Distributed storage and calculation method, system and device for water environment space-time grid model

Abstract

The application discloses a water environment space-time grid model distributed storage calculation method, system and equipment, and relates to the field of data integration; a research area is subjected to multi-level grid dissection and coding processing, and a 4D grid logical model of the research area is obtained; water environment multi-source space-time data of the research area is acquired; the water environment multi-source space-time data is input into the 4D grid logical model, and a 4D grid data model is obtained; and the future attribute change condition and development evolution law of the research area are determined based on the 4D grid data model. The application can realize space-time water environment data fusion, and ensure efficient water environment data management and query service.

Description

Technical Field

[0001] This invention relates to the field of data integration, and in particular to a distributed storage computing method, system and device for a spatiotemporal grid model of a water environment. Background Technology

[0002] Geographic grid technology, with its advantages in hierarchical and segmented data governance, enables unified management and divide-and-conquer, providing an integrated spatial index for multi-source data resource integration. Therefore, it is widely used in large-scale Earth science computing and visualization. For massive spatial data, geographic grids can improve data retrieval efficiency and enhance comprehensive data management capabilities. Chen Shupeng's predictions and estimates of the prospects of geographic grids indicate that they can break down spatial barriers, reduce data memory consumption, and improve data retrieval efficiency, enabling widespread application. Zhou Chenghu et al. defined a geographic grid system, classifying it as the third system in geographical research, characterized by its scientific nature, simplicity, high indexing efficiency, and strong information integration capabilities. However, standardized unit coding operations and indexing are core issues for geographic grid systems. Currently, in-depth exploration of theoretical issues such as grid coding mathematical models, coding methods, and coding operations is still lacking. Particularly for the integration and governance of water environment data with temporal attributes, research on temporal geographic grids is severely lacking. How to manage and integrate multi-source water environment data that changes over time using spatiotemporal grids is a challenging problem that needs to be solved to further utilize grid technology to improve data management capabilities and efficiency. Summary of the Invention

[0003] The purpose of this invention is to provide a distributed storage computing method, system and device for a spatiotemporal grid model of water environment, which can realize the fusion of spatiotemporal water environment data and ensure efficient water environment data management and query services.

[0004] To achieve the above objectives, the present invention provides the following solution:

[0005] A distributed storage computation method for a spatiotemporal grid model of the water environment includes:

[0006] The study area is subjected to multi-level grid subdivision and encoding processing to obtain a 4D grid logical model of the study area; the 4D grid logical model is the gridded study area after grid subdivision and encoding processing.

[0007] Acquire multi-source spatiotemporal data of the water environment in the study area;

[0008] Multi-source spatiotemporal data of the water environment are input into a 4D grid logical model to obtain a 4D grid data model;

[0009] The future attribute changes and evolution patterns of the study area are determined based on a 4D grid data model.

[0010] Optionally, the encoding includes grid hierarchy, latitude and longitude range, depth range, length, width, height, and XYZK-T encoding information;

[0011] Where X represents east-west direction encoding, Y represents north-south direction encoding, Z represents subdivision level, K represents depth direction encoding, and T represents time stamp.

[0012] Optionally, the study area is subjected to multi-level mesh subdivision and encoding to obtain a 4D mesh logical model of the study area, specifically including:

[0013] Determine the grid subdivision level for the study area;

[0014] The quadtree partitioning method is used to perform multi-level grid partitioning and encoding of the study area based on the grid partitioning level, so as to obtain a 4D grid logical model of the study area.

[0015] Optionally, the mesh subdivision level of the study area is determined, specifically including:

[0016] The number of grids, planar resolution, and vertical subdivision of the study area are calculated and hierarchical levels are defined.

[0017] The formula for calculating the number of grid cells is:

[0018] N xy =4 n ;

[0019] The formula for calculating the planar resolution is:

[0020]

[0021] The formula for calculating the hierarchical division during longitudinal partitioning is as follows:

[0022] L z =D z / s xy ;

[0023] Where, N xy The number of meshes in the planar partitioning; n represents the number of times the quadtree partitioning method is used; s xy Indicates the resolution of the grid plane; D x D represents the east-west length of the study area. y L represents the width of the study area in the north-south direction. z Indicates the hierarchical division during vertical partitioning; D z Represents vertical depth information;

[0024] The mesh subdivision level of the study area is determined based on the number of grids, planar resolution, hierarchical division during vertical subdivision, and preset conditions.

