A method and system for soil remediation and digital management

Abstract

The present application relates to the technical field of soil remediation management, and in particular to a method and system for soil remediation and digital management, the method comprising data construction, fusion of multi-source monitoring soil environment parameters and GIS layers, meteorological time series and other spatio-temporal background data to form a data set; state quantification, output of a quantitative remediation urgency and complexity comprehensive state index through a soil health assessment model; scheme generation, matching of a remediation target library and a resource library and generation of a digital remediation scheme containing an expected state trajectory through an optimization algorithm; closed-loop control, continuous collection of feedback data for trajectory comparison, deviation over a threshold value for re-calculation of an index and dynamic adjustment of the scheme, the system comprising a data construction module, a state quantification module, a scheme generation module and a closed-loop control module, and performing the above method. The present application improves data reliability and evaluation scientificity, guarantees scheme adaptability, realizes dynamic controllability of remediation, and promotes digital and efficient management of soil remediation.

Description

Technical Field

[0001] This invention relates to the field of soil remediation management technology, specifically to a method and system for soil remediation and digital management. Background Technology

[0002] In terms of soil condition assessment, existing assessment methods largely rely on human experience and lack a standardized, quantitative assessment system. The assessment process struggles to scientifically differentiate between the urgency and complexity of soil remediation, only providing qualitative descriptions of soil health status without the ability to generate intuitive, comparable quantitative indicators. This leads to a high degree of subjectivity in judging soil remediation needs, resulting in insufficient accuracy and scientific rigor, which in turn affects the rationality of subsequent remediation decisions.

[0003] In the remediation plan generation stage, traditional plan development often lacks a clear mechanism linking remediation target database and resource database. Plan designs are prone to deviating from actual cost limits, timeframe requirements, and expected environmental benefits, or may be mismatched with existing remediation technologies, equipment, and material reserves. Furthermore, the degree of plan optimization is low, making it difficult to achieve a multi-objective balance between remediation costs, remediation timeframes, and environmental benefits. Moreover, the generated plans are mostly static texts, lacking planning for the anticipated trajectory of soil condition changes during the remediation process, thus limiting their guiding role in remediation operations.

[0004] In terms of remediation process management, existing methods generally lack effective closed-loop control mechanisms. After the remediation operation begins, it is difficult to obtain dynamic feedback data on soil condition through continuous monitoring, making it impossible to compare the actual remediation progress with the expected goals in a timely manner. When deviations occur in the remediation process, the process for adjusting the plan cannot be triggered quickly, leading to the continuous accumulation of deviations. Ultimately, this may result in a prolonged remediation cycle, increased remediation costs, or even failure to achieve the preset remediation goals, seriously affecting the overall quality and efficiency of soil remediation work.

[0005] Therefore, a method and system for soil remediation and digital management are proposed to address the above problems. Summary of the Invention

[0006] The purpose of this invention is to provide a method and system for soil remediation and digital management to solve the problems mentioned in the background art.

[0007] To achieve the above objectives, the present invention provides the following technical solution:

[0008] A method for soil remediation and digital management includes the following steps:

[0009] Step S1, Data Construction: Soil environmental parameters collected by multi-source monitoring devices deployed in the target area are acquired and fused with spatiotemporal background data from external sources to form a spatiotemporally correlated soil state dataset; wherein, the spatiotemporal background data includes at least geographic information system layer data and meteorological time series data;

[0010] Step S2, State Quantification: Input the soil state dataset into a trained soil health assessment model. The soil health assessment model outputs a comprehensive state index to quantify the urgency and complexity of soil remediation.

[0011] Step S3, Scheme Generation: Match the comprehensive state index with the preset remediation target library and remediation resource library, and use an optimization algorithm to solve the multi-objective trade-off to generate a digital remediation scheme; the digital remediation scheme shall at least include a combination of remediation measures, implementation path, resource scheduling plan, and expected state trajectory of the remediation process based on the comprehensive state index; wherein, the remediation target library shall at least include the upper limit of remediation cost, remediation cycle requirements, and expected environmental benefit indicators;

[0012] Step S4, Closed-loop control: After the repair operation is started according to the digital repair plan, feedback data is continuously acquired through multi-source monitoring equipment and input into an adaptive controller. The adaptive controller compares the actual state trajectory generated by the feedback data with the expected state trajectory in the digital repair plan. When the deviation exceeds the threshold, it triggers the recalculation of the comprehensive state index and dynamically adjusts the digital repair plan accordingly. At the same time, the adjustment command is output to the repair execution equipment that performs the repair operation.

[0013] As a preferred approach, soil environmental parameters collected by multi-source monitoring devices deployed in the target area are acquired and fused with spatiotemporal background data from external sources to form a spatiotemporally correlated soil state dataset, including:

[0014] Step S1-1, Data Acquisition: Collect soil environmental parameters in real time from multi-source monitoring devices deployed in the target area. Soil environmental parameters include at least soil pH, heavy metal content, organic matter content, and humidity.

[0015] Step S1-2, Data Preprocessing: Clean and standardize the soil environmental parameters to eliminate outliers and unify the data format, and obtain the preprocessed soil environmental parameters.

[0016] Step S1-3, Spatiotemporal Background Data Acquisition: Obtain geographic information system (GIS) layer data and meteorological time series data from external sources. The GIS layer data shall include at least topography, land use type and soil type, and the meteorological time series data shall include at least precipitation, temperature and wind speed.

[0017] Step S1-4, Data Fusion: The preprocessed soil environmental parameters are spatiotemporally aligned and fused with the spatiotemporal background data. By associating timestamps and geographic coordinates, preliminary spatiotemporally correlated data is formed.

[0018] Steps S1-5: Dataset Construction: Organize the initially fused spatiotemporal correlated data into a structured dataset, in which each data point contains soil environmental parameters, corresponding spatiotemporal background data, and spatiotemporal labels to form a spatiotemporally correlated soil state dataset.

[0019] As a preferred approach, the state quantization in step S2 includes the following steps:

[0020] Step S2-1, Data Preparation: Normalize the spatiotemporally correlated soil state dataset to eliminate the influence of units and obtain a standardized dataset.

[0021] Step S2-2, Feature Extraction: Extract key features related to soil health from the standardized dataset. Key features include at least the degree of soil pollution, soil fertility indicators, and soil structure parameters.

[0022] Step S2-3, Model Evaluation: Input the key features into the trained soil health assessment model. The soil health assessment model calculates the soil remediation urgency score and remediation complexity score based on the key features.

[0023] Step S2-4, Index Synthesis: The soil remediation urgency score and remediation complexity score are weighted and fused to generate a comprehensive state index.

