A natural resource global dynamic mapping and planning achievement auxiliary analysis system

Abstract

The natural resource global dynamic surveying and planning achievement auxiliary analysis system provided by the application relates to the field of big data processing, and more particularly to the field of systems or methods particularly suitable for specific commercial fields, and comprises: a first acquisition device and a second acquisition device for acquiring multi-point water level surveying data of a set planning area and three-dimensional dynamic surveying data in a current time segment; a planning issuing device for acquiring the current time segment underground water resource pressure mining volume issued by the natural resource and planning department; and an intelligent prediction device for intelligently predicting the next time interval water resource total amount range by using a deep neural network model of the set planning area. Through the application, the technical problems of the prior art, such as single basic data dimension and fixed prediction model structure, can be solved by dynamically matching the network structure and the planning area, and outputting the future time interval water resource total amount range.

Description

Technical Field

[0001] This invention relates to the field of big data processing, and more specifically to the field of systems or methods particularly applicable to specific business sectors, particularly to a system for dynamic mapping and planning effectiveness analysis of natural resources across the entire region. Background Technology

[0002] Deep learning involves massive amounts of learning. Artificial intelligence models obtained through deep learning have guaranteed stability and reliability in intelligent analysis and can be applied to various sub-fields, such as natural resource planning and water resource management. In particular, groundwater resource prediction and scheduling systems in natural resource planning and water resource management can provide guiding data for the operation mode and control strategy of groundwater resource management.

[0003] With the continuous advancement of digital governance in natural resource planning and algorithm optimization, the performance requirements for groundwater resource prediction systems are becoming increasingly stringent. These requirements are no longer limited to the accuracy and real-time nature of monitoring data. Instead, there is a desire to produce an intelligent prediction system that can integrate multi-source heterogeneous data and adapt to regional mining characteristics. This system should not only improve prediction accuracy but also effectively integrate the extraction constraints issued by natural resources and planning departments, outputting a range of water resource total volume predictions that can be used for decision-making, and directly support the dynamic formulation of scheduling, usage, and early warning strategies.

[0004] For example, Chinese invention patent publication CN116307191A proposes a water resource allocation method, apparatus, and equipment based on artificial intelligence algorithms. The method includes: inputting measured data of the target area into a surface water-groundwater coupled simulation model to generate the available surface water and groundwater quantities under different climates; optimizing the BP neural network model using a particle swarm optimization algorithm, inputting the total water consumption of the target area and the direct surface water consumption coefficients and direct groundwater consumption coefficients of each sub-region into the optimized BP neural network model to generate the predicted water demand of each sub-region; inputting the available surface water quantity, available groundwater quantity, and predicted water demand of each sub-region as decision variables into a multi-objective programming model, so that the multi-objective programming model calculates several objective functions according to constraints to generate water resource optimization allocation results; thereby achieving water resource allocation optimization based on multiple objectives, rationally and accurately allocating water resources, and improving the utilization efficiency of water and soil resources.

[0005] For example, Chinese invention patent publication CN111160430A proposes a water resource optimization allocation method based on artificial intelligence algorithms. The method includes the following steps: using collected hydrological and meteorological data, land use types, DEM (Diagram of Earth's Earth Scale), and hydrogeological data as input data for surface water and groundwater simulation models; calibrating and validating relevant parameters using measured data; determining relevant model parameters; and constructing a surface water-groundwater coupled simulation model. This invention combines future climate scenario data to predict future regional water resource supply and demand, using this as input values ​​for a multi-objective programming model. The constructed multi-objective optimization model is solved using NSGA (Natural Science Basic Algorithm), ultimately yielding a reasonable water resource optimization allocation scheme under future climate change scenarios. This method has significant practical implications and applicability.

[0006] Therefore, it is evident that the various technical solutions for groundwater resource prediction in the existing technologies mentioned above are limited to static monitoring and fixed structure models. They do not address the controlled extraction volume of groundwater resources planned by natural resources and planning departments, nor do they possess an adaptive network structure that can dynamically correlate with extraction characteristics such as the number of groundwater extraction wells in the region. This makes it difficult to adaptively optimize the network structure of deep neural network models under the planning constraints of natural resources and planning departments. Deep neural network models have to use the same network structure in all regions to cope with different extraction intensities that may exist in different regions. Obviously, this network structure preparation strategy can only guarantee the basic use of the prediction model, but cannot guarantee its wide applicability and specificity. It cannot dynamically output prediction results for different regions with different total water resource ranges that can be used for decision-making under controlled extraction constraints.

[0007] In other words, the above technical solutions only involve single-dimensional prediction based on various static monitoring data and the solidification of fixed structure models. Although they can provide some valuable reference data for water resource management, the following technical problems still exist: 1. Existing monitoring methods are unable to simultaneously acquire high spatiotemporal resolution three-dimensional dynamic mapping data and distributed water level data, resulting in a single data dimension.

[0008] 2. Existing intelligent prediction models have a fixed structure and are not dynamically linked to mining characteristics such as the number of groundwater extraction wells in the region, resulting in insufficient applicability and specificity.

[0009] 3. Existing intelligent prediction methods do not effectively integrate the groundwater resource reduction targets for the designated planning area issued by the natural resources and planning departments, as well as multi-source regional correlation information, making it difficult to output intelligent prediction results of the total water resource range that can be used for decision-making. Summary of the Invention

[0010] To address the technical problems in existing technologies, this invention provides a dynamic mapping and planning effectiveness auxiliary analysis system for natural resources across the entire region. This system utilizes a deep learning-based artificial intelligence model—a deep neural network model corresponding to the designated planning area—designed with a customized structure under the planning constraints of natural resources and planning departments. Based on multi-source data including targeted selections of 3D dynamic mapping data, water level mapping data, extraction volume, and multiple regional correlation information, the system performs synchronous intelligent prediction of the maximum and minimum total water resources within the next time segment of the current time segment (which is considered a future time segment). The system then drives the dynamic formulation of scheduling, usage, and early warning strategies based on the synchronous intelligent prediction results. Furthermore, by dynamically linking the number of hidden layer neurons with the number of groundwater extraction wellheads, the network structure adaptively matches the regional extraction intensity, thereby significantly improving prediction accuracy and decision support capabilities.

