A local microclimate particle simulation system based on a Beidou grid

Abstract

The application relates to the technical field of meteorological particle simulation, in particular to a local micro-meteorological particle simulation system based on a Beidou grid, which comprises a Beidou grid positioning module, a multi-source data acquisition module, a data preprocessing module and a micro-meteorological particle simulation module. The Beidou grid positioning module is used for regularly dividing the ground space of a preset local area into grids according to a Beidou satellite navigation system. The multi-source data acquisition module is used for collecting meteorological data of a target area according to a preset meteorological data acquisition device to obtain multi-source meteorological data. The data preprocessing module is used for preprocessing the multi-source meteorological data in a data processing center. The meteorological particle model construction module is used for constructing a meteorological particle model. The micro-meteorological particle simulation module is used for dynamically simulating and analyzing meteorological elements in each three-dimensional grid unit in the meteorological particle model. The application can accurately adapt to the demand of local micro-meteorological simulation for spatial scale fineness, accurately convert meteorological data into particle properties, and improve the accuracy of local micro-meteorological dynamic evolution simulation.

Description

Technical Field

[0001] This invention relates to the field of meteorological particle simulation technology, and in particular to a local micro-meteorological particle simulation system based on the BeiDou grid. Background Technology

[0002] In the field of meteorological research and application, local micro-meteorological simulation has significant application value in urban planning, agricultural production, and disaster prevention and mitigation. The core requirement of local micro-meteorological simulation is to accurately replicate the dynamic evolution process of meteorological elements within a small-scale area. However, current mainstream micro-meteorological simulation technologies are mostly based on traditional latitude and longitude grids or conventional meshing, which are difficult to adapt to the spatial scale precision requirements of local micro-meteorological simulation. Furthermore, there are significant shortcomings in the conversion between meteorological data and particle attributes, directly leading to insufficient accuracy in dynamic evolution simulation. Traditional grid division methods have inherent limitations; latitude and longitude grids are prone to polar distortion. Local micro-meteorological characteristics are often reflected in spatial differences at the meter or even centimeter level. Traditional grids, due to insufficient meshing precision, cannot accurately capture the subtle influences of local environments such as terrain undulations and building distribution on the meteorological field, resulting in spatial scale adaptation bias. The accurate conversion of meteorological data into particle attributes lacks effective technical support. Existing meteorological observation data is mostly macroscopic field data, while particle simulation requires discretizing meteorological parameters such as temperature, humidity, and wind speed into the specific attributes of individual particles. Due to issues with initial data and mesh generation, the constructed meteorological particle model cannot accurately reflect the characteristics of local micro-meteorology. Therefore, it struggles to accurately simulate real-time changes in local micro-meteorology during dynamic simulations. Mesh scale deviations cause the initial meteorological field settings to deviate from the real environment, and inaccurate particle attribute assignments lead to errors in the initial simulation state. These errors accumulate during subsequent evolution, making it impossible to realistically reproduce the dynamic characteristics of local micro-meteorology, such as airflow circulation and temperature field changes, thus failing to meet the simulation accuracy requirements of practical applications.

[0003] Chinese Patent Publication No. CN119862813B discloses a dynamic three-dimensional grid meteorological simulation model based on historical observation data, including: 1) a historical observation data acquisition and integration module: (a) multi-source data acquisition; (b) data cleaning and integration; (c) data storage and management; 2) a three-dimensional meteorological grid construction module: (a) horizontal grid division; (b) vertical hierarchical architecture; (c) grid parameter initialization; (d) smooth transition between grids; 3) a meteorological simulation algorithm module: (a) physics-driven model; (b) data-driven model; (c) dynamic simulation and time-progression algorithm; 4) a simulation result visualization and verification module: (a) simulation result visualization; (b) model verification and optimization. This scheme cannot accurately adapt to the spatial scale precision requirements of local micro-meteorological simulation, making it difficult to achieve accurate conversion of meteorological data into particle attributes, thus reducing the accuracy of local micro-meteorological dynamic evolution simulation. Summary of the Invention

[0004] To address this, the present invention provides a local micro-meteorological particle simulation system based on the BeiDou grid, which overcomes the problems in existing technologies that cannot accurately adapt to the spatial scale precision requirements of local micro-meteorological simulation, make it difficult to achieve accurate conversion of meteorological data into particle attributes, and reduce the accuracy of local micro-meteorological dynamic evolution simulation.

[0005] To achieve the above objectives, this invention provides a local micro-meteorological particle simulation system based on the BeiDou grid, comprising:

[0006] The BeiDou grid positioning module is used to perform regular grid division of the ground space of a preset local area according to the BeiDou satellite navigation system to obtain a ground grid coordinate system, and to perform layer processing on the vertical space covered by the ground grid coordinate system according to a preset height interval to obtain a three-dimensional grid coordinate system.

[0007] The multi-source data acquisition module is used to determine the target area covered by the three-dimensional grid coordinate system according to the three-dimensional grid coordinate system, and to collect meteorological data of the target area according to the preset meteorological data acquisition equipment to obtain multi-source meteorological data, and transmit the multi-source meteorological data to the data processing center through the Beidou communication link;

[0008] The data preprocessing module is used to preprocess the multi-source meteorological data in the data processing center to obtain actual multi-source meteorological data;

[0009] The meteorological particle model building module is used to perform model conversion on actual multi-source meteorological data to obtain meteorological particle models.

[0010] The micro-meteorological particle simulation module is used to perform dynamic simulation analysis of meteorological elements in each three-dimensional grid unit of the meteorological particle model, and obtain local micro-meteorological dynamic simulation results.

[0011] The meteorological particle model construction module includes:

[0012] The data transformation unit is used to perform coordinate system alignment processing on the original coordinate system of actual multi-source meteorological data to obtain a meteorological data set.

[0013] The grid mapping unit is used to map discrete meteorological data in the meteorological dataset to each grid cell in the three-dimensional grid coordinate system, thereby obtaining the meteorological data distribution matrix in each grid cell.

[0014] The model generation unit is used to convert the meteorological data distribution matrix in each grid cell into the corresponding unit particle attribute parameters through attribute mapping rules, and associate the unit particle attribute parameters with the spatial location of the grid cell to obtain the meteorological particle model.

[0015] The micro-meteorological particle simulation module includes:

[0016] The meteorological element initialization unit is used to parameterize and assign values ​​to the meteorological elements of each three-dimensional grid unit in the meteorological particle model to obtain the initial meteorological element dataset.

[0017] The dynamic simulation analysis unit is used to perform stress tensor decomposition between grid cells on the initial meteorological element dataset according to the finite element analysis algorithm, obtain the decomposed stress tensor, and perform spatiotemporal gradient calculation on the decomposed stress tensor based on the fluid dynamics simulation algorithm to obtain the dynamic meteorological element evolution sequence of each three-dimensional grid cell.

[0018] The simulation result generation unit is used to perform data fusion on the evolution sequence of the dynamic meteorological elements through spatial interpolation algorithm and boundary conditions to obtain local micro-meteorological dynamic simulation results.

[0019] Compared with the prior art, the beneficial effects of this application are as follows:

[0020] The meteorological particle model construction module's data conversion unit achieves precise alignment between the original coordinate system of actual multi-source meteorological data and the 3D grid coordinate system, ensuring the spatial matching accuracy of the meteorological data. The grid mapping unit maps discrete meteorological data to each grid cell to form a meteorological data distribution matrix, achieving an ordered arrangement of meteorological data within the local grid. The model generation unit converts the meteorological data distribution matrix into unit particle attribute parameters and associates them with spatial locations, constructing an accurate meteorological particle model and providing a data foundation for fine-grained simulation of local micro-meteorology. The meteorological element initialization unit of the micro-meteorological particle simulation module parameterizes the meteorological elements of each grid cell in the meteorological particle model, obtaining an accurate initial meteorological element dataset. The dynamic simulation analysis unit uses finite element analysis algorithms to achieve stress tensor decomposition between grid cells and combines fluid dynamics simulation algorithms to calculate spatiotemporal gradients, accurately capturing the dynamic evolution of local micro-meteorological elements. The simulation result generation unit uses spatial interpolation algorithms and boundary conditions for data fusion, ensuring smooth transitions between adjacent grid cells and the integrity of simulation results, thereby improving the accuracy and relevance of local micro-meteorological dynamic simulation and meeting the fine-grained requirements of local regional micro-meteorological simulation.

[0021] Furthermore, the BeiDou grid positioning module includes:

[0022] The ground grid division unit is used to calibrate the coordinates of the ground space boundary of a preset local area based on the positioning reference provided by the BeiDou satellite navigation system, and to divide the coordinate-calibrated ground space into grids according to the regular grid division standard to obtain the ground grid coordinate system;

[0023] The three-dimensional mesh construction unit is used to divide the vertical space covered by the ground grid coordinate system into layers according to the preset height interval parameters, using the ground grid coordinate system as the horizontal reference, to obtain the height boundary of each layer, and to spatially associate the ground grid coordinate system with the height boundary of each layer to obtain the three-dimensional mesh coordinate system.

[0024] In this scheme, by using ground grid division units to perform coordinate calibration and regular grid segmentation of a preset local area of ​​ground space based on the positioning benchmark of the BeiDou satellite navigation system, a precise ground grid coordinate system is obtained, which improves the accuracy and standardization of ground positioning. Then, by using three-dimensional grid construction units as a benchmark, vertical space is divided and associated according to preset height intervals to obtain a three-dimensional grid coordinate system, which realizes the precise division and positioning of three-dimensional space, providing more accurate and comprehensive positioning support for various applications that require three-dimensional spatial positioning, and improving the positioning effect.

