Method and system for agricultural meteorological intelligent decision support combined with multi-source heterogeneous data

Abstract

The present application relates to the technical field of agricultural data, in particular to an agricultural meteorological intelligent decision support method and system combined with multi-source heterogeneous data, comprising the following steps: step S1. Multi-source heterogeneous data acquisition and processing: obtaining agricultural multi-source heterogeneous data through official and monitoring equipment, and preprocessing the multi-source heterogeneous data; step S2. Data extraction and fusion: extracting multi-source heterogeneous data according to relevance; the present application fuses meteorological, soil and crop multi-dimensional data through a space-time graph neural network, combines a dynamic adjacency matrix and an attention mechanism, effectively captures the space-time correlation characteristics of farmland areas, solves the data island problem of traditional methods, improves the disaster identification accuracy, and through the introduction of a transfer learning mechanism and a federal learning framework, realizes the rapid migration of the main production area model to the new planting area, simultaneously encrypts and aggregates the farm node data through a parameter server, completes collaborative modeling under the premise of protecting the privacy of agricultural enterprises, and avoids the risk of sensitive data leakage.

Description

Technical Field

[0001] This invention relates to the field of agricultural data technology, specifically to an intelligent decision support method and system for agricultural meteorology that combines multi-source heterogeneous data. Background Technology

[0002] With global population growth, limited land resources, and the challenges posed by climate change to agricultural production, improving agricultural productivity, ensuring food security, and achieving sustainable agricultural development have become major issues of international concern. At the same time, rapid technological advancements, particularly innovations in information technology, the Internet of Things, big data analytics, and artificial intelligence, have provided strong technical support for addressing these challenges.

[0003] For example, a remote sensing monitoring method for crop diseases based on image processing and deep learning, with application number CN202110522015.4 and publication date 20210806, includes the following steps: S1) acquiring remote sensing images of crops; S2) inputting the remote sensing images of crops into a trained YOLOv3 deep learning network based on the Keras and TensorFlow framework to obtain recognition results and realize remote sensing monitoring of crop diseases. This invention realizes remote sensing monitoring of crop diseases and has the advantages of fast recognition speed, high accuracy, high efficiency, and low dependence on farmers' supervision. It works in conjunction with hierarchical detection based on feature extraction algorithms, which not only gives full play to the stability and flexibility of deep learning monitoring effects, but also reflects the advantages of high precision and simple and efficient data annotation of traditional image processing algorithms. It avoids the time-consuming and labor-intensive process of collecting thousands of real object images and manually annotating them before training the neural network for application.

[0004] For example, application number CN201611127292.0, with an authorization announcement date of 20191203, describes an irrigation decision-making system and method based on an agricultural system model, belonging to the field of agricultural irrigation decision-making technology. This system includes a meteorological data interface, a meteorological database, a calibrated agricultural system model, a calibrated agricultural system model library, an irrigation decision algorithm, an irrigation decision algorithm library, and an irrigation control system. The meteorological data interface provides the calibrated agricultural system model with the meteorological data required for calculation. The calibrated agricultural system model calculates based on meteorological conditions to obtain crop growth information and soil and crop moisture status. The irrigation decision algorithm determines whether to irrigate and calculates the irrigation amount based on the model's calculation results. The irrigation control system irrigates according to the irrigation amount calculated by the irrigation decision algorithm. This invention overcomes the shortcomings of existing irrigation decision-making systems and methods, formulating accurate irrigation plans without installing field sensors, and predicting future irrigation plans and effects.

[0005] Traditional agricultural meteorological decision-making relies primarily on a single data source, lacking dynamic assessment and real-time feedback optimization mechanisms for the coupled risks of multiple hazards. Furthermore, existing methods still face technical bottlenecks in areas such as cross-regional adaptability, data privacy protection, and decision interpretability. Therefore, there is an urgent need for an intelligent agricultural meteorological decision support method and system that integrates multi-source heterogeneous data to address these issues. Summary of the Invention

[0006] The purpose of this invention is to provide an intelligent decision support method and system for agricultural meteorology that combines multi-source heterogeneous data, so as to overcome the above-mentioned shortcomings in the prior art.

