Digital twin railway tunnel three-dimensional tile data multi-dimensional spatio-temporal indexing method and device and storage medium

Abstract

The invention relates to the technical field of railway geographic information, particularly discloses a digital twin railway tunnel three-dimensional tile data multi-dimensional spatio-temporal indexing method and equipment and a storage medium, designs a digital twin railway tunnel system-unit-component multi-dimensional spatio-temporal indexing structure, and establishes an internal and external memory collaborative tile data caching algorithm. The method mainly comprises the following steps: firstly, starting from a parallel surrounding space of all component-level twin model axes, realizing OBB surrounding space deviation correction fitting the direction of a railway tunnel strip-shaped space; secondly, sequentially constructing a part, a unit and a system-level index, and realizing association of the three parts by taking the three-dimensional tile as a core; then, dividing cold and hot attributes according to the tile data access frequency, migrating high-frequency access data to a memory, and establishing a data cache; and finally, realizing life cycle-unit mileage-component type multi-dimensional retrieval of the railway tunnel component-level twinborn model in a high-concurrency environment for a multi-dimensional spatio-temporal indexing and caching algorithm.

Description

Technical Field

[0001] The present invention relates to the field of railway geographic information technology, and in particular to a multidimensional spatiotemporal indexing method, device and storage medium for three-dimensional tile data of a digital twin railway tunnel. Background Art

[0002] Aiming at the application requirements of multi-dimensional efficient retrieval and analysis of component-level three-dimensional model data of digital twin railway tunnels, constructing a multi-dimensional spatiotemporal index of three-dimensional tile data is one of the effective ways to improve retrieval performance. On the one hand, the three-dimensional model of the digital twin of the railway tunnel is irregularly distributed in the long strip space along the railway. At the same time, it is affected by complex construction organization and management strategies such as the parallel construction of multiple work sites in the same tunnel project in the same direction or in the opposite direction. The twin model data with local and continuous time is sparsely distributed in space, and an efficient spatiotemporal indexing method is urgently needed to ensure the efficiency of virtual-real interaction of digital twins. The existing spatiotemporal index is mainly designed for continuous and uniformly distributed spatiotemporal data, which is difficult to directly apply to the three-dimensional tile data of railway tunnels with uneven node spatiotemporal distribution and large spatiotemporal overlap, which restricts the multi-dimensional retrieval time performance of railway tunnel twin model data. On the other hand, the three-dimensional tile data of the digital twin railway tunnel is dynamically updated in the whole life cycle stage, and the data access is spatiotemporally local. The data service is easy to fuse in a high concurrent request environment. It is urgent to establish an effective data caching algorithm to improve data retrieval efficiency. The existing caching methods are mainly aimed at data access frequency division, without considering the spatiotemporal multidimensional characteristics of digital twin railway tunnels, and it is difficult to support multidimensional retrieval and access of massive spatiotemporal tile data in a high-concurrency environment.

[0003] The existing three-dimensional tile spatiotemporal indexing methods are mainly divided into multi-layer spatiotemporal indexing and multi-dimensional spatiotemporal indexing. Multi-layer spatiotemporal indexing mainly adopts the time-first or space-first partitioning strategy, which can generally use the convergence of the first-level index to reduce node access and speed up retrieval efficiency. However, the multi-dimensional retrieval mode of this type of method is relatively fixed. Taking the index with time dimension priority partitioning as an example, it is necessary to use the retrieval strategy of time first and space second to ensure high efficiency, which is difficult to support the time performance of diversified retrieval tasks such as single time / space dimension or time-space joint dimension of three-dimensional spatiotemporal tile of railway tunnel; multi-dimensional spatiotemporal indexing aims to index the multi-dimensional space formed by time-space, which is more suitable for diversified retrieval tasks such as single time / space dimension or time-space joint dimension of railway tunnel. This type of method usually directly expands the time dimension in traditional spatial indexes such as R-tree, and faces the long strip irregular distribution characteristics of railway tunnels. Although the overall node distribution is adjusted in the time dimension, the additional time overlap further restricts the retrieval performance. Focusing on the long-band irregular spatial distribution characteristics, several methods and devices are designed for the existing traditional index structures such as grid / KD tree / quadtree / R tree and their improvements, and a variety of single or mixed index structures are designed. However, due to the bottleneck of the index structure itself, it is still difficult to adapt to the long-band irregular distribution characteristics of the railway tunnel digital twin model data, which restricts the multi-dimensional accurate and efficient retrieval of life cycle / unit mileage / component type. Therefore, it is urgent to further develop a multi-dimensional indexing method suitable for digital twin railway tunnel system-unit-component.