[0025] Furthermore, in the model generation unit, the conversion of the meteorological data distribution matrix within each grid cell into corresponding cell particle attribute parameters through attribute mapping rules includes:

[0026] Based on the particle attribute type corresponding to the meteorological particle model, the meteorological data dimensions of the meteorological data distribution matrix in each grid cell are matched and filtered to obtain the target meteorological data dimension set corresponding to the particle attribute type.

[0027] Based on the attribute mapping rules, the meteorological data values ​​under each target meteorological data dimension in the target meteorological data dimension set are numerically transformed to obtain the target meteorological particle attribute parameters corresponding to each target meteorological data dimension.

[0028] Based on the target meteorological particle attribute parameters corresponding to each target meteorological data dimension, the target meteorological particle attribute parameters within the same grid cell are correlated and integrated to obtain the unit particle attribute parameters corresponding to that grid cell.

[0029] In this scheme, the target meteorological data dimension set is screened by matching particle attribute types, which ensures the relevance and accuracy of data selection; the values ​​under the target meteorological data dimension are transformed to obtain basic parameters, realizing the adaptation of data format; the basic parameters within the same grid cell are integrated to obtain complete cell particle attribute parameters, which improves the efficiency and accuracy of data processing and enables the generated particle attribute parameters to more accurately reflect meteorological characteristics.

[0030] Furthermore, the mathematical expression for the unit particle attribute parameters is: In the formula, This indicates the spatial index of a grid cell in the horizontal longitude direction. This indicates the spatial subscript of a grid cell in the horizontal latitudinal direction. This represents the spatial index of a grid cell in the vertical height direction. Indicates time lower coordinate The unit particle attribute parameters corresponding to the mesh unit. This represents the preset spatial gradient weight coefficients. Represents the spatial gradient operator. This represents the total number of preset meteorological data dimensions corresponding to extreme weather. Indicates the first Preset weighting coefficients for each meteorological data dimension Indicates time lower coordinate In the grid cell of the first Preprocessed actual data for each meteorological data dimension This represents the preset time-series change weighting coefficient. Indicates time lower coordinate The unit particle attribute parameters corresponding to the mesh unit. Indicates time lower coordinate The first-order partial derivative of the cell particle property parameters with respect to time for the corresponding grid cell. This represents the preset grid space attribute weight coefficients. Representing coordinates The spatial attribute parameters corresponding to the grid cells.

[0031] In this scheme, by integrating the spatial gradient characteristics of meteorological data, the temporal variation patterns of unit particle attribute parameters, and the spatial attribute characteristics of grid units, appropriate weights are assigned to different influencing factors, thereby improving the accuracy of dynamic representation and prediction of extreme weather data.

[0032] Furthermore, in the dynamic simulation analysis unit, the initial meteorological element dataset is decomposed into inter-grid stress tensors according to the finite element analysis algorithm, and the decomposed stress tensors include:

[0033] Based on the initial meteorological element dataset, the physical attribute parameters of each three-dimensional grid unit are extracted, and the physical attribute matrix corresponding to each three-dimensional grid unit is constructed based on the physical attribute parameters.

[0034] Based on the physical property matrix corresponding to each 3D mesh element and the element partitioning criteria in the finite element analysis algorithm, the contact interface between adjacent mesh elements is discretized to obtain the set of stress application point coordinates at the element interface.

[0035] Based on the coordinate set of stress application points and the physical property matrix of each mesh element, tensor decomposition is performed on the interaction forces between elements to obtain the decomposed stress tensor.

[0036] In this scheme, by extracting the physical attribute parameters of each three-dimensional grid unit from the initial meteorological element dataset and constructing a physical attribute matrix, the characteristics of each grid unit can be accurately characterized, providing a solid foundation for subsequent analysis. Based on the physical attribute matrix and the unit partitioning criteria of the finite element analysis algorithm, the contact interface between adjacent grid units is discretized to obtain a set of stress application point coordinates, which can clearly define the interaction positions. Then, based on the set of stress application point coordinates and the physical attribute matrix, tensor decomposition is performed to obtain the decomposed stress tensor, which can accurately quantify the interaction forces between units and effectively improve the accuracy and reliability of mechanical factor analysis in meteorological simulation.

[0037] Furthermore, the mathematical expression for the decomposed stress tensor is: In the formula, Indicates the first Line number Liede The stress tensor after decomposition of layer mesh elements, Indicates the first The correlation coefficient of physical properties of row grid cells. Indicates the first Line number Eigenvalues ​​of the physical property matrix of column grid cells, Indicates the first Discretization coefficients of the contact interface of column mesh elements. Indicates the first Liede The associated coordinates of the stress application points of the layer mesh element.

[0038] In this scheme, by establishing a decomposition relationship between stress tensor and the physical properties of mesh elements, contact interface characteristics, and stress application points, the stress state of mesh elements at different locations can be accurately characterized.

[0039] Furthermore, in the dynamic simulation analysis unit, the spatiotemporal gradient of the decomposed stress tensor is calculated based on a fluid dynamics simulation algorithm to obtain the dynamic meteorological element evolution sequence of each three-dimensional grid unit, including:

[0040] The spatiotemporal distribution range of the decomposed stress tensor in the three-dimensional grid coordinate system is determined to obtain the spatiotemporal distribution domain of the tensor. The spatiotemporal distribution domain of the tensor is divided into time dimensions according to the preset time step to obtain multiple time segment intervals. The stress tensor in each time segment interval is divided into spatial grids according to the spatial resolution to obtain a discretized stress tensor sequence.

[0041] Tensor data at each time step in the discretized stress tensor sequence are extracted to obtain a single-time tensor dataset. Based on the Navier-Stokes equations in the fluid dynamics simulation algorithm, spatial partial derivatives of the stress tensor components of each grid cell in the single-time tensor dataset are calculated to obtain the cell gradient components of each grid cell. The cell gradient components of each grid cell are then spatially stitched together according to the three-dimensional grid coordinate system to obtain the instantaneous gradient field.

[0042] Gradient data of grid cells in the instantaneous gradient field corresponding to each time step are extracted to obtain a time step gradient dataset. Adjacent time step data in the time step gradient dataset are temporally correlated according to the time axis order to obtain a temporally correlated gradient sequence. The temporally correlated gradient sequence is then mapped to the meteorological element parameter dimension of each three-dimensional grid cell to obtain the dynamic meteorological element evolution sequence of each three-dimensional grid cell.

[0043] In this scheme, by determining the spatiotemporal distribution range of the stress tensor in the three-dimensional grid coordinate system and dividing it into time segments and spatial grids, a discretized stress tensor sequence is obtained, realizing refined discretization of the stress tensor. The spatial partial derivatives of the tensor data at a single moment are calculated and spliced ​​to obtain the instantaneous gradient field, which can accurately capture the stress tensor change characteristics at each moment. The time step gradient dataset is extracted and subjected to time series correlation processing, and mapped to the meteorological element parameter dimension to obtain a dynamic meteorological element evolution sequence, which can accurately reflect the dynamic changes of meteorological elements over time.

[0044] Furthermore, in the simulation result generation unit, the evolution sequence of the dynamic meteorological elements is fused using a spatial interpolation algorithm and boundary conditions to obtain local micro-meteorological dynamic simulation results, including:

[0045] Based on the dynamic meteorological element evolution sequence, meteorological element data and corresponding grid coordinates of the boundary nodes of adjacent three-dimensional grid units are extracted to obtain the boundary data and coordinate set of adjacent units. The transition area between adjacent three-dimensional grid units is determined, and the meteorological element values ​​within the transition area are calculated point by point based on the spatial interpolation algorithm to obtain the first interpolated transition data between adjacent grid units.

[0046] The boundary continuity parameters and their value ranges in the preset boundary condition constraint rules are extracted to obtain the boundary constraint parameter set. The values ​​of each node in the first interpolated transition data are compared with the corresponding parameter value ranges in the boundary constraint parameter set. Abnormal transition data that exceeds the parameter value range is marked. The correction coefficients in the boundary constraint parameter set are used to adjust the values ​​of the abnormal transition data to obtain the second interpolated transition data that meets the boundary conditions.

[0047] The original meteorological element data and corresponding grid index of each three-dimensional grid unit in the dynamic meteorological element evolution sequence are extracted to obtain the original dataset;

[0048] Based on the grid index, the second interpolated transition data is spatially aligned with the corresponding grid positions in the original dataset to obtain the overlapping area. Then, according to the data fusion and integration rules, the original meteorological element data and the second interpolated transition data in the overlapping area are weighted and numerically merged to obtain the local micro-meteorological dynamic simulation results.

[0049] In this scheme, by extracting the boundary data and coordinate set of adjacent three-dimensional grid units based on the evolution sequence of dynamic meteorological elements, and determining the range of the transition area for spatial interpolation calculation, the first interpolated transition data between adjacent grid units can be accurately obtained. The boundary constraint parameter set is extracted, and abnormal transition data is marked and adjusted to obtain the second interpolated transition data that meets the boundary conditions, ensuring the rationality of the data. Then, the original dataset is extracted, and the second interpolated transition data is spatially aligned with the corresponding grid positions in the original dataset. The original meteorological element data and the second interpolated transition data in the overlapping area are weighted and numerically merged, which effectively improves the accuracy and reliability of the local micro-meteorological dynamic simulation results, making the simulation closer to the actual situation. Attached Figure Description

[0050] Figure 1 This is a schematic diagram of the structure of a local micro-meteorological particle simulation system based on the BeiDou grid, according to an embodiment of the present invention.

