A multi-level resource monitoring and evaluation method and system

Abstract

The application relates to the technical field of resource monitoring and evaluation, and discloses a multi-level resource monitoring and evaluation method and system, aiming to solve the problems of incomplete dimensions, data islands, imperfect feedback and insufficient flexibility in existing regional development monitoring and evaluation. The method comprises the following steps: acquiring and integrating multi-source spatio-temporal data to construct a unified database; evaluating resource endowment and governance effectiveness based on the database to generate portraits and reports; dynamically adjusting index weights; performing multi-level evaluation calculation to generate comprehensive results; visualizing the results and providing decision support; and dynamically feeding back and optimizing to propose policy recommendations. The system comprises the following modules: data integration, resource endowment evaluation, governance effectiveness monitoring, index management, evaluation calculation, decision support visualization and dynamic feedback optimization. By adopting the above technical scheme, the application can effectively solve the problems of data islands, incomplete dimensions, complex indexes, insufficient flexibility and imperfect feedback, and realize precise measures and dynamic optimization for high-quality regional development.

Description

Technical Field

[0001] This invention belongs to the field of resource monitoring and evaluation, and specifically relates to a multi-level resource monitoring and evaluation method and system. Background Technology

[0002] The establishment and supervision of the national spatial planning system, as well as the modernization of national spatial governance capabilities, are crucial for promoting regional sustainable development and optimizing resource allocation. This is especially true in regions with relatively lagging economic and social development, such as mountainous counties, which urgently need refined spatial governance and resource element guarantees to help them achieve leapfrog high-quality development. This involves not only macro-policy guidance but also the construction of a scientific and effective monitoring and evaluation mechanism in practice to support precise policy implementation and dynamic management. Objective evaluation of regional resource endowments and comprehensive monitoring of spatial governance effectiveness are fundamental to understanding regional development potential and identifying governance shortcomings.

[0003] Among these, multi-level resource monitoring and evaluation, as a core component of the supervision of territorial spatial planning implementation and regional governance innovation, focuses on conducting systematic resource element analysis and governance performance assessment for specific administrative regions, especially those with developmental differences. This evaluation aims to comprehensively identify the region's multi-dimensional resource advantages and disadvantages (natural, social, and economic), and through quantitative evaluation of the effectiveness of spatial governance policies, accurately reflect the region's current development status and identify potential problems. This provides scientific support for tailored and differentiated policy implementation, and is an important technical pathway to promote high-quality regional development.

[0004] The current regional high-quality development monitoring and evaluation system is mainly dominated by economic and social development indicators, paying insufficient attention to the unique perspectives of land and space governance effectiveness and natural resource element protection. This results in evaluation results that fail to comprehensively reflect the efficiency of spatial resource allocation and environmental sustainability. Constructing a multi-level resource monitoring and evaluation system faces numerous technical bottlenecks: for example, the classification and evaluation of county-level resource endowments lacks a unified content framework adapted to the characteristics of different counties, making it difficult to achieve comprehensive and differentiated profiles; the construction of its indicator system is highly complex, posing challenges to the scientific selection, rational combination, and accurate quantification of multi-dimensional resources; simultaneously, the output format of the evaluation model is too rigid and lacks flexibility, making it difficult to effectively adapt to the decision-making needs of different management entities and the actual work of planning. Furthermore, the monitoring and evaluation of spatial governance effectiveness generally suffers from problems such as scattered data sources, inconsistent standards, and difficulty in effectively integrating cross-departmental information, forming "data silos"; the setting of evaluation indicators struggles to balance multiple dimensions such as economic development, environmental protection, and social equity, and lacks a dynamic adjustment mechanism to adapt to development changes; in addition, the feedback mechanism between evaluation results and policy adjustments is still imperfect, preventing the evaluation effectiveness from being fully transformed into a powerful driving force for decision-making improvement. The aforementioned problems have rendered existing monitoring and evaluation methods inadequate in accurately identifying regional development shortcomings, effectively supporting the formulation of differentiated policies, and promoting high-quality regional development. There is an urgent need to propose a more comprehensive, intelligent, and efficient multi-level resource monitoring and evaluation method and system. Summary of the Invention

[0005] To address the technical problems existing in regional high-quality development monitoring and evaluation systems, such as incomplete evaluation dimensions, lack of a unified framework for county-level resource endowment classification and evaluation, complex indicator system construction, insufficient flexibility in evaluation model output, fragmented data sources leading to data silos, difficulty in setting evaluation indicators to accommodate multi-dimensional objectives, and an imperfect feedback mechanism between evaluation results and policy adjustments, this invention provides a multi-level resource monitoring and evaluation method and system. This invention constructs a complete chain mechanism spanning data collection, feature extraction, multi-scale modeling, intelligent evaluation, decision support, and dynamic feedback. This enables cross-level, multi-dimensional, and adaptive evaluation of regional resource endowments and spatial governance effectiveness, generating refined regional profiles and governance diagnoses, and providing customized decision support to achieve precise policy implementation and dynamic optimization for regional high-quality development.

[0006] According to one aspect of the present invention, a multi-level resource monitoring and evaluation method is provided, comprising:

[0007] Acquire multi-source heterogeneous spatiotemporal data, including remote sensing image data, geographic information data, statistical yearbook data, Internet of Things sensor data, and departmental business data;

[0008] The multi-source heterogeneous spatiotemporal data is cleaned, standardized, and integrated to construct a unified spatiotemporal database for regional resources and governance.

[0009] Based on the aforementioned regional resource and governance spatiotemporal database, a regional resource endowment assessment is constructed and executed to obtain a regional resource endowment classification profile, which includes a natural ecological resource index, a human and social resource index, and an economic development potential index.

[0010] Based on the aforementioned regional resource and governance spatiotemporal database, a spatial governance effectiveness monitoring and evaluation system is constructed and implemented to obtain a spatial governance effectiveness evaluation report. The spatial governance effectiveness evaluation report includes the degree of achievement of planning objectives, the ecological environment improvement index, and the public service equity index.

[0011] The evaluation indicators and their weights are dynamically adjusted through the intelligent indicator system management module to achieve customized configuration of the evaluation model.

