Natural resource database construction method and system

Abstract

The invention relates to the technical field of database construction, in particular to a natural resource database construction method and system, and the method comprises the following steps: collecting multi-source natural resource data and geographic space data having geographic relevance with the multi-source natural resource data; establishing a resource analysis model by using a neural network through the multi-source natural resource data and the geographic space data to obtain a resource analysis model for obtaining a resource analysis result according to the multi-source natural resource data and the geographic space data; and combining the multi-source natural resource data, the geographic space data and the obtained resource analysis result to construct a natural resource database. According to the method, the resource analysis model is constructed, so that the resource analysis result is obtained according to the multi-source natural resource data and the geographic space data, a unified and objective analysis model is established, the data integration analysis efficiency and accuracy of the natural resource database are improved, and the effect of performing data application through the natural resource database is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of database construction, and in particular to a natural resource database construction method and system. Background Art

[0002] Natural resource data include land resources, forest resources, grassland resources, water resources, climate resources, mineral resources, marine resources, energy resources and other resources. The natural resource database is a collection of interconnected natural resource data organized, stored and used in a computer system according to a certain data model to meet the needs of users.

[0003] In the prior art, when constructing a natural resource database, the integration and analysis of multi-source resource data is performed manually, which is highly subjective and inefficient, making it difficult to ensure the credibility and efficiency of the natural resource database construction, and ultimately resulting in poor results in data application of the natural resource database. Summary of the invention

[0004] The purpose of the present invention is to provide a method for constructing a natural resource database to solve the technical problem that the integration and analysis of multi-source resource data in the prior art is carried out manually, which is highly subjective and inefficient, making it difficult to ensure the credibility and efficiency of the natural resource database construction, and ultimately resulting in poor results in data application of the natural resource database.

[0005] In order to solve the above technical problems, the present invention specifically provides the following technical solutions:

[0006] A method for constructing a natural resource database comprises the following steps:

[0007] Collect multi-source natural resource data and geospatial data that are geographically relevant to the multi-source natural resource data;

[0008] By using multi-source natural resource data and geospatial data and using a neural network, a resource analysis model is established to obtain a resource analysis model for obtaining resource analysis results based on the multi-source natural resource data and geospatial data;

[0009] Combine multi-source natural resource data and geospatial data, as well as the obtained resource analysis results, to construct a natural resource database;

[0010] Among them, the resource analysis model is dynamically updated and associated with the natural resource database.

[0011] As a preferred solution of the present invention, the multi-source natural resource data includes: water resource data, forest resource data, land resource data, mineral resource data, climate resource data, and tourism resource data.

[0012] As a preferred solution of the present invention, the geographic spatial data includes: spatial location data, ecological environment data, social economic data, population data, and remote sensing image data.

[0013] As a preferred solution of the present invention, the method for constructing the resource analysis model includes:

[0014] The neural network is trained with multi-source natural resource data and geospatial data as input and resource analysis results as output, thereby obtaining a resource analysis model that encapsulates the mapping relationship between multi-source natural resource data and geospatial data and resource analysis results;

[0015] The resource analysis model is:

[0016] Goal=BP(Resource_data,Space_data);

[0017] Where Goal is the result of resource analysis, Resource_data is the data vector representing multi-source natural resource data, Space_data is the data vector representing geographic space data, and BP is a neural network.

[0018] As a preferred solution of the present invention, the method for dynamically updating and associating the resource analysis model with the natural resource database includes:

[0019] The resource analysis model is trained with multi-source natural resource data as input and resource analysis results as output, so as to migrate the resource analysis model that encapsulates the mapping relationship between the multi-source natural resource data and the geospatial data and the resource analysis results into a first resource analysis migration model that encapsulates the mapping relationship between the multi-source natural resource data and the resource analysis results;

[0020] The first resource analysis migration model is:

[0021] Goal = BP (Resource_data);

[0022] In the formula, Goal is the result of resource analysis, Resource_data is the data vector representing multi-source natural resource data, and BP is a neural network;

[0023] The resource analysis model is trained with the geospatial data as input and the resource analysis results as output, so as to migrate the resource analysis model that encapsulates the mapping relationship between multi-source natural resource data and the geospatial data and the resource analysis results into a second resource analysis migration model that encapsulates the mapping relationship between the geospatial data and the resource analysis results;