[0051] Figure 2 This is a schematic diagram of the structure of a BeiDou grid positioning module in a local micro-meteorological particle simulation system based on a BeiDou grid, according to an embodiment of the present invention.

[0052] Figure 3 This is a schematic diagram of the structure of a meteorological particle model construction module in a local micro-meteorological particle simulation system based on the BeiDou grid, according to an embodiment of the present invention.

[0053] Figure 4 This is a schematic diagram of the micro-meteorological particle simulation module in a local micro-meteorological particle simulation system based on the BeiDou grid, according to an embodiment of the present invention. Detailed Implementation

[0054] The following detailed description illustrates the specific implementation method:

[0055] like Figure 1 As shown, it is a schematic diagram of a local micro-meteorological particle simulation system based on the BeiDou grid according to an embodiment of the present invention, including:

[0056] The BeiDou grid positioning module is used to perform regular grid division of the ground space of a preset local area according to the BeiDou satellite navigation system to obtain a ground grid coordinate system, and to perform layer processing on the vertical space covered by the ground grid coordinate system according to a preset height interval to obtain a three-dimensional grid coordinate system.

[0057] The multi-source data acquisition module is used to determine the target area covered by the three-dimensional grid coordinate system according to the three-dimensional grid coordinate system, and to collect meteorological data of the target area according to the preset meteorological data acquisition equipment to obtain multi-source meteorological data, and transmit the multi-source meteorological data to the data processing center through the Beidou communication link;

[0058] The data preprocessing module is used to preprocess the multi-source meteorological data in the data processing center to obtain actual multi-source meteorological data;

[0059] The meteorological particle model building module is used to perform model conversion on actual multi-source meteorological data to obtain meteorological particle models.

[0060] The micro-meteorological particle simulation module is used to perform dynamic simulation analysis of meteorological elements in each three-dimensional grid unit of the meteorological particle model, and obtain local micro-meteorological dynamic simulation results.

[0061] like Figure 2 As shown, it is a schematic diagram of the structure of the BeiDou grid positioning module in a local micro-meteorological particle simulation system based on the BeiDou grid according to an embodiment of the present invention, including:

[0062] The ground grid division unit is used to calibrate the coordinates of the ground space boundary of a preset local area based on the positioning reference provided by the BeiDou satellite navigation system, and to divide the coordinate-calibrated ground space into grids according to the regular grid division standard to obtain the ground grid coordinate system;

[0063] The three-dimensional mesh construction unit is used to divide the vertical space covered by the ground grid coordinate system into layers according to the preset height interval parameters, using the ground grid coordinate system as the horizontal reference, to obtain the height boundary of each layer, and to spatially associate the ground grid coordinate system with the height boundary of each layer to obtain the three-dimensional mesh coordinate system.

[0064] In this embodiment, the positioning reference refers to the standard used to determine the location reference. For example, when planning agricultural operations in a large farmland, the positioning reference provided by the Beidou satellite navigation system can accurately determine the coordinate positions of the four corners of the farmland. The preset local area refers to a specific ground space range that is predetermined and requires ground grid division. For example, in urban planning, in order to carry out detailed planning and design of a newly built commercial area, the area of ​​approximately 2 square kilometers where the commercial area is located is set as the preset local area. The regularized grid division standard refers to the rules and criteria followed when dividing the coordinate-calibrated ground space into grids. For example, when conducting ecological monitoring of a forest, the regularized grid division standard is used to divide the forest area into square grids with a side length of 100 meters. The preset height interval parameter is the height interval value set in advance when dividing the vertical space covered by the ground grid coordinate system into levels with the ground grid coordinate system as the horizontal reference. Specifically, the preset height interval parameter is set to 50 meters. The pre-set meteorological data acquisition equipment includes temperature and humidity sensors, air pressure sensors, wind speed and direction sensors and other sensing devices carried by the aircraft itself, as well as a distributed radar monitoring network. The sensing devices collect meteorological data around the flight path in real time; the distributed radar monitoring network (such as miniature weather radar and phased array radar) collects meteorological information such as precipitation intensity, airflow speed and cloud height in local areas.

[0065] Based on the positioning benchmark provided by the BeiDou Navigation Satellite System, coordinate calibration of the ground spatial boundary of a pre-defined local area includes: ensuring that the BeiDou Navigation Satellite System receiving equipment is in normal working order and can accurately receive satellite signals; selecting several representative feature points within the pre-defined local area, such as corner points and boundary turning points; using the positioning function of the BeiDou Navigation Satellite System to obtain the precise latitude, longitude, and altitude coordinates of these feature points in a geodetic coordinate system (such as the WGS-84 coordinate system); recording and analyzing the coordinate data of these feature points using professional geographic information processing software; and determining the coordinate range of the entire pre-defined local area's ground spatial boundary based on the coordinates of these feature points, combined with the shape and boundary features of the area. The boundary coordinate data is stored in a corresponding geographic information database to provide an accurate coordinate benchmark for subsequent grid division and other work, ensuring the accurate and reliable coordinate calibration of the ground spatial boundary.

[0066] The process of gridding a coordinate-calibrated ground space according to a standardized grid division standard includes: After completing the coordinate calibration of the ground space boundary, proceeding according to the standardized grid division standard. This involves determining the grid size specifications, such as setting the length of each grid in the longitude and latitude directions. Starting from a corner point of the ground space boundary according to the preset grid size, grids are divided sequentially along the longitude and latitude directions. Using the spatial analysis functions of a geographic information system (GIS), the entire ground space is uniformly divided according to the standardized grid size. During the division process, it is ensured that the boundary of each grid precisely matches the preset coordinate range, avoiding overlaps or omissions. For irregular ground space boundaries, the grids at the edges can be appropriately adjusted to conform to the overall division rules. After division, a unique identifier is assigned to each grid, and a ground grid coordinate system is established. This coordinate system uses the coordinates of the grid's center point or vertex to represent the grid's position information.

[0067] The hierarchical division of the vertical space covered by the ground grid coordinate system based on preset height interval parameters includes: First, defining the preset height interval parameters, which can be set according to actual application needs, such as each height layer being 50 meters. Starting from the ground height represented by the ground grid coordinate system, height layers are sequentially divided upwards according to the preset height interval parameters. Using 3D geographic information modeling technology, clear height boundary values ​​are determined for each height layer in the vertical direction. Specialized software is used to associate the ground grid coordinate system with each height layer, ensuring that each ground grid has a corresponding spatial range at different height layers. During the division process, it is crucial to ensure that the boundaries between each height layer are clear and the height values ​​are accurate, so that the spatial characteristics and attributes of different height layers can be accurately described subsequently.

[0068] Spatially associating the ground grid coordinate system with the height boundaries of each layer involves: after defining the ground grid coordinate system and the height boundaries of each layer in vertical space, performing a spatial association operation. Using a 3D spatial coordinate transformation algorithm, the coordinates of each grid point in the ground grid coordinate system are combined with the height boundary values ​​of each layer. For each ground grid, the height layers it spans in the vertical direction are determined based on its location. By establishing a data association table or spatial database, the identification information and planar coordinate information of the ground grid are bound to the corresponding height layer boundary information. In a 3D visualization environment, the spatial extension of each ground grid at different height layers can be clearly seen, achieving an intuitive spatial association between the ground grid coordinate system and the height boundaries of each layer.

[0069] In this embodiment, the data preprocessing module's preprocessing flow in the data processing center is as follows: First, the received multi-source meteorological data undergoes format standardization and outlier removal, filtering out invalid and redundant data. Then, data model preprocessing is performed, combining a 3D grid coordinate system and adopting a layered meteorological representation using a 3D height map. The data is spatially regularized using a cubic grid with a maximum precision of 1.5cm side length, accurately mapping the multi-source data to the corresponding 3D grid cells. Simultaneously, the data undergoes spatiotemporal alignment, unifying the sampling frequency and coordinate reference, ultimately generating actual multi-source meteorological data suitable for subsequent model conversion. This data can accurately calculate extreme weather parameters such as wind speed, wind direction, air turbulence, and rainfall speed and amount within each 1.5cm side length cubic grid, providing high-quality data support for analyzing the impact of extreme weather parameters on micro-aircraft.

[0070] like Figure 3 As shown, it is a schematic diagram of the structure of a meteorological particle model construction module in a local micro-meteorological particle simulation system based on the Beidou grid according to an embodiment of the present invention, including:

[0071] The data transformation unit is used to perform coordinate system alignment processing on the original coordinate system of the actual multi-source meteorological data according to the spatial dimension attribute of the three-dimensional grid coordinate system, so as to obtain a meteorological data set.