[0004] Existing cache algorithms are divided into high and low water level methods driven by storage space and replacement migration methods driven by access frequency. The high and low water level method aims to migrate data when the storage usage of internal and external memory reaches a threshold. This method is suitable for scenarios where the difference in data access frequency is foreseeable. In the face of massive dynamic updates of railway tunnel 3D tile data, it ignores the multi-dimensional spatiotemporal characteristics; the replacement migration algorithm aims to rate the value of data, high-value data is migrated to memory, and low-value data is migrated to external memory. This method can effectively optimize the storage structure, but it may cause waste of storage space and excessive migration costs. It is still difficult to meet the 3D tile data access in the high concurrency environment of railway tunnels. Therefore, it is urgent to further develop a cache method suitable for digital twin railway tunnel 3D tile data. Summary of the invention

[0005] In order to solve the problems existing in the prior art, the present invention provides a multidimensional spatiotemporal indexing method, device and storage medium for three-dimensional tile data of a digital twin railway tunnel. Aiming at the pain point that the existing three-dimensional tile spatiotemporal indexing algorithm is difficult to adapt to the long strip-shaped irregular spatial distribution characteristics, the present invention closely focuses on the three-dimensional tile data structure characteristics of the digital twin railway tunnel and the multi-dimensional retrieval requirements in a high concurrency environment, and designs a multidimensional spatiotemporal indexing and internal and external memory-coordinated dynamic caching algorithm suitable for the long strip-shaped irregular distribution characteristics of the digital twin railway tunnel, forming a digital twin railway tunnel three-dimensional tile data system-unit-component multidimensional spatiotemporal indexing method with internal and external memory coordination, which solves the problems mentioned in the above background technology.

[0006] To achieve the above object, the present invention provides the following technical solution: a multidimensional spatiotemporal indexing method for three-dimensional tile data of a digital twin railway tunnel, comprising the following steps:

[0007] S1. Calculation of the enclosed space that fits the strip-shaped space orientation of the railway tunnel;

[0008] S2, construction of multidimensional spatiotemporal index of system-unit-component;

[0009] S3, internal and external memory coordinated tile data cache construction;

[0010] S4. Multi-dimensional search for life cycle / unit mileage / component type.

[0011] Preferably, in step S1, the calculation of the enclosing space that fits the strip space trend of the railway tunnel includes: starting from the axis-parallel enclosing space of the railway tunnel component-level twin model data, calculating the global enclosing space that fits the strip space trend of the railway tunnel engineering; first, calculating the axis-parallel enclosing space of the component-level twin model data, and summarizing the geometric vertex sets P of all axis-parallel enclosing spaces AABB ; Secondly, calculate the vertex set P AABB The principal component axis is the global enclosing space axis T that fits the strip space trend of the system-level railway engineering. global ; Finally, the bounding space correction of the global optimal constraint is performed component by component; the specific steps are as follows:

[0012] S11. Calculation of the component-level model axis-parallel enclosing space: Access the component-level 3D model geometric vertex set P model , calculate the minimum and maximum values ​​of the vertex set in the x, y, and z directions, that is, the space range parallel to the twin model data axis;