[0012] Based on the regional resource endowment classification profile, the spatial governance effectiveness assessment report, and the assessment indicators and weights configured in the intelligent indicator system management module, a multi-level assessment model is executed to generate a comprehensive regional assessment result.

[0013] The comprehensive assessment results of the region are analyzed in depth and visualized to provide decision support; and

[0014] Based on the comprehensive regional assessment results and decision support information, dynamic feedback and optimization will be conducted to provide policy adjustment suggestions and planning optimization schemes.

[0015] Furthermore, the acquisition of multi-source heterogeneous spatiotemporal data specifically includes:

[0016] Acquire multispectral high-resolution remote sensing image data and multitemporal remote sensing image data through satellite remote sensing data platforms;

[0017] Digital elevation model data, topographic data, soil type data, hydrological observation data, and land use status data are obtained through geographic information databases.

[0018] Obtain population census data, socio-economic statistics, and fiscal input-output data from statistical yearbook databases;

[0019] Real-time monitoring data on ecological and environmental quality and operational data of public service facilities are acquired through IoT sensor networks; and

[0020] The system obtains land and space planning target texts, land use change survey data, and resident satisfaction survey data through the data interface of local government business systems.

[0021] Furthermore, the data cleaning, standardization, and integration processing of the multi-source heterogeneous spatiotemporal data specifically includes:

[0022] The multi-source heterogeneous spatiotemporal data is collected through a data acquisition agent;

[0023] The collected data is subjected to outlier detection and replacement, and the outlier detection and replacement adopts a rule-based outlier detection algorithm.

[0024] The collected data is subjected to time-series data smoothing processing, which employs the Kalman filter algorithm.

[0025] The collected data is subjected to data standardization, which employs Z-score standardization or min-max normalization; and

[0026] The processed data is stored in a geospatial database whose data model conforms to the ISO19115 standard.

[0027] Furthermore, the construction and execution of the regional resource endowment assessment specifically includes:

[0028] Feature extraction is performed on the multi-temporal remote sensing image data, and the feature extraction adopts a temporal convolutional neural network based on a deep residual network;

[0029] Land use types are finely classified, and the fine classification uses support vector machines;

[0030] The spatial heterogeneity and interactions of various resource elements are analyzed using a geographic detector model; and

[0031] Based on the analysis results, a resource endowment classification profile of the region is calculated and generated.

[0032] Furthermore, the construction and implementation of spatial governance effectiveness monitoring and evaluation specifically includes:

[0033] Land use scenario simulation was conducted by constructing a coupled model based on Markov chains and cellular automata to predict future land use change trends.

[0034] The performance of policy implementation is quantitatively evaluated using a multi-criteria decision analysis method, which includes a weighting method combining the analytic hierarchy process (AHP) and the entropy weighting method.

[0035] Based on the land use scenario simulation results and the quantitative evaluation results of policy implementation performance, the spatial governance effectiveness evaluation report is calculated and generated.

[0036] Furthermore, the dynamic adjustment of evaluation indicators and their weights through the intelligent indicator system management module specifically includes:

[0037] Through the user interface of the intelligent indicator system management module, select and define the name, unit of measurement, data source, calculation formula and evaluation threshold of the evaluation indicator;

[0038] The weights of the evaluation indicators are configured through the intelligent indicator system management module. This weight configuration employs a combination of the analytic hierarchy process (AHP) and entropy-based weighting, and supports expert experience correction.

[0039] The configured evaluation metrics and their weights are stored in a configuration file in Extensible Markup Language (EXPLAIN) or JSON format to enable dynamic loading and updating.

[0040] Furthermore, the calculation of the multi-level evaluation model specifically includes:

[0041] Based on the regional resource endowment classification profile and the spatial governance effectiveness assessment report, a local assessment is performed, which adopts a geographically weighted regression model.

[0042] Based on the regional resource endowment classification profile and the spatial governance effectiveness assessment report, a comprehensive assessment is performed, which uses random forest or gradient boosting decision tree.

[0043] Based on the assessment needs, select a multi-scale assessment strategy based on grid units, township administrative units, or county administrative units; and

[0044] An uncertainty analysis was performed on the evaluation results, which employed Monte Carlo simulation.

[0045] Furthermore, the in-depth analysis and visualization of the comprehensive regional assessment results specifically includes:

[0046] The comprehensive evaluation results of the region are used to generate an interactive map application through the geographic information system rendering engine. The interactive map application supports multi-layer overlay, thematic map production, spatial query and analysis.

[0047] The comprehensive regional assessment results are presented in the form of radar charts, bar charts, line charts, and scatter plots through a data visualization library, supporting data pivoting and drill-down analysis; and

[0048] It provides a customizable report generation feature that allows users to select the evaluation content and the display format.

[0049] Furthermore, the dynamic feedback and optimization specifically includes:

[0050] By constructing a policy intervention effect prediction model, regional shortcomings are identified based on the regional comprehensive assessment results, and targeted policy adjustment suggestions are proposed. The policy intervention effect prediction model adopts a Bayesian network or a deep reinforcement learning model.

[0051] Predict the potential impact of policy adjustments; and

[0052] Based on historical policy effect feedback data, the parameters and weights of the policy intervention effect prediction model are adjusted to achieve closed-loop optimization of evaluation and decision-making.

[0053] According to another aspect of the present invention, a multi-level resource monitoring and evaluation system is provided, comprising:

[0054] The data integration module is used to acquire multi-source heterogeneous spatiotemporal data, including remote sensing image data, geographic information data, statistical yearbook data, IoT sensor data, and departmental business data. The module performs data cleaning, data standardization, and data integration processing on the multi-source heterogeneous spatiotemporal data to construct a unified regional resource and governance spatiotemporal database.

[0055] The resource endowment assessment module is used to construct and execute a regional resource endowment assessment based on the regional resource and governance spatiotemporal database to obtain a regional resource endowment classification profile, which includes a natural ecological resource index, a human and social resource index, and an economic development potential index.

[0056] The governance effectiveness monitoring module is used to construct and execute spatial governance effectiveness monitoring and evaluation based on the regional resource and governance spatiotemporal database to obtain a spatial governance effectiveness evaluation report, which includes the degree of achievement of planning goals, ecological environment improvement index, and public service fairness index.