[0072] The grid mapping unit is used to map discrete meteorological data in the meteorological dataset to each grid cell of the three-dimensional grid coordinate system according to the grid topology of the three-dimensional grid coordinate system, so as to obtain the meteorological data distribution matrix in each grid cell;

[0073] The model generation unit is used to convert the meteorological data distribution matrix within each grid cell into corresponding cell particle attribute parameters through attribute mapping rules, and to associate the cell particle attribute parameters with the spatial location of the grid cell to obtain a meteorological particle model containing spatial location and cell particle attribute parameters. The mathematical expression for the cell particle attribute parameters is as follows: In the formula, This indicates the spatial index of a grid cell in the horizontal longitude direction. This indicates the spatial subscript of a grid cell in the horizontal latitudinal direction. This represents the spatial index of a grid cell in the vertical height direction. Indicates time lower coordinate The unit particle attribute parameters corresponding to the mesh unit. This represents the preset spatial gradient weight coefficients. Represents the spatial gradient operator. This represents the total number of preset meteorological data dimensions corresponding to extreme weather. Indicates the first Preset weighting coefficients for each meteorological data dimension Indicates time lower coordinate In the grid cell of the first Preprocessed actual data for each meteorological data dimension This represents the preset time-series change weighting coefficient. Indicates time lower coordinate The unit particle attribute parameters corresponding to the mesh unit. Indicates time lower coordinate The first-order partial derivative of the cell particle property parameters with respect to time for the corresponding grid cell. This represents the preset grid space attribute weight coefficients. Representing coordinates The spatial attribute parameters corresponding to the grid cells.

[0074] In this embodiment, the attribute mapping rule is the rule used to convert the meteorological data distribution matrix within each grid cell into corresponding particle attribute parameters. The attribute mapping rule is determined based on the physical characteristics and interrelationships of various meteorological elements (such as temperature, humidity, and wind speed). For example, temperature affects the color attribute of particles (different colors represent different temperature ranges in visualization), and humidity affects the size attribute of particles (particles are larger when humidity is high). By studying the physical meaning of meteorological elements and their roles in the meteorological system, a correspondence between these elements and particle attribute parameters is established.

[0075] Preset spatial gradient weight coefficients The determination is based on the spatial resolution of the three-dimensional grid, the spatial correlation of meteorological elements, and the spatial variation characteristics of extreme weather scenarios. The value needs to be determined through multiple sets of comparative experiments using the controlled variable method to test different values. The influence of values ​​on the spatial fit of meteorological particle models is investigated, and the range of values ​​with the smallest fitting error is typically selected. For grids ∈ [0.2, 0.6], especially for high spatial resolution grids (e.g., 1km × 1km × 100m), the spatial gradient of meteorological data varies significantly. Use a value of 0.4-0.6; for low-resolution grids (e.g., 10km×10km×500m), the spatial gradient has a weaker impact. A value of 0.2-0.4 is used to ensure that the contribution of the spatial gradient term to particle properties matches the spatial evolution trend of the actual meteorological field, avoiding model distortion due to weight imbalance. The preset total number of dimensions of meteorological data corresponding to extreme weather events is also used. The core of determining extreme weather is to screen the key meteorological elements involved in the formation and development of extreme weather, which requires combining literature review, extreme weather case statistics, and meteorological theoretical analysis. The values ​​need to be optimized by removing redundant dimensions and retaining core influencing factors. For example, in the case of heavy rain, the values ​​need to cover dimensions such as precipitation, vertical velocity, water vapor flux, CAPE value, temperature, and relative humidity; in the case of typhoons, the values ​​need to cover dimensions such as air pressure, wind speed, wind direction, sea surface temperature, and water vapor content. The value ranges from 4 to 8, and the specific value is dynamically adjusted according to the type of extreme weather. If a meteorological dimension that has a significant effect on the characterization of extreme weather is added (such as lightning intensity), it can be appropriately expanded. The value ensures that the total number of dimensions fully covers the key formation conditions of extreme weather. Preset weighting coefficients for each meteorological data dimension The importance weights of each meteorological dimension are determined based on their contribution priority to the characterization of extreme weather. The Analytic Hierarchy Process (AHP) or Random Forest algorithm is used to quantify the importance weights of each dimension. The value of must satisfy the weight normalization constraint. For example, in severe convective weather scenarios, the precipitation dimension contributes the most to the characterization of extreme weather. Take 0.3; the vertical velocity dimension is next. The weight is set to 0.2; for dimensions such as water vapor flux, CAPE value, and temperature, the weights decrease in order of importance, set to 0.15, 0.15, and 0.1 respectively; the remaining dimensions are assigned a weight of 0.1. These values ​​need to be experimentally verified and the weight ratios adjusted to ensure that the contribution of the dominant dimensions is fully reflected and to avoid secondary dimensions interfering with the accurate characterization of particle properties. Preset time-series variation weight coefficients. The determination of the autocorrelation coefficients is based on the temporal evolution characteristics of meteorological data and the time scale of extreme weather. It requires calculating the autocorrelation coefficients of meteorological data at different time scales through time series analysis, and then combining this with the time series prediction error calibration values ​​of the particle model. For short-term extreme weather events (such as short-duration heavy rainfall) ∈ [0.1, 0.4], meteorological data changes rapidly over time, and historical data has a significant impact on the current state. Use 0.3-0.4; for daily-level medium- to long-term extreme weather (such as continuous heavy rain), the temporal variation is relatively gradual. Take a value of 0.1-0.2. The values ​​of these parameters need to balance the contribution ratios of the temporal partial derivative and spatial gradient terms, ensuring that the model reflects the historical evolution of meteorological data without neglecting the influence of spatial distribution characteristics. Preset grid spatial attribute weight coefficients. The determination is based on the degree of influence of the spatial attributes of grid cells on meteorological elements. Correlation analysis is needed to quantify the relationship between spatial attributes such as terrain height and surface roughness and meteorological data, and the weights are calculated using a multiple linear regression model. For the complex topographic grid in the range [0.05, 0.2], the topography has a significant impact on airflow lifting and water vapor convergence. Use a value of 0.15-0.2; for grids with low terrain complexity such as plains, the influence of spatial attributes is relatively weak. Take a value of 0.05-0.1. The values ​​should be adjusted based on the actual geographical environment to ensure that the spatial attribute terms reflect the modulation effect of the inherent characteristics of the grid on meteorological particles, improve the physical consistency of the model, and avoid unreasonable results where spatial attributes dominate particle attributes due to excessive weight. Coordinates Spatial property parameters corresponding to the mesh cells The determination is based on the inherent spatial characteristics of the grid cells. The values ​​are extracted from high-precision geographic information data and directly mapped to a 3D grid. Specifically, coordinate units are obtained by using a digital elevation model to acquire topographic height values, using land cover data to obtain empirical coefficients for surface roughness, and using a geographic information system to acquire latitude, longitude, and other parameters of the grid units. These are then integrated to form a coordinate system. A collection of multidimensional attributes, such as the terrain height dimension. The value is taken as the actual elevation of the grid cell (unit: m), and the surface roughness dimension is taken as an empirical coefficient of 0.01-0.5 (lower value for bare land, higher value for forest), to ensure... It can accurately characterize the spatial inherent properties of grid cells and provide spatially meaningful constraints for particle property parameters.

[0076] Aligning the original coordinate system of actual multi-source meteorological data with the spatial dimensional attributes of the 3D grid coordinate system involves the following steps: First, clarifying the spatial dimensional attributes of the 3D grid coordinate system, such as longitude, latitude, and altitude in three-dimensional space. Next, analyzing the characteristics of the original coordinate system of the actual multi-source meteorological data to determine its differences from the 3D grid coordinate system in each dimension. For cases where the coordinate origins differ, translation transformations are used to make the origins coincide; if rotation differences exist, rotation matrices are used to adjust the rotation, ensuring that the axes of each dimension are consistent. For scale differences, the original data is scaled according to the scale standard of the 3D grid coordinate system. Through these transformations, the original coordinate system of the actual multi-source meteorological data is aligned with the 3D grid coordinate system, ensuring the data is within a unified spatial framework, providing an accurate foundation for subsequent processing, and resulting in a meteorological data set that meets the requirements.

[0077] Based on the grid topology of the 3D grid coordinate system, mapping discrete meteorological data in the meteorological dataset to each grid cell of the 3D grid coordinate system includes: determining the grid cell division method, size, and adjacency relationship based on the grid topology of the 3D grid coordinate system; determining the spatial location of each discrete meteorological data point in the meteorological dataset, and using this spatial location as the center, determining the grid cell to which the discrete meteorological data point belongs based on the grid cell boundary conditions; if the discrete meteorological data point is near the grid cell boundary, it is assigned according to the proximity principle or based on the distance weight between the discrete meteorological data point and surrounding grid cells; for each grid cell, counting the discrete meteorological data points falling within it, and calculating the meteorological data values ​​at each location within the grid cell based on the values ​​of surrounding data points, thereby obtaining the meteorological data distribution matrix within each grid cell, completing the mapping from discrete meteorological data to grid cells.

[0078] The conversion of the meteorological data distribution matrix within each grid cell into corresponding cell particle attribute parameters using attribute mapping rules involves: Attribute mapping rules act as a bridge connecting the meteorological data distribution matrix and particle attribute parameters. First, the types of particle attribute parameters to be mapped are determined, such as temperature, humidity, and wind speed. For each element in the meteorological data distribution matrix, conversion is performed according to the meteorological element it represents, based on preset attribute mapping rules. For example, for temperature data, a correspondence is established between temperature range and particle attributes such as color and transparency, mapping temperature values ​​to corresponding color codes and transparency values; for wind speed data, based on the mapping relationship between wind speed and particle size and velocity, appropriate particle size and velocity parameters are converted. Through this rule-based mapping, the numerical information in the meteorological data distribution matrix is ​​accurately converted into intuitive particle attribute parameters, providing crucial data support for constructing meteorological particle models.