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

[0008] An intelligent decision support method for agricultural meteorology that combines multi-source heterogeneous data is characterized by the following steps:

[0009] Step S1. Acquisition and processing of multi-source heterogeneous data: Acquire agricultural multi-source heterogeneous data through official channels and monitoring equipment, and preprocess the multi-source heterogeneous data;

[0010] In step S1:

[0011] Official data includes historical temperature / precipitation / wind speed data released by the meteorological bureau, crop growth cycle data and soil moisture reports provided by the agricultural sector;

[0012] Monitoring equipment data is acquired through UAV remote sensing images, ground meteorological sensor arrays, and IoT soil probes;

[0013] Preprocessing includes spatiotemporal interpolation compensation for missing values, wavelet transform filtering of sensor noise data, and normalization of data with different dimensions through min-max standardization.

[0014] Step S2. Data extraction and fusion: Extract multi-source heterogeneous data based on correlation, then construct a spatiotemporal graph neural network, fuse the extracted multi-source heterogeneous data based on spatiotemporal correlation, and introduce an attention mechanism to dynamically weight the contribution of different data sources to the decision objective;

[0015] In step S2, the spatiotemporal graph neural network includes a bidirectional spatiotemporal convolution module. In the spatial dimension, a graph convolutional network is used to model the topological relationship of the fields, and in the temporal dimension, a gated recurrent unit is used to capture the evolution law of meteorological elements. The spatiotemporal graph neural network introduces a transfer learning mechanism to transfer the meteorological early warning model parameters established in the old main production area to the new planting area. The feature shift caused by soil type differences is adjusted through a domain adaptive algorithm.

[0016] The attention mechanism simultaneously calculates the overlap between the meteorological probability distribution and the crop harvest time window to generate a disaster avoidance priority ranking. The attention mechanism includes the following:

[0017] (1) Source Domain Attention Layer: Calculate the feature similarity matrix between satellite remote sensing spectral data and ground meteorological station data.

[0018] (2) Spatiotemporal attention layer: Dynamically capture the spatiotemporal correlation pattern between typhoon path and crop planting distribution through a multi-head self-attention mechanism.

[0019] (3) An interpretable weight visualization module is set at the output end to display the cross-influence factors of soil pH and precipitation;

[0020] The dynamic calculation of edge weights between vertices uses the following formula, which is shown below:

[0021]

[0022] Among them, W ij Let the weight be the edge weight between vertices i and j. The square of the Euclidean spatial distance between vertices i and j, α is the weighting coefficient of the spatial distance term α = 0.6, σ is the bandwidth of the Gaussian kernel function σ = 150m, β is the weighting coefficient of the feature correlation term β = 0.4, Corr(X i ,X j X represents the feature vectors X of nodes i and j. i ,X j The Pearson correlation coefficient;

[0023] The data fusion process employs a federated learning framework to achieve multi-subject data collaboration, and aggregates the local models of each farm node through a parameter server while protecting the privacy data of agricultural enterprises. The fused multidimensional feature vector is stored in a blockchain traceability system, with each feature dimension accompanied by a data source hash value and a timestamp.

[0024] The specific steps of step S2 are as follows:

[0025] Step S1-1. Multi-source data extraction and preprocessing: Based on the Pearson correlation coefficient, the spatiotemporal correlation features of meteorological, soil and crop data are screened, and then a convolutional autoencoder is used to reduce the dimensionality of remote sensing images and sensor data to extract key features;

[0026] Step S1-2. Spatiotemporal graph neural network construction: The farmland area is divided into N×N grid nodes, and a dynamic adjacency matrix is ​​generated based on spatial distance and data correlation. Then, a graph attention network is used to aggregate the features of adjacent 3 layers of nodes to capture local spatial correlation. Bidirectional LSTM is used to extract time series features.

[0027] Step S1-3. Multi-source contribution weighting: The semantic association weights of meteorological, soil, and crop data are calculated using a multi-head attention mechanism, and the allocation rules are determined based on the dynamic weights, as shown below:

[0028] When the meteorological error rate is less than 5%, a weight of 0.4-0.6 is assigned.

[0029] When the correlation of crop data is greater than 0.8, assign a weight of 0.3-0.5;

[0030] When the soil data weight update frequency is > 1 time / h, assign a weight of 0.2-0.4;

[0031] Step S1-4. Spatiotemporal attention optimization: Strengthen the feature weights of key time steps, attenuate the influence of historical redundant data, identify disaster-sensitive areas, and increase local weights.

