Meteorological grid data service method and system fused with three-dimensional terrain

Abstract

The invention provides a meteorological grid data service method and system fused with a three-dimensional terrain, and belongs to the field of meteorological services. The method comprises the following steps: establishing spatial lattice points of meteorological data based on a page end of a B / S structure, obtaining a lattice point data structure, creating a parameter mapping relation between topographic data and meteorological data, integrating the parameter mapping relation, constructing a fusion reconstruction model, obtaining historical meteorological data and topographic data, training the fusion reconstruction model, and obtaining a meteorological data fusion model; and acquiring current meteorological data and topographic data, inputting the acquired data into the trained fusion reconstruction model to output meteorological grid data, performing visual output according to a specific application scene, and providing meteorological service. According to the invention, the extended application service of the meteorological grid data to the three-dimensional virtual space is realized, the fusion and application of the spatial grid point positions of the meteorological grid data and the three-dimensional topographic data are realized, the small-scene and refined application of the meteorological data is improved, and the refined service level is improved.

Description

Technical Field

[0001] The invention belongs to the field of meteorological services, and in particular relates to a meteorological grid data service method and system integrating three-dimensional terrain. Background Art

[0002] With the rapid development of geographic information technology, people's demand for the display of meteorological data on maps is increasing. Meteorological data contains a large amount of meteorological information, such as temperature, humidity, wind speed, air pressure, etc., which can be used in climate prediction, weather forecast, environmental monitoring and other fields. The visualization of meteorological data includes two-dimensional and three-dimensional. Traditional meteorological data display methods such as charts, text, videos, and two-dimensional maps can no longer meet people's needs. A more intuitive, accurate and vivid display method is needed to present meteorological data.

[0003] In the prior art, meteorological data is displayed in a three-dimensional manner, generally by combining meteorological data with a three-dimensional map through a virtual camera, so that the meteorological data is displayed in a three-dimensional map. For example, Chinese patent application number 201510293523.4 discloses a method for displaying and drawing meteorological elements in a three-dimensional earth model, which defines the visible area through a virtual camera and displays the relevant meteorological data in the visible area in three dimensions. However, the three-dimensional conversion process of the above method requires the assistance of a virtual camera, and the process is complicated, the professional requirements are high, and it is not universal; at the same time, the display process cannot be flexibly selected according to different needs, and the applicability of application scenarios is not strong. Summary of the invention

[0004] In view of the above-mentioned defects or deficiencies in the prior art, the present invention aims to provide a meteorological grid data service method and system integrating three-dimensional terrain, so as to realize the fusion and application of the spatial grid position of meteorological grid data and three-dimensional terrain data, and enhance the small-scene and refined application of meteorological data.

[0005] In order to achieve the above purpose, the embodiment of the present invention adopts the following technical solution: In a first aspect, an embodiment of the present invention provides a meteorological grid data service method integrating three-dimensional terrain, the method comprising the following steps: Step S1, establishing a spatial grid of meteorological data on the page end based on the B / S structure to obtain a grid data structure; Step S2, creating a parameter mapping relationship between terrain data and meteorological data based on the grid data structure; Step S3, integrating parameter mapping relationships and constructing a fusion reconstruction model; Step S4, obtaining historical meteorological data and corresponding terrain data, performing corresponding preprocessing, and dividing them into a training data set and a test data set; Step S5, inputting the training data set into the fusion reconstruction model, training the fusion reconstruction model, and then testing the trained fusion reconstruction model with the test data set to obtain a trained fusion reconstruction model; Step S6, obtaining current meteorological data and terrain data, and inputting the obtained data into the trained fusion reconstruction model to output meteorological grid data, and performing visualization output according to specific application scenarios to provide meteorological services.