[0057] The indicator system management module is used to dynamically adjust the evaluation indicators and their weights to achieve customized configuration of the evaluation model;

[0058] The evaluation model calculation module is used to perform multi-level evaluation model calculations based on the regional resource endowment classification profile, the spatial governance effectiveness evaluation report, and the evaluation indicators and weights configured by the indicator system management module, and generate a comprehensive regional evaluation result.

[0059] The decision support and visualization module is used to perform in-depth analysis and visualization of the comprehensive assessment results of the region to provide decision support; and

[0060] The dynamic feedback and optimization module is used to provide dynamic feedback and optimization based on the comprehensive regional assessment results and decision support information, so as to provide policy adjustment suggestions and planning optimization schemes.

[0061] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0062] This invention overcomes the problems of scattered data sources, inconsistent standards, and data silos in existing technologies by introducing multi-source heterogeneous spatiotemporal data integration technology. Through rule-based outlier detection, Kalman filtering for time-series data smoothing, Z-score normalization or min-max normalization for data standardization, and the construction of a geospatial database in accordance with the ISO 19115 standard, unified management and efficient utilization of various types of data are achieved, laying a solid data foundation for subsequent evaluation.

[0063] This invention establishes a multi-level, multi-dimensional framework for assessing regional resource endowments. By utilizing a temporal convolutional neural network with deep residual networks to extract features from multi-temporal remote sensing images, combined with support vector machines for refined land use classification, and employing a geographic detector model to analyze the spatial heterogeneity and interactions of resource elements, it achieves a comprehensive and accurate profile of natural, social, and economic resources. This addresses the shortcomings of existing technologies in classifying and assessing county-level resource endowments, which lack a unified framework adapted to the characteristics of different counties.

[0064] This invention constructs a refined monitoring and evaluation mechanism for the effectiveness of spatial governance. By simulating land use scenarios using a coupled Markov chain and cellular automata model, it predicts land use change trends. Furthermore, it employs a multi-criteria decision analysis method combining the analytic hierarchy process (AHP) and entropy weighting to quantitatively evaluate policy implementation performance. This achieves an objective assessment of the achievement of planning goals, ecological environment improvement, and the fairness of public services, effectively addressing the problem of insufficient attention to the effectiveness of national spatial governance in existing evaluation systems.

[0065] This invention features a dynamic management function for an intelligent indicator system. By providing a configurable user interface, it allows users to customize the name, unit of measurement, data source, calculation formula, evaluation threshold, and weight of evaluation indicators. Furthermore, it employs a combination of analytic hierarchy process (AHP) and entropy-based weighting, along with expert experience correction, to achieve flexible configuration and dynamic adjustment of the evaluation indicator system. This overcomes the technical bottleneck of existing evaluation models having rigid output formats and being difficult to adapt to the decision-making needs of different management entities.

[0066] This invention employs a multi-level, multi-model evaluation and calculation strategy. By combining a geographically weighted regression model for local evaluation, a random forest or gradient boosting decision tree for comprehensive evaluation, and supporting multi-scale evaluation based on grid units, township administrative units, and county administrative units, and supplemented by Monte Carlo simulation for uncertainty analysis, the accuracy and applicability of the evaluation results are greatly improved, ensuring that the evaluation results can accurately reflect the current development status and potential problems of the region.

[0067] This invention provides intuitive and interactive decision support and visualization capabilities. Through interactive map applications based on the WebGIS framework, multi-layer overlay, thematic map creation, spatial query analysis, and various data visualization formats such as radar charts, bar charts, line charts, and scatter plots, it supports data pivoting and drill-down analysis, and provides customizable report generation functionality. This enables evaluation results to be effectively transformed into a powerful driving force for decision support, solving the problem of an imperfect feedback mechanism between existing evaluation results and policy adjustments.

[0068] This invention achieves closed-loop optimization of assessment and decision-making. By constructing a policy intervention effect prediction system based on Bayesian networks or deep reinforcement learning models, it identifies regional shortcomings based on assessment results and provides targeted policy adjustment suggestions. Simultaneously, it predicts the potential impact of policy adjustments and possesses the ability to adaptively adjust model parameters and weights based on historical policy effect feedback data. This enables a continuous cycle of assessment, decision-making, and optimization, significantly improving the intelligence level and efficiency of regional spatial governance. Attached Figure Description

[0069] Figure 1 This is a schematic diagram of the overall technical architecture of the multi-level resource monitoring and evaluation method and system proposed in this invention.

[0070] Figure 2 This is a schematic diagram illustrating the core principle framework of the intelligent indicator system management module in this invention for dynamically adjusting evaluation indicators.

[0071] Figure 3 This is a logical flowchart of the integration of multi-source heterogeneous spatiotemporal data and the construction of a unified database in this invention.

[0072] Figure 4 This is a logical flowchart of the calculation of the multi-level evaluation model in this invention.

[0073] Figure 5 This is a logical flowchart of the dynamic feedback and optimization mechanism in this invention. Detailed Implementation

[0074] See Figure 1 This application proposes a multi-level resource monitoring and evaluation method and system, aiming to address numerous technical bottlenecks faced by existing technologies in the supervision of territorial spatial planning implementation and regional governance innovation, particularly in challenges such as county-level resource endowment classification and assessment, the complexity of indicator system construction, the rigidity of assessment model output, and imperfect data integration and feedback mechanisms. The method in this application integrates multi-source heterogeneous spatiotemporal data, constructs a dynamically managed intelligent indicator system, executes multi-level assessment model calculations, generates multi-dimensional visualized evaluation results, and achieves dynamic feedback and optimization of territorial spatial governance strategies, thereby providing scientific, precise, and intelligent decision support for promoting high-quality regional development.

[0075] According to embodiments of this application, the multi-level resource monitoring and evaluation method includes the following steps:

[0076] S101 is used for the acquisition, integration and preprocessing of multi-source heterogeneous spatiotemporal data.

[0077] S102, Construct and dynamically manage a multi-level resource monitoring and evaluation indicator system;

[0078] S103, execute the calculation of the multi-level resource monitoring and evaluation model;

[0079] S104, Generate and visualize multi-dimensional monitoring and evaluation results;

[0080] S105 enables dynamic feedback and optimization of national land space governance strategies.

