data. The metadata of the obtained preliminary digital twin database is normalized to obtain a standardized digital twin dataset.

4. A digital twin integrated system according to claim 3, characterized in that: The process of building a high-quality visual scene includes: Based on the standardized digital twin dataset, the overall scene in the target city space is partitioned and its boundaries are divided to obtain the corresponding scene partitioning scheme. Based on the scheme, the scene elements in the corresponding overall scene are classified and their attributes are configured to obtain the corresponding scene element library. Based on the streamlined initial data structure tree, the topological relationship between each scene element in the corresponding scene element library is obtained, and the initial scene skeleton is constructed in combination with the obtained scene element library. The constructed initial scene skeleton is then material mapped and illuminated to obtain the corresponding high-quality visual scene.

5. A digital twin integrated system according to claim 4, characterized in that: The process of acquiring a high-precision digital twin scene includes: Configure the physical properties and define the dynamic characteristics of the corresponding high-quality visual scene to obtain the corresponding scene model with physical characteristics; and perform dynamic behavior simulation and state transition rule setting based on the scene model with physical characteristics to obtain the corresponding intelligent interaction scene; Perform historical time series data mapping and future state prediction on the obtained intelligent interaction scenarios to obtain the corresponding multi-temporal scenario sequence; The obtained multi-phase scene sequence is subjected to spatiotemporal continuity verification and data consistency maintenance to obtain the corresponding high-precision digital twin scene.

6. The digital twin integrated system according to claim 1, characterized in that: The process of obtaining a lightweight digital twin scenario package includes: Perform data extraction on the corresponding scene data to obtain the corresponding scene geometry data and scene texture data, and perform mesh simplification and topology optimization on the obtained scene geometry data to obtain the corresponding lightweight geometry model; at the same time, perform adaptive compression and resolution adjustment on the obtained scene texture data to obtain the corresponding optimized texture set; Based on lightweight geometric models and optimized texture sets, rendering pipeline optimization and shader simplification are performed to obtain the corresponding high-performance rendering configuration; and the overall scene packaging and release parameter configuration are performed through high-performance rendering configuration and dynamic scene loading framework to obtain the corresponding lightweight digital twin scene package.

7. A digital twin integrated system according to claim 6, characterized in that: The process of building a digital twin platform includes: Obtain the hardware configuration and performance parameters of various terminal devices, and build corresponding terminal capability classification tables based on them, and obtain corresponding device capabilities based on them; Based on the terminal capability classification table, rendering resources are adaptively allocated and prioritized to obtain a corresponding resource scheduling strategy; based on the obtained resource scheduling strategy, software rendering algorithm optimization and CPU computing efficiency improvement are performed to obtain a corresponding efficient soft rendering engine; and based on the efficient soft rendering engine, multi-threaded parallel processing and task allocation optimization are performed to obtain a corresponding parallel rendering framework; Based on the parallel rendering framework, scene features are simplified and key information is retained to obtain a corresponding terminal adaptation scene model, and the terminal adaptation scene model is adapted to the multi-platform interface layout and the interaction mode is converted to obtain a corresponding cross-platform interaction interface; the obtained cross-platform interaction interface is program packaged and interface standardized to obtain a corresponding initial twin platform, and the obtained initial twin platform is subjected to environmental compatibility testing and performance optimization to obtain a corresponding digital twin platform.

8. The digital twin integrated system according to claim 7, characterized in that: The process of building a standardized digital twin service set includes: Perform two-dimensional and three-dimensional map service configuration and Internet map service integration on the service interface of the corresponding digital twin platform to obtain a corresponding multi-map service set, and perform service registration and directory management on the multi-map service set to obtain a corresponding service resource directory; Based on the service resource directory, user authority level design and access control rule definition are performed to obtain a corresponding service access control matrix, and a corresponding security service access mechanism is constructed based on the service access control matrix; based on the security service access mechanism, service quality monitoring and load balancing configuration are performed on the digital twin platform to obtain a corresponding service quality assurance plan, and service publishing and subscription management are performed through the service quality assurance plan to obtain a corresponding standardized digital twin service set.

9. The digital twin integrated system according to claim 8, characterized in that: The process of generating the corresponding platform adjustment strategy and making feedback adjustments to the corresponding digital twin platform based on it includes: Monitor the operational data involved in the data twin service process of the corresponding digital twin platform in real time to obtain the corresponding performance data stream and abnormal event set; Based on the abnormal event set and combined with the historical operation data of the digital twin platform, performance bottleneck analysis and resource usage evaluation are performed to obtain a corresponding performance bottleneck report; Obtain the target user's user interaction behavior and scene loading performance during the corresponding digital twin service process, and generate corresponding user experience evaluation results based on them; Based on the user experience evaluation results and performance bottleneck report, a multi-objective optimization problem is constructed and solved, and a corresponding comprehensive optimization plan is constructed based on the solution results; based on the comprehensive optimization plan, parameter adjustment instructions and resource allocation strategy updates are generated to obtain a corresponding platform adjustment plan; based on the platform adjustment plan, the constructed data twin platform is optimized.

Images