[0011] According to a first aspect of the present invention, a system for dynamic mapping and planning effectiveness analysis of natural resources across the entire region is provided, the system comprising: The first data acquisition device is used to acquire three-dimensional dynamic mapping data of the designated planning area within the current time segment. The second data acquisition device is used to acquire water level mapping data corresponding to each time point at a uniform interval within the current time segment of the designated planning area. The planning equipment is used to obtain the natural resources of the designated planning area and the groundwater extraction volume of the designated planning area within the current time segment, as issued by the planning department. Area capture device, used to capture multiple sets of area association information for a defined planned area; An object construction device for constructing a deep neural network model corresponding to a defined planned area; The intelligent prediction device is connected to the first acquisition device, the second acquisition device, the planning and distribution device, the regional capture device, and the object construction device, respectively. It is used to use the deep neural network model corresponding to the set planning area to intelligently predict the maximum and minimum total water resources of the set planning area in the next time segment based on the occupation duration of each time segment, the predicted average rainfall of the set planning area in the next time segment of the current time segment, the three-dimensional dynamic mapping data of the set planning area in the current time segment, the water level mapping data corresponding to each time point at even intervals in the set planning area in the current time segment, the groundwater resource extraction volume of the set planning area in the current time segment, and multiple regional association information of the set planning area. Among them, the three-dimensional visual mapping data corresponding to each time point at a uniform interval within the current time segment of the set planning area are used as the three-dimensional dynamic mapping data of the set planning area within the current time segment. Among them, obtaining the compressed extraction volume of groundwater resources in the designated planning area within the current time segment issued by the natural resources and planning department of the designated planning area includes: the compressed extraction volume refers to the upper limit of the total amount of groundwater allowed to be extracted in the designated planning area within the current time segment by the natural resources and planning department of the designated planning area, and its unit is cubic meters, which is a planning constraint value. Among them, the water level mapping data of the planned area at each time point is the real-time water level data of each water level mapping point at a uniform interval within the planned area at that time point. Among them, capturing multiple sets of regional association information of the set planning area includes: the total land area, irrigated area, wetland area, number of groundwater extraction wells, number of permanent residents, total water resources of the set planning area at the beginning of the current time segment, forest coverage rate, type code value corresponding to soil type, and average rainfall of the set planning area in the current time segment, as multiple sets of regional association information of the set planning area. Specifically, the depth-of-field values, grayscale gradient values, red channel values, green channel values, blue channel values, and coordinate values ​​of each pixel in the real-time overhead image acquired by the unmanned aerial vehicle at each time point are used as the three-dimensional visual mapping data corresponding to the time point. The unmanned aerial vehicle is located directly above the center of the set planning area and uses a preset flight altitude when taking the overhead image. The deep neural network model corresponding to the planned area includes a single input layer, a single output layer, and multiple hidden layers located between the single input layer and the single output layer. Each hidden layer has the same number of neurons and uses the GELU function as the activation function. Among them, constructing a deep neural network model corresponding to the set planning area includes: constructing a deep neural network model corresponding to the set planning area, wherein the number of neurons in each hidden layer of the deep neural network model corresponding to the set planning area is the same, and the number of neurons in each hidden layer shows the same numerical change trend as the number of groundwater extraction wells in the set planning area. Among them, the network parameters of the deep neural network model corresponding to the planning area include: the weight vector and bias term of each neuron in each hidden layer, and the weight vector and bias term of each output neuron in the output layer. The network parameters of the deep neural network model corresponding to the designated planning area are obtained through pre-training in the following way: using historical time segment data of the designated planning area as training samples, using the maximum and minimum historical total water resources of each historical time segment as training labels, using the mean squared error loss function to measure the deviation between the model prediction value and the training labels, and iteratively updating the above network parameters through the backpropagation algorithm and gradient descent optimizer until the value of the mean squared error loss function converges to a preset threshold or the number of iterations reaches a preset upper limit, thus obtaining the trained deep neural network model corresponding to the designated planning area; The historical time segment data for the designated planning area includes the duration of each time segment, the predicted average rainfall for the designated planning area in the next time segment of the historical time segment, the three-dimensional dynamic mapping data of the designated planning area in the historical time segment, the water level mapping data corresponding to each time point at even intervals in the historical time segment of the designated planning area, the volume of groundwater resources extracted in the designated planning area in the historical time segment, and multiple regional correlation information of the designated planning area.

[0012] Furthermore, the system also includes: a strategy analysis device, connected to an intelligent prediction device, used to determine, based on the intelligent prediction results, the groundwater resource scheduling strategy, the groundwater resource usage strategy, and the groundwater resource early warning strategy for the planning area in the next time segment of the current time segment; Among them, the determination of the groundwater resource scheduling strategy, groundwater resource use strategy and groundwater resource early warning strategy for the planning area in the next time segment of the current time segment based on the intelligent prediction results includes: when the maximum value of the total water resources in the planning area in the next time segment of the current time segment is less than the preset total water resources threshold, the amount of water resources to be scheduled from the surrounding planning areas to the planning area in the next time segment of the current time segment is determined based on the difference between the two. Among them, the determination of the groundwater resource scheduling strategy, groundwater resource utilization strategy and groundwater resource early warning strategy for the planning area in the next time segment of the current time segment based on the intelligent prediction results includes: when the minimum total water resource value in the planning area in the next time segment of the current time segment is greater than the preset total water resource threshold, the amount of surplus water resources allowed to be used in the next time segment of the current time segment is determined based on the difference between the two. Among them, the determination of the groundwater resource scheduling strategy, groundwater resource utilization strategy and groundwater resource early warning strategy for the planning area in the next time segment based on the intelligent prediction results includes: when the maximum value of the total water resources in the planning area in the next time segment of the current time segment is less than the preset total water resources threshold, a groundwater resource shortage early warning signal for the next time segment of the current time segment is sent; otherwise, the sending of the groundwater resource shortage early warning signal for the next time segment of the current time segment is temporarily suspended.

[0013] Furthermore, the system also includes: a segmentation setting device, which is connected to the first acquisition device, the second acquisition device, and the planning and distribution device respectively, for determining the duration of each time segment; The segmentation setting device is connected to the first acquisition device, the second acquisition device, and the planning and distribution device, respectively, and is used to determine the duration of each time segment. The segmentation setting device sets up a user operation interface to determine the duration of each time segment based on user operations.

[0014] Furthermore, the system also includes: a cloud storage node connected to the planning and distribution equipment, used to store the groundwater extraction volume of each planning area in each time segment pre-configured by the natural resources and planning departments; Among them, the cloud storage node is connected to the planning and distribution equipment and is used to store the groundwater extraction volume of each planning area in each time segment pre-configured by the natural resources and planning department. The cloud storage node is connected to the planning and distribution equipment through a network transmission interface and is used to transmit the groundwater extraction volume of the set planning area in the current time segment issued by the natural resources and planning department to the planning and distribution equipment via the network.

[0015] Furthermore, the system also includes: a content storage node, connected to the intelligent prediction device, for receiving and storing the maximum and minimum total water resources of the planned area in the next time segment of the current time segment; Among them, the content storage node, connected to the intelligent prediction device, is used to receive and store the maximum and minimum total water resources of the planned area in the next time segment of the current time segment. The content storage node is set at the end of the natural resources and planning department.