[0081] Specifically, in the aforementioned multi-level resource monitoring and evaluation method, step S101 involves acquiring, integrating, and preprocessing multi-source heterogeneous spatiotemporal data. This step is fundamental to the entire evaluation system, aiming to address the technical challenges of dispersed data sources, inconsistent standards, and the difficulty in effectively integrating cross-departmental information, thereby constructing a unified data foundation.

[0082] As a further explanation of this embodiment, step S101 further includes:

[0083] S1011, Identify and determine the sources and types of multi-source heterogeneous spatiotemporal data. This sub-step identifies the various types of data used for evaluation. Data sources include, but are not limited to, national and local statistical yearbooks, population census data, economic and social survey data, land use status maps and planning maps from natural resources departments, air quality monitoring data and water quality monitoring data from ecological and environmental departments, forest resource inventory data from forestry departments, crop yield and arable land quality data from agricultural departments, road network distribution data from transportation departments, public service facility distribution data from education and medical departments, satellite remote sensing imagery data, UAV aerial survey data, and real-time environmental monitoring data acquired by ground sensor networks. Data types cover raster data such as digital elevation models, land cover classification maps, and vegetation index maps; vector data such as administrative boundaries, rivers and lakes, road networks, and land ownership parcels; and tabular data such as population statistics, GDP, industrial structure ratios, fiscal revenue and expenditure, pollution emissions, education coverage rate, medical resource allocation, and the number of cultural facilities. Each type of data is assigned a specific timestamp and spatial coordinate information to ensure the traceability of the data's spatiotemporal attributes.

[0084] S1012, Perform data acquisition and preliminary cleaning. This sub-step, based on the data sources and types determined in S1011, collects data using various technical means. For publicly available government data, web crawling and application programming interfaces (APIs) are used for automated data acquisition; for data from internal management systems, data is acquired through direct database connections or file import; for remote sensing imagery and sensor data, data is downloaded and interpreted in batches using data receiving stations and remote sensing image processing software. The acquired data undergoes preliminary cleaning, including removing duplicate records, filling missing values, correcting data format errors, and identifying outliers. Missing value filling methods are based on contextual information, using interpolation algorithms or historical means; outliers are marked and processed by setting statistical thresholds, such as the three-standard-deviation rule. All data records undergo rigorous timestamp verification and spatial reference system conversion to ensure the accuracy of subsequent processing.

[0085] S1013, perform spatiotemporal alignment and data format standardization. This sub-step aligns the collected and cleaned heterogeneous data to a unified spatiotemporal reference. Spatial alignment includes projection transformation, coordinate system transformation, and georegistration, ensuring all spatial data are placed in the same geographic coordinate system and projected reference system. Temporal alignment involves setting a unified sampling frequency and time granularity, resampling or interpolating data at different time resolutions. For example, annual data is refined to quarterly or monthly data, and real-time monitoring data is aggregated to daily or monthly averages. Data format standardization refers to converting data in different formats, such as Shapefile, GeoTIFF, CSV, and JSON, into a unified internal data storage format, such as the table structure of a relational database or the geographic object model of a spatial database, ensuring consistency in naming conventions and data types for all attribute fields.

[0086] S1014, see S1014. Figure 3 This sub-step constructs a unified multimodal spatiotemporal database. All standardized data is integrated into a centralized, high-performance multimodal spatiotemporal database. This database is designed to natively support raster, vector, and tabular data, and can efficiently store and query data with a time dimension. The database includes a detailed metadata management mechanism, recording the source, collection time, processing procedure, quality assessment report, and usage permissions for each data set to ensure data credibility and traceability. The database employs a distributed storage architecture to meet the storage needs of massive amounts of spatiotemporal data, and improves data query and analysis efficiency through index optimization and partition management techniques. The database also provides standardized data access interfaces for easy invocation by subsequent evaluation modules and facilitates data sharing with external systems. This database is a core component for realizing a multi-level resource monitoring and evaluation data foundation.

[0087] Furthermore, in the aforementioned multi-level resource monitoring and evaluation method, step S102 involves constructing and dynamically managing a multi-level resource monitoring and evaluation indicator system. This step aims to address the issues of a lack of a unified content framework for county-level resource endowment classification and evaluation, the high complexity of indicator system construction, and the insufficient rigidity of evaluation model output, thereby ensuring the comprehensiveness and adaptability of the evaluation.

[0088] As a further explanation of this embodiment, step S102 further includes:

[0089] S1021 defines the evaluation objectives and hierarchical scope. This sub-step clarifies the specific objectives of this multi-level resource monitoring and evaluation, such as assessing the effectiveness of regional ecological civilization construction, the quality of economic development, the level of social equity protection, or the capacity for territorial spatial governance. Simultaneously, it clarifies the geographical scope of the evaluation, covering everything from the national macro-scale to the provincial, municipal, county, township, and even village micro-scales. Each level has its own focus and corresponding indicator granularity. For example, county-level evaluations focus on local resource endowments and governance effectiveness, while provincial-level evaluations may focus on regional coordinated development and the overall ecological security pattern.

[0090] S1022, Preliminary Screening and Definition of Candidate Evaluation Indicators. This sub-step, based on the objectives and levels defined in S1021, uses the unified multimodal spatiotemporal database obtained in S101 to initially screen various indicators related to resource monitoring and evaluation. The selection of indicators follows the principles of scientific rigor, availability, representativeness, dynamism, and operability. The preliminary candidate indicators cover five dimensions: ecological environment, social development, economic efficiency, governance capacity, and resource security. For example: the ecological environment dimension includes forest coverage, the proportion of days with good air quality, water resource utilization efficiency, and energy consumption per unit of GDP; the social development dimension includes the level of equal access to education, medical service coverage, life expectancy, and the urban-rural income ratio; the economic efficiency dimension includes regional GDP growth rate, the proportion of the digital economy, the added value of high-tech industries, and land development intensity; the governance capacity dimension includes the implementation rate of national land spatial planning, the area of ​​protected arable land, and the input-output ratio of ecological restoration; and the resource security dimension includes total water resources, arable land reserves, and mineral resource reserves. Each indicator is defined in detail, specifying its calculation formula, data source, statistical scope, and unit.