[0024] As a preferred approach, the solution generation in step S3 includes the following steps:

[0025] Step S3-1, Matching Analysis: Match the comprehensive state index with the preset remediation target library and remediation resource library. The remediation target library shall at least include the upper limit of remediation cost, remediation cycle requirements and expected environmental benefit indicators, and the remediation resource library shall at least include available remediation technologies, equipment and material resources. The matching analysis outputs a set of feasible remediation measures and resource constraints.

[0026] Step S3-2, Multi-objective optimization: Based on the set of feasible remediation measures and resource constraints, a multi-objective trade-off is solved using an optimization algorithm. The optimization algorithm uses remediation cost, remediation cycle, and environmental benefits as objective functions and outputs the optimized remediation strategy. The optimized remediation strategy includes a recommended sequence of remediation measures and a resource allocation scheme.

[0027] Step S3-3, Scheme Synthesis: Based on the optimized repair strategy, a digital repair scheme is generated. The digital repair scheme includes at least a combination of repair measures, an implementation path, and a resource scheduling plan. Based on the comprehensive state index and the combination of repair measures, a predictive model is used to generate the expected state trajectory of the repair process.

[0028] As a preferred embodiment, the closed-loop control in step S4 includes the following steps:

[0029] Step S4-1, Feedback Data Collection: After starting the remediation operation according to the digital remediation plan, soil environmental parameters are continuously collected as feedback data through multi-source monitoring equipment;

[0030] Step S4-2, Generation of Actual State Trajectory: Based on the feedback data collected in step S4-1, the comprehensive state index is calculated through the soil health assessment model, and the actual state trajectory is generated according to the time series.

[0031] Step S4-3, Trajectory Comparison and Deviation Calculation: Compare the actual state trajectory with the expected state trajectory in the digital repair plan, and calculate the deviation value between the two;

[0032] Step S4-4, Deviation Judgment: Determine whether the deviation value exceeds the preset threshold; if it does, trigger the recalculation of the comprehensive status index.

[0033] Step S4-5, Recalculate the comprehensive state index: Based on the latest feedback data collected in step S4-1, recalculate the comprehensive state index using the soil health assessment model to obtain an updated comprehensive state index;

[0034] Step S4-6, Dynamic Adjustment of Repair Scheme: Based on the updated comprehensive status index obtained in step S4-5, the digital repair scheme is re-matched with the repair target library and the repair resource library, and a multi-objective trade-off solution is obtained through optimization algorithm to dynamically adjust the digital repair scheme and generate the adjusted digital repair scheme.

[0035] Step S4-7: Adjustment command output: Output the adjustment command in the adjusted digital repair scheme generated in step S4-6 to the repair execution device that performs the repair operation, so as to control the repair execution device to perform the adjusted repair operation.

[0036] A system for soil remediation and digital management, the system executing methods for soil remediation and digital management, including:

[0037] The data construction module is used to acquire soil environmental parameters collected by multi-source monitoring devices deployed in the target area and integrate them with spatiotemporal background data from the outside to form a spatiotemporally correlated soil state dataset.

[0038] The state quantification module, connected to the data construction module, receives a spatiotemporally correlated soil state dataset and inputs it into a trained soil health assessment model, outputting a comprehensive state index to quantify the urgency and complexity of soil remediation.

[0039] The scheme generation module is connected to the state quantification module. It is used to match the comprehensive state index with the preset repair target library and repair resource library, and to generate a digital repair scheme by solving the multi-objective trade-off through an optimization algorithm.

[0040] The closed-loop control module, connected to the scheme generation module and the multi-source monitoring equipment, is used to continuously acquire feedback data through the multi-source monitoring equipment after the repair operation is started according to the digital repair scheme. It compares the actual state trajectory generated by the feedback data with the expected state trajectory in the digital repair scheme. When the deviation exceeds the threshold, it triggers the recalculation of the comprehensive state index and dynamically adjusts the digital repair scheme accordingly. At the same time, it outputs the adjustment command to the repair execution equipment that performs the repair operation.

[0041] As can be seen from the technical solutions provided by the present invention above, the method and system for soil remediation and digital management provided by the present invention have the following beneficial effects:

[0042] Improve the reliability of soil data support: Collect soil environmental parameters by multi-source monitoring equipment and integrate geographic information system layer data and meteorological time series data to form a spatiotemporally correlated soil state dataset. Eliminate data outliers and format differences to ensure that the data can fully reflect the actual soil state under different times and spaces, and provide accurate basic data for subsequent soil health assessment and remediation plan generation.

[0043] To achieve scientific soil condition assessment: By using a trained soil health assessment model to process the soil condition dataset, key features are extracted and remediation urgency and remediation complexity scores are quantified and calculated. A comprehensive condition index is generated through weighted fusion, avoiding the subjectivity of traditional assessments and making the judgment of soil remediation needs more objective and accurate.

[0044] Ensuring the adaptability and optimization of the remediation plan: Match the comprehensive state index with the preset remediation target library and remediation resource library, and combine the optimization algorithm to solve the multi-objective trade-off between remediation cost, remediation cycle and environmental benefits, and generate a digital remediation plan that includes the implementation path of remediation measures, resource scheduling plan and expected state trajectory, so as to ensure that the plan meets the project target requirements and is suitable for existing resource conditions.

[0045] Enhance the dynamic controllability of repair operations: During the repair operation, feedback data is continuously acquired through multi-source monitoring equipment, the actual state trajectory is constructed and compared with the expected state trajectory, and when the deviation exceeds the threshold, the comprehensive state index is recalculated in a timely manner and the repair plan is dynamically adjusted. The adjustment command is output to the repair execution equipment to avoid the repair process from deviating from the target and improve the stability of the repair effect.

[0046] Promoting the digitalization and efficiency of soil remediation management: The system works collaboratively through data construction modules, status quantification modules, scheme generation modules, and closed-loop control modules, covering the entire soil remediation process. It achieves digital management from data collection and evaluation to scheme generation, execution, and control, reducing manual intervention, minimizing human error, and improving the overall efficiency of soil remediation management. Attached Figure Description

[0047] Figure 1 This is a schematic diagram of the steps of a method for soil remediation and digital management according to the present invention;

[0048] Figure 2 This is a schematic diagram of the overall system framework for soil remediation and digital management according to the present invention. Detailed Implementation

[0049] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0050] To better understand the above technical solutions, the following will provide a detailed description of the technical solutions in conjunction with the accompanying drawings and specific embodiments.

[0051] like Figure 1-2 As shown, this embodiment of the invention provides a method for soil remediation and digital management, comprising the following steps:

[0052] Step S1, Data Construction: Soil environmental parameters collected by multi-source monitoring devices deployed in the target area are acquired and fused with spatiotemporal background data from external sources to form a spatiotemporally correlated soil state dataset; wherein, the spatiotemporal background data includes at least geographic information system layer data and meteorological time series data;

[0053] Step S2, State Quantification: Input the soil state dataset into a trained soil health assessment model. The soil health assessment model outputs a comprehensive state index to quantify the urgency and complexity of soil remediation.