[0006] As a preferred embodiment of the present invention, obtaining the grid data structure specifically includes: Step S11, when establishing the spatial grid of meteorological data, the grid data structure is the same on all height slices. On each data collection plane, the meteorological data and meteorological parameters are converted by formulas (1)-(4) to obtain the data representation of the plane: W = X_{max} - X_{min} (1) H = Y_{max} - Y_{min} (2) Δx=W / (N x -1) (3) Δy=H / (N y -1) (4) In formulas (1)-(4), W is the grid data width, H is the grid data height, X_{max} and X_{min} represent the maximum and minimum values ​​in width, Y_{max} and Y_{min} represent the maximum and minimum values ​​in height, Δx and Δy represent the grid spacing in width and height, respectively, and N x 、N y denote the number of grid points along the width and height directions respectively; Step S12, in the data representation of the plane, assigning the corresponding height value to the grid point position to form a complete grid point data structure, and obtaining the observation value corresponding to the corresponding grid point data structure; In the data representation of the plane, the position coordinates of the meteorological observation value are (x, y), and the height value h corresponding to the observation value is assigned to the three-dimensional position coordinate z of the grid point, forming the three-dimensional spatial coordinates of the meteorological observation value (x, y, z). The formula for the observation value g(x, y, z) at the corresponding spatial grid point is: g(x,y,z)=g_{i,j,k} (5) In formula (5), the grid data structure is g_{i,j,k}, which represents the meteorological observation value at the i-th row and j-th column on the z-height plane, and the k value is the attribute value of the meteorological observation value; x0 and y0 are the spatial width and height value of the origin of the coordinate system, respectively, i=floor(x-x0) / Δx, j=floor(y-y0) / Δy, and floor(x-x0) represents the relative width of the current grid point relative to the origin of the coordinate system, floor(y-y0) represents the relative height of the current grid point relative to the origin of the coordinate system, and x and y represent the absolute longitude and latitude coordinates of the grid point, respectively.

[0007] As a preferred embodiment of the present invention, the parameter mapping relationship in step S2 includes an elevation mapping relationship, a color mapping relationship and an element mapping relationship.

[0008] As a preferred embodiment of the present invention, the elevation mapping relationship, after determining the grid point data structure, obtains the coordinates (x, y, z) of each grid point, and obtains the fused elevation data according to the elevation mapping relationship of formula (6): H m (x,y,z) = H t (x,y,z) + H p (x,y,z) (6) In formula (6), H m (x, y, z) represents the fused elevation data at point (x, y, z), H t (x, y, z) represents the terrain elevation data at point (x, y, z), H p (x,y,z) represents the corresponding height value of the meteorological data at the point (x,y,z).

[0009] As a preferred embodiment of the present invention, the color mapping relationship has different color values ​​for meteorological data and terrain data. In this step, the fused color data is calculated according to the characteristics of the data. The color mapping relationship is shown in formula (7): C m (x,y,z) = C t (x,y,z)×ω t + C p (x,y,z)×ω p (7) In formula (7), C m (x, y, z) represents the fused color data at point (x, y, z), C t (x,y) represents the terrain color data at point (x,y,z), C p (x,y,z) represents the weather color data at point (x,y,z), ω t and ω p They are the weights of terrain color and weather color respectively.

[0010] As a preferred embodiment of the present invention, the element mapping relationship is to grid-structure the meteorological element data and then fuse it with the terrain data. The element mapping relationship formula is: W m (x,y,z) = W t (x,y,z) + W p (x,y,z) (8) In formula (8), W m (x, y, z) represents the fused feature data at point (x, y, z), W t (x, y, z) represents the terrain impact data, W p (x,y,z) represents the meteorological impact data.

[0011] As a preferred embodiment of the present invention, the fusion reconstruction model includes a first convolution layer, a second convolution layer, a sub-pixel convolution layer, a pooling layer and a fully connected layer. The input data is sampled in the first convolution layer, and a multi-scale residual model is used in the second convolution layer to extract the features of meteorological data and terrain data based on a parameter mapping relationship. The high-resolution data is reconstructed in the sub-pixel convolution layer with the features of terrain data and meteorological data. After the high-resolution data is pooled in the pooling layer, it is input into the fully connected layer to obtain fused meteorological grid data including three-dimensional terrain data.

[0012] As a preferred embodiment of the present invention, when training the fusion reconstruction model in step S5, it specifically includes: Step S51, inputting terrain data and meteorological data into the image fusion reconstruction model and sampling through the first convolutional layer; Step S52, input the sampled data into the second convolution layer, reconstruct the high-resolution data with the features of the terrain data and the meteorological data in the sub-pixel convolution layer, input the data into the pooling layer and pass through the fully connected layer, output the fusion result, and use it as the intermediate result for the reconstruction of the next level, and finally fuse the terrain data and the meteorological data into three-dimensional meteorological element data consistent with the three-dimensional terrain image; Step S53, after the features of each layer are pooled in the pooling layer, the fused meteorological grid data including the three-dimensional terrain data is output after passing through the fully connected layer; Step S54, evaluate the fused data, determine whether the result reaches the convergence condition, and reversely transmit the determination result to optimize the model parameters until the convergence condition is reached, thereby obtaining the final trained fusion reconstruction model.