[0079] Associating unit particle attribute parameters with the spatial location of grid cells involves accurately linking them. First, the spatial location information of each grid cell in the 3D grid coordinate system is defined, including its center coordinates and boundary coordinates. For the unit particle attribute parameters transformed within each grid cell, these parameters are bound to their corresponding spatial locations based on the grid cell's spatial location identifier. This can be achieved by establishing a data structure that uses the grid cell's spatial coordinates as an index, storing the unit particle attribute parameters at the corresponding index locations. In practical applications, the corresponding unit particle attribute parameters can be quickly retrieved based on the grid cell's spatial location information, ensuring that the meteorological particle model at each location in 3D space accurately reflects the meteorological characteristics of that location, thus achieving a close association between the meteorological particle model and its spatial location.

[0080] Specifically, in the model generation unit, the conversion of the meteorological data distribution matrix within each grid cell into corresponding cell particle attribute parameters through attribute mapping rules includes:

[0081] Based on the particle attribute type corresponding to the meteorological particle model, the meteorological data dimensions of the meteorological data distribution matrix in each grid cell are matched and filtered to obtain the target meteorological data dimension set corresponding to the particle attribute type.

[0082] Based on the attribute mapping rules, the meteorological data values ​​under each target meteorological data dimension in the target meteorological data dimension set are numerically transformed to obtain the target meteorological particle attribute parameters corresponding to each target meteorological data dimension.

[0083] Based on the target meteorological particle attribute parameters corresponding to each target meteorological data dimension, the target meteorological particle attribute parameters within the same grid cell are correlated and integrated to obtain the unit particle attribute parameters corresponding to that grid cell.

[0084] In this embodiment, matching and filtering the meteorological data dimensions of the meteorological data distribution matrix within each grid cell according to the particle attribute type corresponding to the meteorological particle model includes: clearly understanding the particle attribute type corresponding to the meteorological particle model, such as temperature, humidity, wind speed, and air pressure. These attribute types clarify the key information we need to extract from the meteorological data. Next, for the meteorological data distribution matrix within each grid cell, its meteorological data dimensions are analyzed in detail. Each data dimension in the matrix is ​​compared with the particle attribute type one by one to determine whether each dimension is related to the required particle attribute. Data dimensions that match the particle attribute type are filtered out, while irrelevant dimensions are excluded.

[0085] The numerical transformation of meteorological data values ​​under each target meteorological data dimension in the target meteorological data dimension set according to attribute mapping rules includes: after obtaining the target meteorological data dimension set, transforming the meteorological data values ​​under each target meteorological data dimension according to preset attribute mapping rules. For example, if the particle attribute is the particle color corresponding to temperature, the attribute mapping rules may specify different color values ​​corresponding to different temperature ranges. For each meteorological data value, first determine the temperature range to which the meteorological data value belongs, and then convert the meteorological data value into the corresponding color code according to the rules. If the mapping between wind speed and particle motion velocity is involved, convert the wind speed value into a suitable particle motion velocity parameter according to the correspondence between wind speed magnitude and velocity value. Through this rule-based numerical transformation process, the original meteorological data values ​​are transformed into target meteorological particle attribute parameters that can be used to construct a meteorological particle model.

[0086] Based on the target meteorological particle attribute parameters corresponding to each target meteorological data dimension, the correlation integration of all target meteorological particle attribute parameters within the same grid cell involves: after obtaining all target meteorological particle attribute parameters corresponding to each target meteorological data dimension within the same grid cell, it is necessary to correlate and integrate all target meteorological particle attribute parameters. First, analyze the inherent logical relationships between the various target meteorological particle attribute parameters, such as the potential mutual influence between temperature, humidity, and air pressure. Then, establish an integration model or algorithm based on these relationships. For example, parameters such as temperature and humidity can be combined according to a certain weight ratio to calculate a comprehensive particle state parameter. Alternatively, based on meteorological principles, different parameters can be combined and calculated to obtain unit particle attribute parameters that more accurately reflect the meteorological characteristics of the grid cell.

[0087] like Figure 4 As shown, it is a schematic diagram of the structure of a micro-meteorological particle simulation module in a local micro-meteorological particle simulation system based on the Beidou grid according to an embodiment of the present invention, including:

[0088] The meteorological element initialization unit is used to parameterize and assign values ​​to the meteorological elements of each three-dimensional grid unit in the meteorological particle model according to the three-dimensional grid coordinate system, so as to obtain the initial meteorological element dataset.

[0089] The dynamic simulation analysis unit is used to perform stress tensor decomposition between grid cells on the initial meteorological element dataset according to the finite element analysis algorithm, obtain the decomposed stress tensor, and calculate the spatiotemporal gradient of the decomposed stress tensor based on the fluid dynamics simulation algorithm to obtain the dynamic meteorological element evolution sequence of each three-dimensional grid cell. The mathematical expression of the decomposed stress tensor is as follows: In the formula, Indicates the first Line number Liede The stress tensor after decomposition of layer mesh elements, Indicates the first The correlation coefficient of physical properties of row grid cells. Indicates the first Line number Eigenvalues ​​of the physical property matrix of column grid cells, Indicates the first Discretization coefficients of the contact interface of column mesh elements. Indicates the first Liede The associated coordinates of the stress application points of the layer mesh element;

[0090] The simulation result generation unit is used to perform data fusion on the evolution sequence of the dynamic meteorological elements through spatial interpolation algorithm and boundary conditions to obtain local micro-meteorological dynamic simulation results.

[0091] Specifically, in the dynamic simulation analysis unit, the initial meteorological element dataset is decomposed into inter-grid element stress tensors according to the finite element analysis algorithm, and the decomposed stress tensors include:

[0092] Based on the initial meteorological element dataset, the physical attribute parameters of each three-dimensional grid unit are extracted, and the physical attribute matrix corresponding to each three-dimensional grid unit is constructed based on the physical attribute parameters.

[0093] Based on the physical property matrix corresponding to each 3D mesh element and the element partitioning criteria in the finite element analysis algorithm, the contact interface between adjacent mesh elements is discretized to obtain the set of stress application point coordinates at the element interface.

[0094] Based on the coordinate set of stress application points and the physical property matrix of each mesh element, tensor decomposition is performed on the interaction forces between elements to obtain the decomposed stress tensor.

[0095] In this embodiment, the finite element analysis algorithm is a numerical analysis method used to solve complex physical problems, such as structural mechanics, heat conduction, and fluid mechanics. The basic idea of ​​the finite element analysis algorithm is to discretize the continuous solution domain into a set of finite elements interconnected by nodes. Within each three-dimensional mesh element, the distribution of physical quantities (such as displacement, temperature, and pressure) is assumed, typically represented by a polynomial interpolation function. Then, the equilibrium equations of the three-dimensional mesh element are constructed using variational principles or the weighted residual method, and the equilibrium equations of all three-dimensional mesh elements are assembled into a system of algebraic equations. Finally, this system of equations is solved to obtain the unknown physical quantities at the nodes, and then the distribution of physical quantities throughout the entire solution domain is obtained through the interpolation function.

[0096] No. Physical property correlation coefficient of row grid cell The determination of the first Based on the common physical properties of the row of three-dimensional grid cells, the measured or historical statistical values ​​of core physical parameters such as atmospheric density, specific heat capacity, and thermal conductivity of all grid cells in that row are first extracted. A normalization algorithm is then used to eliminate dimensional differences between different physical quantities. Finally, an attribute weight model is established by combining the distribution characteristics of the grid in the three-dimensional coordinate system (e.g., near-ground row, high-altitude row). The value of needs to balance the stability of physical properties and the continuity between rows. The physical properties of the near-ground row grid fluctuate less. The value range is 1.0-1.2; the physical properties of high-altitude flight exhibit strong dispersion. The value range is adjusted to 0.7-0.9, and it needs to be verified against the meteorological observation dataset of the same period to ensure that adjacent rows are consistent. The gradient difference should not exceed 0.1 to avoid abrupt changes in stress tensor decomposition. Line number Eigenvalues ​​of the physical property matrix of column grid cells The determination is based on the calculation of the multi-dimensional physical property matrix of a single mesh cell. First, the first... Line number The physical attribute matrix of the grid is generated, with the matrix dimensions corresponding to the selected meteorological element types (such as temperature, air pressure, and humidity). The matrix elements are the initial values ​​assigned to the grid or the measured sample values. Then, the QR eigenvalue decomposition algorithm is used to operate on the matrix, extracting the principal eigenvalues ​​that characterize the core physical attributes of the grid. , The value of must satisfy the positive definiteness requirement of the matrix. If the physical properties of the mesh are stable, The values ​​are concentrated in the range of 5-10; if the attribute fluctuates drastically (such as in a strong convection region), the eigenvalue dispersion is increased, and the value range can be extended to 2-15, thereby ensuring the convergence of finite element stress decomposition. Discretization coefficient of the contact interface of column grid element The determination of the determination revolves around the first Unfold the contact interface features between the column grid and adjacent columns, first focusing on the first... The transverse contact interfaces of all mesh elements are finely meshed, and key parameters such as interface roughness, porosity, and frequency of attribute abrupt changes are statistically analyzed. Then, considering the discretization accuracy requirements of finite element analysis, a correlation between interface characteristic parameters and... Quantization mapping model, The value should be appropriate for the complexity of the interface. If the interface is smooth and the meteorological elements transition evenly, the value should be suitable. The value should be between 0.1 and 0.3 to reduce computational redundancy; if complex phenomena such as turbulence or phase transition exist at the interface, The value of is increased to 0.6-0.9 to enhance the accuracy of stress decomposition, while the discretization error needs to be controlled below 5% to meet the accuracy threshold of fluid dynamics simulation. Liede Correlation values ​​of stress application point coordinates of layer mesh elements The determination is based on the three-dimensional coordinate correlation analysis of the three-dimensional mesh coordinate system. First, the first... Liede Geometric center coordinates of layer mesh cells Then locate the actual spatial coordinates of the stress application point of the grid. The coordinate correlation value is obtained by calculating the ratio of the Euclidean distance between two points to the side length of the mesh cell. , The value of is directly related to the location of the stress application point. If the application point is located at the geometric center of the mesh, the ratio approaches 0. The value ranges from 0 to 0.2; if the point of application is close to the mesh edge or vertex, the ratio increases. The value range is 0.5-0.8. At the same time, it is necessary to match the spatiotemporal step size of the fluid dynamics simulation, adjust the coordinate calculation accuracy, and avoid distortion of the dynamic meteorological element evolution sequence caused by the deviation of the correlation value.