[0054] Step S3, Scheme Generation: Match the comprehensive state index with the preset remediation target library and remediation resource library, and use an optimization algorithm to solve the multi-objective trade-off to generate a digital remediation scheme; the digital remediation scheme shall at least include a combination of remediation measures, implementation path, resource scheduling plan, and expected state trajectory of the remediation process based on the comprehensive state index; wherein, the remediation target library shall at least include the upper limit of remediation cost, remediation cycle requirements, and expected environmental benefit indicators;

[0055] Step S4, Closed-loop control: After the repair operation is started according to the digital repair plan, feedback data is continuously acquired through multi-source monitoring equipment and input into an adaptive controller. The adaptive controller compares the actual state trajectory generated by the feedback data with the expected state trajectory in the digital repair plan. When the deviation exceeds the threshold, it triggers the recalculation of the comprehensive state index and dynamically adjusts the digital repair plan accordingly. At the same time, the adjustment command is output to the repair execution equipment that performs the repair operation.

[0056] In this embodiment, step S1 aims to construct a soil state dataset covering the target area and containing spatiotemporal correlation information by fusing multi-source soil environmental parameter collection with external spatiotemporal background data. This provides a comprehensive and accurate data foundation for subsequent soil health assessment model input and comprehensive state index calculation, ensuring that the data accurately reflects the actual state of the soil at different time dimensions and spatial locations. The detailed steps are as follows:

[0057] Step S1-1, Data Acquisition:

[0058] Multi-source monitoring equipment deployment and parameter acquisition: Deploy multi-source monitoring equipment in the target area to collect soil environmental parameters in real time; the collected soil environmental parameters include at least soil pH, heavy metal content, organic matter content, and humidity; the acquisition process must be real-time to ensure that the acquired data can reflect the current soil environmental status in a timely manner and provide the latest basic data for subsequent data processing and fusion;

[0059] Step S1-2, Data Preprocessing:

[0060] Data cleaning operation: The collected soil environmental parameters are cleaned to remove outliers, such as parameter values ​​that are outside the reasonable range due to equipment failure, duplicate data, etc., so as to avoid abnormal data from interfering with the subsequent data fusion and analysis results.

[0061] Data standardization processing: The cleaned soil environmental parameters are standardized to unify the data format of different types of parameters, eliminate the data inconsistency caused by differences in units and dimensions, and obtain the pre-processed soil environmental parameters, which lays the data format foundation for subsequent fusion with spatiotemporal background data.

[0062] Step S1-3: Acquisition of spatiotemporal background data:

[0063] Geographic Information System (GIS) Layer Data Acquisition: Geographic Information System (GIS) layer data is acquired from external sources. This data includes at least the topography, land use type, and soil type of the target area. GIS layer data can provide spatial background information of the soil and reflect the differences in the basic properties of the soil at different spatial locations.

[0064] Meteorological time-series data acquisition: Meteorological time-series data is acquired from external sources. This data includes at least precipitation, temperature and wind speed in the target area. Meteorological time-series data can provide background information on the soil in the time dimension, reflecting the potential impact of meteorological conditions on soil conditions at different time stages.

[0065] Steps S1-4: Data Fusion

[0066] Spatiotemporal alignment operation: The preprocessed soil environmental parameters are spatiotemporally aligned with the acquired geographic information system layer data and meteorological time series data; the timestamp is used as the time dimension association basis to ensure that the soil environmental parameters match the meteorological time series data at the corresponding time; the geographic coordinates are used as the spatial dimension association basis to ensure that the soil environmental parameters match the geographic information system layer data at the corresponding spatial location.

[0067] Data fusion forms preliminary correlated data: By linking the data with timestamps and geographic coordinates, the preprocessed soil environmental parameters are fused with the spatiotemporal background data to form preliminary spatiotemporal correlated data, so that the data simultaneously contains soil environmental attributes, spatial background attributes and temporal background attributes.

[0068] Steps S1-5: Dataset Construction

[0069] Structured data organization: The initially fused spatiotemporal correlated data is organized according to a preset structure to form a structured dataset. In this structured dataset, each data point contains three parts of information: preprocessed soil environmental parameters, spatiotemporal background data corresponding to the soil environmental parameters (i.e., geographic information system layer data and meteorological time series data), and spatiotemporal labels (marking the time information and spatial location information corresponding to the data point).

[0070] Soil condition dataset formation: Through the above structured organization, a spatiotemporally correlated soil condition dataset is finally formed, which can be directly used as input for the soil health assessment model in the subsequent step S2.

[0071] In this embodiment, step S2 processes and analyzes the spatiotemporally correlated soil state dataset. Through a trained soil health assessment model, the soil state is transformed into a quantifiable comprehensive state index. This index simultaneously reflects the urgency and complexity of soil remediation, providing a core decision-making basis for generating a digital remediation plan in subsequent step S3. The detailed steps are as follows:

[0072] Step S2-1, Data Preparation:

[0073] The spatiotemporally correlated soil state dataset constructed in step S1 is normalized. The normalization operation eliminates the influence of different types of parameters (such as soil pH, heavy metal content, organic matter content, etc.) in the dataset due to differences in dimensions, ensuring that each parameter has equal weight in subsequent analysis, and finally obtains a standardized dataset, providing a data foundation with a unified scale for feature extraction and model evaluation.

[0074] Step S2-2, Feature Extraction:

[0075] From the standardized dataset obtained in step S2-1, key features directly related to soil health are extracted. The extracted key features include at least the degree of soil pollution, soil fertility index, and soil structure parameters. Among them, the degree of soil pollution is determined based on parameters such as heavy metal content in the dataset, the soil fertility index is calculated in combination with parameters such as organic matter content, and the soil structure parameters are extracted with reference to relevant attributes in the soil environmental parameters. These key features together constitute the input variables of the soil health assessment model.

[0076] Step S2-3, Model Evaluation:

[0077] The key features extracted in step S2-2 are input into the trained soil health assessment model. Based on the preset assessment logic and training data, the soil health assessment model analyzes and calculates the input key features and outputs two core results: soil remediation urgency score and remediation complexity score. The soil remediation urgency score reflects the degree of urgency of the current state of the soil for remediation operations, while the remediation complexity score reflects the degree of difficulty in overcoming technical, environmental, and other aspects required to carry out remediation operations.

[0078] Step S2-4, Exponent Synthesis:

[0079] The soil remediation urgency score and remediation complexity score obtained in steps S2-3 are weighted and fused. According to the actual needs and objectives of the soil remediation scenario (such as prioritizing pollution risk or remediation feasibility), corresponding weights are assigned to the two scores. The two are then integrated into a unified numerical index, namely the comprehensive state index, through weighted calculation. This index can be directly used to measure the overall health status of the soil and the remediation needs, providing a quantitative basis for the matching and optimization of remediation schemes in the subsequent step S3.