[0013] As a preferred embodiment of the present invention, in step S52, when performing feature extraction in the second convolutional layer, a multi-scale residual model is used to perform multi-scale feature extraction; the multi-scale residual model includes a convolutional layer with a convolution kernel size of 1×1, ensuring that the number of input and output feature layers of the multi-scale residual model remains unchanged; the output of the convolutional layer will be used as input of two completely different network layers, one is to reconstruct the residual information at the same level, and the other is used for feature extraction of the next layer, and parameters of different layers are shared.

[0014] In a second aspect, an embodiment of the present invention further provides a meteorological grid data service system integrating three-dimensional terrain, the system comprising: a spatial grid construction module, a parameter mapping module, a model construction and training module, a data acquisition module and a service module; wherein, The spatial grid point construction module is used to establish the spatial grid points of meteorological data based on the page end of the B / S structure to obtain the grid point data structure; The parameter mapping module is used to create a parameter mapping relationship between terrain data and meteorological data based on a grid data structure; The model building and training module is used to integrate the parameter mapping relationship and build a fusion reconstruction model; and is used to input the training data set into the fusion reconstruction model to train the fusion reconstruction model, and then test the trained fusion reconstruction model with the test data set to obtain a trained fusion reconstruction model; The data acquisition module is used to acquire historical meteorological data and corresponding terrain data, and perform corresponding preprocessing, and divide them into training data sets and test data sets, and obtain current meteorological data and terrain data to be monitored; The service module is used to input the acquired data into the trained fusion reconstruction model to output meteorological grid data, perform visual output according to specific application scenarios, and provide meteorological services.

[0015] The technical solution provided by the embodiment of the present invention has the following beneficial effects: The meteorological grid data service method and system integrated with three-dimensional terrain described in the embodiment of the present invention better meet the needs of rapid mapping of scenes such as rapid response to emergencies, major event support, and review of major historical events, realize the expansion of meteorological data resources from two-dimensional views to three-dimensional geographic scenes, and propose innovative technologies for automatic rapid mapping based on three-dimensional dynamic rendering of weather processes, which solve the problems of information missing and information overload, two-dimensional view limitations, slow update cycles, single map styles, and high map production costs in traditional map making technologies; and realize the extended application service of meteorological grid data to three-dimensional virtual space. By integrating high-resolution meteorological grid data and three-dimensional terrain data, the fusion and application of the spatial grid point positions of meteorological grid data and three-dimensional terrain data are realized, solving the problem of insufficient capacity for small-scene and refined meteorological services, and improving the refined service level of meteorological data.

[0016] Of course, it is not necessary to achieve all of the advantages described above at the same time to implement any product or method of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

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

[0018] Figure 1 It is a flow chart of the meteorological grid data service method integrating three-dimensional terrain according to an embodiment of the present invention; Figure 2 It is a schematic diagram of the structure of the fusion reconstruction model constructed in the embodiment of the present invention. DETAILED DESCRIPTION

[0019] After discovering the above problems, the inventors of this application conducted a detailed study on the existing meteorological data service methods and systems. The study found that after the meteorological data is gridded, it can be well integrated with the three-dimensional terrain grid data. Therefore, using meteorological grid data to provide services can improve the service level and user experience. At the same time, three-dimensional terrain technology is a terrain expression method that has developed rapidly in recent years. It presents terrain data in a three-dimensional form, providing people with more intuitive and accurate terrain information. Three-dimensional technology can provide more realistic and accurate terrain information, which can help people better understand the characteristics and changes of terrain and landforms.

[0020] However, how to effectively integrate meteorological grid data and three-dimensional terrain data is still a problem. If the integration is still carried out by using stand-alone applications such as virtual cameras, such applications usually need to be installed locally. During the installation process, there are certain requirements for the computer environment, such as Java environment, Python dependency packages, etc., which require high professionalism and are not suitable for services for the general public. In addition, these applications can usually only meet tasks in fixed scenarios and lack flexibility and scalability, so the above problems still exist.

[0021] It should be noted that the defects existing in the solutions in the above-mentioned prior art are the results obtained by the inventor after practice and careful research. Therefore, the discovery process of the above-mentioned problems and the solutions proposed in the embodiments of the present invention for the above-mentioned problems below should all be the contributions made by the inventor to the present invention in the process of the present invention.

[0022] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. The components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations. It should be noted that the embodiments of the present invention and the features in the embodiments can also be combined with each other without conflict.

[0023] It should be noted that similar reference numerals and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings. In the description of the present invention, the terms "first", "second", "third", "fourth", etc. are only used to distinguish the description and cannot be understood as indicating or implying relative importance.