[0016] Compared with the prior art, the present invention has at least the following five key inventive points: A deep neural network model with a customized structure for the designated planning area is used to intelligently predict the range of total water resources for the next period. Based on the intelligent prediction results, scheduling, usage, and early warning strategies are automatically generated. In detail, when the maximum value of the intelligently predicted total water resources is less than the preset total water resources threshold, the amount of water resources to be scheduled from the surrounding planning area is determined based on the difference between the two. When the minimum value of the intelligently predicted total water resources is greater than the preset total water resources threshold, the amount of surplus water resources that can be used is determined based on the difference between the two. This realizes the transformation of water resource management from static monitoring to dynamic intelligent prediction, and improves the scientificity and foresight of groundwater resource management.

[0017] The deep neural network model with a customized structure for the designated planning area has a number of neurons in each hidden layer that shows the same numerical trend as the number of groundwater extraction wells in the designated planning area. In addition, the number of time points at uniform intervals within the selected current time segment is proportional to the total area of ​​the designated planning area. The customized structure design of the deep neural network model ensures the stability and reliability of the intelligent prediction results.

[0018] To intelligently predict the total water resources range for the next period, multi-source heterogeneous data is used as the basic data. The basic data includes the duration of each time segment, the predicted average rainfall of the planned area in the next time segment of the current time segment, the three-dimensional dynamic mapping data of the planned area in the current time segment, the water level mapping data corresponding to each time point at even intervals in the planned area in the current time segment, the groundwater extraction volume of the planned area in the current time segment, and multiple regional correlation information of the planned area. The above-mentioned targeted selection of various basic data from multi-source heterogeneous data further ensures the stability and reliability of the intelligent prediction results.

[0019] The three-dimensional dynamic mapping data of the planned area within the current time segment consists of the depth values, grayscale gradient values, red channel values, green channel values, blue channel values, and coordinate values ​​of each pixel in the real-time aerial images acquired by the unmanned aerial vehicle at each time point. Multiple sets of regional correlation information of the planned area include the total land area, irrigated area, wetland area, number of groundwater extraction wells, number of permanent residents, total water resources at the start time, forest coverage rate, type code values ​​corresponding to soil type, and average rainfall. This completes the data structure design for the multi-source heterogeneous data representation of the selected basic data.

[0020] In a deep neural network model, the number of input neurons in a single input layer is equal to the total dimension of the input feature vector. The input feature vector is formed by concatenating data in a fixed order after normalization. Each hidden layer uses the GELU activation function for nonlinear transformation. A single output layer contains two output neurons that output the maximum and minimum total water resources respectively, thus ensuring the prediction effect of each prediction operation of the deep neural network model.

[0021] The technical effects of this invention are as follows: 1. By using unmanned aerial vehicles to capture multi-dimensional aerial images and collect water levels at multiple points simultaneously, the dimensionality and precision of spatiotemporal dynamic data are improved. The three-dimensional dynamic mapping data, which includes depth of field, grayscale gradient, multi-channel color and coordinate information, is integrated with distributed water level data and input into the prediction model to comprehensively reflect the spatiotemporal dynamic evolution characteristics of regional water resources.

[0022] 2. By linking the number of hidden layer neurons in the intelligent prediction model with the number of groundwater extraction wellheads, the network structure is adaptively matched with the regional mining intensity, enhancing the model's structural adaptability and prediction stability in different planning area scenarios.

[0023] 3. By integrating groundwater extraction reduction targets, regional correlation information, and dynamic mapping data issued by the natural resources and planning departments for the designated planning area, the system outputs a forecast of the total water resources volume range for the next period. This directly supports the dynamic formulation of scheduling, utilization, and early warning strategies, enhances the scientific and forward-looking nature of groundwater resource management, and contributes to the digital governance of natural resources planning. Attached Figure Description

[0024] Figure 1 This is a schematic diagram of the working scenario of a dynamic mapping and planning effectiveness auxiliary analysis system for natural resources across the entire region, according to the present invention.

[0025] Figure 2 The core module of the natural resource dynamic mapping and planning effectiveness auxiliary analysis system according to the present invention is a deep neural network model with a customized structure design for the planning area.

[0026] Figure 3 This is an internal structure diagram of a dynamic mapping and planning effectiveness auxiliary analysis system for natural resources across the entire region, as shown in the first embodiment of the present invention.

[0027] Figure 4 This is an internal structure diagram of a dynamic mapping and planning effectiveness auxiliary analysis system for natural resources across the entire region, as shown in the second embodiment of the present invention.

[0028] Figure 5This is an internal structure diagram of a dynamic mapping and planning effectiveness auxiliary analysis system for natural resources across the entire region, as shown in the third embodiment of the present invention.

[0029] Figure 6 This is an internal structure diagram of a dynamic mapping and planning effectiveness auxiliary analysis system for natural resources across the entire region, as shown in the fourth embodiment of the present invention.

[0030] Figure 7 This is an internal structure diagram of a dynamic mapping and planning effectiveness auxiliary analysis system for natural resources across the entire region, as shown in the fifth embodiment of the present invention. Detailed Implementation

[0031] The following will describe in detail the natural resource dynamic mapping and planning effectiveness auxiliary analysis system of the present invention through embodiments. The big data processing involved in the present invention belongs to the field of systems or methods that are particularly applicable to specific business fields.

