Natural resource assessment method and system based on multi-source time series data

Abstract

The present invention discloses a natural resource assessment method and system based on multi-source time series data, and relates to the technical field of data fusion. The method comprises: the target decision task is a task to be assessed determined by resource type and regional characteristics; identifying basic characteristics of the task and performing homologous retrieval to determine a multidimensional data set; hierarchically processing the multidimensional data set to determine a preprocessed data set; performing single assessment analysis and cross-source fitting analysis on the preprocessed data set based on the target decision task to determine a resource assessment result; matching and initializing resource management plans, performing plan expansion and optimization iteration to determine a pre-management plan; and sending the pre-management plan to a natural resource management system, which solves the problem in the prior art that the analysis results of natural resource assessment work are not comprehensive enough due to only performing analysis through a single data source or a single time, and enables more accurate assessment and judgment of natural resources.

Description

Technical Field

[0001] The present invention relates to the technical field of data fusion, and in particular to a natural resource assessment method and system based on multi-source time series data. Background Art

[0002] The assessment and determination of natural resources is crucial for environmental protection, resource management and sustainable development. However, using data from a single data source or a single time period for determination often cannot fully reflect the true status of natural resources. Therefore, in order to more comprehensively and accurately determine natural resources, a method that can span time series and multi-source data is needed. This method can integrate data from different data sources and covering different time spans, comprehensively consider various factors of natural resources, and thus more accurately assess and determine the status of natural resources. This method can help us better understand the status of natural resources and provide a scientific basis for environmental protection, resource management and sustainable development.

[0003] The existing natural resource assessment work in the prior art has the problem that the analysis results are not comprehensive because they are analyzed only through a single data source or at a single time, making it impossible to obtain a more accurate assessment of natural resources. Summary of the invention

[0004] The present application provides a natural resource assessment method and system based on multi-source time series data, which solves the problem in the prior art that the natural resource assessment work is not comprehensive due to analysis only through a single data source or a single time, so that natural resources can be more accurately assessed and determined.

[0005] In view of the above problems, this application provides a natural resource assessment method based on multi-source time series data.

[0006] In a first aspect, the present application provides a natural resource assessment method based on multi-source time series data, the method comprising: reading a target decision task, wherein the target decision task is a task to be assessed determined by resource type and regional characteristics; identifying basic task characteristics and performing homologous retrieval to determine a multidimensional data set, wherein the retrieval criteria are determined based on time and space dimensions; combining a preprocessing model, hierarchically processing the multidimensional data set to determine a preprocessing data set, wherein the preprocessing model includes multiple processing layers and is configured with different data processing algorithms; combining a resource assessment model, performing a single assessment analysis and a cross-source fitting analysis on the preprocessing data set based on the target decision task to determine a resource assessment result; matching an initialized resource management plan for the resource assessment result, performing plan expansion and optimization iteration to determine a pre-management plan; and sending the pre-management plan to a natural resource management system.

[0007] In the second aspect, the present application provides a natural resource assessment system based on multi-source time series data, and the system includes: a target decision task module: reading the target decision task, the target decision task is a task to be assessed determined by resource type and regional characteristics; a multidimensional data set module: identifying the basic characteristics of the task and performing homologous retrieval to determine the multidimensional data set, wherein the retrieval standard is determined based on the time and space dimensions; a preprocessing data set module: combining the preprocessing model, hierarchical processing of the multidimensional data set to determine the preprocessing data set, wherein the preprocessing model includes multiple processing layers and is configured with different data processing algorithms; a resource assessment model module: combining the resource assessment model, performing single assessment analysis and cross-source fitting analysis on the preprocessing data set based on the target decision task to determine the resource assessment result; a resource assessment result module: matching the initialization resource management plan for the resource assessment result, expanding the plan and optimizing iteratively to determine the pre-management plan; a resource management system module: sending the pre-management plan to the natural resource management system.