[0024] After the above in-depth analysis, the embodiment of the present invention aims to propose a meteorological grid data service method and system that integrates three-dimensional terrain. First, the spatial grid of the fused data is constructed based on the page-end technology and the nearest neighbor algorithm of the B / S structure; secondly, the meteorological grid data is fused with the three-dimensional terrain data, and a parameter mapping relationship is established based on the created spatial grid. Finally, a three-dimensional meteorological fusion model is constructed based on the fused new grid data, and meteorological services are provided through the optimized three-dimensional meteorological fusion model.

[0025] like Figure 1 As shown, the meteorological grid data service method integrating three-dimensional terrain includes the following steps: Step S1, establishing a spatial grid of meteorological data based on the page end of the B / S structure to obtain a grid data structure.

[0026] In this step, the page-side technology based on the B / S structure has gradually matured with the continuous development of Internet technology. Compared with traditional professional installed version visualization tools, this technical model has higher flexibility and accessibility, and can better meet the needs of social services. In terms of realizing the non-installed visualization application of meteorological grid data, the page-side technology based on the B / S structure utilizes browser application designs such as WebGIS and HTML5. Among them, WebGIS technology is used to realize the release of dynamic slices, and the front end of the page directly calculates and displays. In this mode, image processing is mainly carried out in the background, and the grid is image processed, and the release of dynamic slices is realized by using WebGIS technology, and the front end of the page directly calculates and displays. This implementation method is efficient, flexible and easy to maintain. The B / S structure provides a technical implementation basis for the fusion of meteorological grid data and three-dimensional terrain data.

[0027] The meteorological data is usually based on plane data at a preset height slice. The spatial grid points of the meteorological data are established by dividing the spatial grids based on the meteorological data (temperature, air pressure, wind speed, etc.) on the plane, reading the corresponding meteorological data observation values ​​at the grid points, and forming the corresponding grid point data structure.

[0028] The forming of the corresponding grid data structure specifically includes: Step S11, when establishing the spatial grid of meteorological data, the grid data structure is the same on all height slices. On each plane of data collection, the meteorological data and meteorological parameters are converted by a resampling algorithm. The formula for data representation is as follows: W = X_{max} - X_{min} (1) H = Y_{max} - Y_{min} (2) Δx=W / (N x -1) (3) Δy=H / (N y -1) (4) In formulas (1)-(4), W is the grid data width, H is the grid data height, X_{max} and X_{min} represent the maximum and minimum values ​​in width, Y_{max} and Y_{min} represent the maximum and minimum values ​​in height, Δx and Δy represent the grid spacing in width and height, respectively, and N x 、N y denote the number of grid points along the width and height directions, respectively. In a preferred embodiment, the resampling algorithm adopts nearest neighbor sampling.

[0029] Step S12, in the plane data representation, assign the corresponding height value to the grid point position to form a complete grid point data structure, and obtain the observation value corresponding to the corresponding grid point data structure.

[0030] In this step, in the data representation of the plane, the position coordinates of the meteorological observation value are (x, y), and the height value h corresponding to the observation value is assigned to the three-dimensional position coordinate z of the grid point, forming the three-dimensional spatial coordinates of the meteorological observation value (x, y, z), and the observation value g (x, y, z) at the corresponding spatial grid point is: g(x,y,z)=g_{i,j,k} (5) In formula (5), the grid data structure is g_{i,j,k}, which represents the meteorological observation value at the i-th row and j-th column on the z-height plane. The k value is generally the attribute value of the meteorological observation value, which is related to the value or layering of the height plane. x0 and y0 are the spatial width and height value of the origin of the coordinate system, i=floor(x-x0) / Δx, j=floor(y-y0) / Δy, where floor(x-x0) represents the relative width of the current grid point relative to the origin of the coordinate system, floor(y-y0) represents the relative height of the current grid point relative to the origin of the coordinate system, and x and y represent the absolute longitude and latitude coordinates of the grid point, respectively. Through nearest neighbor sampling, the corresponding meteorological observation value can be calculated at any position on the plane map at a predetermined height.

[0031] Step S2: creating a parameter mapping relationship between terrain data and meteorological data based on the grid data structure.

[0032] In this step, the parameter mapping relationship includes an elevation mapping relationship, a color mapping relationship, and an element mapping relationship. This step is a basic step for obtaining a fusion data model. According to the obtained parameter mapping relationship, a fusion data model is further obtained. The mapping relationship here is defined according to the characteristics of meteorological data and terrain data, so that a mathematical model can be further established for a certain meteorological parameter or terrain data.