[0024] The second resource analysis migration model is:

[0025] Goal = BP (Space_data);

[0026] In the formula, Goal is the result of resource analysis, Space_data is the data vector representing geographic spatial data, and BP is a neural network;

[0027] When the multi-source natural resource data in the natural resource database is updated, the resource analysis results in the natural resource database are updated using the output of the first resource analysis migration model;

[0028] When the geospatial data in the natural resource database is updated, the resource analysis results in the natural resource database are updated using the output of the second resource analysis migration model;

[0029] When the multi-source natural resource data and geographic spatial data in the natural resource database are updated, the resource analysis results in the natural resource database are updated using the output of the resource analysis model.

[0030] As a preferred solution of the present invention, the method for determining the update of multi-source natural resource data includes:

[0031] Calculate the similarity between the multi-source natural resource data at the current moment in the natural resource database and the multi-source natural resource data at the previous moment in the natural resource database, and mark the multi-source natural resource data at the current moment whose similarity is lower than a preset threshold as multi-source natural resource data update;

[0032] Methods for determining updates to geospatial data include:

[0033] The similarity between the current geospatial data in the natural resource database and the previous geospatial data in the natural resource database is calculated, and the current geospatial data with a similarity lower than a preset threshold is marked as geospatial data update.

[0034] As a preferred solution of the present invention, the function for calculating the similarity includes Euclidean distance, correlation coefficient, and kernel function.

[0035] As a preferred solution of the present invention, the loss function of the first resource analysis migration model during training is:

[0036] Loss1=||BP(Resource_data)-BP(Resource_data,Space_data)||1;

[0037] Wherein, Loss1 is the loss function of the first resource analysis migration model during training, BP(Space_data) is the output of the first resource analysis migration model, BP(Resource_data,Space_data) is the output of the resource analysis model, ||BP(Resource_data)-BP(Resource_data,Space_data)||1 is the L1 norm formula;

[0038] The loss function of the second resource analysis migration model during training is:

[0039] Loss1=||BP(Space_data)-BP(Resource_data,Space_data)||1;

[0040] Where Loss2 is the loss function of the second resource analysis migration model during training, BP(Space_data) is the output of the second resource analysis migration model, BP(Resource_data,Space_data) is the output of the resource analysis model, and ||BP(Resource_data)-BP(Resource_data,Space_data)||1 is the L1 norm formula.

[0041] As a preferred solution of the present invention, multi-source natural resource data and geographic space data, as well as the obtained resource analysis results are subjected to data normalization processing in the natural resource database.

[0042] As a preferred solution of the present invention, the present invention provides a natural resource database construction system, which is applied to the natural resource database construction method described above, and the system includes:

[0043] A data collection unit, used for collecting multi-source natural resource data and geospatial data having geographical relevance to the multi-source natural resource data;

[0044] A model building unit, used to establish a resource analysis model by using a neural network through multi-source natural resource data and geographic space data, and obtain a resource analysis model for obtaining resource analysis results based on the multi-source natural resource data and geographic space data;

[0045] A database construction unit, used to combine multi-source natural resource data and geospatial data, as well as obtained resource analysis results, to construct a natural resource database;

[0046] Among them, the resource analysis model is dynamically updated and associated with the natural resource database.

[0047] Compared with the prior art, the present invention has the following beneficial effects:

[0048] The present invention constructs a resource analysis model through a neural network, so that resource analysis results are obtained based on multi-source natural resource data and geographic spatial data, and a unified and objective analysis model is established, thereby improving the data integration and analysis efficiency and accuracy of the natural resource database, thereby improving the effect of data application through the natural resource database. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the implementation methods of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for the implementation methods or the description of the prior art. Obviously, the drawings in the following description are only exemplary, and for ordinary technicians in this field, other implementation drawings can be derived from the provided drawings without creative work.

[0050] Figure 1 A flow chart of a method for constructing a natural resource database provided by an embodiment of the present invention;

[0051] Figure 2 A block diagram of a natural resource database construction system provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0052] 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.

[0053] like Figure 1 As shown, the present invention provides a method for constructing a natural resource database, comprising the following steps:

[0054] Collect multi-source natural resource data and geospatial data that are geographically relevant to the multi-source natural resource data;

[0055] By using multi-source natural resource data and geospatial data and using a neural network, a resource analysis model is established to obtain a resource analysis model for obtaining resource analysis results based on the multi-source natural resource data and geospatial data;

[0056] Combine multi-source natural resource data and geospatial data, as well as the obtained resource analysis results, to construct a natural resource database;

[0057] Among them, the resource analysis model is dynamically updated and associated with the natural resource database.

[0058] The present invention constructs a resource analysis model, analyzes and integrates multi-source natural resource data and geographic spatial data, obtains resource analysis results, and combines the multi-source natural resource data, geographic spatial data and resource analysis results to construct a natural resource database, so that the natural resource database directly includes resource analysis data, and the data can be directly applied or interpreted, thereby realizing a natural resource database that improves data application effects, rather than a natural resource database that only serves as a single record.

[0059] The present invention adopts a modeled data analysis and integration method in constructing a natural resource database, which improves data processing efficiency. At the same time, a unified model is used for data analysis and integration, and the standardization of data processing is improved, thereby ensuring that the constructed natural resource database is more standardized and credible.

[0060] Furthermore, in order to obtain resource analysis results more efficiently when data in the natural resource database is updated, the present invention performs transfer learning on the resource analysis model so as to dynamically associate with the data update in the natural resource database. After the natural resource database is established, when only multi-source natural resource data exists in the natural resource database for updating, the geospatial data does not change. If multi-source natural resource data and geospatial data, or the resource analysis model, are used to obtain resource analysis results, the analysis and processing process of the geospatial data is redundant, which will only cause the computational burden of the resource analysis model and prolong the computational time. Therefore, the present invention migrates the resource analysis model of the mapping relationship between multi-source natural resource data and geospatial data and resource analysis results to a first resource analysis migration model that encapsulates the mapping relationship between geospatial data and resource analysis results, so as to update the resource analysis results in the natural resource database only according to the updated multi-source natural resource data, thereby improving the data analysis efficiency while updating the resource analysis results, thereby realizing rapid and efficient updating of the natural resource database.

[0061] Similarly, when the geospatial data in the natural resource database is updated, the multi-source natural resource data does not change. If the resource analysis results are obtained by using multi-source natural resource data and geospatial data, or by using a resource analysis model, the analysis and processing process of the multi-source natural resource data is redundant, which will only cause the computational burden of the resource analysis model and extend the computational time. Therefore, the present invention migrates the resource analysis model of the mapping relationship between multi-source natural resource data and geospatial data and resource analysis results to a second resource analysis migration model that encapsulates the mapping relationship between geospatial data and resource analysis results, so as to update the resource analysis results in the natural resource database only according to the updated geospatial data, thereby improving the data analysis efficiency while updating the resource analysis results, thereby realizing rapid and efficient updating of the natural resource database.

[0062] Moreover, when the present invention migrates and trains the resource analysis model into the first resource analysis migration model and the second resource analysis migration model, the difference between the output of the first resource analysis migration model and the second resource analysis migration model and the output of the resource analysis model is used as the loss function, so that the first resource analysis migration model and the second resource analysis migration model have the output performance of the resource analysis model, and can obtain the resource analysis results analyzed by multi-source natural resource data and geospatial data based on single-category data (multi-source natural resource data or geospatial data), that is, to reduce the amount of data processing and still obtain the same accurate analysis results. Therefore, the present invention can use the first resource analysis migration model and the second resource analysis migration model to accurately analyze resources, while achieving efficient analysis of resources and improving the updating efficiency of the natural resource database.

[0063] Resource analysis in the present invention may include multiple types of analysis, such as analysis of resource quantity and quality: mainly including the types, quantity, distribution of natural resources, and analysis of indicators such as resource quality, quality, purity, etc. Analysis of sustainable utilization capacity: mainly including analysis of indicators such as resource re-production capacity, renewable capacity, and scientific and technological tolerance to evaluate the sustainable utilization of resources. Analysis of economic value: mainly including analysis of indicators such as market value, commodity value, and productivity value of resources to evaluate the value and contribution of resources in economic development. Analysis of ecological environment: mainly including analysis of indicators such as ecological environmental protection of resources, ecosystem stability, and biodiversity to evaluate the ecological value and protection of resources. Analysis of social benefits: mainly including analysis of indicators such as social benefits of resources, quality of life of the people, and cultural value to evaluate the contribution and impact of resources on society.

[0064] Therefore, the results of resource analysis include the analysis results of resource quantity and quality, the analysis results of sustainable utilization capacity, the analysis results of economic value, the analysis results of the ecological environment, and the analysis results of social benefits, etc.

[0065] Multi-source natural resource data include: water resource data, forest resource data, land resource data, mineral resource data, climate resource data, and tourism resource data.

[0066] Land resources: including cultivated land, forest land, grassland, construction land, etc.

[0067] Forest resources: including natural forests and plantations, as well as indicators such as forest coverage and timber stock. Grassland resources: divided into natural pastures and artificial pastures, as well as other grasslands.

[0068] Water resources: including surface water and groundwater, as well as water quality and total water resources.

[0069] Climate resources: information related to climate conditions, meteorological disasters, etc.

[0070] Mineral resources: including metal minerals, non-metallic minerals and energy minerals, etc.

[0071] Marine resources: including marine biological resources, marine mineral resources and marine energy, etc.

[0072] Energy resources: including traditional energy sources such as coal, oil, and natural gas, as well as renewable energy sources such as solar energy and wind energy;

[0073] Tourism resources: including man-made tourism resources (human buildings) and natural tourism resources (unique mountains and waters).

[0074] Geographical data include: spatial location data, ecological environment data, socio-economic data, population data, and remote sensing image data.

[0075] The present invention constructs a resource analysis model, analyzes and integrates multi-source natural resource data and geographic spatial data, obtains resource analysis results, and combines the multi-source natural resource data, geographic spatial data and resource analysis results to construct a natural resource database, so that the natural resource database directly contains resource analysis data, and data application or interpretation can be directly performed, thereby realizing a natural resource database that improves the data application effect, rather than a natural resource database with a single record function, as follows:

[0076] The construction method of resource analysis model includes:

[0077] The neural network is trained with multi-source natural resource data and geospatial data as input and resource analysis results as output, thereby obtaining a resource analysis model that encapsulates the mapping relationship between multi-source natural resource data and geospatial data and resource analysis results;

[0078] The resource analysis model is:

[0079] Goal=BP(Resource_data,Space_data);

[0080] Where Goal is the result of resource analysis, Resource_data is the data vector representing multi-source natural resource data, Space_data is the data vector representing geographic space data, and BP is a neural network.

[0081] The resource analysis model loss function is the error between the output of the resource analysis model and the true value.

[0082] Methods for dynamically updating and associating resource analysis models with natural resource databases include:

[0083] The resource analysis model is trained with multi-source natural resource data as input and resource analysis results as output, so as to migrate the resource analysis model that encapsulates the mapping relationship between the multi-source natural resource data and the geospatial data and the resource analysis results into a first resource analysis migration model that encapsulates the mapping relationship between the multi-source natural resource data and the resource analysis results;

[0084] The first resource analysis migration model is:

[0085] Goal = BP (Resource_data);

[0086] In the formula, Goal is the result of resource analysis, Resource_data is the data vector representing multi-source natural resource data, and BP is a neural network;

[0087] The resource analysis model is trained with the geospatial data as input and the resource analysis results as output, so as to migrate the resource analysis model that encapsulates the mapping relationship between multi-source natural resource data and the geospatial data and the resource analysis results into a second resource analysis migration model that encapsulates the mapping relationship between the geospatial data and the resource analysis results;

[0088] The second resource analysis migration model is:

[0089] Goal = BP (Space_data);

[0090] In the formula, Goal is the result of resource analysis, Space_data is the data vector representing geographic spatial data, and BP is the neural network.

[0091] In order to obtain resource analysis results more efficiently when data in the natural resource database is updated, the present invention performs transfer learning on the resource analysis model to dynamically associate with the data update in the natural resource database. After the natural resource database is established, when only multi-source natural resource data exists in the natural resource database for update, the geographic spatial data does not change. If multi-source natural resource data and geographic spatial data, or the resource analysis model, are used to obtain resource analysis results, the analysis and processing process of the geospatial data is redundant, which will only cause the computational burden of the resource analysis model and prolong the computational time. Therefore, the present invention migrates the resource analysis model of the mapping relationship between multi-source natural resource data and geographic spatial data and resource analysis results to a first resource analysis migration model that encapsulates the mapping relationship between geographic spatial data and resource analysis results, so as to update the resource analysis results in the natural resource database only according to the updated multi-source natural resource data, thereby improving the data analysis efficiency while updating the resource analysis results, thereby realizing rapid and efficient updating of the natural resource database.

[0092] Similarly, when the geospatial data in the natural resource database is updated, the multi-source natural resource data does not change. If the resource analysis results are obtained by using multi-source natural resource data and geospatial data, or by using a resource analysis model, the analysis and processing process of the multi-source natural resource data is redundant, which will only cause the computational burden of the resource analysis model and extend the computational time. Therefore, the present invention migrates the resource analysis model of the mapping relationship between multi-source natural resource data and geospatial data and resource analysis results to a second resource analysis migration model that encapsulates the mapping relationship between geospatial data and resource analysis results, so as to update the resource analysis results in the natural resource database only according to the updated geospatial data, thereby improving the data analysis efficiency while updating the resource analysis results, thereby realizing rapid and efficient updating of the natural resource database.

[0093] When the multi-source natural resource data in the natural resource database is updated, the resource analysis results in the natural resource database are updated using the output of the first resource analysis migration model;

[0094] When the geospatial data in the natural resource database is updated, the resource analysis results in the natural resource database are updated using the output of the second resource analysis migration model;

[0095] When the multi-source natural resource data and geographic spatial data in the natural resource database are updated, the resource analysis results in the natural resource database are updated using the output of the resource analysis model.

[0096] When both multi-source natural resource data and geographic spatial data are changed or updated, the original resource analysis model is used to calculate their configuration parameters to ensure the accuracy of the resource analysis results.

[0097] Methods for determining updates to multi-source natural resource data include:

[0098] Calculate the similarity between the multi-source natural resource data at the current moment in the natural resource database and the multi-source natural resource data at the previous moment in the natural resource database, and mark the multi-source natural resource data at the current moment whose similarity is lower than a preset threshold as multi-source natural resource data update;

[0099] Methods for determining updates to geospatial data include:

[0100] The similarity between the current geospatial data in the natural resource database and the previous geospatial data in the natural resource database is calculated, and the current geospatial data with a similarity lower than a preset threshold is marked as geospatial data update.

[0101] Functions for similarity calculation include Euclidean distance, correlation coefficient, and kernel function.

[0102] The loss function of the first resource analysis migration model during training is:

[0103] Loss1=||BP(Resource_data)-BP(Resource_data,Space_data)||1;

[0104] Wherein, Loss1 is the loss function of the first resource analysis migration model during training, BP(Space_data) is the output of the first resource analysis migration model, BP(Resource_data,Space_data) is the output of the resource analysis model, ||BP(Resource_data)-BP(Resource_data,Space_data)||1 is the L1 norm formula;

[0105] The loss function of the second resource analysis migration model during training is:

[0106] Loss1=||BP(Space_data)-BP(Resource_data,Space_data)||1;

[0107] Where Loss2 is the loss function of the second resource analysis migration model during training, BP(Space_data) is the output of the second resource analysis migration model, BP(Resource_data,Space_data) is the output of the resource analysis model, and ||BP(Resource_data)-BP(Resource_data,Space_data)||1 is the L1 norm formula.

[0108] Moreover, when the present invention migrates and trains the resource analysis model into the first resource analysis migration model and the second resource analysis migration model, the difference between the output of the first resource analysis migration model and the second resource analysis migration model and the output of the resource analysis model is used as the loss function, so that the first resource analysis migration model and the second resource analysis migration model have the output performance of the resource analysis model, and can obtain the resource analysis results analyzed by multi-source natural resource data and geospatial data based on single-category data (multi-source natural resource data or geospatial data), that is, to reduce the amount of data processing and still obtain the same accurate analysis results. Therefore, the present invention can use the first resource analysis migration model and the second resource analysis migration model to accurately analyze resources, while achieving efficient analysis of resources and improving the updating efficiency of the natural resource database.