[0025] Optionally, multi-source spatiotemporal data of the water environment are input into a 4D grid logical model to obtain a 4D grid data model, specifically including:

[0026] The multi-source spatiotemporal data of the water environment are uniformly transformed to obtain the transformed multi-source spatiotemporal data of the water environment.

[0027] The transformed multi-source spatiotemporal data of the water environment are input into a 4D grid logic model for slicing processing to obtain a spatial data cube model and a time slice dataset.

[0028] The data at each moment in the cyclic time-slice dataset is mapped to a spatial data cube model to form a spatiotemporal dataset;

[0029] Based on the spatiotemporal dataset, a dual index of time and space is constructed to obtain a spatial data cube model that changes over time as a 4D grid data model.

[0030] Optionally, the slicing process includes temporal slicing and spatial slicing.

[0031] Optionally, the future attribute changes and evolution patterns of the study area can be determined based on a 4D grid data model, specifically including:

[0032] In the time dimension, the 4D mesh data model is used to fit the change curve of the entity itself using the attribute values ​​of the time series;

[0033] Predicting future attribute changes in the study area based on the entity's own change curve;

[0034] Spatial trend surfaces are constructed based on 4D mesh data models by using proximity relationship description information and attribute data.

[0035] In the spatiotemporal dual dimensions, visual analysis technology and human-interactive heuristic analysis technology are used to determine the spatiotemporal evolution laws of data, phenomena and environment in the study area based on spatial trend surfaces, thereby obtaining the development and evolution laws of spatiotemporal data in the study area.

[0036] Optionally, multi-source spatiotemporal data of the water environment are input into a 4D grid logical model to obtain a 4D grid data model, which then includes:

[0037] A database sharding and table partitioning approach is used to store spatiotemporal tile data and construct indexes for the 4D grid data model.

[0038] A distributed storage and computing system for a spatiotemporal grid model of the water environment is applied to the aforementioned distributed storage and computing method for a spatiotemporal grid model of the water environment. The distributed storage and computing system for the spatiotemporal grid model of the water environment includes:

[0039] The 4D mesh logical model determination module is used to perform multi-level mesh subdivision and encoding processing on the study area to obtain a 4D mesh logical model of the study area; the 4D mesh logical model is the meshed study area after mesh subdivision and encoding processing.

[0040] The acquisition module is used to acquire multi-source spatiotemporal data of the water environment in the study area;

[0041] The 4D grid data model determination module is used to input multi-source spatiotemporal data of the water environment into the 4D grid logical model to obtain the 4D grid data model.

[0042] The prediction module is used to determine the future attribute changes and development patterns of the study area based on a 4D grid data model.

[0043] An electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the distributed storage computation method for a spatiotemporal grid model of a water environment as described above.

[0044] According to specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0045] This invention discloses a distributed storage computing method, system, and device for a spatiotemporal grid model of water environment. By processing the study area, a 4D grid logical model corresponding to the study area is obtained; multi-source spatiotemporal data of the water environment in the study area are input into the 4D grid logical model to obtain a 4D grid data model; based on the 4D grid data model, the future attribute changes and development evolution laws of the study area are determined, which can realize the fusion of spatiotemporal water environment data and ensure efficient water environment data management and query services. Attached Figure Description

[0046] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0047] Figure 1 This is a schematic diagram of multi-source spatiotemporal data integration of the water environment in an embodiment of the present invention;

[0048] Figure 2 This is a schematic diagram of a spatial data cube model that changes over time in an embodiment of the present invention;

[0049] Figure 3 This is a structural diagram of the distributed storage and computation method for the spatiotemporal grid model of the water environment in this embodiment of the invention;

[0050] Figure 4 This is a schematic diagram of the distributed storage and computation method for the spatiotemporal grid model of the water environment in an embodiment of the present invention;

[0051] Figure 5 This is a flowchart of the distributed storage and computation method for the spatiotemporal grid model of the water environment in an embodiment of the present invention. Detailed Implementation

[0052] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0053] The purpose of this invention is to provide a distributed storage computing method, system and device for a spatiotemporal grid model of water environment, which can realize the fusion of spatiotemporal water environment data and ensure efficient water environment data management and query services.

[0054] This invention's spatiotemporal data organization method based on a grid model can effectively improve the efficiency of spatiotemporal data fusion and achieve efficient management of spatiotemporal data. Specifically, it adopts the "time + space" integrated management approach in the new generation of spatiotemporal big data GIS systems, organizes water environment data based on spatiotemporal grid technology, constructs a multi-temporal level of detail (LOD) model, realizes partition management and spatial quadtree (octree) index construction through spatiotemporal grid, achieves spatiotemporal water environment data fusion, ensures efficient water environment data management and query services, and provides a basic guarantee for realizing three-dimensional visualization of water environment models on the Web.