[0008] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0009] The embodiment of the present application provides a natural resource assessment method and system based on multi-source time series data, which reads the target decision task, which is a task to be assessed determined by resource type and regional characteristics, and then identifies the basic characteristics of the task and performs homologous retrieval to determine a multidimensional data set. In combination with a preprocessing model, the multidimensional data set is hierarchically processed to determine a preprocessing data set, and then in combination with a resource assessment model, a single assessment analysis and a cross-source fitting analysis are performed on the preprocessing data set based on the target decision task to determine the resource assessment result. Based on the resource assessment result, an initial resource management plan is matched, the plan is expanded and iterated for optimization, a pre-management plan is determined, and finally the pre-management plan is sent to the natural resource management system, thereby solving the problem in the prior art that the analysis results of natural resource assessment work are not comprehensive enough due to analysis only through a single data source or a single time, so that more accurate assessment and judgment of natural resources can be obtained. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1 A flow chart of a natural resource assessment method based on multi-source time series data is provided for this application;

[0011] Figure 2 A structural diagram of a natural resource assessment system based on multi-source time series data is provided for this application.

[0012] Explanation of the accompanying drawings: target decision task module 11, multi-dimensional data set module 12, pre-processing data set module 13, resource evaluation model module 14, resource evaluation result module 15, resource management system module 16. DETAILED DESCRIPTION

[0013] The present application provides a natural resource assessment method and system based on multi-source time series data, by reading the target decision task, which is a task to be assessed determined by resource type and regional characteristics, then identifying the basic characteristics of the task and performing homologous retrieval, determining a multidimensional data set, and combining the preprocessing model, hierarchically processing the multidimensional data set, determining the preprocessing data set, and then combining the resource assessment model, performing single assessment analysis and cross-source fitting analysis on the preprocessing data set based on the target decision task, determining the resource assessment result, matching the initialization resource management plan for the resource assessment result, performing plan expansion and optimization iteration, determining the pre-management plan, and finally sending the pre-management plan to the natural resource management system. The problem that the analysis results of natural resource assessment work in the prior art are not comprehensive enough due to only analyzing through a single data source or a single time is solved, so that more accurate assessment and judgment on natural resources can be obtained.

[0014] Embodiment 1

[0015] like Figure 1 As shown, the present application provides a natural resource assessment method and system based on multi-source time series data, the method comprising:

[0016] Reading a target decision task, wherein the target decision task is a task to be evaluated determined by resource type and regional characteristics;

[0017] Resource types include mineral resources, water resources, land resources, etc. Regional characteristics refer to the specific attributes of different regions, including geographical location, climate conditions, etc. Obtain tasks to be evaluated. Tasks to be evaluated are tasks that need to be evaluated. Generate corresponding target decision tasks based on the resource type and geographical location of the tasks to be evaluated. Provide a data basis for subsequent identification of basic task features and homology retrieval to determine multi-dimensional data sets.

[0018] Identify the basic features of the task and perform homology search to determine the multi-dimensional data set, where the search criteria are determined based on the time and space dimensions;

[0019] The basic features of a task are the time and space features of the task. The basic features of the task are identified and extracted, and the retrieval criteria are determined according to the time and space dimensions. The retrieval criteria include the retrieval criteria and scope, and a determined retrieval criteria is obtained. According to the retrieval criteria, the data related to the task are retrieved in the same data source to obtain a multi-dimensional data set. Determining the retrieval criteria and scope according to the time and space features of the task can more accurately obtain the data related to the task.

[0020] hierarchically processing the multidimensional data set in combination with a preprocessing model to determine a preprocessed data set, wherein the preprocessing model includes a plurality of processing layers and is configured with different data processing algorithms;

[0021] Training a first processing layer and configuring a first preprocessing algorithm, wherein the first processing layer is provided with a first data screening checkpoint;

[0022] Complete the construction of the Nth processing layer, wherein the Nth processing layer is configured with an Nth preprocessing algorithm and is provided with an Nth data screening checkpoint;

[0023] The first processing layer to the Nth processing layer are associated to generate the preprocessing model.