[0109] In the natural resource database, multi-source natural resource data, geospatial data, and resource analysis results are normalized.

[0110] The neural network can choose BP neural network. BP (backpropagation) neural network is a concept proposed by scientists led by Rumelhart and McClelland. It is a multi-layer feedforward neural network trained according to the error back propagation algorithm and is one of the most widely used neural network models.

[0111] BP neural network has the ability to classify arbitrarily complex patterns and excellent multi-dimensional function mapping capabilities. It can learn certain rules only through its own training and obtain the result closest to the expected output value when the input value is given, that is, it determines the mapping relationship between input and output. After repeated learning and training, it can process the input information of similar samples and obtain the output information with the smallest output error through nonlinear transformation.

[0112] Utilizing the function mapping properties of the BP neural network, the present invention applies the BP neural network to the resource analysis process, trains the BP neural network to learn the mapping relationship between multi-source natural resource data and geographic spatial data, and resource analysis results, thereby being able to obtain resource analysis results based on the multi-source natural resource data and geographic spatial data.

[0113] Transfer learning is a machine learning method that uses the model developed for task A as a starting point and reuses it in the process of developing a model for task B. By using a model suitable for different but related tasks, the search range of possible models is narrowed in an advantageous way. Therefore, the present invention uses it to determine the resource analysis results and reduce the amount of data processing.

[0114] like Figure 2 As shown, the present invention provides a natural resource database construction system, which is applied to a natural resource database construction method, and the system includes:

[0115] A data collection unit, used for collecting multi-source natural resource data and geospatial data having geographical relevance to the multi-source natural resource data;

[0116] A model building unit, used to establish a resource analysis model by using a neural network through multi-source natural resource data and geographic space data, and obtain a resource analysis model for obtaining resource analysis results based on the multi-source natural resource data and geographic space data;

[0117] A database construction unit, used to combine multi-source natural resource data and geospatial data, as well as obtained resource analysis results, to construct a natural resource database;

[0118] Among them, the resource analysis model is dynamically updated and associated with the natural resource database.

[0119] The present invention constructs a resource analysis model through a neural network, so that resource analysis results are obtained based on multi-source natural resource data and geographic spatial data, and a unified and objective analysis model is established, thereby improving the data integration and analysis efficiency and accuracy of the natural resource database, thereby improving the effect of data application through the natural resource database.

[0120] The above embodiments are only exemplary embodiments of the present application and are not intended to limit the present application. The protection scope of the present application is defined by the claims. Those skilled in the art may make various modifications or equivalent substitutions to the present application within the essence and protection scope of the present application, and such modifications or equivalent substitutions shall also be deemed to fall within the protection scope of the present application.

Claims

1. A method for constructing a natural resource database. It is characterized in that The following steps are involved: Collect multi-source natural resource data and geospatial data that are geographically relevant to the multi-source natural resource data; By using multi-source natural resource data and geospatial data and using a neural network, a resource analysis model is established to obtain a resource analysis model for obtaining resource analysis results based on the multi-source natural resource data and geospatial data; Combine multi-source natural resource data and geospatial data, as well as the obtained resource analysis results, to construct a natural resource database; Among them, the resource analysis model is dynamically updated and associated with the natural resource database.

2. A method for constructing a natural resource database according to claim 1, Features: The multi-source natural resource data includes: water resource data, forest resource data, land resource data, mineral resource data, climate resource data, and tourism resource data.

3. A method for constructing a natural resource database according to claim 1, Features: The geographic spatial data includes: spatial location data, ecological environment data, social and economic data, population data, and remote sensing image data.

4. A method for constructing a natural resource database according to claim 1, Features: The method for constructing the resource analysis model includes: The neural network is trained with multi-source natural resource data and geospatial data as input and resource analysis results as output, thereby obtaining a resource analysis model that encapsulates the mapping relationship between multi-source natural resource data and geospatial data and resource analysis results; The resource analysis model is: Goal=BP(Resource_data,Space_data); Where Goal is the result of resource analysis, Resource_data is the data vector representing multi-source natural resource data, Space_data is the data vector representing geographic space data, and BP is a neural network.