[0032] The specific technical process of this invention is as follows: Technical Process A: For the current planning area, in order to perform synchronous intelligent prediction of the maximum and minimum total water resources in the next time segment of the current time segment, a deep neural network model with a customized structure for the current planning area is introduced. like Figure 1 As shown, a working scenario for predicting the total water resources range using a deep neural network model in the current planning area is presented. exist Figure 1 The text provides a deep neural network model for customizing structural designs for the currently defined planning area. Figure 1 The deep neural network model corresponding to the planning area in the text is actually a deep neural network model with different customized structure for different planning areas. The corresponding deep neural network model is designed based on the mining characteristics, such as the number of groundwater extraction wells in the currently planned area. Please describe in detail, such as Figure 2 As shown, the structural customization of the deep neural network model is mainly reflected in the following aspects: First: The deep neural network model is a feedforward fully connected deep neural network, which includes a single input layer, multiple hidden layers and a single output layer in sequence along the data flow direction, wherein the multiple hidden layers are connected in series between the single input layer and the single output layer. Second: The number of neurons in each hidden layer and the number of groundwater extraction wells in the planned area show the same numerical trend. That is, the dynamic linkage between the number of neurons in the hidden layer and the number of groundwater extraction wells is achieved through the first conversion formula H=k×N_well, where H is the number of neurons in each hidden layer, which is a positive integer greater than or equal to 6, N_well is the number of groundwater extraction wells in the planned area, which is a positive integer greater than or equal to 1, and k is a preset proportional coefficient. For example, when the number of groundwater extraction wells in the planned area is 10, the number of neurons in each hidden layer is 30; when the number of groundwater extraction wells in the planned area is 15, the number of neurons in each hidden layer is 45; when the number of groundwater extraction wells in the planned area is 20, the number of neurons in each hidden layer is 60; when the number of groundwater extraction wells in the planned area is 25, the number of neurons in each hidden layer is 75, and so on. Third: Each hidden layer uses the GELU function as the activation function. The GELU activation function is a non-linear mapping function that outputs the input value after multiplying the input value with the cumulative distribution function value of the input value under the standard normal distribution. That is, for any input value, the output value of the GELU activation function is equal to the input value multiplied by the cumulative probability value of the input value under the standard normal distribution with a mean of zero and a standard deviation of one. Fourth: During the pre-training process of the network parameters of the deep neural network model corresponding to the designated planning area, the historical time segment data of the designated planning area is used as training samples, and the maximum and minimum historical total water resources of each historical time segment are used as training labels. The mean squared error loss function is used to measure the deviation between the model prediction value and the training label. The network parameters are iteratively updated through the backpropagation algorithm and the gradient descent optimizer until the value of the mean squared error loss function converges to a preset threshold or the number of iterations reaches a preset upper limit. The trained deep neural network model corresponding to the designated planning area is obtained, thereby ensuring the training effect of each training session. In this way, through the customized structural design of the above aspects, deep learning artificial neural network models with different structures were designed for different planned areas, thereby ensuring the stability and reliability of synchronous intelligent prediction results. Technical Process B: To perform synchronous intelligent prediction of the maximum and minimum total water resources in the planned area within the next time segment of the current time segment, multi-source data involving 3D dynamic mapping data, water level mapping data, pressure extraction volume, and multiple regional correlation information are introduced, such as... Figure 1 As shown; In detail, Figure 1The multi-source data includes three-dimensional dynamic mapping data, namely, three-dimensional dynamic mapping data of the planned area in the current time segment; water level mapping data, namely, water level mapping data corresponding to each time point at a uniform interval in the planned area in the current time segment; planned extraction volume, namely, the extraction volume of groundwater resources in the planned area in the current time segment; and other auxiliary data, namely, the occupation duration of each time segment, the predicted average rainfall of the planned area in the next time segment of the current time segment, and multiple regional correlation information of the planned area. For example, the number of time points at uniform intervals within the selected current time segment is proportional to the total area of ​​the planned area. This is achieved through the second conversion formula N_t=⌈c×S_area⌉, where N_t is the number of time points at uniform intervals within the current time segment, which is a positive integer greater than or equal to 10; S_area is the total area of ​​the planned area, in square kilometers; c is a preset proportional coefficient, c=0.5 time points / square kilometer; and ⌈ ⌉ is the round-up sign, indicating that the calculation result is rounded up to the nearest positive integer. For example, when the total area of ​​the planned area is 20 square kilometers, the number of time points with even intervals selected within the current time segment is 10; when the total area of ​​the planned area is 40 square kilometers, the number of time points with even intervals selected within the current time segment is 20; when the total area of ​​the planned area is 60 square kilometers, the number of time points with even intervals selected within the current time segment is 30; when the total area of ​​the planned area is 80 square kilometers, the number of time points with even intervals selected within the current time segment is 40, and so on. In a more specific description, the 3D visual mapping data of the planned area at each time point consists of the depth values, grayscale gradient values, red channel values, green channel values, blue channel values, and coordinate values ​​of each pixel in the real-time overhead image obtained by the unmanned aerial vehicle at that time point. The unmanned aerial vehicle is located directly above the center of the planned area and uses a preset flight altitude when taking the overhead image. The water level mapping data of the planned area at each time point consists of the real-time water level data corresponding to each water level mapping point at a uniform interval within the planned area at that time point. The multiple regional association information of the planned area includes the total area of ​​the planned area, irrigated area, wetland area, number of groundwater extraction wells, number of permanent residents, total water resources of the planned area at the beginning of the current time segment, forest coverage rate, type code value corresponding to soil type, and average rainfall of the planned area within the current time segment. This provides a customized data structure for multi-source data used for synchronous intelligent prediction. In this way, the targeted selection of multi-source data further ensures the stability and reliability of the synchronous intelligent prediction results; Technical Process C: Using a deep neural network model with a customized structure for the current planning area based on Technical Process A, and using multi-source data specifically selected based on Technical Process B, synchronous intelligent prediction of the maximum and minimum total water resources for the planning area in the next time segment of the current time segment is completed. Figure 1 As shown; In detail, the next time segment of the current time segment is the future time segment relative to the current time segment; Technical Process D: Based on the synchronous intelligent prediction results of Technical Process C, drive the dynamic formulation of scheduling, usage and early warning strategies; Specifically, when the maximum total water resources in the planned area predicted by the intelligent system are less than a preset total water resources threshold in the next time segment, the amount of water resources to be allocated from surrounding planned areas to the planned area in the next time segment is determined based on the difference between the two, and a groundwater shortage warning signal is sent for the next time segment. When the minimum total water resources in the planned area predicted by the intelligent system are greater than the preset total water resources threshold in the next time segment, the amount of surplus water resources allowed to be used in the next time segment is determined based on the difference between the two; otherwise, the groundwater shortage warning signal for the next time segment is temporarily suspended, thus completing the dynamic formulation of future time segment water resources management strategies based on synchronous intelligent prediction. Therefore, through the coordinated operation of the above-mentioned technical processes, a deep neural network model with different customized structural designs for different planned areas can be introduced. Based on multi-source data involving three-dimensional dynamic mapping data, water level mapping data, extraction volume, and multiple regional correlation information, the maximum and minimum total water resources in the current planned area are simultaneously predicted. The results of the simultaneous intelligent prediction drive the dynamic formulation of scheduling, usage, and early warning strategies. Thus, while ensuring the accuracy of prediction, it effectively integrates the groundwater extraction targets for the planned areas issued by the natural resources and planning departments, greatly improving the scientific nature and foresight of groundwater resource management.

[0033] The key points of this invention are: the dynamic formulation of water resource scheduling, usage and early warning strategies based on synchronous intelligent prediction results; the design of deep neural network models with different customized structures for different planned areas; the targeted selection of multi-source data involving three-dimensional dynamic mapping data, water level mapping data, pressure extraction volume and multiple regional correlation information; the specific design of each training of the deep neural network model; and the dynamic linkage design of the number of hidden layer neurons and the number of groundwater extraction wellheads.

[0034] Figure 2 The core module of the natural resource dynamic mapping and planning effectiveness auxiliary analysis system according to the present invention is a deep neural network model with a customized structure design for the planning area.

[0035] In this invention, deep neural network models with different customized structures can be designed for different planning areas. For example, a deep neural network model with a customized structure can be designed for a specific planning area. The following will describe in detail the various technical solutions of this invention and the deep neural network model with a customized structure designed for a specific planning area in conjunction with various embodiments of this invention.

[0036] First Embodiment Figure 3 This is an internal structure diagram of a dynamic mapping and planning effectiveness auxiliary analysis system for natural resources across the entire region, as shown in the first embodiment of the present invention.