[0055] like Figure 1 As shown, this invention achieves integrated management of water environment data through three steps: multi-level grid partitioning, grid data integration, and 4D grid model calculation. It utilizes spatiotemporal grids to achieve efficient hierarchical and partitioned management and fusion processing of massive water environment models. Grid partitioning includes partition slices and time slices, such as... Figure 1 As shown, the partitioned slices yield fluid, water surface, map, imagery, topography, and water depth data. Based on large-scale 3D geospatial data processing methods in GIS, such as hierarchical classification, asynchronous scheduling, and parallel processing, a spatiotemporal index is constructed by indexing the spatiotemporal geographic and real-time monitoring data of the water environment using a spatiotemporal grid method. This enables distributed storage and rapid extraction of multi-source data, and the construction of a large-scale hierarchical spatiotemporal data model for the water environment, thus solving the problem of low efficiency in the storage, query, and analysis of spatiotemporal data of the water environment.

[0056] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0057] Example 1

[0058] like Figure 3 and Figure 5 As shown, the distributed storage and computation method for a spatiotemporal grid model of a water environment according to the present invention includes:

[0059] Based on a unified coordinate system, the study area is spatiotemporally meshed. The key to multi-level spatiotemporal meshing lies in the selection of the minimum granularity. Too small a granularity can lead to redundant data storage, increased computational load, and is detrimental to efficient online visual computation and analysis over a large area. Conversely, too large a granularity results in reduced precision; although computational load is significantly reduced, the required visual computation accuracy and visualization effects are not achieved. Therefore, a scientific meshing and encoding system is needed to achieve integrated and efficient management of multi-source spatial data, and to balance visualization effects and computational efficiency in 3D visual computation through multi-level meshing.

[0060] For multi-level mesh partitioning and encoding, a unified spatiotemporal reference needs to be constructed to create a unified spatiotemporal mesh. Multi-level mesh partitioning includes both temporal and spatial partitioning. For terrestrial lake environments, the Z-axis water depth information is generally much smaller than the XY plane spatial range. Therefore, when performing three-dimensional mesh partitioning, different mesh partitioning strategies need to be set for the planar and vertical directions. This invention uses the relatively mature quadtree for spatial LOD, combined with the vertical Z-axis partitioning method for hierarchical partitioning. For the planar XY space partitioning, a multi-level quadtree is used for hierarchical construction. For the vertical Z-axis, there are two methods: one is to set a fixed spacing for mesh partitioning, and the other is to partition according to the planar resolution. If fixed spacing partitioning is used, the number of meshes can be well adjusted and the computational cost can be evaluated, but the mesh is generally not a regular cube. If planar resolution partitioning is used, the mesh is divided into small cubes. For large lakes, the number of meshes will become huge, and it is necessary to evaluate the mesh partitioning level based on the balance between spatial resolution and computational cost.

[0061] Step 101: Perform multi-level grid subdivision and coding on the study area to obtain a 4D grid logical model of the study area. The 4D grid logical model is the gridded study area after grid subdivision and coding. The coding includes grid level, latitude and longitude range, depth range, length, width and height, and XYZK-T coding information. Wherein, X represents the east-west direction coding, Y represents the north-south direction coding, Z represents the subdivision level, K represents the depth direction coding, and T represents the time stamp.

[0062] Step 101 specifically includes:

[0063] Determine the grid subdivision level for the study area, specifically including:

[0064] like Figure 4 As shown, the specific steps are as follows:

[0065] (1) Time Labeling. Based on the time resolution of the data in the study area, the smallest granular time interval is selected as the grid system for division and labeling. For example, the water environment model is calculated hourly to obtain scalar or vector field data, while other imagery, topographic, and other data are organized by year or month, and the spatiotemporal grid is labeled according to the smallest granularity (hour). For data that does not reach the minimum time resolution, the basic geographic data is managed using a constant mode, that is, the data remains unchanged from the previous node to the next node.

[0066] (2) Spatial partitioning calculation. The number of times the quadtree mesh is partitioned according to planar control is denoted by n, D x D represents the east-west length of the study area. y D represents the width of the study area in the north-south direction. z L represents vertical depth information. z The vertical division hierarchy is represented by N, and the total number of planar meshes is represented by N. xy The resolution of the grid plane is represented by s. xy This indicates that the size of the actual area is represented by the resolution of the grid.

[0067] The number of grids, planar resolution, and hierarchical division of the study area are calculated.

[0068] The formula for calculating the number of grid cells is:

[0069] N xy =4 n ;

[0070] The formula for calculating the planar resolution is:

[0071]

[0072] The formula for calculating the hierarchical division during longitudinal partitioning is as follows:

[0073] L z =D z / s xy ;

[0074] Where, N xy The number of meshes in the planar partitioning; n represents the number of times the quadtree partitioning method is used; s xy Indicates the resolution of the grid plane; D xD represents the east-west length of the study area. y L represents the width of the study area in the north-south direction. z Indicates the hierarchical division during vertical partitioning; D z Represents vertical depth information.

[0075] (3) Spatial resolution calculation. Based on the above grid partitioning formula, taking global geospatial space as the partitioning object, the global grid partitioning level, number of grids, and ground resolution correspondence are obtained (as shown in Table 1), and the display scale of the study area in the geographic environment is evaluated. According to the size of the study area, the number of grids is calculated, and the computational load in subsequent visual computing is evaluated and analyzed. The overall computational load is directly proportional to the number of grids, while the grid data and partitioning level have an exponential relationship. On the basis of taking into account both spatial resolution and grid computational load, the partitioning level of the study area under the global grid system is calculated.

[0076] The grid subdivision level of the study area is determined based on the number of grids, planar resolution, hierarchical division during vertical subdivision, and preset conditions. The preset conditions are the correspondence between global grid subdivision level, number of grids, and ground resolution, as shown in Table 1.

[0077] Table 1. Correspondence between global grid subdivision levels, grid quantity, and ground resolution.

[0078]

[0079]

[0080] (4) Initial partitioning level assessment. Based on the spatial scale of the study area, the initial partitioning level of the study area is assessed and determined according to the globally unified grid partitioning system. Under the global grid quadtree partitioning system, the area that can be covered is calculated based on the grid plane resolution and the number of pixels used in the grid graphic representation.

[0081] (5) Determine the maximum meshing level. Based on the planar resolution of the study area, determine the maximum meshing level for the study area. Assess the computational load based on the number of meshes, and then adjust the mesh spacing of the water environment calculation model accordingly, and determine the maximum meshing level.

[0082] (6) Multi-level grid subdivision coding. Starting from the initial grid subdivision level of the study area, a multi-level grid subdivision system for the study area is constructed under the globally unified geocoding and grid framework. Starting from the next level, the study area is subdivided for the first time according to the quadtree subdivision method. The next level grid divides the space of the previous level grid into four parts, and the number of parts is four times that of the previous level. This subdivision is repeated until the maximum grid subdivision level determined in the fifth step is reached, thus completing the grid subdivision.

[0083] (7) 4D Mesh Logical Model. Construct a mesh logical model according to the research mesh subdivision system. Output the hierarchy, latitude and longitude range, depth range, length, width and height, and XYZK-T encoding information of each mesh (where X represents the east-west direction encoding, Y represents the north-south direction encoding, Z represents the subdivision level, K represents the depth direction encoding, and T represents the time identifier). Z and T can be batch encoded to construct a collection of output meshes with the combination of Z and T as the unique identifier. For example, output a mesh description file according to the combination of time identifier and hierarchy. The file name or table name of the output mesh model can be constructed with Z and T.

[0084] The study area is divided into grids to form a 4D grid logical model, resulting in a three-dimensional grid with spatial location and geometric attributes. At the same time, the time dimension is introduced to manage the study area in a grid-based manner, constructing spatiotemporal slice data managed in a time-sharing, layered, and hierarchical manner, thus forming a grid-based analysis and research system for the target area.

[0085] The quadtree partitioning method is used to perform multi-level grid partitioning and encoding of the study area based on the grid partitioning level, so as to obtain a 4D grid logical model of the study area.

[0086] like Figure 5 As shown, step 102: Obtain multi-source spatiotemporal data of the water environment in the study area.

[0087] Step 103: Input the multi-source spatiotemporal data of the water environment into the 4D grid logical model to obtain the 4D grid data model.

[0088] The multi-source spatiotemporal data of the water environment are uniformly transformed to obtain the transformed multi-source spatiotemporal data of the water environment.

[0089] The transformed multi-source spatiotemporal data of the water environment are input into a 4D grid logic model for slicing processing to obtain a spatial data cube model and a time slice dataset.