[0091] S1023, Determine the weights of the indicator system based on expert knowledge and data-driven methods. This sub-step is crucial to ensuring the scientific rigor and objectivity of the evaluation. The determination of indicator weights comprehensively employs expert decision-making methods and data-driven approaches. Expert decision-making methods include the Delphi method and the analytic hierarchy process (AHP), inviting experts from fields such as planning, economics, environment, and geographic information to score and rank the importance of each indicator, iteratively converging to form preliminary weights. Data-driven methods, through principal component analysis, entropy weighting, and the coefficient of variation method, objectively calculate indicator weights from the inherent distribution and correlation of the data, avoiding subjective bias. For example, the entropy weighting method determines weights based on the degree of variation in indicator values; the greater the variation, the greater the information content, and the higher the weight. The final weights are formed by integrating expert weights and objective weights, for example, by using a multiplicative synthesis method or a least squares weighted average, resulting in a scientifically sound and comprehensive weight system. This weight system requires sensitivity analysis to assess its impact on the stability of the evaluation results.

[0092] S1024, Establish a dynamic adjustment mechanism for the indicator system. This sub-step ensures that the indicator system can adapt to changes in regional development and policy updates. The dynamic adjustment mechanism is based on the following triggering conditions: major adjustments to national or local policies; changes in regional development goals; significant changes in the availability of new data sources or the quality of existing data sources; historical assessment results showing that certain indicators cannot accurately reflect the actual situation; and feedback from stakeholders. See also Figure 2 The adjustment process is implemented through an intelligent indicator system management module, which incorporates a rule engine and a machine learning model. The rule engine automatically introduces or adjusts relevant indicators based on preset adjustment rules, such as "when a policy is released." The machine learning model identifies potential shortcomings in the current indicator system by performing trend prediction and anomaly detection on historical data, and recommends new indicators or adjusts the weights of existing indicators. All adjustments undergo an approval process and are logged in detail to ensure the transparency and traceability of the indicator system.

[0093] Furthermore, in the aforementioned multi-level resource monitoring and evaluation method, step S103 involves calculating the multi-level resource monitoring and evaluation model. This step is the core of the evaluation method, aiming to combine preprocessed data with a pre-constructed indicator system and, through scientific model algorithms, quantitatively assess the regional resource endowment and governance effectiveness.

[0094] Step S103 further includes:

[0095] S1031 involves feature engineering and normalization of the preprocessed data. This sub-step transforms the raw data in the unified multimodal spatiotemporal database constructed in S101 into features suitable for the evaluation model. Feature engineering includes aggregating, decomposing, transforming, and combining the raw data to extract more meaningful information. For example, it involves generating continuous raster surfaces from scattered ecological environment monitoring point data through spatial interpolation, or calculating derived indicators such as population density and road network density. Subsequently, all indicator data are normalized to eliminate dimensional differences and ensure the comparability of different indicators in model calculations. Commonly used normalization methods include min-max normalization and Z-score normalization.

[0096] The minimum-maximum normalization formula is:

[0097]

[0098] in, X These are the original indicator values. X min It is the minimum value of the indicator. X max It is the maximum value of the indicator. X normalized It is the normalized index value.

[0099] S1032, see S1032. Figure 4 The core calculation is performed using a multi-level evaluation model. This sub-step, combining the indicator system and weights determined in S102, uses a multi-level evaluation model to comprehensively quantify the resource status and governance effectiveness at each level and dimension. The evaluation model includes, but is not limited to: the comprehensive index method, which calculates a comprehensive score through weighted summation; the fuzzy comprehensive evaluation method, which handles the uncertainty and fuzziness of indicators; the grey relational analysis method, used for data analysis of small samples and information-poor systems; and a spatial regression model based on machine learning, which models the nonlinear relationships between various elements. Model calculations are performed independently at different administrative levels. For example, basic elements are first evaluated at the township level, and then aggregated upwards to the county and city levels. The aggregation method uses weighted average or geometric average to ensure logical consistency between levels.

[0100] S1033, Perform spatial heterogeneity analysis and temporal evolution assessment. This sub-step delves into the spatial distribution characteristics and temporal variation patterns of the evaluation results. Spatial heterogeneity analysis employs spatial statistical tools from Geographic Information Systems (GIS), such as the local Moran's index and hotspot analysis, to identify spatial clusters and discrete areas of resource endowment and governance effectiveness, discovering "low-lying areas" and "high-lying areas" in regional development. Temporal evolution assessment compares and analyzes evaluation results at consecutive time points to identify trends such as growth, decline, and fluctuations in various indicators and comprehensive evaluation scores, revealing the dynamic trajectory and potential risks of regional development. For example, it uses time series forecasting models to assess future changes in resource carrying capacity.

[0101] S1034, perform reliability and sensitivity analysis on the evaluation results. This sub-step verifies the robustness of the evaluation results and the degree to which they are affected by indicator weights and model parameters. Reliability analysis uses statistical methods such as cross-validation and leave-one-out method to assess the model's generalization ability on different datasets. Sensitivity analysis observes the magnitude of changes in evaluation results by fine-tuning indicator weights or model parameters, identifying the key indicators or parameters that have the most significant impact on the results. For example, by changing the weight of a core indicator by 5%, the change in the overall evaluation score can be monitored to determine the sensitivity of that indicator. This analysis helps improve the credibility and decision-making reference value of the evaluation results.

[0102] Based on this, step S104 generates and visualizes the multi-dimensional monitoring and evaluation results. This step aims to transform the complex model calculation results into an intuitive and easy-to-understand form, so that different management entities can review, analyze, and make decisions.

[0103] As a further explanation of this embodiment, step S104 further includes:

[0104] S1041, Integrating Assessment Results from Different Levels and Dimensions. This sub-step logically categorizes and structures the multi-level and multi-dimensional calculation results generated in step S103. The integrated content includes, but is not limited to: the overall regional evaluation score, the scores of each primary and secondary indicator, the ranking of each administrative unit, the development potential index of a specific region, the resource carrying capacity assessment level, and the land space governance performance level. Data is presented in the form of structured reports, cross-analysis matrices, and key performance indicator dashboards, ensuring clear correlation and traceability among all results.