[0080] Furthermore, the soil health assessment model is a multi-indicator assessment model based on machine learning or statistics. Its core function is to quantify the urgency and complexity of soil remediation by analyzing soil environmental parameters and spatiotemporal background data. The specific structure and working principle of the soil health assessment model are as follows:

[0081] Model Input: The model input is a spatiotemporally correlated soil state dataset, including preprocessed soil environmental parameters (such as soil pH, heavy metal content, organic matter content, and humidity) and corresponding spatiotemporal background data (such as topography, land use type, soil type, precipitation, temperature, and wind speed). This data is transformed into key features through a feature extraction module, including at least:

[0082] Soil pollution level: calculated based on parameters such as heavy metal content, quantified by pollution indices (such as the Nemerow pollution index) or single-factor pollution indices;

[0083] Soil fertility indicators: calculated based on parameters such as organic matter content and pH value, and quantified by, for example, a comprehensive soil fertility index (such as a soil fertility score);

[0084] Soil structure parameters: calculated based on parameters such as soil moisture and particle composition, and quantified by soil porosity or aggregate stability index;

[0085] Model Structure: Soil health assessment models can be implemented using various machine learning algorithms, such as:

[0086] Random Forest model: It learns by ensemble learning from multiple decision trees, performs regression or classification on input features, and outputs urgency score and complexity score;

[0087] Neural Network Model: Uses multilayer perceptron (MLP) or convolutional neural network (CNN) to learn the mapping relationship between features and scores through nonlinear transformation;

[0088] Support Vector Machine (SVM) model: It uses kernel functions to process high-dimensional features and achieves regression prediction of scores;

[0089] Alternatively, the model can be based on a weighted scoring method, using expert experience or historical data to set the weights of each feature and calculate a comprehensive score.

[0090] Model output:

[0091] Soil remediation urgency score This score reflects the urgency of soil remediation efforts based on its current condition; a higher score indicates greater urgency. For example, the urgency score can be calculated by weighting pollution levels and ecological risk indicators, using the following formula: ,in, To score the urgency of soil remediation, As an index of soil pollution level, An ecological risk index calculated based on pollution levels and land use types. and The weighting coefficients are determined through the expert Delphi method or historical data regression analysis, and satisfy the following conditions: ;

[0092] Repair complexity score The complexity score reflects the degree of technical and environmental difficulties that need to be overcome to carry out remediation operations; a higher score indicates greater remediation complexity. For example, the complexity score can be calculated by weighting soil heterogeneity, meteorological conditions, and resource availability, using the following formula: ,in, This is a soil heterogeneity index calculated based on soil type and topographic data. Meteorological influencing factors, This is the resource availability factor (normalized to the range of 0-1). , and Let be the weight coefficient, and satisfy... ;

[0093] Comprehensive Status Index The urgency score and complexity score are weighted and fused together, and the formula is as follows: ,in, and Let be the weight coefficient, and satisfy... Adjustments can be made based on the remediation scenario (e.g., increasing the risk of contamination when prioritizing pollution risk). ;

[0094] Model training: The soil health assessment model is trained using historical soil data, which includes soil samples labeled with urgency and complexity. The training process includes feature selection, model parameter optimization, and validation to ensure the model's generalization ability. Alternatively, appropriate algorithms and training methods can be selected based on the actual data.

[0095] In this embodiment, step S3 involves matching a comprehensive state index quantifying the urgency and complexity of soil remediation with a pre-defined remediation target database and remediation resource database. This is combined with an optimization algorithm to perform multi-objective trade-offs, generating a digital remediation plan that includes a combination of remediation measures, implementation paths, resource scheduling plans, and the expected trajectory of the remediation process. This provides clear guidance for the precise implementation of subsequent remediation operations while ensuring that the plan meets pre-defined requirements for cost, timeframe, and environmental benefits. The detailed steps are as follows:

[0096] Step S3-1, Matching Analysis:

[0097] The remediation target database and resource database are organized as follows: The pre-set remediation target database stores the core requirements that the remediation operation must meet, including at least the upper limit of remediation cost, the remediation cycle requirement, and the expected environmental benefit indicators. These indicators are pre-set based on the soil remediation needs of the target area, the project budget, and environmental protection standards. The remediation resource database stores the resource information that can be used for the remediation operation, including at least the available remediation technologies, equipment, and material resources. The resource information will be dynamically updated according to the actual reserve situation.

[0098] Comprehensive State Index Matching Operation: The comprehensive state index generated in step S2 is matched with the remediation target library and the remediation resource library; based on the urgency and complexity of soil remediation reflected by the comprehensive state index, the remediation cost, cycle and environmental benefit requirements that match the index are screened in the remediation target library. At the same time, the remediation technologies, equipment and material resources that can meet the remediation requirements are searched in the remediation resource library, and options that exceed the resource reserve capacity or do not meet the target requirements are excluded.

[0099] Matching results output: Through the above matching process, the final output is a set of feasible remediation measures and resource constraints; the set of feasible remediation measures is all combinations of remediation technologies that meet the remediation target requirements and have corresponding resource support, and the resource constraints clearly indicate the resource usage restrictions that must be followed during the remediation process, such as the available time of specific equipment and the maximum supply of materials;

[0100] Step S3-2, Multi-objective optimization:

[0101] Determining the optimization boundary and target: The set of feasible remediation measures output in step S3-1 is used as the optimization target, and the resource constraints are used as the optimization boundary to ensure that the optimization process does not exceed the current resource support range, while meeting the preset cost, cycle and environmental benefit requirements in the remediation target library;

[0102] Optimization algorithm and objective function construction: An optimization algorithm is used to solve multi-objective trade-offs. The algorithm takes repair cost, repair cycle and environmental benefits as the core objective functions and constructs a multi-dimensional optimization model. During the solution process, the algorithm will coordinate the trade-offs of the three objective functions. For example, under the premise of controlling the repair cost to not exceed the preset upper limit and the repair cycle to not exceed the required duration, it will maximize the expected environmental benefits of the repair operation, or find the optimal balance among multiple objectives.

[0103] Optimization result generation: Through the solution calculation of the optimization algorithm, the optimized repair strategy is finally output; the repair strategy includes at least a recommended sequence of repair measures and a resource allocation scheme. The sequence of repair measures clarifies the implementation order of various repair technologies, while the resource allocation scheme specifies in detail the specific amount and time allocation of equipment and materials required for each repair measure.

[0104] Step S3-3, Scheme Synthesis:

[0105] Digital restoration solution integration: Based on the optimized restoration strategy output in step S3-2, a complete digital restoration solution is integrated; this solution includes at least a combination of restoration measures, an implementation path, and a resource scheduling plan. The combination of restoration measures is a specific presentation of the optimized sequence of restoration measures; the implementation path, combined with the geographic information system layer data (such as topography and land use type) of the target area, plans the spatial execution route of the restoration operation; the resource scheduling plan, based on the implementation progress of the restoration measures and the resource allocation plan, formulates a schedule for the transportation, allocation, and use of equipment and materials.