[0032] Step S3. Data Calculation and Analysis: Calculate the risk of agricultural meteorology based on the contribution level, and then conduct a severity analysis of the risk;

[0033] The risk calculation in step S3 includes the following steps:

[0034] Step S2-1. Establish a multi-hazard coupling assessment model, and weight and fuse drought index, accumulated temperature deviation, and rainstorm disaster-causing factors;

[0035] Step S2-2. Generate 1000 sets of meteorological disturbance scenarios using Monte Carlo simulation;

[0036] Step S2-3. Output a five-level risk warning using the fuzzy comprehensive evaluation method.

[0037] In another embodiment provided by the present invention, the specific steps of the risk severity analysis in step S3 are as follows:

[0038] Step S3-1. Disaster probability classification and spatial location: Based on the risk level distribution map, identify high-risk areas (such as drought areas and low-lying flood-prone areas), overlay farmland boundary data to locate affected fields, and combine with historical disaster databases (10-year / 50-year / 100-year disaster records) to assess the historical recurrence period and extreme nature of the current risk;

[0039] Step S3-2. Crop Loss Prediction: Input a crop growth model (such as the WOFOST model) to simulate yield loss rates under different risk scenarios:

[0040] Arid scenario: Soil moisture content < 0.15m 3 / m 3 At that time, crop yield loss can reach 25%-40%.

[0041] Flood scenario: When rainfall exceeds 100mm for 3 consecutive days, crop yield loss rate increases by 50%;

[0042] Step S3-2. Economic Loss and Ecological Impact Assessment: Calculate direct economic losses (such as reduced yield and increased irrigation costs) and indirect losses (such as market fluctuations and insurance payouts), assess the risks of secondary disasters such as soil salinization and the spread of pests and diseases, and generate an ecological impact index (0-1 range).

[0043] Step S3-3. Expert consultation and verification: Conduct multidisciplinary consultations with meteorological and agricultural departments, use a digital twin system to simulate the disaster evolution process, verify the accuracy of the risk model, then revise the early warning threshold and optimize the severity classification rules.

[0044] Step S4. Decision generation and verification: Based on the severity analysis results, the AI ​​generates multiple decision options, inputs these options into the virtual verification model for verification, and selects the optimal decision option based on the verification results.

[0045] The specific operation steps of the virtual verification model are as follows:

[0046] (1) First, the changes in yield and soil salinity after the implementation of the decision-making scheme were simulated by coupling the WOFOST crop model and the SWAT hydrological model;

[0047] (2) When the simulated yield reduction rate is >8% or the soil electrical conductivity increase is >0.5dS / m, the marking scheme is high risk;

[0048] In step S4, Pareto front analysis is used to select the optimal solution.

[0049] Step S5. Solution Verification and Optimization: Apply the decision-making solution to adjust agricultural planting, and collect actual data during the adjustment process. Optimize the decision-making solution based on the actual data.

[0050] In step S5, the scheme optimization adopts a closed-loop feedback mechanism. When collecting data in step S5, field sensors are deployed to monitor soil moisture content and canopy temperature data in real time after the decision is implemented. When the deviation between the monitored data and the predicted value is >15%, the Bayesian network is triggered to dynamically adjust the model parameters. When applying the decision scheme for planting in step S5, an incremental learning dataset is generated every quarter, and the disaster identification feature space is optimized through comparative learning.

[0051] An intelligent decision support system for agricultural meteorology, integrating multi-source heterogeneous data, includes a data acquisition and processing module, a data fusion module, a risk calculation and analysis module, a decision generation and verification module, and a feedback optimization module. The data acquisition and processing module collects multi-source heterogeneous agricultural data through official data interfaces and monitoring equipment, and also preprocesses the data. The data fusion module fuses multi-source heterogeneous data based on a spatiotemporal graph neural network, and also dynamically weights the contributions of different data sources using an attention mechanism. The risk calculation and analysis module calculates agricultural meteorological risks based on the data fusion results and assesses the severity of the risks. The decision generation and verification module generates multiple decision schemes and simulates and verifies them using a virtual verification model to select the optimal scheme. The feedback optimization module executes the decision schemes and monitors the execution effect in real time, dynamically optimizing model parameters through incremental learning.