[0033] After determining the grid point data structure, the coordinates (x, y, z) of each grid point are obtained, and the fused elevation data is obtained according to the elevation mapping relationship of formula (6): H m (x,y,z) = H t (x,y,z) + H p (x,y,z) (6) In formula (6), H m (x, y, z) represents the fused elevation data at point (x, y, z), H t (x, y, z) represents the terrain elevation data at point (x, y, z), Hp (x,y,z) represents the corresponding height value of the meteorological data at the point (x,y,z).

[0034] The color mapping relationship is that the color values ​​of meteorological data and terrain data are different. In this step, the fused color data is calculated according to the characteristics of the data. The color mapping relationship is shown in formula (7): C m (x,y,z) = C t (x,y,z)×ω t + C p (x,y,z)×ω p (7) In formula (7), C m (x, y, z) represents the fused color data at point (x, y, z), C t (x,y) represents the terrain color data at point (x,y,z), C p (x,y,z) represents the weather color data at point (x,y,z), ω t and ω p They are the weights of terrain color and weather color respectively.

[0035] The element mapping relationship is to grid-structure the meteorological element data and then fuse it with the terrain data. The element mapping relationship formula is: W m (x,y,z) = W t (x,y,z) + W p (x,y,z) (8) In formula (8), W m (x, y, z) represents the fused feature data at point (x, y, z), W t (x, y, z) represents the terrain impact data, W p (x,y,z) represents the meteorological impact data.

[0036] Step S3: Integrate the parameter mapping relationship and construct a fusion reconstruction model.

[0037] In this step, if Figure 2 As shown, the fusion reconstruction model includes a first convolution layer, a second convolution layer, a sub-pixel convolution layer, a pooling layer and a fully connected layer. The input data is sampled in the first convolution layer, and a multi-scale residual model is used in the second convolution layer to extract the features of meteorological data and terrain data based on a parameter mapping relationship. The high-resolution data is reconstructed with the features of terrain data and meteorological data in the sub-pixel convolution layer. After the high-resolution data is pooled in the pooling layer, it is input into the fully connected layer to obtain fused meteorological grid data including three-dimensional terrain data.

[0038] The constructed fusion reconstruction model is a three-dimensional meteorological fusion model with spatiotemporal characteristics based on meteorological data and terrain data. Through the fusion reconstruction model, based on three-dimensional visualization technology, the elevation extremes of three-dimensional terrain data are obtained, and various meteorological spatial elements such as points, lines, and surfaces are virtually stretched into three-dimensional shapes to form a three-dimensional terrain fusion display and application of various elements. The defined parameter mapping relationship is integrated while ensuring the spatial consistency of the data, and the meteorological data and terrain data are fused to generate a three-dimensional meteorological fusion model with time and space.

[0039] Step S4, obtaining historical meteorological data and corresponding terrain data, performing corresponding preprocessing, and dividing them into a training data set and a test data set.

[0040] Step S5, inputting the training data set into the fusion reconstruction model, training the fusion reconstruction model, and then testing the trained fusion reconstruction model with the test data set to obtain a trained fusion reconstruction model.

[0041] In this step, when the fusion reconstruction model is trained, it specifically includes: Step S51, inputting terrain data and meteorological data into the image fusion reconstruction model and sampling through the first convolutional layer.

[0042] In step S52, the sampled data is input into the second convolution layer, and the high-resolution data is reconstructed with the features of the terrain data and the meteorological data in the sub-pixel convolution layer. The data is input into the pooling layer and passes through the fully connected layer, and the fusion result is output. The data is used as an intermediate result for reconstruction at the next level, and finally the terrain data and the meteorological data are fused into three-dimensional meteorological element data consistent with the three-dimensional terrain image.

[0043] In this step, when performing feature extraction in the second convolutional layer, the multi-scale residual model is used for multi-scale feature extraction, which can effectively avoid model degradation and accelerate convergence. The multi-scale residual model first includes a convolution layer with a convolution kernel size of 1×1, which can greatly reduce the amount of feature parameters and ensure that the number of input and output feature layers of the multi-scale residual model remains unchanged. The output of the convolution layer will be used as the input of two completely different network layers, one is to reconstruct the residual information at the same level, and the other is used for feature extraction of the next layer. In order to effectively reduce the computational complexity, parameters of different layers are shared. The feature value extraction analysis mainly realizes the spatial analysis of meteorological data based on three-dimensional terrain scenes. When performing special feature extraction, image fusion reconstruction uses high-resolution terrain data and dynamic vector meteorological data as input. For each input (grid data and high-resolution terrain data), the features of each layer are extracted through the convolution algorithm.

[0044] Step S53, after the features of each layer are pooled in the pooling layer, the fused meteorological grid data including the three-dimensional terrain data is output after passing through the fully connected layer.

[0045] Step S54, evaluate the fused data, determine whether the result reaches the convergence condition, and reversely transmit the determination result to optimize the model parameters until the convergence condition is reached, thereby obtaining the final trained fusion reconstruction model.

[0046] Step S6, obtaining current meteorological data and terrain data, and inputting the obtained data into the trained fusion reconstruction model to output meteorological grid data, and performing visualization output according to specific application scenarios to provide meteorological services.

[0047] In this step, three-dimensional meteorological data is output through the three-dimensional meteorological fusion model. In terms of presentation form, two-dimensional visualization is difficult to express complex three-dimensional spatial information and cannot present the three-dimensional sense and depth of the data. At the same time, two-dimensional graphics may be misleading when expressing three-dimensional data, requiring users to have a certain spatial imagination ability; the expression of meteorological data combined with three-dimensional terrain is intuitive: three-dimensional visualization can intuitively present the three-dimensional spatial distribution of data and is easy to understand. Rich expression of details: three-dimensional visualization can better express the details and changes of data. Strong interactivity: three-dimensional visualization supports user interactive operations, such as rotation, zooming, etc., so that users can have a deeper understanding of the data.

[0048] Based on the same idea, an embodiment of the present invention also provides a meteorological grid data service system integrating three-dimensional terrain, the system comprising: a spatial grid construction module, a parameter mapping module, a model construction and training module, a data acquisition module and a service module; wherein, The spatial grid point construction module is used to establish the spatial grid points of meteorological data based on the page end of the B / S structure to obtain the grid point data structure; The parameter mapping module is used to create a parameter mapping relationship between terrain data and meteorological data based on a grid data structure; The model building and training module is used to integrate the parameter mapping relationship and build a fusion reconstruction model; and is used to input the training data set into the fusion reconstruction model to train the fusion reconstruction model, and then test the trained fusion reconstruction model with the test data set to obtain a trained fusion reconstruction model; The data acquisition module is used to acquire historical meteorological data and corresponding terrain data, and perform corresponding preprocessing, and divide them into training data sets and test data sets, and obtain current meteorological data and terrain data to be monitored; The service module is used to input the acquired data into the trained fusion reconstruction model to output meteorological grid data, perform visual output according to specific application scenarios, and provide meteorological services.

[0049] In this embodiment, each module is implemented by a processor, and a memory is appropriately added when storage is required. Among them, the processor can be but is not limited to a microprocessor MPU, a central processing unit (CPU), a network processor (NP), a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA), other programmable logic devices, discrete gates, transistor logic devices, discrete hardware components, etc. The memory may include a random access memory (RAM) and may also include a non-volatile memory (NVM), such as at least one disk storage. Optionally, the memory may also be at least one storage device located away from the aforementioned processor.

[0050] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented by software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of the present invention is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from a website site, computer, server or data center to another website site, computer, server or data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) mode.

[0051] It should also be noted that the meteorological grid data service system integrating three-dimensional terrain described in this embodiment corresponds to the meteorological grid data service method integrating three-dimensional terrain. The description and limitation of the method are also applicable to the system and will not be repeated here.

[0052] It can be seen that the meteorological grid data service method and system for integrating three-dimensional terrain described in the embodiment of the present invention establishes the spatial grid of meteorological data based on the page end of the B / S structure, obtains the grid data structure, and creates the parameter mapping relationship between the terrain data and the meteorological data, and then integrates the parameter mapping relationship to build a fusion reconstruction model, obtains historical meteorological data and terrain data, and after training the fusion reconstruction model, obtains the current meteorological data and terrain data, and inputs the obtained data into the trained fusion reconstruction model to output meteorological grid data, and outputs it visually according to the specific application scenario to provide meteorological services. The present invention better meets the needs of rapid mapping of scenes such as rapid response to emergencies, major event security, and major historical event review, realizes the expansion of meteorological data resources from two-dimensional views to three-dimensional geographic scenes, and proposes an innovative technology for automatic rapid mapping based on three-dimensional dynamic rendering of weather processes, which solves the problems of information missing and information overload, two-dimensional view limitations, slow update cycle, single map style, and high cost of drawing in traditional map mapping technology; and realizes the extended application service of meteorological grid data to three-dimensional virtual space. By integrating high-resolution meteorological grid data and three-dimensional terrain data, the fusion and application of the spatial grid positions of meteorological grid data and three-dimensional terrain data are realized, the problem of insufficient capabilities for small-scale and refined meteorological services is solved, and the level of refined meteorological data services is improved.