[0024] The preprocessing model is a model for preprocessing multidimensional data sets. The preprocessing model has multiple processing layers, and each processing layer has different processing algorithms. These algorithms can clean, filter, convert, and other operations on the data to optimize the quality and structure of the data. When constructing the preprocessing model, it is constructed according to the hierarchical structure, that is, it is constructed from the first layer in order until the Nth layer. And the data first enters the first layer for processing, and enters the second layer after the first layer is processed, and enters the third layer after the second layer is processed, and so on, until the data enters the Nth layer and is processed, which is the data level flow. When training the first layer, the preprocessing algorithm is configured for the first layer, that is, the first data screening level. The level screens some data with data defects and processes based on the algorithm of this layer. The algorithm of this layer can be for the noise reduction processing layer, identifying data above the noise threshold, etc., to complete the construction of the first data layer, and perform corresponding preprocessing configuration according to the multidimensional data set, and it needs to be determined according to the specific application scenario and requirements. Different application scenarios and tasks may require different preprocessing strategies and algorithms. Constructing a preprocessing model based on a multi-layer structure can improve data processing efficiency and accuracy, thereby providing better support for subsequent decision-making or analysis.

[0025] In combination with the resource evaluation model, a single evaluation analysis and a cross-source fitting analysis based on the target decision task are performed on the preprocessed data set to determine a resource evaluation result;

[0026] Wherein, the resource assessment model includes a single item analysis module and a cross-source fitting module;

[0027] Based on the single item analysis module, identify the single dimension data group of the preprocessed data set, perform independent decision analysis based on the target decision task, and determine X single item evaluation results;

[0028] The X individual evaluation results are transferred to the cross-source fitting module to perform feature complementation and result fitting to generate the resource evaluation result.

[0029] The single-item analysis module is mainly used to process the single-dimensional data group in the preprocessing data set, and processes the single-dimensional data group without considering the mutual influence relationship between multiple data. And perform independent decision analysis based on the target decision task to determine X single-item evaluation results. The resource evaluation model first identifies the single-dimensional data group in the preprocessing data set, and independently analyzes each source data in the first data group and each source data in the second data group. And perform independent decision analysis based on the target decision task for these single-dimensional data groups. Independent decision analysis refers to independent evaluation and analysis of each single-dimensional data group to determine the characteristics and relationships related to the target decision task. This analysis can help the system better understand the attributes and laws of the data, and provide a basis for subsequent feature complementation and result fitting. After completing the independent decision analysis, the obtained X single-item evaluation results are transferred to the cross-source fitting module. The cross-source fitting module can use the feature differences between different single-dimensional data groups to improve the description and evaluation of resources in a complementary way to obtain resource evaluation results. For example, if a single-dimensional data set lacks a detailed description of a specific resource, but other data sets contain detailed information about the resource, then the evaluation of the resource can be improved by complementing the features across the data sets. By performing feature complementation and result fitting, the differences and connections between different data sets can be taken into account, thereby obtaining a more representative and reliable evaluation result.

[0030] According to the resource evaluation results, match the initial resource management plan, expand the plan and iterate the optimization, and determine the preliminary management plan;

[0031] The pre-management plan is sent to the natural resource management system.

[0032] Determine a solution index based on the initialization resource management solution, and determine an optimization space based on the index degree of freedom;

[0033] Based on the optimization space, randomly perturb the initialization resource management scheme to determine multiple expansion schemes;

[0034] For the multiple expansion plans, determine an optimization plan based on fitness, perform optimization adjustment on each expansion plan, and determine multiple optimization plans, wherein the optimization adjustment target is any expansion plan with a fitness higher than the current expansion plan;

[0035] Iterate and optimize the multiple optimization schemes until the convergence condition is met, and select the optimization scheme corresponding to the highest fitness as the pre-management scheme.

[0036] The optimization space is a space built by multiple indicator thresholds that meet ecological standards. The initial resource management scheme is the initial resource management scheme that meets ecological standards, and the corresponding multiple indicators that meet ecological standards are the scheme indicators of the initial resource management scheme. The indicators are key factors for evaluating and measuring resource management schemes, such as resource utilization efficiency, cost-effectiveness, response speed, etc. Random perturbations can be random changes or mutations within a certain range, with the purpose of exploring more possible solutions. Through random perturbations, the number and diversity of solutions can be increased, providing more options for subsequent optimization. In the optimization space, the initial resource management scheme is randomly perturbed to generate multiple expansion schemes. The fitness of multiple expansion schemes is calculated to obtain the fitness of multiple expansion schemes. Among the multiple schemes with higher fitness than the current expansion scheme, one is randomly determined as the adjustment direction, that is, an arbitrary expansion scheme, and multiple arbitrary expansion schemes are used as multiple optimization schemes. Repeat the above steps for multiple optimization schemes, and perform iterative optimization until the convergence condition is met. The convergence condition can be reaching a preset number of iterations, the fitness reaching a preset threshold or other conditions, indicating that the optimization process has gradually converged to the optimal solution, and select the optimization scheme with the highest fitness as the pre-management scheme for output. By matching the initialization resource management scheme, expanding the scheme and iterating the optimization, it can be ensured that the final pre-management scheme is the best.