[0052] The data acquisition and processing module includes an official data interface unit, an equipment acquisition unit, and a preprocessing unit. The official data interface unit accesses historical temperature / precipitation / wind speed data from the meteorological bureau, crop growth cycle data from the agricultural department, and soil moisture reports. The equipment acquisition unit integrates a UAV remote sensing image acquisition module, a ground meteorological sensor array, and an IoT soil probe. The preprocessing unit is configured with a spatiotemporal interpolation algorithm to compensate for missing values, a wavelet transform filter to eliminate sensor noise, and a min-max standardization module to unify data dimensions.

[0053] The decision generation and verification module couples the WOFOST crop model and the SWAT hydrological model to simulate changes in yield and soil salinity after the implementation of the decision scheme, and also selects the optimal scheme based on a multi-objective optimization algorithm.

[0054] The data fusion module includes a spatiotemporal graph neural network construction unit, a transfer learning unit, an attention mechanism unit, a dynamic edge weight calculation unit, a federated learning unit, and a blockchain storage unit. The spatiotemporal graph neural network construction unit is used to deploy a bidirectional spatiotemporal convolution module. The transfer learning unit is used to transfer the parameters of the meteorological early warning model established in the old main production area to the new planting area. The attention mechanism unit is used to simultaneously calculate the overlap between the meteorological probability distribution and the crop harvest time window, and generate a disaster avoidance priority ranking. The dynamic edge weight calculation unit dynamically adjusts the edge weights between vertices according to a formula. The federated learning unit aggregates the local model gradients of each farm node through a parameter server and performs privacy data localization training. The blockchain storage unit is used to collect the fused data.

[0055] The risk calculation and analysis module includes a multi-hazard coupling assessment unit, a Monte Carlo simulation unit, and a fuzzy comprehensive evaluation unit. The multi-hazard coupling assessment unit is used to weight and fuse drought index, accumulated temperature deviation, and rainstorm disaster factors. The Monte Carlo simulation unit is used to generate 1,000 sets of meteorological disturbance scenarios. The fuzzy comprehensive evaluation unit outputs a five-level risk warning based on the results generated by the Monte Carlo simulation unit.

[0056] The feedback optimization module includes a closed-loop feedback unit, an incremental learning unit, and a storage unit. The closed-loop feedback unit monitors soil moisture content and canopy temperature data in real time by deploying field sensors. The incremental learning unit generates an incremental dataset every quarter and optimizes the disaster identification feature space through comparative learning.

[0057] In the above technical solution, the agricultural meteorological intelligent decision support method and system that combines multi-source heterogeneous data provided by the present invention has the following beneficial effects:

[0058] (1) This invention integrates multi-dimensional data of meteorology, soil and crops through spatiotemporal graph neural networks, and combines dynamic adjacency matrix and attention mechanism to effectively capture the spatiotemporal correlation features of farmland areas, solve the data island problem of traditional methods, improve the accuracy of disaster identification, and realize the rapid migration of the main production area model to the new planting area by introducing transfer learning mechanism and federated learning framework. At the same time, the farm node data is encrypted and aggregated through parameter server, and collaborative modeling is completed under the premise of protecting the privacy of agricultural enterprises, avoiding the risk of sensitive data leakage.

[0059] (2) This invention uses Monte Carlo simulation to generate 1,000 sets of meteorological disturbance scenarios, and combines fuzzy comprehensive evaluation method to output five-level risk warnings. It supports the quantitative assessment of the coupling effects of multiple disasters such as drought and flood. It simulates the implementation effect of decision-making schemes through digital twin system, and uses Pareto front analysis to screen the optimal solution that balances economic benefits and ecological losses, so that the yield loss rate after the implementation of irrigation schemes is reduced. Furthermore, it monitors the implementation effect of decision-making in real time by deploying field sensors. When the deviation between the monitoring data and the predicted value is >15%, it triggers the Bayesian network to dynamically adjust the model parameters. It updates the dataset quarterly through incremental learning and optimizes the feature space by using contrastive learning, so that the warning accuracy of the model under extreme climate scenarios continues to improve.

[0060] (3) This invention writes the fused multidimensional feature vector (with data source hash value and timestamp) into the blockchain to ensure that the decision-making process is traceable and tamper-proof. