Plain area groundwater prediction model construction method and application platform

Abstract

The invention discloses a plain area groundwater prediction model construction method and an application platform, and relates to the field of groundwater overdraft area prediction. The method comprises the following steps: collecting underground water related data, preprocessing the underground water related data, and establishing a multivariate database which comprises a basic information library, a model intermediate result library, a dynamic database, a prediction database and a model result library; calling example drill holes, earth surface DEM, hydrogeological profile maps and hydrogeological parameter maps in the multivariate database, determining information characteristics of stratums and environments of the stratums by applying a Shannon information theory on the basis of drill hole and profile data, and editing a three-dimensional scene by utilizing the information characteristics of the stratums and the environments of the stratums; and establishing an initial groundwater prediction model based on the neural network model and the time sequence model, performing three-dimensional description on the initial groundwater prediction model by the three-dimensional scene, and establishing an initial groundwater model. The method is efficient in data integration, accurate in environment description, scientific and advanced in model construction and reliable and practical in prediction result.

Description

Technical Field

[0001] The present invention relates to the field of groundwater over-exploitation area prediction, and more specifically to a method for constructing a groundwater prediction model in a plain area and an application platform. Background Art

[0002] The development of groundwater dynamic prediction technology has gone through a process from qualitative to quantitative, from one-dimensional to three-dimensional, and from stable to unstable. In the early days, groundwater dynamic prediction mainly relied on analytical methods, such as Darcy's law, Joubert equation, and Theis formula. These methods are suitable for study areas with regular geometric shapes, simple conditions, and single boundaries. However, with the increasing complexity of groundwater systems, these methods have gradually revealed their limitations.

[0003] Since the mid-20th century, with the rapid development of computer technology, groundwater dynamic prediction models have gradually shifted to numerical simulation methods. Numerical simulation methods can more accurately simulate the dynamic changes of groundwater systems by constructing mathematical models of groundwater systems and solving them using computers. At the same time, groundwater dynamic prediction methods have gradually developed into two categories: deterministic models and stochastic models to meet the prediction needs under different conditions.

[0004] Although the groundwater dynamic prediction model has made significant progress in numerical simulation methods, it still has defects such as complex model construction and parameter setting, limitations of numerical simulation methods, practical application challenges, and technical and resource constraints. Therefore, how to provide a groundwater prediction model for plain areas and put it into application is urgently needed for technical personnel in this field. Summary of the invention

[0005] In view of this, the present invention provides a method for constructing a groundwater prediction model in plain areas and an application platform, which has efficient data integration, accurate environmental characterization, scientific and advanced model construction, and reliable and practical prediction results, thereby solving the problems existing in the background technology.

[0006] In order to achieve the above object, the present invention adopts the following technical solution:

[0007] A method for constructing a groundwater prediction model in a plain area comprises the following steps:

[0008] Collect and pre-process groundwater-related data to establish a multivariate database, which includes a basic information database, a model intermediate results database, a dynamic database, a prediction database, and a model results database;

[0009] Call the sample boreholes, surface DEM, hydrogeological profiles, and hydrogeological parameter maps in the multivariate database, use the borehole and profile data as the basis, use Shannon information theory to determine the information characteristics of the stratum and its environment, and use the information characteristics of the stratum and its environment to edit the three-dimensional scene;

[0010] Based on the neural network model and time series model, an initial groundwater prediction model is established. The three-dimensional scene depicts the initial groundwater prediction model in three dimensions to establish the initial groundwater model.

[0011] Optionally, it also includes fitting and verification of the groundwater prediction model, specifically: using long-hole observation and flow field data, through groundwater level process lines and groundwater flow field fitting correction, identifying hydrogeological parameters and inverting source and sink items; delineating groundwater equilibrium zones, and determining the equilibrium items of the regional groundwater system; establishing prediction models according to different schemes, and predicting the changing trends of groundwater flow fields under different schemes.

[0012] Optionally, the three-dimensional characterization of the initial groundwater prediction model is specifically as follows: using the borehole code as a keyword, establishing the relationship between various data tables, using the keyword as a link, calling data from various data tables to form a physical geological model.

[0013] Optionally, groundwater-related data include hydrogeological condition boundaries, initial conditions, and groundwater system source and sink items.