[0097] Element meshing criteria refer to the rules and standards followed when discretizing a continuum into a finite number of elements in finite element analysis. Common element meshing criteria include: Geometric shape criteria: The shape of the elements should be as regular as possible, avoiding excessively long or twisted elements to improve computational accuracy and stability. Size transition criteria: The dimensions of adjacent elements should gradually transition to avoid abrupt size changes and reduce numerical errors. Mesh density criteria: In critical areas such as stress concentration and drastic changes in boundary conditions, the mesh density should be appropriately increased to improve computational accuracy. Element type criteria: Based on the physical characteristics of the problem and the solution requirements, an appropriate element type should be selected, such as one-dimensional elements, two-dimensional elements, three-dimensional elements, etc.

[0098] Constructing the physical attribute matrix for each 3D grid cell based on the aforementioned physical attribute parameters involves: extracting physical attribute parameters, such as density, elastic modulus, Poisson's ratio, and thermal conductivity, from the initial meteorological element dataset. These physical attribute parameters describe the characteristics of the grid cell in different physical phenomena. Depending on the nature and geometry of the problem, a suitable 3D grid cell type is selected, such as tetrahedral or hexahedral cells. Different types of cells have different numbers of nodes and interpolation function forms. An interpolation function is established for each cell to describe the distribution of physical quantities within the cell. The interpolation function is typically in polynomial form, with coefficients related to the physical attribute parameters of the cell nodes. Based on the interpolation function and physical attribute parameters, a physical attribute matrix is ​​constructed for each 3D grid cell. The elements of the physical attribute matrix represent the relationships between physical attributes between different nodes within the cell. For example, in structural mechanics problems, the physical attribute matrix may include a mass matrix, stiffness matrix, and damping matrix. The mass matrix describes the mass distribution of the cell, the stiffness matrix describes the stiffness characteristics of the cell, and the damping matrix describes the damping characteristics of the cell.

[0099] Based on the physical property matrices corresponding to each 3D mesh element and the element partitioning criteria in the finite element analysis algorithm, the contact interface between adjacent mesh elements is discretized. This includes: identifying the contact interface between adjacent mesh elements according to the element partitioning criteria. The contact interface is where two or more elements interact, and its geometry and physical properties have a significant impact on the calculation results. A suitable discretization method is selected based on the geometry and physical properties of the contact interface. Common discretization methods include node matching, node mismatch, and boundary element method. The node matching method requires that the nodes of adjacent elements on the contact interface completely coincide; the node mismatch method allows that the nodes of adjacent elements on the contact interface do not coincide, ensuring the continuity of the interface by introducing constraints; the boundary element method treats the contact interface as a boundary and describes the interaction of the interface through boundary integral equations. Based on the selected discretization method, a discretized model of the contact interface between adjacent mesh elements is established. In the node matching method, the nodes of adjacent elements on the contact interface are directly connected. In the node mismatch method, connections between adjacent elements are established by introducing virtual nodes or constraint equations. In the boundary element method, the contact interface is divided into a series of boundary elements, and boundary integral equations are established on these boundary elements. In the discretized model, the boundary conditions of the contact interface are handled correctly. Boundary conditions may include displacement boundary conditions, force boundary conditions, and mixed boundary conditions. Appropriate boundary conditions are applied according to the actual situation of the problem to ensure the accuracy of the calculation results.

[0100] Based on the set of stress application point coordinates and the physical property matrix of each mesh element, tensor decomposition calculation of the interaction force between elements includes: obtaining the set of stress application point coordinates at the element interface based on the discretization results of the contact interface between adjacent mesh elements. The stress application point is a specific point on the element interface used to calculate stress, and its coordinates can be determined through the discretization model. Based on the physical property matrix of each mesh element and the set of stress application point coordinates, the interaction force between elements is calculated using the equilibrium equations and constitutive relations in the finite element analysis algorithm. The equilibrium equations describe the equilibrium relationship of internal forces within the element, and the constitutive relations describe the relationship between stress and strain. By solving the equilibrium equations and constitutive relations, the magnitude and direction of the interaction force between elements can be obtained. The interaction force between elements is then represented as a tensor. A tensor is a multidimensional array that can describe the direction and magnitude of physical quantities. In three-dimensional space, the stress tensor is a second-order tensor, which can be decomposed into normal stress and shear stress components. Tensor decomposition methods are used to decompose the stress tensor into normal stress and shear stress components. Common tensor decomposition methods include principal stress decomposition and stress invariant decomposition. The principal stress decomposition method obtains the principal stresses and principal directions by solving for the eigenvalues ​​and eigenvectors of the stress tensor; the stress invariant decomposition method decomposes the stress tensor into isotropic and skew parts by calculating the invariants of the stress tensor.

[0101] Specifically, in the dynamic simulation analysis unit, the spatiotemporal gradient of the decomposed stress tensor is calculated based on a fluid dynamics simulation algorithm to obtain the dynamic meteorological element evolution sequence of each three-dimensional grid unit, including:

[0102] The spatiotemporal distribution range of the decomposed stress tensor in the three-dimensional grid coordinate system is determined to obtain the spatiotemporal distribution domain of the tensor. The spatiotemporal distribution domain of the tensor is divided into time dimensions according to the preset time step to obtain multiple time segment intervals. The stress tensor in each time segment interval is divided into spatial grids according to the spatial resolution to obtain a discretized stress tensor sequence.

[0103] Tensor data at each time step in the discretized stress tensor sequence are extracted to obtain a single-time tensor dataset. Based on the Navier-Stokes equations in the fluid dynamics simulation algorithm, spatial partial derivatives of the stress tensor components of each grid cell in the single-time tensor dataset are calculated to obtain the cell gradient components of each grid cell. The cell gradient components of each grid cell are then spatially stitched together according to the three-dimensional grid coordinate system to obtain the instantaneous gradient field.

[0104] Gradient data of grid cells in the instantaneous gradient field corresponding to each time step are extracted to obtain a time step gradient dataset. Adjacent time step data in the time step gradient dataset are temporally correlated according to the time axis order to obtain a temporally correlated gradient sequence. The temporally correlated gradient sequence is then mapped to the meteorological element parameter dimension of each three-dimensional grid cell to obtain the dynamic meteorological element evolution sequence of each three-dimensional grid cell.

[0105] In this embodiment, the preset time step is a pre-set time interval during time-related simulations or analyses. It determines the accuracy and computational efficiency of discretizing the continuous process in the time dimension. A time step that is too small increases computation but allows for more accurate capture of the process's dynamic changes; a time step that is too large improves computational efficiency but may lead to the loss of important details and distortion of simulation results. Typically, the preset time step needs to be determined comprehensively based on the characteristics of the specific problem, the limitations of computational resources, and the required accuracy of the results. For example, in meteorological simulations, the time step might be set to a few minutes if studying short-term severe convective weather; while for long-term climate simulations, the time step might be set to several hours or even longer.

[0106] Fluid dynamics simulation algorithms are numerical computation methods used to simulate and analyze the motion of fluids (such as gases and liquids). Based on fundamental equations of fluid mechanics, such as the continuity equation, momentum equation, and energy equation, these algorithms discretize the continuous fluid domain into a finite number of elements. Within each element, the equations are solved to obtain the distribution and changes in physical quantities such as velocity, pressure, and temperature of the fluid.

[0107] Spatial meshing of the stress tensor within each time segment interval based on spatial resolution includes: determining the spatial resolution, which refers to the size of the spatial mesh and determines the ability to capture spatial details. A suitable spatial resolution is determined based on the specific problem requirements and computational resource limitations. For example, in weather simulations, if small-scale weather phenomena need to be studied, the spatial resolution is set to a few kilometers; while for large-scale climate simulations, the spatial resolution is set to tens of kilometers or even larger. Within each time segment interval, according to the determined spatial resolution, the space in the three-dimensional grid coordinate system is divided into a series of regular or irregular grid cells. The stress tensor data within each time segment interval are assigned to the corresponding spatial grid cells. Based on the location information of the stress tensor and the spatial mesh division results, the grid cell to which each stress tensor point belongs is determined, and the stress tensor value is assigned to that grid cell.