[0037] like Figure 3 As shown, the natural resource dynamic mapping and planning effectiveness auxiliary analysis system includes the following components: The first data acquisition device is used to acquire three-dimensional dynamic mapping data of the designated planning area within the current time segment. In specific terms, the 3D dynamic mapping data for different planned areas may be different. The spatiotemporal resolution of the 3D dynamic mapping data is dynamically matched with the total area of ​​different areas. This can ensure the accuracy of the monitoring data while avoiding the waste of a lot of storage and computing resources due to the collection of too many data points. Specifically, the first data acquisition device controls an unmanned aerial vehicle to take a picture from above the center of the planned area at a preset flight altitude, and acquires the depth value, grayscale gradient value, red channel value, green channel value, blue channel value and coordinate value of each pixel in the real-time overhead image. After being unfolded in time sequence, it forms three-dimensional dynamic mapping data. To be more specific, the values ​​of the red channel, green channel, and blue channel are all between 0 and 255. The coordinate values ​​include horizontal and vertical coordinate values. The grayscale gradient value of each pixel is the standard deviation of the grayscale values ​​of the pixels surrounding that pixel. For example, different unmanned aerial vehicles can be used to take aerial photos of the planned area at different times, and the aerial photos taken at each time point need to be taken simultaneously. The second data acquisition device is used to acquire water level mapping data corresponding to each time point at a uniform interval within the current time segment of the designated planning area. Generally, the spatial distribution of water level monitoring points at uniform intervals within a planning area is relatively fixed, while the real-time water level data corresponding to each water level monitoring point at each time point changes dynamically over time. The planning equipment is used to obtain the natural resources of the designated planning area and the groundwater extraction volume of the designated planning area within the current time segment, as issued by the planning department. Specifically, the natural resources and planning department pre-configures the groundwater extraction volume of each planning area in each time segment and stores it through cloud storage nodes. The planning distribution equipment connects to the cloud storage nodes through a network transmission interface and transmits the groundwater extraction volume of the set planning area in the current time segment issued by the natural resources and planning department to the planning distribution equipment via the network. Area capture device, used to capture multiple sets of area association information for a defined planned area; Specifically, the area capture device will set the total area of ​​the planned area, the irrigated area, the wetland area, the number of groundwater extraction wells, the number of permanent residents, the total water resources of the planned area at the beginning of the current time segment, the forest coverage rate, the type code value corresponding to the soil type, and the average rainfall of the planned area in the current time segment as multiple sets of area association information for the planned area. An object construction device for constructing a deep neural network model corresponding to a defined planned area; In detail, the deep neural network model corresponding to the planned area built by the object construction device can be regarded as performing a structural customization design on the deep neural network model, thereby ensuring the stability and reliability of the transformation from input content to output content of the obtained deep neural network model; The intelligent prediction device is connected to the first acquisition device, the second acquisition device, the planning and distribution device, the regional capture device, and the object construction device, respectively. It is used to use the deep neural network model corresponding to the set planning area to intelligently predict the maximum and minimum total water resources of the set planning area in the next time segment based on the occupation duration of each time segment, the predicted average rainfall of the set planning area in the next time segment of the current time segment, the three-dimensional dynamic mapping data of the set planning area in the current time segment, the water level mapping data corresponding to each time point at even intervals in the set planning area in the current time segment, the groundwater resource extraction volume of the set planning area in the current time segment, and multiple regional association information of the set planning area. For example, if the current time segment is from 8:00 AM to 12:00 PM on August 1, 2026, and the next time segment is from 12:00 PM to 4:00 PM on August 1, 2026, the intelligent prediction device predicts the maximum and minimum total water resources of the planned area during the period from 12:00 PM to 4:00 PM on August 1, 2026. For example, the number of time points at uniform intervals within the selected current time segment is proportional to the total area of ​​the planned area. This is achieved through the second conversion formula N_t=⌈c×S_area⌉, where N_t is the number of time points at uniform intervals within the current time segment, which is a positive integer greater than or equal to 10; S_area is the total area of ​​the planned area, in square kilometers; c is a preset proportional coefficient, c=0.5 time points / square kilometer; and ⌈ ⌉ is the round-up sign, indicating that the calculation result is rounded up to the nearest positive integer. To further illustrate, the number of time points at even intervals within the selected current time segment is proportional to the total area of ​​the planned area, including: when the total area of ​​the planned area is 50 square kilometers, the number of time points at even intervals within the selected current time segment is 25; when the total area of ​​the planned area is 100 square kilometers, the number of time points at even intervals within the selected current time segment is 50; when the total area of ​​the planned area is 150 square kilometers, the number of time points at even intervals within the selected current time segment is 75, and so on. The deep neural network model corresponding to the planning area is a feedforward fully connected deep neural network, which includes a single input layer, multiple hidden layers and a single output layer along the data flow direction. The multiple hidden layers are connected in series between the single input layer and the single output layer. In this context, the number of input neurons in a single input layer of the deep neural network model corresponding to the designated planning area is equal to the total dimension of the input feature vector. The input feature vector is composed of the following items after normalization and concatenation in a fixed order: the occupation duration of the current time segment, the predicted average rainfall of the designated planning area in the next time segment of the current time segment, the three-dimensional dynamic mapping data of the designated planning area in the current time segment, the water level mapping data corresponding to each time point at even intervals in the current time segment of the designated planning area, the groundwater resource extraction volume of the designated planning area in the current time segment, and multiple regional association information of the designated planning area. Each hidden layer processes its received input data as follows: First, the weight vector corresponding to each neuron in the hidden layer is multiplied by the input data, and the result of the multiplication is added to the bias term corresponding to the neuron to obtain the weighted sum of the neuron. Then, the weighted sum is input to the GELU activation function to obtain the activation output value of the neuron. The activation output values ​​of all neurons in the hidden layer constitute the output feature vector of the hidden layer, and are passed as the input data of the next layer to the next hidden layer or output layer. The deep neural network model corresponding to the planning area contains two output neurons in a single output layer. These neurons are used to output the maximum and minimum total water resources of the planning area in the next time segment of the current time segment. The output layer performs a linear transformation on the input data it receives. That is, it performs a vector dot product operation between the input data and the weight vectors corresponding to each output neuron in the output layer. The dot product result is then added to the bias terms corresponding to each output neuron, and the maximum and minimum total water resources are directly output. The numerical trend of the number of neurons in each hidden layer and the number of groundwater extraction wells in the planned area is as follows: The first conversion formula reflecting the numerical trend is: H=k×N_well, where H is the number of neurons in each hidden layer, which is a positive integer greater than or equal to 6, N_well is the number of groundwater extraction wells in the planned area, which is a positive integer greater than or equal to 1, and k is a preset proportional coefficient. The network parameters of the deep neural network model corresponding to the designated planning area are obtained through pre-training in the following way: using historical time segment data of the designated planning area as training samples, using the maximum and minimum historical total water resources of each historical time segment as training labels, using the mean squared error loss function to measure the deviation between the model prediction value and the training labels, and iteratively updating the above network parameters through the backpropagation algorithm and gradient descent optimizer until the value of the mean squared error loss function converges to a preset threshold or the number of iterations reaches a preset upper limit, thus obtaining the trained deep neural network model corresponding to the designated planning area; Among them, the historical time segment data of the planning area includes the duration of each time segment, the predicted average rainfall of the planning area in the next time segment of the historical time segment, the three-dimensional dynamic mapping data of the planning area in the historical time segment, the water level mapping data corresponding to each time point at even intervals in the historical time segment of the planning area, the volume of groundwater resources extracted in the planning area in the historical time segment, and multiple regional correlation information of the planning area. Specifically, the depth-of-field values, grayscale gradient values, red channel values, green channel values, blue channel values, and coordinate values ​​of each pixel in the real-time overhead image acquired by the unmanned aerial vehicle at each time point are used as the three-dimensional visual mapping data corresponding to the time point. The unmanned aerial vehicle is located directly above the center of the set planning area and uses a preset flight altitude when taking the overhead image. Among them, obtaining the compressed extraction volume of groundwater resources in the designated planning area within the current time segment issued by the natural resources and planning department of the designated planning area includes: the compressed extraction volume refers to the upper limit of the total amount of groundwater allowed to be extracted in the designated planning area within the current time segment by the natural resources and planning department of the designated planning area, and its unit is cubic meters, which is a planning constraint value. Among them, the water level mapping data of the planned area at each time point is the real-time water level data of each water level mapping point at a uniform interval within the planned area at that time point. Among them, capturing multiple sets of regional association information of the set planning area includes: the total land area, irrigated area, wetland area, number of groundwater extraction wells, number of permanent residents, total water resources of the set planning area at the beginning of the current time segment, forest coverage rate, type code value corresponding to soil type, and average rainfall of the set planning area in the current time segment, as multiple sets of regional association information of the set planning area. Specifically, there are various soil types, such as loam, sandy soil, clay, and gravelly soil. Different soil types correspond to different type code values, all of which are represented in binary numerical form. For example, different binary values ​​can be used to represent the type-coded values ​​of different soil types; The duration of each time segment is determined by the segment setting device through the user operation interface based on user operation. For example, when setting up a segmented device user operation interface to determine the duration of each time segment based on the user's operation, the user can choose 4 hours or a longer time span. Example

[0038] Figure 4 This is an internal structure diagram of a dynamic mapping and planning effectiveness auxiliary analysis system for natural resources across the entire region, as shown in the second embodiment of the present invention.

[0039] like Figure 4 As shown, compared to Figure 3 The natural resource dynamic mapping and planning effectiveness auxiliary analysis system also includes: The strategy analysis device, connected to the intelligent prediction device, is used to determine the groundwater resource scheduling strategy, groundwater resource utilization strategy, and groundwater resource early warning strategy for the planning area in the next time segment based on the intelligent prediction results. For example, if the maximum total water resources in the planned area in the next time segment of the current time segment is 8 million cubic meters and the preset total water resources threshold is 10 million cubic meters, based on the difference of 2 million cubic meters between the two, it is determined that the amount of water resources to be dispatched from the surrounding planned areas to the planned area in the next time segment of the current time segment is 2 million cubic meters, and a groundwater shortage warning signal is sent for the next time segment of the current time segment. For example, if the minimum total water resources in the planned area of ​​the current time segment in the next time segment is 12 million cubic meters and the preset total water resources threshold is 10 million cubic meters, the amount of surplus water resources allowed to be used in the next time segment in the current time segment is determined to be 2 million cubic meters based on the difference of 2 million cubic meters between the two. For example, if the maximum total water resources in the planned area of ​​the current time segment is 11 million cubic meters and the preset total water resources threshold is 10 million cubic meters, the warning signal for insufficient groundwater resources in the next time segment of the current time segment will be temporarily suspended.

[0040] Third Embodiment Figure 5This is an internal structure diagram of a dynamic mapping and planning effectiveness auxiliary analysis system for natural resources across the entire region, as shown in the third embodiment of the present invention.

[0041] like Figure 5 As shown, compared to Figure 3 The natural resource dynamic mapping and planning effectiveness auxiliary analysis system also includes: The equipment is set up in segments and connected to the first acquisition device, the second acquisition device, and the planning and distribution device respectively, in order to determine the duration of each time segment. Among them, the segmented device is equipped with a user operation interface, which is used to determine the duration of each time segment based on user operation. For example, if a user selects 4 hours as the duration of a time segment through the user interface, then each time segment will be 4 hours long. Alternatively, a longer time span can be selected.

[0042] Fourth embodiment Figure 6 This is an internal structure diagram of a dynamic mapping and planning effectiveness auxiliary analysis system for natural resources across the entire region, as shown in the fourth embodiment of the present invention.

[0043] like Figure 6 As shown, compared to Figure 3 The natural resource dynamic mapping and planning effectiveness auxiliary analysis system also includes: The cloud storage node is connected to the planning and distribution equipment and is used to store the groundwater extraction volume of each planning area in each time segment pre-configured by the natural resources and planning departments. Among them, the cloud storage node is connected to the planning distribution equipment through the network transmission interface, and is used to transmit the compressed extraction volume of groundwater resources in the current time segment of the designated planning area issued by the natural resources and planning department to the planning distribution equipment. For example, blockchain service network elements or big data service network elements can be used to replace cloud storage nodes. Example

[0044] Figure 7 This is an internal structure diagram of a dynamic mapping and planning effectiveness auxiliary analysis system for natural resources across the entire region, as shown in the fifth embodiment of the present invention.

[0045] like Figure 7 As shown, compared to Figure 3 The natural resource dynamic mapping and planning effectiveness auxiliary analysis system also includes: Content storage nodes, connected to intelligent prediction devices, are used to receive and store the maximum and minimum total water resources of a set planning area in the next time segment of the current time segment. Specifically, the content storage node is set up at the natural resources and planning department to centrally store the maximum and minimum total water resources of the designated planning area in the next time segment of the current time segment. Among them, the content storage node is connected to the intelligent prediction device and is used to receive and store the maximum and minimum total water resources of the planned area in the next time segment of the current time segment. The content storage node is connected to the intelligent prediction device through a network transmission interface.

[0046] Next, various embodiments of the present invention will be further described.

[0047] Optionally, within the above embodiments, in the natural resource dynamic mapping and planning effectiveness auxiliary analysis system: Using a deep neural network model corresponding to a defined planning area, based on the occupation duration of each time segment, the predicted average rainfall of the defined planning area in the next time segment of the current time segment, the 3D dynamic mapping data of the defined planning area in the current time segment, the water level mapping data corresponding to each time point at even intervals in the defined planning area in the current time segment, the groundwater extraction volume of the defined planning area in the current time segment, and multiple regional association information of the defined planning area, the model intelligently predicts the maximum and minimum total water resources of the defined planning area in the next time segment of the current time segment. This includes inputting the occupation duration of each time segment, the predicted average rainfall of the defined planning area in the next time segment of the current time segment, the 3D dynamic mapping data of the defined planning area in the current time segment, the water level mapping data corresponding to each time point at even intervals in the defined planning area in the current time segment, the groundwater extraction volume of the defined planning area in the current time segment, and multiple regional association information of the defined planning area into the deep neural network model corresponding to the defined planning area in parallel. For example, the following can be described: the duration of each time segment, the predicted average rainfall of the planned area in the next time segment of the current time segment, the three-dimensional dynamic mapping data of the planned area in the current time segment, the water level mapping data corresponding to each time point at even intervals in the planned area in the current time segment, the groundwater extraction volume of the planned area in the current time segment, and multiple regional correlation information of the planned area are input in parallel into the deep neural network model corresponding to the planned area. The parallel input can be accomplished by using a programmable logic device. The method involves using a deep neural network model corresponding to a designated planning area. Based on the duration of each time segment, the predicted average rainfall of the designated planning area in the next time segment of the current time segment, the three-dimensional dynamic mapping data of the designated planning area in the current time segment, the water level mapping data corresponding to each time point at even intervals in the designated planning area in the current time segment, the groundwater extraction volume of the designated planning area in the current time segment, and multiple regional correlation information of the designated planning area, the method intelligently predicts the maximum and minimum total water resources of the designated planning area in the next time segment of the current time segment. This also includes running the deep neural network model corresponding to the designated planning area to obtain the maximum and minimum total water resources of the designated planning area in the next time segment of the current time segment output by the deep neural network model corresponding to the designated planning area.

[0048] Optionally, within the above embodiments, in the natural resource dynamic mapping and planning effectiveness auxiliary analysis system: Using a deep neural network model corresponding to the designated planning area, based on the occupation duration of each time segment, the predicted average rainfall of the designated planning area in the next time segment, the 3D dynamic mapping data of the designated planning area in the current time segment, the water level mapping data corresponding to each time point at even intervals in the designated planning area in the current time segment, the groundwater extraction volume of the designated planning area in the current time segment, and multiple regional correlation information of the designated planning area, the system intelligently predicts the maximum total water resources and water resources of the designated planning area in the next time segment. The total minimum value includes: the duration of each time segment, the predicted average rainfall of the planned area in the next time segment of the current time segment, the three-dimensional dynamic mapping data of the planned area in the current time segment, the water level mapping data corresponding to each time point at even intervals in the planned area in the current time segment, the volume of groundwater resources extracted in the planned area in the current time segment, multiple regional correlation information of the planned area, and the maximum and minimum total water resources of the planned area in the next time segment of the current time segment are all in a numerically normalized representation. For example, the duration of each time segment, the predicted average rainfall of the planned area in the next time segment of the current time segment, the three-dimensional dynamic mapping data of the planned area in the current time segment, the water level mapping data corresponding to each time point at even intervals in the planned area in the current time segment, the groundwater resource extraction volume of the planned area in the current time segment, multiple regional association information of the planned area, and the maximum and minimum total water resources of the planned area in the next time segment of the current time segment are all represented in a normalized numerical form, including: the normalized numerical form is a binary numerical form; The method involves using a deep neural network model corresponding to a defined planning area. Based on the duration of each time segment, the predicted average rainfall of the defined planning area in the next time segment, the three-dimensional dynamic mapping data of the defined planning area in the current time segment, the water level mapping data corresponding to each time point at even intervals in the defined planning area in the current time segment, the groundwater extraction volume of the defined planning area in the current time segment, and multiple regional correlation information of the defined planning area, the method intelligently predicts the maximum and minimum total water resources of the defined planning area in the next time segment. It also includes using the MATLAB toolbox to simulate and test the intelligent prediction of the maximum and minimum total water resources of the defined planning area in the next time segment.

[0049] And, optionally, within the above embodiments, in the natural resource dynamic mapping and planning effectiveness auxiliary analysis system: The number of neurons in each hidden layer and the number of groundwater extraction wells in the set planning area show the same numerical change trend, including: the first conversion formula reflecting the numerical change trend is: H=k×N_well, where H is the number of neurons in each hidden layer, which is a positive integer greater than or equal to 6, N_well is the number of groundwater extraction wells in the set planning area, which is a positive integer greater than or equal to 1, and k is a preset proportional coefficient. For example, you can choose the MATLAB toolbox to perform tests and simulations of the conversion process of the first conversion formula; The correlation between the number of time points at uniform intervals within the selected current time segment and the total area of ​​the planned area includes the following: The second conversion formula reflecting the correlation is: N_t = ⌈c × S_area⌉, where N_t is the number of time points at uniform intervals within the current time segment, which is a positive integer greater than or equal to 10, S_area is the total area of ​​the planned area in square kilometers, c is a preset proportional coefficient, c = 0.5 time points / square kilometer, and ⌈ ⌉ is the round-up sign, which indicates that the calculation result is rounded up to the nearest positive integer; For example, you can choose the MATLAB toolbox to perform tests and simulations of the conversion process of the second conversion formula.

[0050] Furthermore, in a natural resource dynamic mapping and planning effectiveness auxiliary analysis system according to the present invention: The fact that the number of neurons in each hidden layer and the number of groundwater extraction wells in the designated planning area show the same numerical trend also includes: using a first numerical change curve to represent the numerical change trend of the number of groundwater extraction wells in the designated planning area, and using a second numerical change curve to represent the numerical change trend of the number of neurons in each hidden layer. The fact that the number of neurons in each hidden layer and the number of groundwater extraction wells in the planned area show the same numerical variation trend also includes: normalizing the first and second numerical variation curves to obtain a first normalized curve and a second normalized curve, wherein the curvature values ​​of the first and second normalized curves are equal in length and evenly spaced.

[0051] The foregoing description of exemplary embodiments of the invention is provided for illustrative and descriptive purposes. It is not intended to be exhaustive or to limit the invention to the precise forms disclosed. Many modifications and variations will obviously be apparent to those skilled in the art. Exemplary embodiments were chosen and described in order to best illustrate the principles of the invention and its practical application, thereby enabling others skilled in the art to understand the various embodiments and variations of the invention suitable for the contemplated particular purpose. The scope of the invention is intended to be defined by the appended claims and their equivalents.

Claims

1. A system for dynamic mapping and planning effectiveness analysis of natural resources across the entire region, characterized in that, The system includes: The first data acquisition device is used to acquire three-dimensional dynamic mapping data of the designated planning area within the current time segment. The second data acquisition device is used to acquire water level mapping data corresponding to each time point at even intervals within the current time segment of the designated planning area. The planning equipment is used to obtain the natural resources of the designated planning area and the groundwater extraction volume of the designated planning area within the current time segment, as issued by the planning department. Area capture device, used to capture multiple sets of area association information for a defined planned area; An object construction device for constructing a deep neural network model corresponding to a defined planned area; The intelligent prediction device is connected to the first acquisition device, the second acquisition device, the planning and distribution device, the regional capture device, and the object construction device, respectively. It is used to use a deep neural network model corresponding to the set planning area to intelligently predict the maximum and minimum total water resources of the set planning area in the next time segment, based on the occupation duration of each time segment, the predicted average rainfall of the set planning area in the next time segment of the current time segment, the three-dimensional dynamic mapping data of the set planning area in the current time segment, the water level mapping data corresponding to each time point at even intervals in the set planning area in the current time segment, the groundwater resource extraction volume of the set planning area in the current time segment, and multiple regional correlation information of the set planning area.

2. The natural resource dynamic mapping and planning effectiveness auxiliary analysis system according to claim 1, characterized in that: Each set of three-dimensional visual mapping data corresponding to each time point at a uniform interval within the current time segment of the designated planning area is used as the three-dimensional dynamic mapping data of the designated planning area within the current time segment.

3. The system for dynamic mapping and planning effectiveness analysis of natural resources across the entire region according to claim 1, characterized in that: The reduction volume of groundwater resources in the designated planning area within the current time segment, as issued by the natural resources and planning department of the designated planning area, includes: the reduction volume refers to the upper limit of the total amount of groundwater allowed to be extracted in the designated planning area within the current time segment by the natural resources and planning department of the designated planning area, and its unit is cubic meters, which is a planning constraint value. Among them, the water level mapping data corresponding to each time point in the designated planning area are the real-time water level data corresponding to each water level mapping point at a uniform interval within the designated planning area at that time point.

4. The natural resource dynamic mapping and planning effectiveness auxiliary analysis system according to claim 3, characterized in that: Capturing multiple sets of regional association information for the designated planning area includes: the total land area, irrigated area, wetland area, number of groundwater extraction wells, number of permanent residents, total water resources, forest coverage, type code value corresponding to soil type, and average rainfall in the designated planning area during the current time segment. Specifically, the depth-of-field values, grayscale gradient values, red channel values, green channel values, blue channel values, and coordinate values ​​of each pixel in the real-time overhead image acquired by the unmanned aerial vehicle at each time point are used as the three-dimensional visual mapping data corresponding to the time point. The unmanned aerial vehicle is located directly above the center of the set planning area and uses a preset flight altitude when taking the overhead image.

5. The natural resource dynamic mapping and planning effectiveness auxiliary analysis system according to claim 4, characterized in that: The deep neural network model corresponding to the planning region includes a single input layer, a single output layer, and multiple hidden layers located between the single input layer and the single output layer. Each hidden layer has the same number of neurons and uses the GELU function as the activation function. The process of constructing a deep neural network model corresponding to the designated planning area includes: constructing a deep neural network model corresponding to the designated planning area, wherein the number of neurons in each hidden layer of the deep neural network model corresponding to the designated planning area is the same, and the number of neurons in each hidden layer shows the same numerical trend as the number of groundwater extraction wells in the designated planning area.

6. The system for dynamic mapping and planning effectiveness analysis of natural resources across the entire region according to claim 5, characterized in that: The network parameters of the deep neural network model corresponding to the planning area include: the weight vector and bias term of each neuron in each hidden layer, and the weight vector and bias term of each output neuron in the output layer. The network parameters of the deep neural network model corresponding to the designated planning area are obtained through pre-training in the following way: using historical time segment data of the designated planning area as training samples, using the maximum and minimum historical total water resources of each historical time segment as training labels, using the mean squared error loss function to measure the deviation between the model's predicted value and the training labels, and iteratively updating the above network parameters through the backpropagation algorithm and gradient descent optimizer until the value of the mean squared error loss function converges to a preset threshold or the number of iterations reaches a preset upper limit, thus obtaining the trained deep neural network model corresponding to the designated planning area.

7. The system for dynamic mapping and planning effectiveness analysis of natural resources across the entire region according to claim 6, characterized in that: The historical time segment data of the planning area includes the duration of each time segment, the predicted average rainfall of the planning area in the next time segment of the historical time segment, the three-dimensional dynamic mapping data of the planning area in the historical time segment, the water level mapping data corresponding to each time point at even intervals in the historical time segment of the planning area, the volume of groundwater resources under extraction in the planning area in the historical time segment, and multiple regional correlation information of the planning area.

8. The system for dynamic mapping and planning effectiveness analysis of natural resources across the entire region according to claim 7, characterized in that, The system also includes: The strategy analysis device, connected to the intelligent prediction device, is used to determine the groundwater resource scheduling strategy, groundwater resource utilization strategy, and groundwater resource early warning strategy for the planning area in the next time segment based on the intelligent prediction results. Among them, the determination of the groundwater resource scheduling strategy, groundwater resource use strategy and groundwater resource early warning strategy for the planning area in the next time segment of the current time segment based on the intelligent prediction results includes: when the maximum value of the total water resources in the planning area in the next time segment of the current time segment is less than the preset total water resources threshold, the amount of water resources to be scheduled from the surrounding planning areas to the planning area in the next time segment of the current time segment is determined based on the difference between the two. Among them, the determination of the groundwater resource scheduling strategy, groundwater resource utilization strategy and groundwater resource early warning strategy for the planning area in the next time segment of the current time segment based on the intelligent prediction results includes: when the minimum total water resource value in the planning area in the next time segment of the current time segment is greater than the preset total water resource threshold, the amount of surplus water resources allowed to be used in the next time segment of the current time segment is determined based on the difference between the two. Among them, the determination of the groundwater resource scheduling strategy, groundwater resource utilization strategy and groundwater resource early warning strategy for the planning area in the next time segment based on the intelligent prediction results includes: when the maximum value of the total water resources in the planning area in the next time segment of the current time segment is less than the preset total water resources threshold, a groundwater resource shortage early warning signal for the next time segment of the current time segment is sent; otherwise, the sending of the groundwater resource shortage early warning signal for the next time segment of the current time segment is temporarily suspended.

9. The system for dynamic mapping and planning effectiveness analysis of natural resources across the entire region according to claim 7, characterized in that, The system also includes: The equipment is set up in segments and connected to the first acquisition device, the second acquisition device, and the planning and distribution device respectively, in order to determine the duration of each time segment. The segmentation setting device is connected to the first acquisition device, the second acquisition device, and the planning and distribution device, respectively, and is used to determine the duration of each time segment. The segmentation setting device sets up a user operation interface to determine the duration of each time segment based on user operations.

10. The natural resource dynamic mapping and planning effectiveness auxiliary analysis system according to claim 7, characterized in that, The system also includes: The cloud storage node is connected to the planning and distribution equipment and is used to store the groundwater extraction volume of each planning area in each time segment pre-configured by the natural resources and planning departments. Among them, the cloud storage node is connected to the planning and distribution equipment and is used to store the groundwater extraction volume of each planning area in each time segment pre-configured by the natural resources and planning department. The cloud storage node is connected to the planning and distribution equipment through a network transmission interface and is used to transmit the groundwater extraction volume of the set planning area in the current time segment issued by the natural resources and planning department to the planning and distribution equipment via the network.

Images