[0014] Optionally, the three-dimensional scene is edited using information features of the stratum and its environment, specifically including the following steps:

[0015] Store and manage data representing three-dimensional scenes in a tree structure, and provide focus information for spatial objects;

[0016] Edit and analyze the stored 3D scene data, and provide registration tools to accurately place the 3D scene on the earth's surface data.

[0017] Reading drawing data according to the specified display area, searching the scene graph data for the specified spatial object on the three-dimensional scene, and reading the focus point information of the searched spatial object;

[0018] The read three-dimensional scene data is subjected to display processing.

[0019] Optionally, the stored three-dimensional scene data is edited and analyzed specifically including: basic data processing of the three-dimensional scene, data correction, data compilation, data stitching, data clipping, data overlay, annotation rendering, scene rendering, and scene integration.

[0020] A plain area groundwater prediction model application platform, using any one of the plain area groundwater prediction model construction methods, comprising:

[0021] Database module: used to collect data information including geospatial data, monitoring data, model data, basic data, thematic results data and metadata;

[0022] Support engine module: used to support business applications and manage data resources, provide data exchange and management, data management, and 3D display functions for groundwater models and 3D platforms, and provide support for efficient application of browsing, querying, analysis, and decision-making functions of the platform;

[0023] Groundwater prediction model module: used to predict the impact of different environmental factors on groundwater change trends;

[0024] Construction of groundwater three-dimensional platform: used for three-dimensional scene establishment and business application construction, including geological structure management, groundwater level management and groundwater simulation prediction.

[0025] It can be seen from the above technical solutions that, compared with the prior art, the present invention provides a method for constructing a groundwater prediction model in a plain area and an application platform, which has the following beneficial effects:

[0026] 1. Editing 3D scenes using information features of the strata and their environment enables accurate depiction of the groundwater environment. This 3D visualization method helps to intuitively understand the spatial distribution and dynamic changes of the groundwater system, providing an intuitive reference for model construction.

[0027] 2. The information characteristics of the strata and their environment are determined by Shannon information theory, providing a scientific basis for the construction of three-dimensional scenes. The extraction of such information characteristics helps to capture the key elements and characteristics of the groundwater system and improve the accuracy and reliability of the model.

[0028] 3. Establish the initial groundwater prediction model based on the neural network model and the time series model. These two models have significant advantages in dealing with complex nonlinear problems and time series data. Through multi-model fusion, the advantages of each model can be comprehensively utilized to improve the prediction accuracy and robustness of the model. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0030] Figure 1 is a flow chart of the method of the present invention;

[0031] Figure 2 Schematic diagram of the application platform of the present invention. DETAILED DESCRIPTION

[0032] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0033] The embodiment of the present invention discloses a method for constructing a groundwater prediction model in a plain area. Figure 1 As shown, the following steps are included:

[0034] Step 1: Collect and pre-process groundwater-related data to establish a multivariate database, which includes a basic information database, a model intermediate results database, a dynamic database, a prediction database, and a model results database;

[0035] Step 2: Call the sample boreholes, surface DEM, hydrogeological profiles, and hydrogeological parameter maps in the multivariate database, use the borehole and profile data as the basis, use Shannon information theory to determine the information characteristics of the stratum and its environment, and use the information characteristics of the stratum and its environment to edit the three-dimensional scene;

[0036] Step 3: Establish an initial groundwater prediction model based on the neural network model and the time series model. Use the three-dimensional scene to depict the initial groundwater prediction model in three dimensions and establish the initial groundwater model.

[0037] Furthermore, in step one, the groundwater-related data include hydrogeological condition boundaries, initial conditions, and groundwater system source and sink items.

[0038] Among them, the model basic information database mainly includes basic information of administrative regions, basic information of river basins, water level observation stations, rainwater monitoring stations, hydrogeological overview, groundwater storage overview, groundwater exploitation overview, etc.

[0039] The model intermediate results library refers to the processed raw data, which is used to establish the groundwater model but is not the final result, that is, various information of the model discretized to the grid points. It mainly includes the top and bottom plate information of the model aquifer group, hydrogeological parameter information, boundary condition information, equilibrium partition information, etc. Among them, the hydrogeological parameter information includes the permeability coefficient of each aquifer group, the water supply degree of the phreatic aquifer group, the water storage coefficient of the confined aquifer group, the rainfall infiltration coefficient of the surface, etc. The boundary condition information includes the lateral recharge boundary and the universal head boundary.

[0040] The dynamic database refers to the data that changes continuously during the process of groundwater numerical simulation. It mainly includes two parts: one is the dynamic information of groundwater or surface water level; the other is the dynamic information of source and sink items. The dynamic information of source and sink items includes information such as groundwater exploitation, boundary lateral infiltration, river infiltration, evapotranspiration, rainfall, etc.

[0041] The prediction database mainly stores data related to model prediction, including basic information of the prediction scheme, specified simulation period, and various source and sink data within the simulation prediction period.

[0042] The model results library mainly stores model results, including the calculation results of each input item in the model calculation process, the water level, drawdown, equilibrium data, model fitting data, etc. generated after the model calculation is completed.

[0043] The special results database is the data analyzed and managed based on the data obtained from investigation, monitoring and exploration, mainly including the groundwater resources evaluation database.

[0044] Define and describe the data items, data structures, etc. of the database data, and provide detailed explanations for each element in the database.

[0045] Furthermore, in step three, the three-dimensional characterization of the initial groundwater prediction model is specifically as follows: using the borehole code as a keyword, establishing the relationship between various data tables, using the keyword as a link, calling data from various data tables to form a physical geological model.

[0046] More specifically, the inner boundary of the borehole distribution is used as the calculation area, that is, the closed area surrounded by the outermost borehole connection line in the borehole distribution map. In the calculation area, a specific algorithm is used to calculate and construct the three-dimensional line file 3DS and the three-dimensional elevation model file DTM to form a surface model and a stratum model.

[0047] Surface generation: The main data for generating the surface is the surface discrete point data. The plane segments that make up the contour lines correspond to their elevation values ​​one by one, so there is a three-dimensional coordinate set of the surface. When the system is input, the points that make up the contour lines, that is, the points of each line segment, are read in. Combined with the elevation values ​​corresponding to the line segments, the discrete points required for generating the model are formed. The system uses these discrete points for interpolation to generate the surface.

[0048] Loading surface information: Geographic information data: Geographic information data includes point, line and surface information of rivers, railways, roads, lakes and cities. This information plays a positive role in enhancing the connotation and display effect of 3D visualization models. After each geographic information data is read in, 3D display technology is used to realize the display status of these graphics in the 3D model, and the corresponding attribute information content is given.

[0049] Remote sensing imagery: After the surface grid is divided, the system uses 3D simulation texture mapping technology to paste the satellite image onto the surface, which makes the entire model look more realistic and has better effects. In order to reduce the amount of memory occupied, the satellite image is converted into a bitmap image format (.BMP) and read in while ensuring the visual resolution.

[0050] In order to make the remote sensing image fit well with the generated surface grid and position more accurately, the system uses the three-point coordinate positioning method to calibrate the image. The calibration method is to select three points from the image, ensure that the three points are not on the same straight line, and record them as P1, P2 and P3 respectively. Users need to know the true coordinate values ​​of these three points, so they can select those iconic geographical locations.

[0051] Top or bottom surface model: First, interpolate the data points on the upper surface of each rock layer to form a line file related to these points, and then use this line file to make a three-dimensional elevation model of the top layer of the rock mass; the method for making a three-dimensional elevation model of the bottom surface of the rock mass is the same as that of the top surface.

[0052] Single-layer model: The lithology of the coded boreholes is generalized and the three-dimensional models of the top surface, bottom surface and boundary are seamlessly connected to form a solid model of the rock layer.

[0053] Multi-layer model: All rock layers are combined to form a solid geological model. This model can combine strata as needed, or multiple layers can be combined.

[0054] Furthermore, the three-dimensional scene is edited using the information characteristics of the stratum and its environment, which specifically includes the following steps:

[0055] Step 3.1, storing and managing data representing a three-dimensional scene using a tree structure, and providing focus information for spatial objects;

[0056] Step 3.2: Edit and analyze the stored three-dimensional scene data, and provide a registration tool to accurately place the three-dimensional scene on the earth's surface data.

[0057] Step 3.3, reading drawing data according to the specified display area, searching the scene graph data for the specified spatial object on the three-dimensional scene, and reading the focus point information of the searched spatial object;

[0058] Step 3.4: Perform display processing on the read three-dimensional scene data.

[0059] Furthermore, the stored three-dimensional scene data is edited and analyzed specifically including: basic data processing of the three-dimensional scene, data correction, data compilation, data splicing, data cropping, data overlay, annotation rendering, scene rendering, and scene integration.

[0060] Furthermore, the present embodiment also includes fitting and verification of the groundwater prediction model, specifically: using long observation holes and flow field data, through groundwater level process lines and groundwater flow field fitting correction, identifying hydrogeological parameters and inverting source and sink items; delineating groundwater equilibrium zones, and determining equilibrium items of the regional groundwater system; establishing prediction models according to different schemes, and predicting the changing trends of groundwater flow fields under different schemes.

[0061] In the present embodiment, BP neural network and time series analysis method are applied to the dynamic prediction of groundwater level in conjunction with MATLAB software, and two methods more accurately predict the changing trend of groundwater level, and time series method can also help people understand the changing trend and periodicity of groundwater level. Both methods are to historical data analysis, and no consideration is given to future mutation situations, such as the impact of human activities on groundwater. Therefore, short-term prediction accuracy is higher, but for long-term prediction, its error is susceptible to the influence of mutation terms, and the progressive accumulation of errors makes it possible for the final model prediction error to increase, and the latest data should be added to the sequence in time to establish a new model for rolling prediction.

[0062] According to the hydrogeological conceptual model, the heterogeneous anisotropic, three-dimensional structure, and unstable groundwater flow system can be described by the following equation:

[0063]

[0064] Where: Ω - seepage area; h - water level of aquifer (m); K x , K y , K z - horizontal and vertical permeability coefficients (m / d); ε - source and sink terms of the aquifer (1 / d); h0 - initial water level distribution of the aquifer (m); S - water storage rate of the aquifer (1 / m); K n - Permeability coefficient in the normal direction of the boundary surface (m / d); τ1 - lateral and lower boundaries of the seepage area; q - flow rate at the τ1 boundary, inflow is positive, outflow is negative, and the water-blocking boundary is 0 (m / d); n - normal direction of the boundary surface. Q - flow rate; K - permeability coefficient; H - external boundary head; H0 - model boundary head; a - single cell (grid) length; b - single cell (grid) width; L - seepage path length.

[0065] A groundwater prediction model application platform in plain areas, such as Figure 2 As shown, a method for constructing a groundwater prediction model in a plain area using any one of the above-mentioned methods includes:

[0066] Database module: used to collect data information including geospatial data, monitoring data, model data, basic data, thematic results data and metadata;

[0067] Support engine module: used to support business applications and manage data resources, provide data exchange and management, data management, and 3D display functions for groundwater models and 3D platforms, and provide support for efficient application of browsing, querying, analysis, and decision-making functions of the platform;

[0068] Groundwater prediction model module: used to predict the impact of different environmental factors on groundwater change trends;

[0069] Construction of groundwater three-dimensional platform: used for three-dimensional scene establishment and business application construction, including geological structure management, groundwater level management and groundwater simulation prediction.

[0070] Furthermore, in the process of building the groundwater 3D platform, the GIS 3D scene editing part is responsible for the data editing and generation of the 3D scene, and has basic spatial data display and spatial data processing functions, such as data acquisition, data organization, data manipulation, data analysis and data presentation, etc. Specifically, it includes scene data storage unit, scene data editing unit and scene data display unit.

[0071] Scene data storage unit, used to store and manage data representing three-dimensional scenes in tree structure. Including the focus information provided for spatial objects. File storage is based on files and databases. The database is based on a unified coordinate system and a hybrid modeling method. It uses object-oriented database OODB technology to organize and manage digital elevation models DEM, orthophotos DOM, digital line data DLG, three-dimensional city models 3DCM and attribute data on a relational database; it can also manage various basic and business data.

[0072] The scene data editing unit is used to edit and analyze the stored 3D scene data, and provides registration tools to accurately place the 3D scene on the earth's surface data. The editing unit is the core part of making 3D map scenes, and is responsible for basic data processing, data correction, data compilation, data splicing, data clipping, data overlay, annotation rendering, scene rendering, scene integration, etc. of the 3D scene.

[0073] The data reading unit is used to read the drawing data according to the specified display area, search the scene graph data for the spatial objects specified by the user on the three-dimensional scene displayed by the display processing unit, and read the focus point information of the searched spatial objects.

[0074] The scene data display unit performs display processing for the user based on the read three-dimensional scene data, specifies a display area with reference to the scene graph, and reads and displays the rendering data according to the specified display area.

[0075] Through the 3D GIS scene browsing part, you can call the established 3D GIS scene and develop the program according to the project requirements. There are two ways to browse the 3D GIS scene. One is the local browsing method; the other is the publishing method. The scene data is published on a server. By building a wide area network distributed deployment solution, the network connected to the server can access the scene data, and multi-user access is possible. At the same time, the server architecture can aggregate third-party services and provide rich server-side management functions such as data management, log management, traffic monitoring, and permission management, which facilitates the management of the browsed 3D scene data.

[0076] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0077] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for constructing a groundwater prediction model in a plain area, characterized in that: The following steps are involved: Collect and pre-process groundwater-related data to establish a multivariate database, which includes a basic information database, a model intermediate results database, a dynamic database, a prediction database, and a model results database; Call the sample boreholes, surface DEM, hydrogeological profiles, and hydrogeological parameter maps in the multivariate database, use the borehole and profile data as the basis, use Shannon information theory to determine the information characteristics of the stratum and its environment, and use the information characteristics of the stratum and its environment to edit the three-dimensional scene; Based on the neural network model and time series model, an initial groundwater prediction model is established. The three-dimensional scene depicts the initial groundwater prediction model in three dimensions to establish the initial groundwater model.

2. The method for constructing a groundwater prediction model in a plain area according to claim 1, characterized in that: It also includes fitting and verification of groundwater prediction models, specifically: using long-hole observation and flow field data, through groundwater level process lines and groundwater flow field fitting correction, identifying hydrogeological parameters and inverting source and sink items; delineating groundwater equilibrium zones, and determining the equilibrium items of the regional groundwater system; establishing prediction models according to different schemes, and predicting the changing trends of groundwater flow fields under different schemes.

3. The method for constructing a groundwater prediction model in a plain area according to claim 1, characterized in that: The three-dimensional characterization of the initial groundwater prediction model is as follows: using the borehole code as the keyword, establishing the relationship between various data tables, using the keyword as the connection, calling data from various data tables to form a physical geological model.

4. The method for constructing a groundwater prediction model in a plain area according to claim 1, characterized in that: Groundwater-related data include hydrogeological condition boundaries, initial conditions, and groundwater system source and sink items.

5. The method for constructing a groundwater prediction model in a plain area according to claim 1, characterized in that: The three-dimensional scene is edited using the information characteristics of the stratum and its environment, which specifically includes the following steps: Store and manage data representing three-dimensional scenes in a tree structure, and provide focus information for spatial objects; Edit and analyze the stored 3D scene data, and provide registration tools to accurately place the 3D scene on the earth's surface data. Reading drawing data according to the specified display area, searching the scene graph data for the specified spatial object on the three-dimensional scene, and reading the focus point information of the searched spatial object; The read three-dimensional scene data is subjected to display processing.

6. The method for constructing a groundwater prediction model in a plain area according to claim 5, characterized in that: The editing and analysis of the stored 3D scene data include: basic data processing of 3D scenes, data correction, data compilation, data splicing, data cropping, data overlay, annotation rendering, scene rendering, and scene integration.

7. A groundwater prediction model application platform in plain areas, characterized by: A method for constructing a groundwater prediction model in a plain area according to any one of claims 1 to 6, comprising: Database module: used to collect data information including geospatial data, monitoring data, model data, basic data, thematic results data and metadata; Support engine module: used to support business applications and manage data resources, provide data exchange and management, data management, and 3D display functions for groundwater models and 3D platforms, and provide support for efficient application of browsing, querying, analysis, and decision-making functions of the platform; Groundwater prediction model module: used to predict the impact of different environmental factors on groundwater change trends; Construction of groundwater three-dimensional platform: used for three-dimensional scene establishment and business application construction, including geological structure management, groundwater level management and groundwater simulation prediction.

Images