[0105] S1042, Generate Visualized Reports and Interactive Dashboards. This sub-step utilizes professional visualization tools and platforms to present the integrated evaluation results graphically. The visualized reports come in various formats, including thematic maps showing the spatial distribution of resources and evaluation scores, bar charts and line charts displaying the time-series trends of indicators, radar charts or word clouds showing the balance of performance across dimensions, and scatter plots and bubble charts revealing the correlations between indicators. The interactive dashboard provides flexible data drill-down and query functions. Users can dynamically generate customized visualized views by filtering administrative divisions, time ranges, evaluation dimensions, and other conditions. For example, users can click on a county on the map to immediately view a detailed evaluation report and historical trends for that county. The visualization interface design prioritizes user experience, ensuring the accuracy and intuitiveness of information delivery.

[0106] S1043 provides result output interfaces and customization services. This sub-step offers flexible result acquisition methods for different user groups and external systems. The system provides a standardized application programming interface, allowing other business systems or third-party applications to access evaluation result data, achieving data sharing and collaboration. For specific management departments or decision-makers, the system provides customized report generation services, automatically generating exclusive reports containing specific charts, data summaries, and policy recommendations based on their specific needs and concerns. Reports support multiple output formats, such as PDF, Excel, and images. Simultaneously, the system also provides an early warning function; when the evaluation score of a key indicator or region falls below a preset threshold, it automatically sends warning messages to relevant personnel, triggering an early intervention mechanism.

[0107] Finally, in step S105, see... Figure 5 This enables dynamic feedback and optimization of land space governance strategies. This step is key to achieving closed-loop management in this invention, aiming to transform evaluation results into specific policy recommendations and action plans, and to optimize land space governance practices through continuous monitoring.

[0108] As a further explanation of this embodiment, step S105 further includes:

[0109] S1051 analyzes key issues and development trends in the evaluation results. This sub-step provides an in-depth interpretation of the evaluation report and visualization results generated in step S104, identifying prominent problems, potential risks, and development opportunities in current regional land space governance. The analysis focuses not only on the absolute values ​​of evaluation scores but also on spatial patterns, temporal changes, and the interactions between indicators. For example, discovering that a region experiences rapid economic growth but a continuous decline in its ecological environment index reveals an imbalance between economic development and environmental protection; identifying persistent lags in the equalization of public services in some remote mountainous counties points to issues of unequal resource allocation. The analysis process combines the knowledge of domain experts with data analysis tools to generate a problem diagnostic report, providing clear guidance for subsequent strategy formulation.

[0110] S1052 proposes optimization suggestions for territorial spatial governance strategies based on problem diagnosis. This sub-step, based on the problem diagnosis results of S1051, and combined with national and local policy orientations, laws and regulations, and the actual situation of regional development, proposes specific and actionable optimization suggestions for territorial spatial governance strategies. The suggestions cover multiple aspects, such as: for ecological and environmental issues, strengthening ecological restoration, optimizing industrial structure, and implementing a strict environmental access system; for unbalanced economic development, increasing fiscal transfer payments to underdeveloped regions, guiding the gradient transfer of industries, and developing characteristic and advantageous industries; for shortcomings in public services, optimizing infrastructure layout, increasing investment in education and healthcare, and improving the social security system. The strategy optimization suggestions are generated with the assistance of expert systems and decision support models, and preliminary policy effect pre-assessments are conducted, such as analyzing the potential impact of different policy options through scenario simulation analysis.

[0111] S1053, Establish a mechanism for monitoring and continuously evaluating the effectiveness of strategies. This sub-step ensures that the proposed territorial spatial governance strategies can be effectively tracked and evaluated after implementation. A monitoring mechanism is established to quantify the key objectives and expected results of the implemented strategies into measurable indicators and incorporate them into the intelligent indicator system constructed in S102. Through data acquisition in S101 and evaluation calculations in S103, the regional development status after strategy implementation is periodically re-evaluated, resulting in a strategy effectiveness evaluation report. The continuous evaluation mechanism requires long-term tracking of strategy effectiveness and dynamic adjustment and iterative optimization of the strategies based on evaluation results, forming a closed-loop management process of "evaluation-decision-implementation-re-evaluation." This mechanism ensures the effectiveness and adaptability of territorial spatial governance strategies, supporting the continuous advancement of high-quality regional development.

[0112] This application also proposes a multi-level resource monitoring and evaluation system. This system implements the aforementioned multi-level resource monitoring and evaluation method. The system includes: a data acquisition and preprocessing module, an intelligent indicator system management module, a multi-level evaluation model calculation module, an evaluation result generation and visualization module, a dynamic feedback and optimization decision support module, and a unified multimodal spatiotemporal database. The overall system architecture employs a modular design to ensure the independence and synergy of each component, supporting flexible configuration and expansion.

[0113] The data acquisition and preprocessing module is responsible for implementing the functions of step S101. This module integrates multiple data interfaces, including an application programming interface for connecting to the government data sharing platform, an IoT communication protocol module for accessing real-time sensor data, and a spatial data processing engine for processing remote sensing imagery and geographic information data. Internally, it includes a data cleaning unit that uses machine learning algorithms, such as anomaly detection models, to identify and process errors and inconsistencies in the data; a spatiotemporal alignment unit that uses georeferencing algorithms and time-series resampling techniques to ensure the integration of heterogeneous data on a unified spatiotemporal reference; and a data standardization unit that converts data of different formats and data types into a unified internal representation.

[0114] See Figure 2 The intelligent indicator system management module is responsible for implementing the functions of step S102. This module includes an indicator definition and management interface, allowing users or administrators to add, delete, modify, and query various evaluation indicators, and define their calculation logic and data sources. Its core function lies in the dynamic weight allocation engine, which combines an expert knowledge rule base with a data-driven weight calculation model, such as the entropy weight calculator, to achieve automatic or semi-automatic updates of indicator weights. The module also includes a strategy adjustment suggestion unit, which can intelligently recommend adjustment schemes for the indicator system based on changes in the external environment or feedback from evaluation results, and record all change logs to ensure the transparency and traceability of the indicator system.