5. A method for constructing a natural resource database according to claim 4, Features: Methods for dynamically updating and associating resource analysis models with natural resource databases include: The resource analysis model is trained with multi-source natural resource data as input and resource analysis results as output, so as to migrate the resource analysis model that encapsulates the mapping relationship between the multi-source natural resource data and the geospatial data and the resource analysis results into a first resource analysis migration model that encapsulates the mapping relationship between the multi-source natural resource data and the resource analysis results; The first resource analysis migration model is: Goal = BP (Resource_data); In the formula, Goal is the result of resource analysis, Resource_data is the data vector representing multi-source natural resource data, and BP is a neural network; The resource analysis model is trained with the geospatial data as input and the resource analysis results as output, so as to migrate the resource analysis model that encapsulates the mapping relationship between multi-source natural resource data and the geospatial data and the resource analysis results into a second resource analysis migration model that encapsulates the mapping relationship between the geospatial data and the resource analysis results; The second resource analysis migration model is: Goal = BP (Space_data); In the formula, Goal is the result of resource analysis, Space_data is the data vector representing geographic spatial data, and BP is a neural network; When the multi-source natural resource data in the natural resource database is updated, the resource analysis results in the natural resource database are updated using the output of the first resource analysis migration model; When the geospatial data in the natural resource database is updated, the resource analysis results in the natural resource database are updated using the output of the second resource analysis migration model; When the multi-source natural resource data and geographic spatial data in the natural resource database are updated, the resource analysis results in the natural resource database are updated using the output of the resource analysis model.

6. A method for constructing a natural resource database according to claim 5, Features: Methods for determining updates to multi-source natural resource data include: Calculate the similarity between the multi-source natural resource data at the current moment in the natural resource database and the multi-source natural resource data at the previous moment in the natural resource database, and mark the multi-source natural resource data at the current moment whose similarity is lower than a preset threshold as multi-source natural resource data update; Methods for determining updates to geospatial data include: The similarity between the current geospatial data in the natural resource database and the previous geospatial data in the natural resource database is calculated, and the current geospatial data with a similarity lower than a preset threshold is marked as geospatial data update.

7. A method for constructing a natural resource database according to claim 6, Features: The functions for similarity calculation include Euclidean distance, correlation coefficient, and kernel function.

8. A method for constructing a natural resource database according to claim 5, Features: The loss function of the first resource analysis migration model during training is: Loss1=||BP(Resource_data)-BP(Resource_data,Space_data)|| 1 ; Where Loss1 is the loss function of the first resource analysis migration model during training, BP(Space_data) is the output of the first resource analysis migration model, BP(Resource_data,Space_data) is the output of the resource analysis model, ||BP(Resource_data)-BP(Resource_data,Space_data)|| 1 is the L1 norm; The loss function of the second resource analysis migration model during training is: Loss1=||BP(Space_data)-BP(Resource_data,Space_data)|| 1 ; Where Loss2 is the loss function of the second resource analysis migration model during training, BP(Space_data) is the output of the second resource analysis migration model, BP(Resource_data,Space_data) is the output of the resource analysis model, ||BP(Resource_data)-BP(Resource_data,Space_data)|| 1 is the L1 norm.

9. A method for constructing a natural resource database according to claim 1, Features: In the natural resource database, multi-source natural resource data, geospatial data, and resource analysis results are normalized.

10. A natural resource database construction system, It is characterized in that A method for constructing a natural resource database according to any one of claims 1 to 9, the system comprising: A data collection unit, used for collecting multi-source natural resource data and geospatial data having geographical relevance to the multi-source natural resource data; A model building unit, used to establish a resource analysis model by using a neural network through multi-source natural resource data and geographic space data, and obtain a resource analysis model for obtaining resource analysis results based on the multi-source natural resource data and geographic space data; A database construction unit, used to combine multi-source natural resource data and geospatial data, as well as obtained resource analysis results, to construct a natural resource database; Among them, the resource analysis model is dynamically updated and associated with the natural resource database.

Images