[0108] Based on the Navier-Stokes equations in fluid dynamics simulation algorithms, calculating the spatial partial derivatives of the stress tensor components of each grid cell in a single-time tensor dataset involves: extracting stress tensor-related terms from the Navier-Stokes equations and understanding their spatial variation. A suitable numerical method is selected to calculate the spatial partial derivatives based on the spatial grid division and the distribution of the stress tensor data. Common numerical methods include central difference, forward difference, and backward difference. Depending on the selected method, the partial derivatives of the stress tensor components of each grid cell in the single-time tensor dataset are calculated along the spatial coordinate axes. For example, for stress tensor components in three-dimensional space, the partial derivatives in the x, y, and z directions are calculated respectively.

[0109] The temporal correlation processing of adjacent time step data in the time step gradient dataset according to the time axis order includes: extracting grid cell gradient data of adjacent time steps (such as step t and step t+1) from the time step gradient dataset to ensure that the spatial coordinates of the corresponding 3D grid cells of the two sets of data are completely matched to avoid spatial misalignment; then, performing consistency verification on the two sets of data to check whether the number of grid cells and the gradient component dimensions in the x / y / z directions are consistent, and removing invalid data caused by spatial grid division errors. A linear interpolation method is used to establish temporal correlation: for each grid cell with the same spatial coordinates, the gradient component values ​​of step t and step t+1 are extracted, and the gradient change rate between the two steps is calculated. This change rate is used to quantify the gradient evolution correlation between adjacent time steps; simultaneously, based on the physical constraints of fluid dynamics simulation (conforming to the physical laws of the Navier-Stokes equations), a gradient change threshold is preset. If the gradient difference of a certain grid cell exceeds the threshold, the gradient change trend of three or more adjacent spatial cells around that grid cell is used to correct it, eliminating abnormal fluctuations and making the time series more continuous and smooth.

[0110] Mapping the temporally correlated gradient sequence to the meteorological element parameter dimensions of each 3D grid cell includes: determining the meteorological element parameter dimensions, which refer to various physical quantities describing meteorological phenomena, such as temperature, humidity, wind speed, and wind direction; determining the meteorological element parameter dimensions to be mapped based on the specific problem requirements; establishing a mapping relationship between the temporally correlated gradient sequence and the meteorological element parameters based on the relationship between them. For example, the stress tensor components in the gradient sequence can be linked to the meteorological element parameters through physical models or empirical formulas. For each 3D grid cell, the temporally correlated gradient sequence is mapped to the corresponding meteorological element parameter dimension according to the established mapping relationship. Through this mapping process, the dynamic meteorological element evolution sequence of each 3D grid cell can be obtained, thereby better understanding and predicting changes in meteorological phenomena.

[0111] Specifically, in the simulation result generation unit, the evolution sequence of the dynamic meteorological elements is fused using a spatial interpolation algorithm and boundary conditions to obtain local micro-meteorological dynamic simulation results, including:

[0112] Based on the dynamic meteorological element evolution sequence, meteorological element data and corresponding grid coordinates of the boundary nodes of adjacent three-dimensional grid units are extracted to obtain the boundary data and coordinate set of adjacent units. The transition area between adjacent three-dimensional grid units is determined, and the meteorological element values ​​within the transition area are calculated point by point based on the spatial interpolation algorithm to obtain the first interpolated transition data between adjacent grid units.

[0113] The boundary continuity parameters and their value ranges in the preset boundary condition constraint rules are extracted to obtain the boundary constraint parameter set. The values ​​of each node in the first interpolated transition data are compared with the corresponding parameter value ranges in the boundary constraint parameter set. Abnormal transition data that exceeds the parameter value range is marked. The correction coefficients in the boundary constraint parameter set are used to adjust the values ​​of the abnormal transition data to obtain the second interpolated transition data that meets the boundary conditions.

[0114] The original meteorological element data and corresponding grid index of each three-dimensional grid unit in the dynamic meteorological element evolution sequence are extracted to obtain the original dataset;

[0115] Based on the grid index, the second interpolated transition data is spatially aligned with the corresponding grid positions in the original dataset to obtain the overlapping area. Then, according to the data fusion and integration rules, the original meteorological element data and the second interpolated transition data in the overlapping area are weighted and numerically merged to obtain the local micro-meteorological dynamic simulation results.

[0116] In this embodiment, the preset boundary condition constraint rules are rules pre-set during meteorological simulation or related calculations to ensure the rationality and physical consistency of the calculation results. It mainly includes boundary continuity parameters and their value ranges. Boundary continuity parameters describe the continuity characteristics that meteorological elements should satisfy at the boundary, such as the gradient or difference relationship between temperature, wind speed, and air pressure on both sides of the boundary. The value range limits the reasonable numerical range of these meteorological elements at the boundary, preventing extreme values ​​that do not conform to actual physical phenomena.

[0117] The point-by-point calculation of meteorological element values ​​within the transition region based on spatial interpolation algorithms includes: determining the range of the transition region between adjacent cells based on the boundary data and coordinate sets of adjacent 3D grid cells, combined with grid division information. The transition region is the area where data transitions and merges between two adjacent grid cells; its size and shape depend on the grid division method and the relative positions of adjacent cells. Based on the characteristics of the transition region and the distribution patterns of meteorological elements, a suitable spatial interpolation algorithm is selected, such as linear interpolation, inverse distance weighted interpolation, or Kriging interpolation. Different interpolation algorithms have different applicable scenarios and accuracy characteristics, requiring selection based on specific circumstances. For each point within the transition region, the meteorological element value at that point is calculated using the selected interpolation algorithm based on its location information and the boundary data and coordinate sets of adjacent grid cells. For example, in linear interpolation, the values ​​of points within the transition region are calculated through a linear relationship based on the meteorological element values ​​and distances of adjacent boundary nodes.

[0118] The numerical adjustment of anomalous transition data using correction coefficients from the boundary constraint parameter set includes: extracting boundary continuity parameters and their value ranges from preset boundary condition constraint rules to obtain a boundary constraint parameter set. This set clarifies the conditions that meteorological elements should meet at the boundary and the parameter value ranges. The values ​​of each node in the first interpolated transition data are compared with the corresponding parameter value ranges in the boundary constraint parameter set. If the value of a node exceeds the parameter value range, that node is marked as anomalous transition data. Correction coefficients from the boundary constraint parameter set are then used to numerically adjust the marked anomalous transition data. These correction coefficients are determined based on preset boundary condition constraint rules and the actual physical process. By multiplying the anomalous data by the correction coefficients, the corrected values ​​fall within the parameter value range, thus obtaining the second interpolated transition data that meets the boundary conditions.

[0119] Spatially aligning the second interpolated transition data with the corresponding grid positions in the original dataset based on the grid index involves: extracting the original dataset: extracting the original meteorological element data and corresponding grid indices for each 3D grid unit from the dynamic meteorological element evolution sequence. The grid index is information used to uniquely identify the position of each grid unit, typically including the grid row number, column number, etc. Based on the grid index, establishing a spatial mapping relationship between the second interpolated transition data and the grid positions in the original dataset. Determining their corresponding spatial positions by comparing their grid indices. Spatially aligning the second interpolated transition data with the corresponding grid positions in the original dataset based on the established spatial mapping relationship. This ensures that the second interpolated transition data can be accurately placed at the grid positions where the original data was located.

[0120] The weighting and numerical merging of the original meteorological element data and the second interpolated transition data in the overlapping area according to the data fusion and integration rules includes: Data fusion and integration rules guide how to merge the original meteorological element data and the second interpolated transition data in the overlapping area. These rules typically include weighting allocation methods and numerical merging methods. Weighting allocation methods determine the weights of the original data and the second interpolated transition data in the merging process based on factors such as data reliability and accuracy. Numerical merging methods can include weighted averaging, taking the maximum value, taking the minimum value, etc. According to the weighting allocation method in the data fusion and integration rules, appropriate weights are assigned to the original meteorological element data and the second interpolated transition data in the overlapping area. For example, if the original data has higher accuracy, a larger weight is assigned to it; if the second interpolated transition data better reflects the current dynamic changes, a larger weight is assigned to it. Based on the assigned weights, the original meteorological element data and the second interpolated transition data in the overlapping area are numerically merged. Following the selected numerical merging method, such as weighted averaging, the values ​​of the original meteorological element data and the second interpolated transition data are calculated to obtain the final local micro-meteorological dynamic simulation results.

[0121] The above are merely embodiments of the present invention. Commonly known structures and characteristics are not described in detail here. Those skilled in the art are aware of all common technical knowledge in the field prior to the application date or priority date, are aware of all existing technologies in that field, and have the ability to apply conventional experimental methods prior to that date. Those skilled in the art can, under the guidance of this application, improve and implement this solution in combination with their own capabilities. Some typical known structures or methods should not be obstacles for those skilled in the art to implement this application. It should be noted that those skilled in the art can make several modifications and improvements without departing from the structure of the present invention. These should also be considered within the scope of protection of the present invention, and will not affect the effectiveness of the implementation of the present invention or the practicality of the patent. The scope of protection claimed in this application should be determined by the content of its claims, and the specific embodiments described in the specification can be used to interpret the content of the claims.