[0115] The multi-level evaluation model calculation module is responsible for implementing the functions of step S103. This module is the core computing engine of the system, with built-in libraries of various evaluation algorithms, such as a comprehensive index calculator, a fuzzy comprehensive evaluation model, a spatial regression analysis tool, and a time series prediction model. It receives the indicator system and weights from the intelligent indicator system management module, as well as preprocessed data from a unified multimodal spatiotemporal database. The module includes a parallel computing framework, capable of efficiently processing large-scale spatiotemporal data and simultaneously performing evaluation calculations at multiple administrative levels. Its result verification unit is responsible for performing reliability and sensitivity analyses of the model to ensure the robustness of the evaluation results.

[0116] The evaluation result generation and visualization module is responsible for implementing the functions of step S104. This module includes a report generator and a visualization engine. The report generator can automatically generate multi-dimensional and multi-level evaluation reports based on preset templates or user-defined needs, supporting output in various file formats. The visualization engine provides rich interactive charts and geographic information system map rendering functions. Users can configure dashboards through a drag-and-drop interface to view the spatial distribution, temporal evolution trend, and key indicator performance of regional evaluation results in real time. The module also provides an early warning notification service, promptly notifying relevant users via email or instant messaging when abnormal situations are detected.

[0117] The dynamic feedback and optimization decision support module is responsible for implementing the functions of step S105. This module acts as a bridge between evaluation and decision-making, incorporating a problem diagnosis expert system and a strategy optimization recommendation engine. The problem diagnosis expert system identifies shortcomings and contradictions in regional development by analyzing evaluation reports and combining them with preset knowledge rules. The strategy optimization recommendation engine, based on artificial intelligence algorithms such as reinforcement learning models and scenario simulation tools, generates multiple optimized land space governance strategies for the diagnosed problems and predicts their potential impact, providing decision-makers with a scientific reference. The module also tracks the effectiveness of implemented strategies and feeds back to the intelligent indicator system management module through evaluation reports, forming a closed-loop management system for continuous optimization.

[0118] The unified multimodal spatiotemporal database serves as the foundational data storage layer for the entire system, responsible for storing all raw data, preprocessed data, indicator system configuration information, intermediate model calculation results, and final evaluation reports. This database employs distributed storage and indexing technologies to support the management and efficient querying of large-scale spatiotemporal data. It provides standardized data access interfaces to ensure seamless data interaction between system modules and guarantees data integrity, consistency, and security.

[0119] The above are merely specific embodiments of the present invention, but the technical features of the present invention are not limited thereto. Any simple changes, equivalent substitutions, or modifications made based on the present invention to solve essentially the same technical problems and achieve essentially the same technical effects are all covered within the protection scope of the present invention.

Claims

1. A multi-level resource monitoring and evaluation method, characterized in that, include: Acquire multi-source heterogeneous spatiotemporal data, including remote sensing imagery, geographic information, statistical yearbooks, IoT sensor data, and departmental business data; The multi-source heterogeneous spatiotemporal data are cleaned, standardized, and integrated to construct a unified regional resource and governance spatiotemporal database; Based on the aforementioned regional resource and governance spatiotemporal database, a regional resource endowment assessment is constructed and executed to obtain a regional resource endowment classification profile, including natural ecology, human society and economic development potential indices. The construction and execution of the regional resource endowment assessment includes: Feature extraction is performed on the remote sensing image data, and the feature extraction adopts a temporal convolutional neural network based on a deep residual network; Land use types are finely classified, and the fine classification uses support vector machines; The spatial heterogeneity and interactions of various resource elements are analyzed using a geographic detector model; and Based on the analysis results, calculate and generate a resource endowment classification profile for the region; Based on the aforementioned regional resource and governance spatiotemporal database, a spatial governance effectiveness monitoring and evaluation system is constructed and implemented to obtain a spatial governance effectiveness evaluation report, including the degree of achievement of planning goals, ecological environment improvement, and public service fairness index. The construction and implementation of spatial governance effectiveness monitoring and evaluation includes: Land use scenario simulation was conducted by constructing a coupled model based on Markov chains and cellular automata to predict future land use change trends. The performance of policy implementation is quantitatively evaluated using a multi-criteria decision analysis method, which includes a weighting method combining the analytic hierarchy process (AHP) and the entropy weighting method. Based on the results of land use scenario simulation and the quantitative evaluation results of policy implementation performance, the aforementioned spatial governance effectiveness evaluation report is calculated and generated; The evaluation indicators and their weights can be dynamically adjusted through the intelligent indicator system management module to achieve customized configuration of the evaluation model; Based on the regional resource endowment classification profile, the spatial governance effectiveness assessment report, and the assessment indicators and weights configured in the intelligent indicator system management module, a multi-level assessment model is executed to generate a comprehensive regional assessment result. The calculation of the multi-level evaluation model includes: Based on the regional resource endowment classification profile and the spatial governance effectiveness assessment report, a local assessment is performed, which adopts a geographically weighted regression model. Based on the regional resource endowment classification profile and the spatial governance effectiveness assessment report, a comprehensive assessment is performed, which uses random forest or gradient boosting decision tree. Based on the assessment needs, select a multi-scale assessment strategy based on grid units, township administrative units, or county administrative units; and An uncertainty analysis was performed on the evaluation results, using Monte Carlo simulation. The comprehensive assessment results of the region are analyzed in depth and visualized to provide decision support; and Based on the comprehensive regional assessment results and decision support information, dynamic feedback and optimization will be conducted to provide policy adjustment suggestions and planning optimization schemes. The dynamic feedback and optimization specifically include: By constructing a policy intervention effect prediction model, regional shortcomings are identified based on the regional comprehensive assessment results, and targeted policy adjustment suggestions are proposed. The policy intervention effect prediction model adopts a Bayesian network or a deep reinforcement learning model. Predicting the potential impact of policy adjustments; and Based on historical policy effect feedback data, the parameters and weights of the policy intervention effect prediction model are adjusted to achieve closed-loop optimization of evaluation and decision-making.

2. The multi-level resource monitoring and evaluation method according to claim 1, characterized in that, Acquiring multi-source heterogeneous spatiotemporal data, including: Acquire multispectral high-resolution remote sensing image data and multitemporal remote sensing image data through satellite remote sensing data platform; Digital elevation model data, topographic data, soil type data, hydrological observation data, and land use status data are obtained through geographic information databases. Obtain population census data, socio-economic statistics, and fiscal input-output data from statistical yearbook databases; Real-time monitoring data on ecological and environmental quality and operational data of public service facilities are acquired through IoT sensor networks; and The system obtains land and space planning target texts, land use change survey data, and resident satisfaction survey data through the data interface of local government business systems.

3. The multi-level resource monitoring and evaluation method according to claim 1, characterized in that, The cleaning, standardization, and integration processing of the multi-source heterogeneous spatiotemporal data includes: The multi-source heterogeneous spatiotemporal data is collected through a data acquisition agent; Outlier detection and replacement are performed on the collected data, and the outlier detection and replacement adopts a rule-based outlier detection algorithm; The collected data undergoes time-series data smoothing processing, which employs the Kalman filtering algorithm. The collected data is standardized using Z-score standardization or min-max normalization; and The processed data is stored in a geospatial database whose data model conforms to the ISO 19115 standard.

4. The multi-level resource monitoring and evaluation method according to claim 1, characterized in that, The method of dynamically adjusting evaluation indicators and their weights through the intelligent indicator system management module includes: Through the user interface of the intelligent indicator system management module, select and define the name, unit of measurement, data source, calculation formula and evaluation threshold of the evaluation indicator; The weights of the evaluation indicators are configured through the intelligent indicator system management module. The weight configuration employs a combination of the analytic hierarchy process (AHP) and entropy-based weighting, and supports expert experience correction. The configured evaluation metrics and their weights are stored in a configuration file in Extensible Markup Language (EXPLAIN) or JSON format to enable dynamic loading and updating.

5. The multi-level resource monitoring and evaluation method according to claim 1, characterized in that, The in-depth analysis and visualization of the comprehensive assessment results of the region includes: The comprehensive evaluation results of the region are used to generate an interactive map application through the geographic information system rendering engine. The interactive map application supports multi-layer overlay, thematic map production, spatial query and analysis. The comprehensive regional assessment results are presented in the form of radar charts, bar charts, line charts, and scatter plots through a data visualization library, supporting data pivoting and drill-down analysis; and It provides a customizable report generation function, which allows users to select the evaluation content and display format.

6. A multi-level resource monitoring and evaluation system, characterized in that, include: The data integration module is used to acquire multi-source heterogeneous spatiotemporal data, including remote sensing images, geographic information, statistical yearbooks, IoT sensor data and departmental business data, and to clean, standardize and integrate the multi-source heterogeneous spatiotemporal data to build a unified regional resource and governance spatiotemporal database. The resource endowment assessment module is used to construct and execute a regional resource endowment assessment based on the aforementioned regional resource and governance spatiotemporal database, so as to obtain a regional resource endowment classification profile including natural ecology, human society and economic development potential index. The construction and execution of the regional resource endowment assessment includes: Feature extraction is performed on the remote sensing image data, and the feature extraction adopts a temporal convolutional neural network based on a deep residual network; Land use types are finely classified, and the fine classification uses support vector machines; The spatial heterogeneity and interactions of various resource elements are analyzed using a geographic detector model; and Based on the analysis results, calculate and generate a resource endowment classification profile for the region; The governance effectiveness monitoring module is used to construct and execute spatial governance effectiveness monitoring and evaluation based on the regional resource and governance spatiotemporal database, so as to obtain a spatial governance effectiveness evaluation report including the degree of achievement of planning goals, ecological environment improvement and public service fairness index; The construction and implementation of spatial governance effectiveness monitoring and evaluation includes: Land use scenario simulation was conducted by constructing a coupled model based on Markov chains and cellular automata to predict future land use change trends. The performance of policy implementation is quantitatively evaluated using a multi-criteria decision analysis method, which includes a weighting method combining the analytic hierarchy process (AHP) and the entropy weighting method. Based on the results of land use scenario simulation and the quantitative evaluation results of policy implementation performance, the aforementioned spatial governance effectiveness evaluation report is calculated and generated; The indicator system management module is used to dynamically adjust the evaluation indicators and their weights, enabling customized configuration of the evaluation model; The evaluation model calculation module is used to perform multi-level evaluation model calculations based on the regional resource endowment classification profile, the spatial governance effectiveness evaluation report, and the evaluation indicators and weights configured by the indicator system management module, and generate a comprehensive regional evaluation result. The calculation of the multi-level evaluation model includes: Based on the regional resource endowment classification profile and the spatial governance effectiveness assessment report, a local assessment is performed, which adopts a geographically weighted regression model. Based on the regional resource endowment classification profile and the spatial governance effectiveness assessment report, a comprehensive assessment is performed, which uses random forest or gradient boosting decision tree. Based on the assessment needs, select a multi-scale assessment strategy based on grid units, township administrative units, or county administrative units; and An uncertainty analysis was performed on the evaluation results, using Monte Carlo simulation. The decision support and visualization module is used to perform in-depth analysis and visualization of the comprehensive assessment results of the region, providing decision support; and The dynamic feedback and optimization module is used to provide dynamic feedback and optimization based on the comprehensive evaluation results and decision support information of the region, and to provide policy adjustment suggestions and planning optimization schemes. The dynamic feedback and optimization specifically include: By constructing a policy intervention effect prediction model, regional shortcomings are identified based on the regional comprehensive assessment results, and targeted policy adjustment suggestions are proposed. The policy intervention effect prediction model adopts a Bayesian network or a deep reinforcement learning model. Predicting the potential impact of policy adjustments; and Based on historical policy effect feedback data, the parameters and weights of the policy intervention effect prediction model are adjusted to achieve closed-loop optimization of evaluation and decision-making.

7. The multi-level resource monitoring and evaluation system according to claim 6, characterized in that, The data integration module is used for: It integrates multiple data interfaces, including an application programming interface for connecting to a government data sharing platform, an IoT communication protocol module for accessing real-time sensor data, and a spatial data processing engine for processing remote sensing imagery and geographic information data. It contains a data cleaning unit that uses machine learning algorithms to identify and process errors and inconsistencies in the data. Internally, it includes a spatiotemporal alignment unit, which uses georeferencing algorithms and time-series resampling technology to ensure the integration of heterogeneous data on a unified spatiotemporal reference; and It contains a data standardization unit that converts data of different formats and data types into a unified internal representation.

Images