[0106] Generate the expected state trajectory of the remediation process: Based on the comprehensive state index generated in step S2 and the combination of remediation measures determined above, calculations are carried out through a prediction model. The prediction model simulates the changes in the comprehensive state of the soil at different time points during the remediation operation based on the initial soil state reflected by the comprehensive state index and the improvement law of the soil state by the remediation measures, and generates the expected state trajectory of the remediation process. This trajectory intuitively reflects the expected evolution trend of the soil health status from the start to the completion of the remediation operation.

[0107] Furthermore, the predictive model is used to generate the expected state trajectory of the remediation process. The core function of the predictive model is to predict the changes in the comprehensive state index at future time points based on the current soil condition and the combination of remediation measures. The specific structure and working principle of the predictive model are as follows:

[0108] Model input: The model input includes:

[0109] Comprehensive State Index: Output from the soil health assessment model;

[0110] The remediation measures package includes the remediation technologies used, the order of implementation, and the allocation of resources.

[0111] Spatiotemporal background data: such as meteorological time series data (precipitation, temperature, etc.) and geographic information system layer data (topography, soil type, etc.).

[0112] Model structure: The prediction model can employ time series forecasting algorithms or dynamic models, for example:

[0113] Autoregressive Integral Moving Average (ARIMA) model: used to predict future trends based on historical composite state index data;

[0114] Long Short-Term Memory Network (LSTM): A type of recurrent neural network (RNN) suitable for processing time series data and capable of capturing long-term dependencies;

[0115] System dynamics model: Differential equations are constructed based on the physicochemical principles of the soil remediation process to simulate state changes;

[0116] Regression models, such as multiple linear regression or random forest regression, use remediation measures and environmental impacts as features to predict state indices;

[0117] Model Output: The predictive model outputs the expected state trajectory during the repair process, i.e., the predicted comprehensive state index values ​​at a series of time points; the trajectory generation process includes:

[0118] Initialization: Starting from the current comprehensive state index;

[0119] Iterative prediction: For each time step, the state index is updated based on the impact of remediation measures (such as pollution removal rate and soil improvement effect) and environmental changes (such as the impact of rainfall on soil moisture);

[0120] Trajectory smoothing: The predicted values ​​are processed using moving average or filtering techniques to generate a smooth trajectory curve;

[0121] Model training: The prediction model is trained using historical restoration process data, which includes restoration measures, environmental conditions, and corresponding state index change sequences; the training objective is to minimize the prediction error (such as mean squared error).

[0122] In this embodiment, step S4 serves to continuously acquire feedback data through multi-source monitoring devices after the repair operation starts. It compares the actual state trajectory with the expected state trajectory using an adaptive controller. When the deviation exceeds a threshold, it triggers a recalculation of the comprehensive state index and dynamically adjusts the digital repair scheme. Simultaneously, it outputs adjustment commands to the repair execution equipment, achieving closed-loop control of the repair process. This ensures that the repair operation always conforms to the preset target, improving repair accuracy and efficiency. The detailed steps are as follows:

[0123] Step S4-1, Feedback Data Collection:

[0124] Monitoring equipment deployment and parameter acquisition: After the remediation operation is launched according to the digital remediation plan, the multi-source monitoring equipment already deployed in the target area is used to continuously collect soil environmental parameters and use them as feedback data. The collected soil environmental parameters include at least soil pH, heavy metal content, organic matter content and humidity. The collection process must be continuous to ensure that the dynamic changes of the soil environment during the remediation operation can be reflected in real time, so as to provide continuous data support for the subsequent generation of actual state trajectory.

[0125] Data transmission and storage: The collected feedback data needs to be transmitted to the data storage unit associated with the adaptive controller in real time. During the storage process, the time series integrity of the data must be maintained to avoid data loss or transmission delay, and to ensure that the complete feedback data sequence can be called up in subsequent calculations.

[0126] Step S4-2, Generation of actual state trajectory:

[0127] Comprehensive state index calculation: Based on the feedback data collected in step S4-1, it is input into the soil health assessment model consistent with step S2; the model first normalizes the feedback data to eliminate the influence of dimensions, then extracts key features such as soil pollution degree, soil fertility index, and soil structure parameters, and then calculates the soil remediation urgency score and remediation complexity score of the corresponding feedback data, and finally generates the comprehensive state index through weighted fusion.

[0128] Time series trajectory construction: According to the chronological order of the collection time of the feedback data, the comprehensive state index calculated at different time nodes is sequentially correlated to form the actual state trajectory with time as the horizontal axis and comprehensive state index as the vertical axis, which intuitively presents the dynamic change trend of soil state during the remediation process.

[0129] Step S4-3, Trajectory Comparison and Deviation Calculation:

[0130] Trajectory Correspondence Comparison: The actual state trajectory generated in step S4-2 is compared with the expected state trajectory included in the digital repair scheme in step S3 on a time-by-time basis; ensure that the actual comprehensive state index at each time point corresponds one-to-one with the expected comprehensive state index, and avoid deviation in comparison results due to time misalignment.

[0131] Deviation value calculation: For each corresponding time node, calculate the difference between the actual comprehensive status index and the expected comprehensive status index, and then obtain the overall deviation value through the preset deviation calculation rules (such as summing the absolute values ​​of the differences at each time node, or taking the square root of the sum of the squares of the differences at each time node) to quantitatively reflect the degree of difference between the actual repair process and the expected process.

[0132] Step S4-4, Deviation Judgment:

[0133] Threshold setting basis: The preset threshold is set in advance based on the expected environmental benefit indicators and repair accuracy requirements of the repair cycle in the repair target library. This threshold represents the maximum acceptable deviation range of the actual repair process. If it exceeds this range, the repair may fail to achieve the expected goal or exceed the preset cycle.

[0134] Deviation comparison judgment: Compare the deviation value calculated in step S4-3 with the preset threshold; if the deviation value does not exceed the preset threshold, it is determined that the current repair operation meets the expectations and continues to be executed according to the original digital repair plan; if the deviation value exceeds the preset threshold, the subsequent comprehensive status index recalculation process is triggered.

[0135] Step S4-5: Recalculate the comprehensive state index:

[0136] Latest data selection: Select the latest feedback data from the continuously collected feedback data in step S4-1 when the deviation judgment is triggered, to ensure that the data used can truly reflect the actual state of the soil and avoid using lagging data that could lead to inaccurate calculation results;

[0137] Index recalculation: Input the latest selected feedback data into the soil health assessment model, repeat the data preparation, feature extraction, model assessment, and index synthesis process in step S2, and recalculate the updated comprehensive state index; this updated index will serve as the core basis for subsequent remediation plan adjustments to ensure that the adjustment direction is in line with the current soil conditions.