[0090] like Figure 2 As shown, the spatiotemporal data of the water environment includes three levels of data: first, near-surface atmospheric data, including temperature, wind speed, wind direction, and topographic data; second, three-dimensional water body data, including water temperature, flow, water quality, hydrodynamics, depth, and biomass; and third, shallow geology, including underwater topography and bottom pollutants. Each type of spatiotemporal data is matched and transformed according to coordinates to ensure all data are in a unified coordinate system (this invention selects the transverse Mercator projection system). Data of various formats undergoes unified spatial transformation and is sliced ​​according to grid partitioning rules (the grid partitioning level used when constructing the 4D grid logical model), including temporal and spatial slicing. Temporal slicing mainly assigns time encoding, while spatial slicing mainly constructs a standardized LOD model within the global grid system.

[0091] The data at each moment in the cyclic time slice dataset is mapped onto a spatial data cube model to form a spatiotemporal dataset.

[0092] Based on a spatiotemporal dataset, dimensionality partitioning is used to divide the spatiotemporal data into a temporal attribute set and a spatial dataset. Dimensional partitioning divides data according to the temporal and spatial variations of its attributes. The temporal attribute set reveals the interrelationships of the same attribute among multiple things at the same point in time, enabling prediction of the development trend of the time series.

[0093] Based on the spatiotemporal dataset, a dual index of time and space is constructed to obtain a spatial data cube model that changes over time as a 4D grid data model.

[0094] The resulting spatial data cube model is further divided into a static geometric dataset and a temporal geometric dataset. Temporal data, or time series data, is a data column recording the same indicator in chronological order. All data points within the same column must be of the same caliber and comparable. Temporal data can be either period data or point-in-time data. Finally, time series analysis is performed on the temporal data columns composed of the temporal attribute set and the temporal geometric dataset to identify the statistical characteristics and development patterns of the time series within the sample, thus constructing a spatial data cube model that changes over time.

[0095] The main steps in the slice processing of the converted multi-source spatiotemporal water environment data are as follows:

[0096] (1) Based on the association information of water environment data slices under dimensional segmentation, construct a spatial data cube model, which supports both the three-dimensional model description of the empty physical space and the data cube model description of the data space, and obtain the water environment data cube.

[0097] (2) Divide the entire space into a three-dimensional cubic grid with a certain size, that is, each grid has length, width and height, and its position is represented by grid XYZK-T encoding information (where X represents the east-west direction encoding, Y represents the north-south direction encoding, Z represents the subdivision level, K represents the depth direction encoding, and T represents the time identifier); map the time series attribute data v into each grid, and store the mapped cubic grid in a unified manner.

[0098] (3) The data at each time step in the cyclic time series dataset are mapped to the mapped cube grid to form the XYZ-DT dataset; a time and space dual index is constructed to obtain a spatial data cube model that changes with time as a 4D grid data model.

[0099] A database sharding and table partitioning approach is used to store spatiotemporal tile data and construct indexes for the 4D grid data model.

[0100] Based on spatiotemporal tile data storage, parallel computing is carried out, including spatial interpolation computing, spatial autoregression and spatial correlation computing, to support large-scale water environment hydrodynamic model analysis and computing, including hydrodynamic model operators, water quality model operators, grid flow field computing and grid particle force computing.

[0101] To achieve effective storage and efficient management of massive amounts of data, this invention designs an adaptive gridded distributed storage method to achieve efficient storage of massive spatiotemporal data. A database sharding and table partitioning approach is adopted for spatiotemporal tile data storage and index construction. Database query performance and data storage methods are optimized according to the 3D spatiotemporal geographic environment display level. After analysis and comparison, it is determined that a single table can store a maximum of 47 = 16384 data entries; tiles at 3D visualization level 7 and below can be placed in a single table. After database performance evaluation, it is determined that a single database node can hold a maximum of 43 = 64 tables, with a maximum of 43 * 47 = 410 = 1048576 data entries. Therefore, to enable data queries below level 10, all tile data at 3D map display levels 10 and below are compressed and stored in a single file library. According to different grid partitioning levels, levels 0 to 7 are set to a single-database, single-edge mode; levels 8 to 10 to a single-database, multi-table mode; and levels above 10 to a multi-database, multi-table mode. Table 2 shows the correspondence between grid partitioning levels and the database and table names in the distributed database.

[0102] Table 2. Correlation between Grid Partitioning Levels and Distributed Database Storage Design

[0103] z＝0～7 Single database, single table z+string1 z + string 2 xyz z＝8～10 Single database with multiple tables z+string1 f1(x,y,z) xyz z＝11～22 Multiple databases and multiple tables f2(x,y,z) f1(x,y,z) xyz

[0104] Based on the grid xyz encoding, the data of the grid slice is determined to be stored in the distributed database. This requires constructing the association between the grid spatial encoding xyz (hierarchy K and time T are not involved in the database sharding logic design but are stored as grid attributes) and the distributed database. A mapping rule from grid encoding to distributed database names and table names is proposed to achieve automatic database sharding and table partitioning. First, the parameters are defined:

[0105] D: The maximum number of records a table can store.

[0106] T: The maximum number of tables a database can hold.

[0107] Q: The maximum number of records a database can store, i.e., equal to D*T.

[0108] R: The row number of the database containing a particular tile within the entire square two-dimensional table (the entire Earth map).

[0109] C: The column number of the database containing a particular tile within the entire square two-dimensional table (the entire Earth map).

[0110] r: The row number of the table containing a certain tile in its database (which can also be viewed as a square two-dimensional table).

[0111] c: The column number of the table containing a certain tile in its database (which can also be viewed as a square two-dimensional table).

[0112] After deriving R, C, r, and c from x, y, and z, we concatenate R, C, and z into a string to obtain the database name, and concatenate r, c, and z into a string to obtain the table name. We construct mapping functions f1(x, y, z) between grid coding and table names, and f2(x, y, z) between x, y, and z and database names. The mapping rules between grid coding and distributed database storage are shown in Table 3.

[0113] Table 3 Mapping Functions Between Grid Coding and Distributed Database Storage

[0114]

[0115] Based on the construction of a spatiotemporal grid system and data integration, calculations based on a 4D grid data model are carried out from both temporal and spatial dimensions.

[0116] Step 104: Determine the future attribute changes and development patterns of the study area based on the 4D grid data model.

[0117] In the time dimension, the 4D grid data model uses time series attribute values ​​to fit the entity's own change curve.

[0118] In the time dimension, the entity's own change curve is fitted using the numerical attributes of the time series, and nonlinear fitting methods such as the Autoregressive Moving Average Model (ARAM) are used to fit the entity's own change curve.

[0119] Predict future attribute changes in the study area based on the entity's own change curve.

[0120] Spatial trend surfaces are constructed based on 4D grid data models by using proximity relationship description information and attribute data.

[0121] In the spatiotemporal dual dimensions, visual analysis technology and human-interactive heuristic analysis technology are used to determine the spatiotemporal evolution laws of data, phenomena and environment in the study area based on spatial trend surfaces, thereby obtaining the development and evolution laws of spatiotemporal data in the study area.

[0122] In the spatiotemporal dimensions, spatial trend surfaces are constructed based on entity proximity descriptions and attribute data. Orientation descriptions are used to express spatial location relationships between entities. Spatial change trends are calculated using spatial interpolation methods, neighbor entity retrieval, semivariance functions, and trend surface fitting. Iterative growth methods are then used to extract spatial development paths.

[0123] Calculation of 4D mesh data model:

[0124] On the one hand, it provides an algorithmic foundation for multi-level LOD construction, especially in preserving mesh data characteristics during mesh merging and simplification, and supporting interpolation calculations, including spatial interpolation and attribute interpolation calculations, during mesh refinement. On the other hand, it supports 3D water body modeling, water quality rendering, spatial interpolation calculations, spatial analysis, and spatial topology calculations during the coupled motion of hydrodynamic particle systems.

[0125] This invention mainly includes: inputting multi-source spatiotemporal data (flow field, water surface, map, image, topography, and water depth), performing multi-level grid subdivision on the study area, and then integrating the multi-source spatiotemporal data with the subdivided study area into 4D grid data. This invention studies the integration technology of spatiotemporal data with inconsistent spatial scale, time dimension, and data accuracy, realizes the standardization of distributed multi-scale, multi-format, and multi-temporal data, forms a comprehensive and systematic multi-source data integration tool, and constructs a 4D grid data model under a unified spatiotemporal framework.

[0126] This invention performs grid subdivision and encoding under a unified spatiotemporal benchmark, and then constructs a spatiotemporal grid method system using a three-dimensional geographic grid with continuous encoding in the time dimension. On this basis, data slicing management and distributed storage are carried out, and a multi-source vector raster integrated fusion method is used to realize the spatiotemporal integration of multi-layer and multi-regional data.

[0127] First, spatial vector data is integrated, fusing multi-source geospatial vector data with multiple semantic, spatiotemporal, scale, acquisition methods, storage formats, and spatial benchmark inconsistencies. The basic theories and methods include the fusion of geographic data models, the fusion of geographic element semantics, and the unification of spatial geographic data projection and coordinate systems. This involves the fusion of simple to complex data models, the fusion of complex to simple data models, and the fusion between complex data models. Within a big data framework, multi-source thematic spatial libraries are processed and created, including various types of spatial data such as vector, imagery, raster, and 3D models. The main functions implemented include vector geographic data format conversion, data import, editing, symbolization, projection conversion, and data compression.