Claims

1. A local micro-meteorological particle simulation system based on the BeiDou grid, characterized in that: include: The BeiDou grid positioning module is used to perform regular grid division of the ground space of a preset local area according to the BeiDou satellite navigation system to obtain a ground grid coordinate system, and to perform layer processing on the vertical space covered by the ground grid coordinate system according to a preset height interval to obtain a three-dimensional grid coordinate system. The multi-source data acquisition module is used to determine the target area covered by the three-dimensional grid coordinate system according to the three-dimensional grid coordinate system, and to collect meteorological data of the target area according to the preset meteorological data acquisition equipment to obtain multi-source meteorological data, and transmit the multi-source meteorological data to the data processing center through the Beidou communication link; The data preprocessing module is used to preprocess the multi-source meteorological data in the data processing center to obtain actual multi-source meteorological data; The meteorological particle model building module is used to perform model conversion on actual multi-source meteorological data to obtain meteorological particle models. The micro-meteorological particle simulation module is used to perform dynamic simulation analysis of meteorological elements in each three-dimensional grid unit of the meteorological particle model, and obtain local micro-meteorological dynamic simulation results. The meteorological particle model construction module includes: The data transformation unit is used to perform coordinate system alignment processing on the original coordinate system of actual multi-source meteorological data to obtain a meteorological data set. The grid mapping unit is used to map discrete meteorological data in the meteorological dataset to each grid cell in the three-dimensional grid coordinate system, thereby obtaining the meteorological data distribution matrix in each grid cell. The model generation unit is used to convert the meteorological data distribution matrix in each grid cell into the corresponding unit particle attribute parameters through attribute mapping rules, and associate the unit particle attribute parameters with the spatial location of the grid cell to obtain the meteorological particle model. The micro-meteorological particle simulation module includes: The meteorological element initialization unit is used to parameterize and assign values ​​to the meteorological elements of each three-dimensional grid unit in the meteorological particle model to obtain the initial meteorological element dataset. The dynamic simulation analysis unit is used to perform stress tensor decomposition between grid cells on the initial meteorological element dataset according to the finite element analysis algorithm to obtain the decomposed stress tensor. The process includes: extracting the physical property parameters of each three-dimensional grid cell according to the initial meteorological element dataset, constructing the physical property matrix corresponding to each three-dimensional grid cell based on the physical property parameters, discretizing the contact interface of adjacent grid cells according to the physical property matrix corresponding to each three-dimensional grid cell and the cell partitioning criteria in the finite element analysis algorithm to obtain the set of stress application point coordinates of the cell interface, and performing tensor decomposition operation on the interaction force between cells based on the set of stress application point coordinates and the physical property matrix of each grid cell to obtain the decomposed stress tensor. Based on the fluid dynamics simulation algorithm, the spatiotemporal gradient of the decomposed stress tensor is calculated to obtain the dynamic meteorological element evolution sequence of each three-dimensional grid unit. The simulation result generation unit is used to perform data fusion on the evolution sequence of the dynamic meteorological elements through spatial interpolation algorithm and boundary conditions to obtain local micro-meteorological dynamic simulation results.

2. The local micro-meteorological particle simulation system based on BeiDou grid according to claim 1, characterized in that: The BeiDou grid positioning module includes: The ground grid division unit is used to calibrate the coordinates of the ground space boundary of a preset local area based on the positioning reference provided by the BeiDou satellite navigation system, and to divide the coordinate-calibrated ground space into grids according to the regular grid division standard to obtain the ground grid coordinate system; The three-dimensional mesh construction unit is used to divide the vertical space covered by the ground grid coordinate system into layers according to the preset height interval parameters, using the ground grid coordinate system as the horizontal reference, to obtain the height boundary of each layer, and to spatially associate the ground grid coordinate system with the height boundary of each layer to obtain the three-dimensional mesh coordinate system.

3. The local micro-meteorological particle simulation system based on the BeiDou grid according to claim 1, characterized in that: In the model generation unit, the meteorological data distribution matrix within each grid cell is converted into corresponding cell particle attribute parameters through attribute mapping rules, including: Based on the particle attribute type corresponding to the meteorological particle model, the meteorological data dimensions of the meteorological data distribution matrix in each grid cell are matched and filtered to obtain the target meteorological data dimension set corresponding to the particle attribute type; Based on the attribute mapping rules, the meteorological data values ​​under each target meteorological data dimension in the target meteorological data dimension set are numerically transformed to obtain the target meteorological particle attribute parameters corresponding to each target meteorological data dimension. Based on the target meteorological particle attribute parameters corresponding to each target meteorological data dimension, the target meteorological particle attribute parameters within the same grid cell are correlated and integrated to obtain the unit particle attribute parameters corresponding to that grid cell.

4. The local micro-meteorological particle simulation system based on the BeiDou grid according to claim 3, characterized in that: The mathematical expression for the unit particle attribute parameters is: In the formula, This indicates the spatial index of a grid cell in the horizontal longitude direction. This indicates the spatial subscript of a grid cell in the horizontal latitudinal direction. This represents the spatial index of a grid cell in the vertical height direction. Indicates time lower coordinate The unit particle attribute parameters corresponding to the mesh unit. This represents the preset spatial gradient weight coefficients. Represents the spatial gradient operator, This represents the total number of preset meteorological data dimensions corresponding to extreme weather. Indicates the first Preset weighting coefficients for each meteorological data dimension Indicates time lower coordinate In the grid cell of the first Preprocessed actual data for each meteorological data dimension This represents the preset time-series change weighting coefficient. Indicates time lower coordinate The unit particle attribute parameters corresponding to the mesh unit. Indicates time lower coordinate The first-order partial derivative of the cell particle property parameters with respect to time for the corresponding grid cell. This represents the preset grid space attribute weight coefficients. Representing coordinates The spatial attribute parameters corresponding to the grid cells.

5. A local micro-meteorological particle simulation system based on BeiDou grid according to claim 1, characterized in that: The mathematical expression for the decomposed stress tensor is: In the formula, Indicates the first Line number Liede The stress tensor after decomposition of layer mesh elements, Indicates the first The physical property correlation coefficient of the row grid cell. Indicates the first Line number Eigenvalues ​​of the physical property matrix of column grid cells Indicates the first Discretization coefficients of the contact interface of column mesh elements. Indicates the first Liede The associated coordinates of the stress application points of the layer mesh element.

6. The local micro-meteorological particle simulation system based on the BeiDou grid according to claim 1, characterized in that: In the dynamic simulation analysis unit, the spatiotemporal gradient of the decomposed stress tensor is calculated based on the fluid dynamics simulation algorithm to obtain the dynamic meteorological element evolution sequence of each three-dimensional grid unit, including: The spatiotemporal distribution range of the decomposed stress tensor in the three-dimensional grid coordinate system is determined to obtain the spatiotemporal distribution domain of the tensor. The spatiotemporal distribution domain of the tensor is divided into time dimensions according to the preset time step to obtain multiple time segment intervals. The stress tensor in each time segment interval is divided into spatial grids according to the spatial resolution to obtain a discretized stress tensor sequence. Tensor data at each time step in the discretized stress tensor sequence are extracted to obtain a single-time tensor dataset. Based on the Navier-Stokes equations in the fluid dynamics simulation algorithm, spatial partial derivatives of the stress tensor components of each grid cell in the single-time tensor dataset are calculated to obtain the cell gradient components of each grid cell. The cell gradient components of each grid cell are then spatially stitched together according to the three-dimensional grid coordinate system to obtain the instantaneous gradient field. Gradient data of grid cells in the instantaneous gradient field corresponding to each time step are extracted to obtain a time step gradient dataset. Adjacent time step data in the time step gradient dataset are temporally correlated according to the time axis order to obtain a temporally correlated gradient sequence. The temporally correlated gradient sequence is then mapped to the meteorological element parameter dimension of each three-dimensional grid cell to obtain the dynamic meteorological element evolution sequence of each three-dimensional grid cell.

7. A local micro-meteorological particle simulation system based on BeiDou grid according to claim 1, characterized in that: In the simulation result generation unit, the evolution sequence of the dynamic meteorological elements is fused using a spatial interpolation algorithm and boundary conditions to obtain local micro-meteorological dynamic simulation results, including: Based on the dynamic meteorological element evolution sequence, meteorological element data and corresponding grid coordinates of the boundary nodes of adjacent three-dimensional grid units are extracted to obtain the boundary data and coordinate set of adjacent units. The transition area between adjacent three-dimensional grid units is determined, and the meteorological element values ​​within the transition area are calculated point by point based on the spatial interpolation algorithm to obtain the first interpolated transition data between adjacent grid units. The boundary continuity parameters and their value ranges in the preset boundary condition constraint rules are extracted to obtain the boundary constraint parameter set. The values ​​of each node in the first interpolated transition data are compared with the corresponding parameter value ranges in the boundary constraint parameter set. Abnormal transition data that exceeds the parameter value range is marked. The correction coefficients in the boundary constraint parameter set are used to adjust the values ​​of the abnormal transition data to obtain the second interpolated transition data that meets the boundary conditions. The original meteorological element data and corresponding grid index of each three-dimensional grid unit in the dynamic meteorological element evolution sequence are extracted to obtain the original dataset; Based on the grid index, the second interpolated transition data is spatially aligned with the corresponding grid positions in the original dataset to obtain the overlapping area. Then, according to the data fusion and integration rules, the original meteorological element data and the second interpolated transition data in the overlapping area are weighted and numerically merged to obtain the local micro-meteorological dynamic simulation results.

Images