[0037] Furthermore, the method further comprises:

[0038] Connecting a multi-source data port, performing multi-source data collection based on a time span, and determining a first data group;

[0039] Based on the big data platform, multi-source data collection based on spatial span is performed to determine the second data group;

[0040] The step of integrating the first data group and the second data group includes:

[0041] Perform periodic division of the time span, and perform cross-period and same-period attribution integration for the first data group;

[0042] The second data group is weighted and identified based on the similarity of the basic features of the task.

[0043] Multi-source data ports include data ports based on remote sensing, radar, etc., and multi-source data collection based on time span. For example, periodic division is performed based on time periods such as seasons and climate, and multi-port data in the same period is associated. Multi-source data collection is performed based on multi-ports to obtain a first data group. Multi-source data based on spatial span refers to reference data located in different regions and homologous to task characteristics. Since it is data obtained based on big data, it is necessary to screen the acquired data, weight the acquired data, and determine the reference degree based on the assigned weights. Then, multi-source data collection is performed based on the reference degree to obtain a second data group. The first data group is subjected to cross-period and same-period attribution integration, that is, the first data group is divided according to the time period, and the data in different periods are integrated and attributed to complete the periodic division of the time span. Based on the approximation based on the basic characteristics of the task, the second data group is weighted and identified, that is, the data is evaluated and classified according to the basic characteristics of the task, and different weights or identifications are assigned to determine the subsequent reference degree. By integrating the first data group and the second data group, subsequent processing can be facilitated.

[0044] Furthermore, the method further comprises:

[0045] Inputting the multidimensional data set into the preprocessing model;

[0046] Performing screening based on the multi-dimensional data set based on the first data screening level, processing the screened data based on the first pre-processing algorithm, and determining a layer of processed data;

[0047] The first-layer processed data is transferred to the second processing layer for level screening and algorithm processing to determine the second-layer processed data;

[0048] The data flow is iterated sequentially until the level screening and algorithm processing of the Nth processing layer are completed, and the pre-processed data set is output.

[0049] The multidimensional data set is input into the preprocessing model as input data. A screening operation is performed based on the multidimensional data set to remove noise, redundant data and defective data in the multidimensional data set, thereby improving the quality and accuracy of the data. Then the first preprocessing algorithm is used to process the screened data, further screen and extract useful information, and perform necessary conversion or sorting. After the processing is completed, the processed data of one layer is transferred to the second processing layer for level screening and algorithm processing. This step is similar to the operation of the first layer, and is also to further improve the quality and accuracy of the data and extract more useful information. Repeat the above steps and iterate the data flow in sequence until the level screening and algorithm processing of the Nth processing layer is completed. Each layer will process and screen the data, and screen the multidimensional data set layer by layer, and gradually extract more useful information. After processing and screening by all processing layers, the preprocessed data set is output. By performing hierarchical processing on the multidimensional data set, the multidimensional data set contains more comprehensive, accurate and useful information, which is convenient for subsequent analysis, decision-making or model training tasks.

[0050] Furthermore, the method further comprises:

[0051] Performing a dispersion analysis on the X individual evaluation results, and extracting Y individual evaluation results that conform to the overall trend, where Y is a positive integer less than or equal to X;

[0052] Identify the Y individual evaluation results, calculate the mean of the individual evaluation results based on the first data group, perform weighted summation on the individual evaluation results based on the second data group, and determine two evaluation results;

[0053] The two evaluation results are subjected to feature complementation and result fitting to generate the resource evaluation result.

[0054] Dispersion analysis is a statistical method used to measure the dispersion or central tendency of a set of data. Through dispersion analysis, the degree of conformity of each single evaluation result with the overall trend can be evaluated. Dispersion analysis is performed on X single evaluation results to obtain multiple dispersion analysis results, and single evaluation results that conform to the overall trend are extracted from multiple dispersion analysis results to obtain Y single evaluation results, where Y is a positive integer that must be less than or equal to X. Mean calculation is a commonly used statistical method used to measure the average level of a set of data. By calculating the mean of each single evaluation result, an overall evaluation result can be obtained, and a single evaluation result can be obtained. Weighted summation is also a commonly used statistical method. By assigning different weights to different data, a comprehensive evaluation result can be obtained. That is, the single evaluation results based on the second data group are weighted summed to obtain a weighted summation result. Feature complementation refers to complementing the features of different data groups to improve the description and evaluation of resources. Result fitting refers to fitting different evaluation results to obtain more comprehensive and accurate resource evaluation results. Finally, the two assessment results are complemented with features and fitted with results. For example, the existence differences of the two assessment results are merged, and the differences in magnitude are averaged to generate the final resource assessment results. Through feature complementation and result fitting, the resource assessment results can be made more accurate and reliable.

[0055] Embodiment 2

[0056] Based on the same inventive concept as the natural resource assessment method based on multi-source time series data in the aforementioned embodiment, Figure 2 As shown, the present application provides a natural resource assessment system based on multi-source time series data, the system comprising:

[0057] Target decision task module 11: The target decision task module 11 is used to read the target decision task, which is a task to be evaluated determined by resource type and regional characteristics;

[0058] Multidimensional data set module 12: The multidimensional data set module 12 is used to identify basic features of the task and perform homology search to determine a multidimensional data set, wherein the search criteria are determined based on the spatiotemporal dimension;

[0059] Preprocessing data set module 13: the preprocessing data set module 13 is used to perform hierarchical processing on the multi-dimensional data set in combination with a preprocessing model to determine a preprocessing data set, wherein the preprocessing model includes multiple processing layers and is configured with different data processing algorithms;

[0060] Resource evaluation model module 14: the resource evaluation model module 14 is used to combine the resource evaluation model, perform single evaluation analysis and cross-source fitting analysis based on the target decision task on the preprocessed data set, and determine the resource evaluation result;

[0061] Resource evaluation result module 15: The resource evaluation result module 15 is used to match the initial resource management plan according to the resource evaluation result, perform plan expansion and optimization iteration, and determine the pre-management plan;

[0062] Resource management system module 16: The resource management system module 16 is used to send the pre-management plan to the natural resource management system.

[0063] Furthermore, the system further comprises:

[0064] Connecting a multi-source data port, performing multi-source data collection based on a time span, and determining a first data group;

[0065] Based on the big data platform, multi-source data collection based on spatial span is performed to determine the second data group;

[0066] The step of integrating the first data group and the second data group includes:

[0067] Perform periodic division of the time span, and perform cross-period and same-period attribution integration for the first data group;

[0068] The second data group is weighted and identified based on the similarity of the basic features of the task.

[0069] Furthermore, the pre-processing data set module 13 also includes:

[0070] Training a first processing layer and configuring a first preprocessing algorithm, wherein the first processing layer is provided with a first data screening checkpoint;

[0071] Complete the construction of the Nth processing layer, wherein the Nth processing layer is configured with an Nth preprocessing algorithm and is provided with an Nth data screening checkpoint;

[0072] The first processing layer to the Nth processing layer are associated to generate the preprocessing model.

[0073] Furthermore, the pre-processing data set module 13 also includes:

[0074] Inputting the multidimensional data set into the preprocessing model;

[0075] Performing screening based on the multi-dimensional data set based on the first data screening level, processing the screened data based on the first pre-processing algorithm, and determining a layer of processed data;

[0076] The first-layer processed data is transferred to the second processing layer for level screening and algorithm processing to determine the second-layer processed data;

[0077] The data flow iteration is performed in sequence until the level screening and algorithm processing of the Nth processing layer are completed, and the pre-processed data set is output.

[0078] Furthermore, the resource assessment model module 14 also includes:

[0079] Wherein, the resource assessment model includes a single item analysis module and a cross-source fitting module;

[0080] Based on the single item analysis module, identify the single dimension data group of the preprocessed data set, perform independent decision analysis based on the target decision task, and determine X single item evaluation results;

[0081] The X individual evaluation results are transferred to the cross-source fitting module to perform feature complementation and result fitting to generate the resource evaluation result.

[0082] Furthermore, the resource assessment model module 14 also includes:

[0083] Performing a dispersion analysis on the X individual evaluation results, and extracting Y individual evaluation results that conform to the overall trend, where Y is a positive integer less than or equal to X;

[0084] Identify the Y individual evaluation results, calculate the mean of the individual evaluation results based on the first data group, perform weighted summation on the individual evaluation results based on the second data group, and determine two evaluation results;

[0085] The two evaluation results are subjected to feature complementation and result fitting to generate the resource evaluation result.

[0086] Furthermore, the resource assessment result module 15 also includes:

[0087] Determine a solution index based on the initialization resource management solution, and determine an optimization space based on the index degree of freedom;

[0088] Based on the optimization space, randomly perturb the initialization resource management scheme to determine multiple expansion schemes;

[0089] For the multiple expansion plans, determine an optimization plan based on fitness, perform optimization adjustment on each expansion plan, and determine multiple optimization plans, wherein the optimization adjustment target is any expansion plan with a fitness higher than the current expansion plan;

[0090] Iterate and optimize the multiple optimization schemes until the convergence condition is met, and select the optimization scheme corresponding to the highest fitness as the pre-management scheme.

[0091] Through the above-mentioned detailed description of the natural resource assessment method based on multi-source time series data in this specification, those skilled in the art can clearly understand the natural resource assessment method based on multi-source time series data in this embodiment. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the description of the method part.

[0092] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A natural resource assessment method based on multi-source time series data, characterized in that: The method comprises: Reading a target decision task, wherein the target decision task is a task to be evaluated determined by resource type and regional characteristics; Identify the basic features of the task and perform homology search to determine the multi-dimensional data set, where the search criteria are determined based on the time and space dimensions; hierarchically processing the multidimensional data set in combination with a preprocessing model to determine a preprocessed data set, wherein the preprocessing model includes a plurality of processing layers and is configured with different data processing algorithms; In combination with the resource evaluation model, a single evaluation analysis and a cross-source fitting analysis based on the target decision task are performed on the preprocessed data set to determine a resource evaluation result; According to the resource evaluation results, match the initial resource management plan, expand the plan and iterate the optimization, and determine the preliminary management plan; sending the pre-management plan to a natural resource management system; Performing a single evaluation analysis and a cross-source fitting analysis based on the target decision task on the preprocessed data set, the method comprising: Wherein, the resource assessment model includes a single item analysis module and a cross-source fitting module; Based on the single item analysis module, identify the single dimension data group of the preprocessed data set, perform independent decision analysis based on the target decision task, and determine X single item evaluation results; The X individual evaluation results are transferred to the cross-source fitting module to perform feature complementation and result fitting to generate the resource evaluation result; Performing feature complementation and result fitting to generate the resource assessment result, the method includes: Performing a dispersion analysis on the X individual evaluation results, and extracting Y individual evaluation results that conform to the overall trend, where Y is a positive integer less than or equal to X; Identify the Y individual evaluation results, calculate the mean of the individual evaluation results based on the first data group, perform weighted summation on the individual evaluation results based on the second data group, and determine two evaluation results; Performing feature complementation and result fitting on the two evaluation results to generate the resource evaluation result, wherein feature complementation refers to complementing the features of different data groups, and result fitting refers to fitting different evaluation results, and averaging or weighted summing the differences in magnitude; Matching the initialization resource management plan, expanding the plan and optimizing iteratively, the method includes: Determine a solution index based on the initialization resource management solution, and determine an optimization space based on the index degree of freedom; Based on the optimization space, randomly perturb the initialization resource management scheme to determine multiple expansion schemes; For the multiple expansion plans, determine an optimization plan based on fitness, perform optimization adjustment on each expansion plan, and determine multiple optimization plans, wherein the optimization adjustment target is any expansion plan with a fitness higher than the current expansion plan; Iterate and optimize the multiple optimization schemes until the convergence condition is met, and select the optimization scheme corresponding to the highest fitness as the pre-management scheme.