[0013] S12. Calculation of the principal component axis in the strip space of the railway tunnel project: First, all the vertices of the space surrounded by the axis are summarized to form a geometric vertex set P AABB ; Secondly, calculate the vertex set P AABBThe covariance matrix and eigenvectors in the x, y, and z directions, each eigenvector forms a matrix T in column priority global That is, the main component axis that fits the strip space of the railway tunnel project;

[0014] S13. Calculation of the enclosed space of each component in the strip space of the railway tunnel project: Calculate the enclosed space of each component in the direction of the global optimal axial T global The axis-parallel bounding box in the parallel relative coordinate system converts its center point Center, x-axis Vx, y-axis Vy, and z-axis Vz back to the absolute coordinate system, which is the optimal bounding space Box for the component to fit the strip space of the railway tunnel. global .

[0015] Preferably, in step S2, the system-unit-component multidimensional spatiotemporal index construction includes: first, constructing a component-level three-dimensional R* tree index to generate a three-dimensional tile data set and a tile local multi-layer semantic index; then, constructing a unit-level R-tree mileage space index for the tile mileage space, and associating it with the component-level semantic index; finally, dividing the time index for the system-level life cycle, and associating it with the unit-level mileage index; the specific steps are as follows:

[0016] S21. Implement component-level index construction and 3D tile generation: First, construct an R* tree index for the component-level twin model data enclosing space; second, cluster the component-level twin model data stored in the leaf nodes into 3D tiles based on spatial proximity; finally, construct a local multi-level semantic index inside each tile, and output binary geometric information and multi-dimensional attribute information to a 3D tile file for storage;

[0017] S22, implement unit-level index construction: first, obtain the three-dimensional enclosing space, life cycle stage, and unit mileage information of each tile; second, construct a three-dimensional R-tree index for the tile enclosing space; finally, extract the mileage information of each node unit, generate the corresponding R-tree unit-level mileage index, and output the unit-level index file;

[0018] S23. Implement system-level index construction: First, filter out a portion of the time periods of the life cycle stages from each node of the unit-level index; second, merge the overlapping time periods and insert them into a linear linked list to form a system-level index; finally, associate it with the unit-level index and output the system-unit index file.

[0019] Preferably, in step S23, the screening criteria are: (1) the number of unit mileage segments contained does not overflow; (2) the node is at a level as high as possible.

[0020] Preferably, in step S3, the internal and external memory coordinated tile data cache construction includes: first, establishing a knowledge-guided tile data multi-dimensional access frequency division model, and dividing the data into cold, warm and hot data according to the division model; then, based on the data cache algorithm, migrating the hot data and warm data to the internal memory, and migrating the cold data and warm data to the external memory; finally, structuring the data in the memory and establishing a Hash data index; the specific steps are as follows:

[0021] S31. Construction of cold, warm and hot partition model for tile data: Construct tile data access frequency partition model for system life cycle, unit mileage space and component classification structure dimensions respectively. Finally, integrate the factors of each dimension to realize the construction of cold, warm and hot data partition model.

[0022] S32, cache data migration: first, the accessed data is classified in real time based on the cold-warm-hot partition model calculation; then the hot and warm data obtained by the partition are migrated from the external memory to the internal memory based on the FIFO queue, and the cold data in the internal memory is migrated to the external memory; finally, the hot data cache in the internal memory and the cold data persistence in the external memory are realized;

[0023] S33, in-memory data cache index construction: First, add a concatenation prefix to the data and encrypt the data with random salt; second, use the data as the key and random characters as the value to generate a Hash structure index; finally, hot data is oriented to the data cache index to achieve efficient retrieval response.

[0024] Preferably, in step S4, the multi-dimensional retrieval for life cycle / unit mileage / component type includes: first, filtering the unit mileage space of the current life cycle at the system level index; second, filtering the tiles that intersect or are included in the target unit mileage space window at its associated unit level index; then, filtering the three-dimensional model that matches the target component type at the associated component level index; finally, obtaining the model data entity from the data persistence node that collaborates with the internal and external memories.