[0128] Secondly, raster data fusion involves fusing image data and vector information from multiple remote sensors. This process processes multi-source data that are spatially or temporally redundant or complementary, according to certain rules (or algorithms), to obtain more accurate and richer information than any single data source, generating a synthetic image with new spatial, spectral, and temporal characteristics. Due to the differences in attributes, space, and time, remote sensing image data fusion should begin with advanced data preprocessing. This includes spatial geometric correction, noise reduction, absolute registration (geographic coordinate registration) or relative registration, and quantization of non-remote sensing data from images of different sources and resolutions, to form digital images, image cubes, and feature maps from each sensor. Figure 3 A spatial dataset or database composed of auxiliary data such as 3D topographic maps and various geophysical data distribution maps.

[0129] Example 2

[0130] A distributed storage and computing system for a spatiotemporal grid model of a water environment is applied to the distributed storage and computing method for a spatiotemporal grid model of a water environment described in Example 1. The distributed storage and computing system for a spatiotemporal grid model of a water environment includes:

[0131] The 4D mesh logical model determination module is used to perform multi-level mesh subdivision and encoding processing on the study area to obtain a 4D mesh logical model of the study area. The 4D mesh logical model is the meshed study area after mesh subdivision and encoding processing.

[0132] The acquisition module is used to acquire multi-source spatiotemporal data of the water environment in the study area.

[0133] The 4D grid data model determination module is used to input multi-source spatiotemporal data of the water environment into the 4D grid logical model to obtain the 4D grid data model.

[0134] The prediction module is used to determine the future attribute changes and development patterns of the study area based on a 4D grid data model.

[0135] An electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the distributed storage computing method for a spatiotemporal grid model of a water environment as described in Embodiment 1.

[0136] This invention utilizes large-scale 3D geospatial data processing methods in GIS, such as hierarchical and asynchronous scheduling and parallel processing, to construct a hierarchical spatiotemporal data model of a large-scale water environment through a multi-level spatiotemporal grid, thereby solving the problems of data storage, querying and calculation efficiency of water environment spatiotemporal data.

[0137] This invention mainly includes the following aspects in the spatiotemporal regional analysis and visual computation of geographic environmental entities:

[0138] (1) Supports the dimensional segmentation calculation of spatiotemporal datasets, and describes the geometric and non-geometric attributes of geographic entities in two dimensions of space and time. That is, it expresses the change information of the entity itself by describing the time series numerical data of the entity attribute data, and divides the spatial geometric data into static geometric datasets and temporal geometric datasets according to the entity's location and shape change.

[0139] (2) Based on dimensional segmentation, the loading and unloading of time series datasets are realized through time axis management, and the loading and unloading of geometric space datasets are realized through spatial rapid linkage.

[0140] (3) Supports computational simulation and predictive early warning. By fitting the entity's own change curve with the attribute values ​​of the time series, the future attribute changes of the entity can be predicted. By constructing spatial trend surfaces with the entity's proximity relationship description information and attribute data, the trend calculation and development direction extraction of spatial data can be completed to achieve predictive early warning of future spatial attributes.

[0141] (4) Visual computing and analysis. In the two dimensions of time and space, combining the changes of the entity itself and the global spatial trend, based on visual analysis technology and human interactive heuristic analysis technology, we use the continuously changing and fluctuating spatial trend surface to discover the spatiotemporal evolution law of data, phenomena and environment, extract and summarize the development and evolution law of spatiotemporal data, and detect the spatial development trend in the next time period from a global perspective.

[0142] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the systems disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the descriptions are relatively simple; relevant parts can be referred to the method section.

[0143] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. Furthermore, those skilled in the art will recognize that, based on the ideas of the present invention, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A distributed storage and computation method for a spatiotemporal grid model of a water environment, characterized in that, The distributed storage and computation method for the spatiotemporal grid model of the water environment includes: The study area is subjected to multi-level grid subdivision and encoding processing to obtain a 4D grid logical model of the study area; the 4D grid logical model is the gridded study area after grid subdivision and encoding processing. Acquire multi-source spatiotemporal data of the water environment in the study area; Multi-source spatiotemporal data of the water environment are input into a 4D grid logical model to obtain a 4D grid data model; Determine the future attribute changes and development patterns of the study area based on a 4D grid data model; The encoding includes grid hierarchy, latitude and longitude range, depth range, length, width, height, and XYZK-T encoding information; where X represents east-west direction encoding, Y represents north-south direction encoding, Z represents subdivision hierarchy, K represents depth direction encoding, and T represents time identifier; The study area is subjected to multi-level mesh subdivision and encoding to obtain a 4D mesh logical model of the study area, specifically including: Determine the grid subdivision level for the study area; The quadtree partitioning method is used to perform multi-level grid partitioning and encoding of the study area based on the grid partitioning level, so as to obtain a 4D grid logical model of the study area. Specifically, determining the grid subdivision level for the study area includes: The number of grids, planar resolution, and vertical subdivision of the study area are calculated and hierarchical levels are defined. The formula for calculating the number of grid cells is: = ； The formula for calculating the planar resolution is: = ； The formula for calculating the hierarchical division during longitudinal partitioning is as follows: = / ； in, Indicates the number of meshes in the planar subdivision; n This indicates the number of times the quadtree partitioning method is used; Indicates the resolution of the grid plane; Indicates the east-west length of the study area; Indicates the width of the study area in the north-south direction; This indicates the hierarchical division during vertical partitioning; Represents vertical depth information; The mesh subdivision level of the study area is determined based on the number of grids, planar resolution, hierarchical division during vertical subdivision, and preset conditions.

2. The distributed storage and computation method for a spatiotemporal grid model of the water environment according to claim 1, characterized in that, Multi-source spatiotemporal data of the water environment are input into a 4D grid logical model to obtain a 4D grid data model, which specifically includes: The multi-source spatiotemporal data of the water environment are uniformly transformed to obtain the transformed multi-source spatiotemporal data of the water environment. Based on the grid subdivision level corresponding to the 4D grid logic model, the transformed multi-source spatiotemporal data of the water environment is sliced ​​to obtain sliced ​​multi-source spatiotemporal data of the water environment. The multi-source spatiotemporal data of the water environment after slice processing are input into the 4D grid logic model to obtain a spatial data cube model and a time slice dataset. The data at each moment in the cyclic time-slice dataset is mapped to a spatial data cube model to form a spatiotemporal dataset; Based on the spatiotemporal dataset, a dual index of time and space is constructed to obtain a spatial data cube model that changes over time as a 4D grid data model.

3. The distributed storage and computation method for a spatiotemporal grid model of the water environment according to claim 2, characterized in that, The slicing process includes time slicing and spatial slicing.

4. The distributed storage and computation method for a spatiotemporal grid model of the water environment according to claim 1, characterized in that, Based on a 4D grid data model, the future attribute changes and evolution patterns of the study area are determined, specifically including: In the time dimension, the 4D mesh data model is used to fit the change curve of the entity itself using the attribute values ​​of the time series; Predicting future attribute changes in the study area based on the entity's own change curve; Spatial trend surfaces are constructed based on 4D mesh data models by using proximity relationship description information and attribute data. In the spatiotemporal dual dimensions, visual analysis technology and human-interactive heuristic analysis technology are used to determine the spatiotemporal evolution laws of data, phenomena and environment in the study area based on spatial trend surfaces, thereby obtaining the development and evolution laws of spatiotemporal data in the study area.

5. The distributed storage and computation method for a spatiotemporal grid model of the water environment according to claim 1, characterized in that, Multi-source spatiotemporal data of the water environment are input into a 4D mesh logical model to obtain a 4D mesh data model, which then includes: A database sharding and table partitioning approach is used to store spatiotemporal tile data and construct indexes for the 4D grid data model.

6. A distributed storage and computing system for a spatiotemporal grid model of a water environment, characterized in that, The distributed storage and computing system for the spatiotemporal grid model of the water environment is applied to the distributed storage and computing method for the spatiotemporal grid model of the water environment as described in any one of claims 1-5, and the distributed storage and computing system for the spatiotemporal grid model of the water environment comprises: The 4D mesh logical model determination module is used to perform multi-level mesh subdivision and encoding processing on the study area to obtain a 4D mesh logical model of the study area; the 4D mesh logical model is the meshed study area after mesh subdivision and encoding processing. The acquisition module is used to acquire multi-source spatiotemporal data of the water environment in the study area; The 4D grid data model determination module is used to input multi-source spatiotemporal data of the water environment into the 4D grid logical model to obtain the 4D grid data model. The prediction module is used to determine the future attribute changes and development patterns of the study area based on a 4D grid data model.

7. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the distributed storage computing method for the spatiotemporal grid model of the water environment as described in any one of claims 1-5.

Images