2. The method according to claim 1, characterized in that The method includes: Connecting a multi-source data port, performing multi-source data collection based on a time span, and determining a first data group; Based on the big data platform, multi-source data collection based on spatial span is performed to determine the second data group; The step of integrating the first data group and the second data group includes: Perform periodic division of the time span, and perform cross-period and same-period attribution integration for the first data group; The second data group is weighted and identified based on the similarity of the basic features of the task.

3. The method according to claim 1, characterized in that Constructing a preprocessing model, the method includes: Training a first processing layer and configuring a first preprocessing algorithm, wherein the first processing layer is provided with a first data screening checkpoint; Complete the construction of the Nth processing layer, wherein the Nth processing layer is configured with an Nth preprocessing algorithm and is provided with an Nth data screening checkpoint; The first processing layer to the Nth processing layer are associated to generate the preprocessing model.

4. The method according to claim 3, characterized in that Performing hierarchical processing on the multidimensional data set, the method comprising: Inputting the multidimensional data set into the preprocessing model; Performing screening based on the multi-dimensional data set based on the first data screening level, processing the screened data based on the first pre-processing algorithm, and determining a layer of processed data; The first-layer processed data is transferred to the second processing layer for level screening and algorithm processing to determine the second-layer processed data; The data flow iteration is performed in sequence until the level screening and algorithm processing of the Nth processing layer are completed, and the pre-processed data set is output.

5. A natural resource assessment system based on multi-source time series data, characterized in that: The system comprises: Target decision task module: reads the target decision task, which is a task to be evaluated determined by resource type and regional characteristics; Multi-dimensional data set module: Identify the basic features of the task and perform homology retrieval to determine the multi-dimensional data set, where the retrieval criteria are determined based on the spatiotemporal dimensions; Preprocessing data set module: hierarchically processing the multi-dimensional data set in combination with a preprocessing model to determine a preprocessing data set, wherein the preprocessing model includes multiple processing layers and is configured with different data processing algorithms; Resource evaluation model module: combining the resource evaluation model, performing single evaluation analysis and cross-source fitting analysis based on the target decision task on the preprocessed data set to determine the resource evaluation result; Resource evaluation result module: based on the resource evaluation results, match and initialize the resource management plan, expand the plan and iterate the optimization, and determine the pre-management plan; Resource management system module: sending the pre-management plan to the natural resource management system; The resource assessment model module also includes: Wherein, the resource assessment model includes a single item analysis module and a cross-source fitting module; Based on the single item analysis module, identify the single dimension data group of the preprocessed data set, perform independent decision analysis based on the target decision task, and determine X single item evaluation results; The X individual evaluation results are transferred to the cross-source fitting module to perform feature complementation and result fitting to generate the resource evaluation result; The resource assessment model module also includes: Performing a dispersion analysis on the X individual evaluation results, and extracting Y individual evaluation results that conform to the overall trend, where Y is a positive integer less than or equal to X; Identify the Y individual evaluation results, calculate the mean of the individual evaluation results based on the first data group, perform weighted summation on the individual evaluation results based on the second data group, and determine two evaluation results; Performing feature complementation and result fitting on the two evaluation results to generate the resource evaluation result, wherein feature complementation refers to complementing the features of different data groups, and result fitting refers to fitting different evaluation results, and averaging or weighted summing the differences in magnitude; The resource assessment result module also includes: Determine a solution index based on the initialization resource management solution, and determine an optimization space based on the index degree of freedom; Based on the optimization space, randomly perturb the initialization resource management scheme to determine multiple expansion schemes; For the multiple expansion plans, determine an optimization plan based on fitness, perform optimization adjustment on each expansion plan, and determine multiple optimization plans, wherein the optimization adjustment target is any expansion plan with a fitness higher than the current expansion plan; Iterate and optimize the multiple optimization schemes until the convergence condition is met, and select the optimization scheme corresponding to the highest fitness as the pre-management scheme.

Images