[0025] On the other hand, to achieve the above-mentioned purpose, the present invention also provides the following technical solution: an electronic device, the electronic device comprising: a processor; and a memory for storing one or more programs;

[0026] When the one or more programs are executed by the processor, the processor executes the multi-dimensional spatiotemporal indexing method for three-dimensional tile data of a digital twin railway tunnel.

[0027] On the other hand, to achieve the above-mentioned purpose, the present invention also provides the following technical solution: a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the multi-dimensional spatiotemporal indexing method of three-dimensional tile data of a digital twin railway tunnel.

[0028] The beneficial effects of the present invention are as follows: the present invention designs a digital twin railway tunnel life cycle / unit mileage / component type multidimensional three-dimensional tile data structure, and constructs a digital twin railway tunnel system-unit-component multidimensional spatiotemporal index and internal and external memory collaborative dynamic cache algorithm. The multidimensional spatiotemporal index is oriented to the global optimal directed enclosing space and linear time index to adjust the uneven distribution of nodes, reduce the spatiotemporal overlap of nodes, and speed up the retrieval efficiency; the dynamic cache algorithm coordinated by internal and external memory is oriented to the multidimensional tile data access frequency division model, establishes a cache of high-frequency access tile data in the memory space, and improves the data access performance in a high-concurrency environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 It is a schematic diagram of the overall process of the multi-dimensional spatiotemporal indexing method of the three-dimensional tile data of the digital twin railway tunnel of the present invention;

[0030] Figure 2 Schematic diagram of a multi-dimensional spatiotemporal index structure of a digital twin railway tunnel system-unit-component in an embodiment of the present invention;

[0031] Figure 3 It is a schematic diagram of a multi-dimensional three-dimensional tile structure of a digital twin railway tunnel life cycle / unit mileage / component type in an embodiment of the present invention;

[0032] Figure 4 It is a schematic diagram of the enclosed space calculation process that fits the direction of the railway strip space in an embodiment of the present invention;

[0033] Figure 5 A schematic diagram of a process for constructing a system-unit-component multidimensional spatiotemporal index in an embodiment of the present invention;

[0034] Figure 6 A schematic diagram of a tile data caching process for coordinated internal and external memory in an embodiment of the present invention;

[0035] Figure 7 It is a schematic diagram of a multi-dimensional search process of life cycle / unit mileage / component type in an embodiment of the present invention;

[0036] Figure 8 This is a schematic diagram of the structure of an electronic device in an embodiment of the present invention;

[0037] In the figure, 210 is a processor; 220 is a storage. DETAILED DESCRIPTION

[0038] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0039] See also Figure 1-5 The present invention provides a technical solution: a multi-dimensional spatiotemporal indexing method for three-dimensional tile data of a digital twin railway tunnel, the process is as follows Figure 1 As shown. First, starting from the axis-parallel enclosing space of the component-level twin model data, the enclosing space that fits the system-level railway tunnel project is calculated; secondly, the component-level, unit-level, and system-level indexes are constructed in turn for the component classification structure, unit mileage, and life cycle dimensions, and the three-dimensional tiles are used as the core to realize the association between the three, and the life cycle / unit mileage / component type multidimensional three-dimensional tile data structure is output; then, facing the tile data access frequency partition model, the high-frequency access tile data is migrated to the memory space, and a hash data index is established in the memory. Finally, the railway tunnel component-level twin model data life cycle-unit mileage-component type multidimensional retrieval is realized.