[0138] Step S4-6: Dynamic adjustment of the repair plan:

[0139] Rematching analysis: The updated comprehensive state index obtained in steps S4-5 is rematched with the preset remediation target library and remediation resource library. The remediation target library is still based on the remediation cost ceiling, remediation cycle requirements and expected environmental benefit indicators. The remediation resource library is still based on the available remediation technology, equipment and material resources. By matching, remediation measures and resource allocations that cannot be met at present are excluded, and a new set of feasible remediation measures and resource constraints are output.

[0140] Multi-objective optimization solution: Based on the new set of feasible remediation measures and resource constraints, the multi-objective trade-off solution is performed again through the optimization algorithm; the optimization algorithm still uses remediation cost, remediation cycle and environmental benefits as objective functions to calculate the optimized new remediation strategy, including the new remediation measure sequence and resource allocation scheme;

[0141] Solution Adjustment Generation: Based on the new optimized repair strategy, the original digital repair solution is adjusted, the implementation path resource scheduling plan of the repair measures combination is updated, and based on the updated comprehensive status index and the new repair measures combination, the expected status trajectory of the repair process is regenerated through the prediction model, and finally the adjusted digital repair solution is formed.

[0142] Step S4-7: Adjust the instruction output:

[0143] Instruction Extraction and Organization: Extract adjustment instructions related to the repair execution equipment from the adjusted digital repair plan generated in steps S4-6, including instructions for switching repair technologies, instructions for adjusting equipment operating parameters, and instructions for scheduling material supply, etc., to ensure that the instructions are clear, unambiguous and unambiguous.

[0144] Command transmission and execution: The adjusted commands are output to the repair execution equipment that performs the repair work. The transmission process must ensure real-time performance and stability to ensure that the repair execution equipment can receive the commands in a timely manner and perform the repair work according to the adjusted requirements, so as to realize the dynamic implementation of the repair plan and complete the execution link of closed-loop control.

[0145] A system for soil remediation and digital management is disclosed. The core function of the system is to execute the aforementioned methods for soil remediation and digital management (i.e., steps S1 to S4 and their sub-steps). Through the collaborative work of four functional modules, it achieves full-process digital management, from soil data acquisition and fusion, quantitative assessment of soil condition, and generation of digital remediation plans to closed-loop control of the remediation process. This ensures that soil remediation operations are accurate, efficient, and aligned with preset objectives. The specific functions, operating logic, and inter-module relationships of each module are as follows:

[0146] I. Data Construction Module:

[0147] The data construction module is the system's fundamental data support unit. Its main function is to acquire soil environmental parameters of the target area and fuse them with external spatiotemporal background data to form a spatiotemporally correlated soil state dataset, providing comprehensive and accurate input data for subsequent modules. Its specific workflow and operational details are as follows:

[0148] Data acquisition sub-function: This module establishes a real-time data interaction channel with multi-source monitoring equipment deployed in the target area, and continuously collects soil environmental parameters through the multi-source monitoring equipment; the collected soil environmental parameters include at least soil pH value, heavy metal content, organic matter content and humidity. The acquisition process is kept in real time to ensure that the data can reflect the current soil environmental status in a timely manner and avoid the impact of data lag on subsequent analysis.

[0149] The data preprocessing sub-function performs data cleaning and standardization on the collected soil environmental parameters. The data cleaning step mainly removes outliers (such as parameter values ​​that are outside the reasonable range due to equipment failure or duplicate data records) to eliminate the interference of outlier data on subsequent analysis. The standardization step unifies the data format of different types of parameters, eliminates data inconsistencies caused by differences in units and dimensions, and finally obtains the preprocessed soil environmental parameters.

[0150] Spatiotemporal background data acquisition sub-function: Acquire spatiotemporal background data from external data sources (such as geographic information platforms and meteorological monitoring platforms); among which, geographic information system layer data includes at least the topography, land use type and soil type of the target area, which is used to provide spatial background information of the soil; meteorological time series data includes at least the precipitation, temperature and wind speed of the target area, which is used to provide temporal background information of the soil and reflect the potential impact of meteorological conditions on soil conditions.

[0151] Data fusion sub-function: It performs spatiotemporal alignment and fusion of preprocessed soil environmental parameters with acquired spatiotemporal background data; it uses timestamps as the basis for time dimension association to ensure that soil environmental parameters match meteorological time series data at the corresponding time; it uses geographic coordinates as the basis for spatial dimension association to ensure that soil environmental parameters match geographic information system layer data at the corresponding spatial location, and forms preliminary fused spatiotemporal associated data through dual association.

[0152] The dataset construction sub-function organizes the initially fused spatiotemporal correlated data into a structured dataset. Each data point contains three core pieces of information: preprocessed soil environmental parameters, corresponding spatiotemporal background data (GIS layer data and meteorological time series data), and spatiotemporal labels (time information and spatial location information corresponding to the labeled data points). Finally, a spatiotemporally correlated soil state dataset is formed, and this dataset is transmitted to the state quantification module.

[0153] II. State Quantization Module:

[0154] The soil condition quantification module, connected to the data construction module, is the core unit of the system's soil condition assessment. Its main function is to receive spatiotemporally correlated soil condition datasets and, through a trained soil health assessment model, transform the soil condition into a quantifiable comprehensive condition index, providing a decision-making basis for the solution generation module. Its specific workflow and operational details are as follows:

[0155] Data preparation sub-function: Receives the spatiotemporally correlated soil state dataset transmitted by the data construction module, performs normalization processing on the dataset; eliminates the influence of different parameters (such as pH value and heavy metal content) in the dataset due to differences in units through normalization operation, ensures that the weights of each parameter are consistent in subsequent analysis, and finally obtains a standardized dataset.

[0156] Feature extraction sub-function: Extract key features directly related to soil health from the standardized dataset; the extracted key features include at least the degree of soil pollution (determined based on parameters such as heavy metal content in the dataset), soil fertility indicators (calculated in combination with parameters such as organic matter content), and soil structure parameters (extracted by referring to relevant attributes in soil environmental parameters). These key features together constitute the input variables of the soil health assessment model.

[0157] Model evaluation sub-function: The extracted key features are input into the trained soil health assessment model; the model analyzes and calculates the key features based on the preset assessment logic and training data, and outputs two core results: soil remediation urgency score (reflecting the degree of urgency of the current soil condition for remediation work) and remediation complexity score (reflecting the degree of technical and environmental difficulties that need to be overcome to carry out remediation work).

[0158] Index Synthesis Sub-function: Weighted fusion of soil remediation urgency score and remediation complexity score; Set the weights of the two types of scores according to the actual needs of the soil remediation scenario (such as prioritizing pollution risk or remediation feasibility), integrate the two into a unified numerical index through weighted calculation - comprehensive state index, and transmit the comprehensive state index to the scheme generation module.

[0159] III. Solution Generation Module:

[0160] The solution generation module, connected to the state quantification module, is the core unit for designing repair solutions. Its main function is to match the comprehensive state index with a preset repair target library and repair resource library, and then use optimization algorithms to solve multi-objective trade-offs, generating a digital repair solution to provide clear guidance for repair operations. Its specific workflow and operational details are as follows:

[0161] Matching Analysis Sub-function: This function calls upon preset remediation target and remediation resource libraries and correlates the comprehensive status index transmitted by the status quantification module with these two libraries. The remediation target library includes at least the upper limit of remediation cost, remediation cycle requirements, and expected environmental benefit indicators (preset based on project needs and environmental standards). The remediation resource library includes at least available remediation technologies, equipment, and material resources (dynamically updated based on actual reserves). By matching and eliminating options that exceed resource capacity or do not meet target requirements, it outputs a set of feasible remediation measures and resource constraints (such as equipment availability time and maximum material supply).

[0162] Multi-objective optimization sub-function: Taking the set of feasible remediation measures as the optimization object and resource constraints as the optimization boundary, the optimization algorithm is used to solve the multi-objective trade-off. The optimization algorithm takes remediation cost, remediation period, and environmental benefits as objective functions. Under the premise of controlling the remediation cost to not exceed the upper limit and the remediation period to not exceed the required duration, it maximizes the environmental benefits and finally outputs the optimized remediation strategy, including the recommended sequence of remediation measures (clearly defining the order of technical implementation) and the resource allocation scheme (specifying the amount of equipment and materials and the allocation of time periods).

[0163] The scheme synthesis sub-function integrates the optimized remediation strategy to form a complete digital remediation scheme. This scheme includes at least a combination of remediation measures (a specific presentation of the optimized remediation measure sequence), an implementation path (a spatial execution route planned based on geographic information system layer data), and a resource scheduling plan (an equipment and material allocation schedule based on the resource allocation plan). Simultaneously, based on the comprehensive state index and the combination of remediation measures, a predictive model is used to simulate soil state changes during the remediation process, generating the expected state trajectory of the remediation process (a change curve with time as the horizontal axis and the comprehensive state index as the vertical axis). Finally, the digital remediation scheme is transmitted to the closed-loop control module.

[0164] IV. Closed-loop control module:

[0165] The closed-loop control module is connected to both the scheme generation module and the multi-source monitoring equipment. It is the core unit for dynamic control of the system's repair process. Its main function is to continuously acquire feedback data during repair and operation, compare the actual and expected trajectory, and dynamically adjust the scheme and output instructions when deviations exceed thresholds, ensuring that the repair operation conforms to the preset goals. Its specific workflow and operational details are as follows:

[0166] Feedback data acquisition sub-function: After the remediation operation is started according to the digital remediation plan transmitted by the scheme generation module, a continuous data interaction is established with the multi-source monitoring equipment. The soil environmental parameters (the same parameter types as in step S1-1) are collected in real time by the multi-source monitoring equipment as feedback data, and the feedback data is temporarily stored in the data storage unit built into the module to ensure the integrity of the data time series.

[0167] The actual state trajectory generation sub-function calls feedback data from the data storage unit and inputs it into the soil health assessment model (which is consistent with the model used in the state quantification module). It repeats the index calculation process of the state quantification module (normalization, feature extraction, model evaluation, and index synthesis) to obtain the comprehensive state index at different time points. Then, according to the order of the feedback data collection time, it associates the comprehensive state indices of each time point to form an actual state trajectory with time as the horizontal axis and comprehensive state index as the vertical axis.

[0168] The trajectory comparison and deviation calculation sub-function calls the expected state trajectory in the digital repair plan transmitted by the plan generation module, compares the actual state trajectory with the expected state trajectory at each time node; for each corresponding time node, it calculates the difference between the actual comprehensive state index and the expected comprehensive state index, and then calculates the overall deviation value through preset rules (such as summing the absolute values ​​of the differences at each node, taking the square root of the sum of squares of the differences), quantitatively reflecting the difference between the actual repair progress and the expected progress.

[0169] Deviation Judgment and Index Recalculation Sub-function: The calculated deviation value is compared with the preset threshold (set according to the cycle requirements, environmental benefit indicators and remediation accuracy requirements in the remediation target library); if the deviation value does not exceed the threshold, the remediation operation is determined to meet expectations and continues to be executed according to the original plan; if the deviation value exceeds the threshold, the latest feedback data is retrieved from the data storage unit and re-entered into the soil health assessment model to calculate the comprehensive state index, and an updated comprehensive state index is obtained.

[0170] The dynamic adjustment sub-function of the repair plan: The updated comprehensive status index is rematched with the repair target library and the repair resource library. The matching analysis and multi-objective optimization process of the plan generation module are repeated to output a new set of feasible repair measures, resource constraints and optimized repair strategies. Then, the original digital repair plan is adjusted according to the new repair strategy, updating the combination of repair measures, implementation path and resource scheduling plan. The expected state trajectory of the repair process is regenerated through the prediction model to form the adjusted digital repair plan.

[0171] Adjustment instruction output sub-function: Extract adjustment instructions (such as repair technology switching instructions, equipment operating parameter adjustment instructions, and material supply scheduling instructions) related to the repair execution equipment from the adjusted digital repair plan, ensuring that the instructions are clear and unambiguous; then transmit the adjustment instructions to the repair execution equipment that performs the repair operation, ensuring the real-time and stable transmission, controlling the repair execution equipment to perform the operation according to the adjusted plan, and completing the closed-loop control of the repair process;

[0172] V. Module Collaboration Mechanism:

[0173] The system's four main modules form a complete workflow through data interaction and functional integration: the data construction module provides basic data for the state quantification module, the state quantification module provides core decision indices for the solution generation module, the solution generation module provides initial remediation solutions for the closed-loop control module, and the closed-loop control module, after acquiring feedback data from multi-source monitoring equipment, not only achieves its own dynamic regulation but also reverse-links the core logic of the state quantification module by recalculating the comprehensive state index, ensuring the seamless flow of the entire system from data input to solution execution and dynamic adjustment, ultimately achieving digital, precise, and closed-loop management of soil remediation.