[0053] The above description is only a preferred embodiment of the present invention and an explanation of the technical principles used. It is not intended to limit the scope of the invention claimed for protection, but only represents the preferred embodiment of the present invention. Those skilled in the art should understand that the scope of the invention involved in the present invention is not limited to the technical solution formed by a specific combination of the above-mentioned technical features, but should also cover other technical solutions formed by any combination of the above-mentioned technical features or their equivalent features without departing from the inventive concept. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making creative work belong to the scope of protection of the present invention.

Claims

1. A meteorological grid data service method integrating three-dimensional terrain, characterized in that: The method comprises the following steps: Step S1, establishing a spatial grid of meteorological data on the page end based on the B / S structure to obtain a grid data structure; Step S2, creating a parameter mapping relationship between terrain data and meteorological data based on the grid data structure; Step S3, integrating parameter mapping relationships and constructing a fusion reconstruction model; Step S4, obtaining historical meteorological data and corresponding terrain data, performing corresponding preprocessing, and dividing them into a training data set and a test data set; Step S5, inputting the training data set into the fusion reconstruction model, training the fusion reconstruction model, and then testing the trained fusion reconstruction model with the test data set to obtain a trained fusion reconstruction model; Step S6, obtaining current meteorological data and terrain data, and inputting the obtained data into the trained fusion reconstruction model to output meteorological grid data, and performing visualization output according to specific application scenarios to provide meteorological services.

2. The meteorological grid data service method for integrating three-dimensional terrain according to claim 1 is characterized in that: The obtaining of the grid data structure specifically includes: Step S11, when establishing the spatial grid of meteorological data, the grid data structure is the same on all height slices. On each data collection plane, the meteorological data and meteorological parameters are converted by formulas (1)-(4) to obtain the data representation of the plane: W = X_{max} - X_{min} (1) H = Y_{max} - Y_{min} (2) Δx=W / (N x -1) (3) Δy=H / (N y -1) (4) In formulas (1)-(4), W is the grid data width, H is the grid data height, X_{max} and X_{min} represent the maximum and minimum values ​​in width, Y_{max} and Y_{min} represent the maximum and minimum values ​​in height, Δx and Δy represent the grid spacing in width and height, respectively, and N x 、N y denote the number of grid points along the width and height directions respectively; Step S12, in the data representation of the plane, assigning the corresponding height value to the grid point position to form a complete grid point data structure, and obtaining the observation value corresponding to the corresponding grid point data structure; In the data representation of the plane, the position coordinates of the meteorological observation value are (x, y), and the height value h corresponding to the observation value is assigned to the three-dimensional position coordinate z of the grid point, forming the three-dimensional spatial coordinates of the meteorological observation value (x, y, z). The formula for the observation value g(x, y, z) at the corresponding spatial grid point is: g(x,y,z)=g_{i,j,k} (5) In formula (5), the grid data structure is g_{i,j,k}, which represents the meteorological observation value at the i-th row and j-th column on the z-height plane, and the k value is the attribute value of the meteorological observation value; x0 and y0 are the spatial width and height value of the origin of the coordinate system, respectively, i=floor(x-x0) / Δx, j=floor(y-y0) / Δy, and floor(x-x0) represents the relative width of the current grid point relative to the origin of the coordinate system, floor(y-y0) represents the relative height of the current grid point relative to the origin of the coordinate system, and x and y represent the absolute longitude and latitude coordinates of the grid point, respectively.

3. The meteorological grid data service method for integrating three-dimensional terrain according to claim 1 is characterized in that: The parameter mapping relationship in step S2 includes an elevation mapping relationship, a color mapping relationship and an element mapping relationship.

4. The meteorological grid data service method for integrating three-dimensional terrain according to claim 3 is characterized in that: After determining the grid point data structure, the coordinates (x, y, z) of each grid point are obtained, and the fused elevation data is obtained according to the elevation mapping relationship of formula (6): H m (x,y,z) = H t (x,y,z) + H p (x,y,z) (6) In formula (6), H m (x, y, z) represents the fused elevation data at point (x, y, z), H t (x, y, z) represents the terrain elevation data at point (x, y, z), H p (x,y,z) represents the corresponding height value of the meteorological data at the point (x,y,z).