[0040] The present invention has the following three improvements and breakthroughs: First, in view of the multi-dimensional retrieval requirements of component-level three-dimensional models of long-belt irregularly distributed digital twin railway tunnels, a digital twin railway tunnel system-unit-component multi-dimensional spatiotemporal index is constructed, such as Figure 2 As shown in the figure. The globally optimal bounded space and one-layer time dimension index reduce the access to nodes with large time-space overlap and speed up the retrieval efficiency. Secondly, a multi-dimensional three-dimensional tile data structure of railway tunnel life cycle / unit mileage / component type is designed, such as Figure 3 As shown in the figure, it includes the internal organization structure of the tile and the tile organization structure, which correspond to the component-level index and the system-unit-level index respectively. The tile includes multi-dimensional attribute information of life cycle / unit mileage / component type, as well as geometric information such as vertices, normals, and textures. The multi-dimensional attribute information mainly stores the corresponding time, mileage, and semantic fields in the attribute table structure. The geometric information adopts the mixed form of Json and binary, and stores metadata and actual data respectively. The tile organization structure is mainly presented in a Json file organized in a tree structure. The tree parent-child nodes are stored through the root node and its nested child nodes. For each node, the bounding box field indicating the spatial position, the uri field of the data link, the geometric error field specifying the node loading conditions, and the detailed form field of the loading method are stored. In addition, the start and end point information of the life cycle period / unit mileage section is stored in the extension field. Third, a tile data caching algorithm that cooperates with internal and external memory is designed to establish a cache of high-frequency access tile data in the memory space to improve data access performance in a high-concurrency environment.

[0041] The implementation process of the present invention is realized by computer automation, including the following detailed steps:

[0042] Step 1: This step implements the global enclosing space calculation applicable to the strip space orientation of the railway tunnel, such as Figure 4 Starting from the axis-parallel enclosing space of the railway tunnel component-level twin model data, the global enclosing space that fits the strip-shaped space direction of the railway tunnel project is calculated; first, the axis-parallel enclosing space of the component-level twin model data is calculated, and the geometric vertex sets P of all axis-parallel enclosing spaces are summarized. AABB ; Secondly, calculate the vertex set P AABB The principal component axis is the global enclosing space axis T that fits the strip space trend of the system-level railway engineering. global ; Finally, the bounding space correction of the global optimal constraint is performed component by component.

[0043] Furthermore, step 1 includes the following sub-steps:

[0044] Step 1.1: This step implements the calculation of the axis-parallel enclosing space of the component-level model, such as Figure 4 As shown. Access the component-level 3D model geometric vertex set P model , calculate the minimum and maximum values ​​of the vertex set in the x, y, and z directions, which corresponds to the spatial range parallel to the twin model data axis.

[0045] Step 1.2: This step realizes the calculation of the principal component axis that fits the strip space of the railway tunnel project. First, all the vertices of the parallel enclosed space are summarized to form a geometric vertex set P AABB ; Secondly, calculate the vertex set P AABB The covariance matrix and eigenvectors in the x, y, and z directions, each eigenvector forms a matrix T in column priority global That is, it is the main component axis that fits the strip space of the railway tunnel project.

[0046] Step 1.3: This step realizes the calculation of the enclosed space of each component in the strip space direction of the railway tunnel project. The calculation component is in the direction of the global optimal axial T global The axis-parallel bounding box in the parallel relative coordinate system converts its center point Center, x-axis Vx, y-axis V, and z-axis Vz back to the absolute coordinate system, which is the optimal bounding space Box for the component to fit the strip space of the railway tunnel. global .

[0047] Step 2: This step implements the construction of system-unit-component multidimensional index, such as Figure 5As shown. Component, unit, and system dimension indexes are constructed in sequence, and the system-unit index is associated with the mileage space consistency, and the unit-component index is associated with the 3D geographic space consistency. First, a component-level 3D R* tree index is constructed to generate a 3D tile dataset and a tile local multi-layer semantic index; then, a unit-level R-tree mileage space index is constructed for the tile mileage space, and associated with the component-level semantic index; finally, a time index for the system-level life cycle is divided and associated with the unit-level mileage index.

[0048] Furthermore, step 2 includes the following sub-steps:

[0049] Step 2.1: This step implements component-level index construction and 3D tile generation. First, an R* tree index is constructed for the component-level twin model data enclosing space; second, the component-level twin model data stored in the leaf nodes are clustered into 3D tiles based on spatial proximity; finally, a local multi-level semantic index is constructed inside each tile, and binary geometric information and multi-dimensional attribute information are output to the 3D tile file for storage.

[0050] Step 2.2: This step implements the construction of the unit-level index. First, obtain the three-dimensional enclosing space, life cycle stage, and unit mileage information of each tile; second, build a three-dimensional R-tree index for the tile enclosing space; finally, extract the mileage information of each node unit, generate the corresponding R-tree unit-level mileage index, and output the unit-level index file.

[0051] Step 2.3: This step implements the construction of the system-level index. First, select a portion of the time periods of the life cycle stages from each node of the unit-level index. The selection criteria are: (1) the number of unit mileage segments contained does not overflow; (2) the node level is as high as possible; secondly, merge the overlapping parts of the time periods and insert them into a linear linked list to form a system-level index; finally, associate it with the unit-level index and output the system-unit index file.

[0052] Step 3: This step implements the construction of tile data cache algorithm that cooperates with internal and external memory, such as Figure 6 First, a knowledge-guided multi-dimensional access frequency partition model for tile data is established, and the data is divided into hot, warm, and cold data according to the partition model; then, based on the data cache algorithm, hot data and warm data are migrated to the memory, and cold data and warm data are migrated to the external memory; finally, the data is structured in the memory and a hash data index is established.

[0053] Furthermore, step 3 includes the following sub-steps:

[0054] Step 3.1: This step realizes the construction of the cold, warm and hot partition model of tile data. The tile data access frequency partition model is constructed for the system life cycle, unit mileage space, and component classification structure dimensions respectively. Finally, the cold, warm and hot data partition model is constructed by integrating the factors of each dimension.

[0055] Step 3.2: This step implements cache data migration. First, the accessed data is classified in real time based on the cold-warm-hot partition model calculation; then the hot and warm data obtained by the partition are migrated from the external memory to the internal memory based on the FIFO queue, and the cold data in the internal memory is migrated to the external memory; finally, the hot data cache in the internal memory and the cold data persistence in the external memory are implemented.

[0056] Step 3.3: This step implements the construction of the in-memory data cache index. First, add a concatenated prefix to the data and encrypt the data with random salt; second, use the data as the key and random characters as the value to generate a Hash structure index; finally, hot data is oriented to the data cache index to achieve efficient retrieval response.

[0057] Step 4: This step implements multi-dimensional retrieval of railway tunnel digital twin model data life cycle-unit mileage-component type, such as Figure 7 As shown. First, the system-level index is used to filter the current life cycle period (unit mileage space); second, the associated unit-level index is used to filter the tiles that intersect (or contain) the target unit mileage space window; then, the associated component-level index is used to filter the 3D model that matches the target component type; finally, the model data entity is obtained from the data persistence node that coordinates the internal and external memory.

[0058] Based on the same inventive concept as the above method embodiment, the present application embodiment also provides an electronic device, such as Figure 8 As shown, the device includes: a processor 210; and a memory 220, for storing one or more programs;

[0059] When the one or more programs are executed by the processor 210, the processor executes the multi-dimensional spatiotemporal indexing method for three-dimensional tile data of a digital twin railway tunnel.

[0060] The multi-dimensional spatiotemporal indexing method of the three-dimensional tile data of the digital twin railway tunnel specifically includes the following steps:

[0061] Calculation of enclosed space that fits the strip-shaped space orientation of railway tunnels;

[0062] Construction of multidimensional spatiotemporal index of systems-units-components;

[0063] Tile data cache construction with coordinated internal and external memory;

[0064] Multi-dimensional search based on life cycle / unit mileage / component type.

[0065] Based on the same inventive concept as the above-mentioned method embodiment, the embodiment of the present application also provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by the processor 210, it implements the multi-dimensional spatiotemporal indexing method of three-dimensional tile data of the digital twin railway tunnel.

[0066] The multi-dimensional spatiotemporal indexing method of the three-dimensional tile data of the digital twin railway tunnel specifically includes the following steps:

[0067] Calculation of enclosed space that fits the strip-shaped space orientation of railway tunnels;

[0068] Construction of multidimensional spatiotemporal index of systems-units-components;

[0069] Tile data cache construction with coordinated internal and external memory;

[0070] Multi-dimensional search based on life cycle / unit mileage / component type.

[0071] The present invention designs a digital twin railway tunnel life cycle / unit mileage / component type multidimensional three-dimensional tile data structure, and constructs a digital twin railway tunnel system-unit-component multidimensional spatiotemporal index and internal and external memory collaborative dynamic cache algorithm. The multidimensional spatiotemporal index is oriented to the global optimal directed enclosing space and linear time index to adjust the uneven distribution of nodes, reduce the spatiotemporal overlap of nodes, and speed up the retrieval efficiency; the dynamic cache algorithm coordinated by internal and external memory is oriented to the multidimensional tile data access frequency division model, establishes a cache of high-frequency access tile data in the memory space, and improves the data access performance in a high-concurrency environment.

[0072] Although the present invention has been described in detail with reference to the aforementioned embodiments, it is still possible for those skilled in the art to modify the technical solutions described in the aforementioned embodiments, or to make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A multi-dimensional spatiotemporal indexing method for three-dimensional tile data of a digital twin railway tunnel, characterized in that: The steps include: S1. Calculation of the enclosed space that fits the strip-shaped space orientation of the railway tunnel; S2, construction of multidimensional spatiotemporal index of system-unit-component; S3, internal and external memory coordinated tile data cache construction; S4. Multi-dimensional search for life cycle / unit mileage / component type.

2. The multidimensional spatiotemporal indexing method for three-dimensional tile data of a digital twin railway tunnel according to claim 1 is characterized in that: In step S1, the calculation of the enclosing space that fits the strip space trend of the railway tunnel includes: starting from the axis-parallel enclosing space of the railway tunnel component-level twin model data, calculating the global enclosing space that fits the strip space trend of the railway tunnel project; first, calculating the axis-parallel enclosing space of the component-level twin model data, and summarizing the geometric vertex sets P of all axis-parallel enclosing spaces AABB ; Secondly, calculate the vertex set P AABB The principal component axis is the global enclosing space axis T that fits the strip space trend of the system-level railway engineering. global ; Finally, the bounding space correction of the global optimal constraint is performed component by component; the specific steps are as follows: S11. Calculation of the component-level model axis-parallel enclosing space: Access the component-level 3D model geometric vertex set P model , calculate the minimum and maximum values ​​of the vertex set in the x, y, and z directions, that is, the space range parallel to the twin model data axis; S12. Calculation of the principal component axis in the strip space of the railway tunnel project: First, all the vertices of the space surrounded by the axis are summarized to form a geometric vertex set P AABB ; Secondly, calculate the vertex set P AABB The covariance matrix and eigenvectors in the x, y, and z directions, each eigenvector forms a matrix T in column priority global That is, the main component axis that fits the strip space of the railway tunnel project; S13. Calculation of the enclosed space of each component in the strip space of the railway tunnel project: Calculate the enclosed space of each component in the direction of the global optimal axial T global The axis-parallel bounding box in the parallel relative coordinate system converts its center point Center, x-axis Vx, y-axis Vy, and z-axis Vz back to the absolute coordinate system, which is the optimal bounding space Box for the component to fit the strip space of the railway tunnel. global .