[0174] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for soil remediation and digital management, characterized in that: Includes the following steps: Step S1, Data Construction: Soil environmental parameters collected by multi-source monitoring devices deployed in the target area are acquired and fused with spatiotemporal background data from external sources to form a spatiotemporally correlated soil state dataset; wherein, the spatiotemporal background data includes at least geographic information system layer data and meteorological time series data; Step S2, State Quantification: Input the soil state dataset into a trained soil health assessment model. The soil health assessment model outputs a comprehensive state index to quantify the urgency and complexity of soil remediation. Step S3, Scheme Generation: Match the comprehensive state index with the preset remediation target library and remediation resource library, and use an optimization algorithm to solve for multi-objective trade-offs to generate a digital remediation scheme. The digital remediation scheme should at least include a combination of remediation measures, implementation path, resource scheduling plan, and the expected state trajectory of the remediation process based on the comprehensive state index prediction. The remediation target library should at least include the upper limit of remediation cost, remediation cycle requirements, and expected environmental benefit indicators. The scheme generation in step S3 includes the following steps: Step S3-1, Matching Analysis: Match the comprehensive state index with the preset remediation target library and remediation resource library. The remediation target library shall at least include the upper limit of remediation cost, remediation cycle requirements and expected environmental benefit indicators, and the remediation resource library shall at least include available remediation technologies, equipment and material resources. The matching analysis outputs a set of feasible remediation measures and resource constraints. Step S3-2, Multi-objective optimization: Based on the set of feasible remediation measures and resource constraints, a multi-objective trade-off is solved using an optimization algorithm. The optimization algorithm uses remediation cost, remediation cycle, and environmental benefits as objective functions and outputs the optimized remediation strategy. The optimized remediation strategy includes a recommended sequence of remediation measures and a resource allocation scheme. Step S3-3, Scheme Synthesis: Based on the optimized repair strategy, a digital repair scheme is generated. The digital repair scheme includes at least a combination of repair measures, an implementation path, and a resource scheduling plan. Based on the comprehensive state index and the combination of repair measures, a predictive model is used to generate the expected state trajectory of the repair process. Step S4, Closed-loop control: After the repair operation is started according to the digital repair plan, feedback data is continuously acquired through multi-source monitoring equipment and input into an adaptive controller. The adaptive controller compares the actual state trajectory generated by the feedback data with the expected state trajectory in the digital repair plan. When the deviation exceeds the threshold, it triggers the recalculation of the comprehensive state index and dynamically adjusts the digital repair plan accordingly. At the same time, the adjustment command is output to the repair execution equipment that performs the repair operation.

2. The method for soil remediation and digital management according to claim 1, characterized in that: Soil environmental parameters collected by multi-source monitoring devices deployed in the target area were acquired and fused with spatiotemporal background data from external sources to form a spatiotemporally correlated soil state dataset, including: Step S1-1, Data Acquisition: Collect soil environmental parameters in real time from multi-source monitoring devices deployed in the target area. Soil environmental parameters include at least soil pH, heavy metal content, organic matter content, and humidity. Step S1-2, Data Preprocessing: Clean and standardize the soil environmental parameters to eliminate outliers and unify the data format, and obtain the preprocessed soil environmental parameters. Step S1-3, Spatiotemporal Background Data Acquisition: Obtain geographic information system (GIS) layer data and meteorological time series data from external sources. The GIS layer data shall include at least topography, land use type and soil type, and the meteorological time series data shall include at least precipitation, temperature and wind speed. Step S1-4, Data Fusion: The preprocessed soil environmental parameters are spatiotemporally aligned and fused with the spatiotemporal background data. By associating timestamps and geographic coordinates, preliminary spatiotemporally correlated data is formed. Steps S1-5: Dataset Construction: Organize the initially fused spatiotemporal correlated data into a structured dataset, in which each data point contains soil environmental parameters, corresponding spatiotemporal background data, and spatiotemporal labels to form a spatiotemporally correlated soil state dataset.

3. The method for soil remediation and digital management according to claim 1, characterized in that: The state quantization in step S2 includes the following steps: Step S2-1, Data Preparation: Normalize the spatiotemporally correlated soil state dataset to eliminate the influence of units and obtain a standardized dataset. Step S2-2, Feature Extraction: Extract key features related to soil health from the standardized dataset. Key features include at least the degree of soil pollution, soil fertility indicators, and soil structure parameters. Step S2-3, Model Evaluation: Input the key features into the trained soil health assessment model. The soil health assessment model calculates the soil remediation urgency score and remediation complexity score based on the key features. Step S2-4, Index Synthesis: The soil remediation urgency score and remediation complexity score are weighted and fused to generate a comprehensive state index.

4. The method for soil remediation and digital management according to claim 1, characterized in that: The closed-loop control in step S4 includes the following steps: Step S4-1, Feedback Data Collection: After starting the remediation operation according to the digital remediation plan, soil environmental parameters are continuously collected as feedback data through multi-source monitoring equipment; Step S4-2, Generation of Actual State Trajectory: Based on the feedback data collected in step S4-1, the comprehensive state index is calculated through the soil health assessment model, and the actual state trajectory is generated according to the time series. Step S4-3, Trajectory Comparison and Deviation Calculation: Compare the actual state trajectory with the expected state trajectory in the digital repair plan, and calculate the deviation value between the two; Step S4-4, Deviation Judgment: Determine whether the deviation value exceeds the preset threshold; if it does, trigger the recalculation of the comprehensive status index. Step S4-5, Recalculate the comprehensive state index: Based on the latest feedback data collected in step S4-1, recalculate the comprehensive state index using the soil health assessment model to obtain an updated comprehensive state index; Step S4-6, Dynamic Adjustment of Repair Scheme: Based on the updated comprehensive status index obtained in step S4-5, the digital repair scheme is re-matched with the repair target library and the repair resource library, and a multi-objective trade-off solution is obtained through optimization algorithm to dynamically adjust the digital repair scheme and generate the adjusted digital repair scheme. Step S4-7: Adjustment command output: Output the adjustment command in the adjusted digital repair scheme generated in step S4-6 to the repair execution device that performs the repair operation, so as to control the repair execution device to perform the adjusted repair operation.

5. A system for soil remediation and digital management, characterized in that: The system performs the method for soil remediation and digital management as described in any one of claims 1-4, comprising: The data construction module is used to acquire soil environmental parameters collected by multi-source monitoring devices deployed in the target area and integrate them with spatiotemporal background data from the outside to form a spatiotemporally correlated soil state dataset. The state quantification module, connected to the data construction module, receives a spatiotemporally correlated soil state dataset and inputs it into a trained soil health assessment model, outputting a comprehensive state index to quantify the urgency and complexity of soil remediation. The scheme generation module is connected to the state quantification module. It is used to match the comprehensive state index with the preset repair target library and repair resource library, and to generate a digital repair scheme by solving the multi-objective trade-off through an optimization algorithm. The closed-loop control module, connected to the scheme generation module and the multi-source monitoring equipment, is used to continuously acquire feedback data through the multi-source monitoring equipment after the repair operation is started according to the digital repair scheme. It compares the actual state trajectory generated by the feedback data with the expected state trajectory in the digital repair scheme. When the deviation exceeds the threshold, it triggers the recalculation of the comprehensive state index and dynamically adjusts the digital repair scheme accordingly. At the same time, it outputs the adjustment command to the repair execution equipment that performs the repair operation.

Images