5. The meteorological grid data service method for integrating three-dimensional terrain according to claim 3 is characterized in that: The color mapping relationship is that the color values ​​of meteorological data and terrain data are different. In this step, the fused color data is calculated according to the characteristics of the data. The color mapping relationship is shown in formula (7): C m (x,y,z) = C t (x,y,z)×ω t + C p (x,y,z)×ω p (7) In formula (7), C m (x, y, z) represents the fused color data at point (x, y, z), C t (x,y) represents the terrain color data at point (x,y,z), C p (x, y, z) represents the weather color data at point (x, y, z), ω t and ω p They are the weights of terrain color and weather color respectively.

6. The meteorological grid data service method for integrating three-dimensional terrain according to claim 3 is characterized in that: The element mapping relationship is to grid-structure the meteorological element data and then fuse it with the terrain data. The element mapping relationship formula is: W m (x,y,z) = W t (x,y,z) + W p (x,y,z) (8) In formula (8), W m (x, y, z) represents the fused feature data at point (x, y, z), W t (x, y, z) represents the terrain impact data, W p (x,y,z) represents the meteorological impact data.

7. The meteorological grid data service method for integrating three-dimensional terrain according to claim 1 is characterized in that: The fusion reconstruction model includes a first convolution layer, a second convolution layer, a sub-pixel convolution layer, a pooling layer and a fully connected layer. The input data is sampled in the first convolution layer. A multi-scale residual model is used in the second convolution layer to extract the features of meteorological data and terrain data based on a parameter mapping relationship. High-resolution data is reconstructed in the sub-pixel convolution layer using the features of terrain data and meteorological data. After the high-resolution data is pooled in the pooling layer, it is input into the fully connected layer to obtain fused meteorological grid data including three-dimensional terrain data.

8. The meteorological grid data service method for integrating three-dimensional terrain according to claim 7 is characterized in that: When the fusion reconstruction model is trained in step S5, it specifically includes: Step S51, inputting terrain data and meteorological data into the image fusion reconstruction model and sampling through the first convolutional layer; Step S52, input the sampled data into the second convolution layer, reconstruct the high-resolution data with the features of the terrain data and the meteorological data in the sub-pixel convolution layer, input the data into the pooling layer and pass through the fully connected layer, output the fusion result, and use it as the intermediate result for the reconstruction of the next level, and finally fuse the terrain data and the meteorological data into three-dimensional meteorological element data consistent with the three-dimensional terrain image; Step S53, after the features of each layer are pooled in the pooling layer, the fused meteorological grid data including the three-dimensional terrain data is output after passing through the fully connected layer; Step S54, evaluate the fused data, determine whether the result reaches the convergence condition, and reversely transmit the determination result to optimize the model parameters until the convergence condition is reached, thereby obtaining the final trained fusion reconstruction model.

9. The meteorological grid data service method for integrating three-dimensional terrain according to claim 8 is characterized in that: In step S52, when feature extraction is performed in the second convolutional layer, a multi-scale residual model is used to perform multi-scale feature extraction; the multi-scale residual model includes a convolutional layer with a convolution kernel size of 1×1, ensuring that the number of input and output feature layers of the multi-scale residual model remains unchanged; the output of the convolutional layer will be used as input of two completely different network layers, one is to reconstruct the residual information at the same level, and the other is used for feature extraction of the next layer, and parameters of different layers are shared.

10. A meteorological grid data service system integrating three-dimensional terrain, characterized in that: The system includes: a spatial grid construction module, a parameter mapping module, a model construction and training module, a data acquisition module and a service module; wherein, The spatial grid point construction module is used to establish the spatial grid points of meteorological data based on the page end of the B / S structure to obtain the grid point data structure; The parameter mapping module is used to create a parameter mapping relationship between terrain data and meteorological data based on a grid data structure; The model building and training module is used to integrate the parameter mapping relationship and build a fusion reconstruction model; and is used to input the training data set into the fusion reconstruction model to train the fusion reconstruction model, and then test the trained fusion reconstruction model with the test data set to obtain a trained fusion reconstruction model; The data acquisition module is used to acquire historical meteorological data and corresponding terrain data, and perform corresponding preprocessing, and divide them into training data sets and test data sets, and obtain current meteorological data and terrain data to be monitored; The service module is used to input the acquired data into the trained fusion reconstruction model to output meteorological grid data, perform visual output according to specific application scenarios, and provide meteorological services.

Images