3. The multi-dimensional spatiotemporal indexing method of three-dimensional tile data of a digital twin railway tunnel according to claim 1 is characterized in that: In step S2, the system-unit-component multidimensional spatiotemporal index construction includes: first, constructing a component-level three-dimensional R* tree index to generate a three-dimensional tile data set and a tile local multi-layer semantic index; then, constructing a unit-level R-tree mileage space index for the tile mileage space and associating it with the component-level semantic index; finally, dividing the time index for the system-level life cycle and associating it with the unit-level mileage index; the specific steps are as follows: S21. Implement component-level index construction and 3D tile generation: First, construct an R* tree index for the component-level twin model data enclosing space; second, cluster the component-level twin model data stored in the leaf nodes into 3D tiles based on spatial proximity; finally, construct a local multi-level semantic index inside each tile, and output binary geometric information and multi-dimensional attribute information to a 3D tile file for storage; S22, implement unit-level index construction: first, obtain the three-dimensional enclosing space, life cycle stage, and unit mileage information of each tile; second, construct a three-dimensional R-tree index for the tile enclosing space; finally, extract the mileage information of each node unit, generate the corresponding R-tree unit-level mileage index, and output the unit-level index file; S23. Implement system-level index construction: First, filter out a portion of the time periods of the life cycle stages from each node of the unit-level index; second, merge the overlapping time periods and insert them into a linear linked list to form a system-level index; finally, associate it with the unit-level index and output the system-unit index file.

4. The multi-dimensional spatiotemporal indexing method for three-dimensional tile data of a digital twin railway tunnel according to claim 3 is characterized in that: In step S23, the screening criteria are: (1) the number of unit mileage segments contained does not overflow; (2) the node is at the highest possible level.

5. The multi-dimensional spatiotemporal indexing method of three-dimensional tile data of a digital twin railway tunnel according to claim 1 is characterized in that: In step S3, the internal and external memory coordinated tile data cache construction includes: first, establishing a knowledge-guided tile data multi-dimensional access frequency division model, and dividing the data into cold, warm and hot data according to the division model; then, based on the data cache algorithm, migrating the hot data and warm data to the internal memory, and migrating the cold data and warm data to the external memory; finally, structuring the data in the memory and establishing a Hash data index; the specific steps are as follows: S31. Construction of cold, warm and hot partition model for tile data: Construct tile data access frequency partition model for system life cycle, unit mileage space and component classification structure dimensions respectively. Finally, integrate the factors of each dimension to realize the construction of cold, warm and hot data partition model. S32, cache data migration: first, the accessed data is classified in real time based on the cold-warm-hot partition model calculation; then the hot and warm data obtained by the partition are migrated from the external memory to the internal memory based on the FIFO queue, and the cold data in the internal memory is migrated to the external memory; finally, the hot data cache in the internal memory and the cold data persistence in the external memory are realized; S33, in-memory data cache index construction: First, add a concatenation prefix to the data and encrypt the data with random salt; second, use the data as the key and random characters as the value to generate a Hash structure index; finally, hot data is oriented to the data cache index to achieve efficient retrieval response.

6. The multi-dimensional spatiotemporal indexing method for three-dimensional tile data of a digital twin railway tunnel according to claim 1 is characterized in that: In step S4, the multi-dimensional retrieval for life cycle / unit mileage / component type includes: first, filtering the unit mileage space of the current life cycle at the system level index; second, filtering the tiles that intersect or are included in the target unit mileage space window at its associated unit level index; then, filtering the three-dimensional model that matches the target component type at the associated component level index; finally, obtaining the model data entity from the data persistence node that collaborates with the internal and external memories.

7. An electronic device, characterized in that: The electronic device comprises: a processor (210); and a memory (220) for storing one or more programs; When the one or more programs are executed by the processor (210), the processor executes the multi-dimensional spatiotemporal indexing method for three-dimensional tile data of a digital twin railway tunnel as described in any one of claims 1 to 6.

8. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and when the computer program is executed by the processor (210), the method for multi-dimensional spatiotemporal indexing of three-dimensional tile data of a digital twin railway tunnel as described in any one of claims 1 to 6 is implemented.

Images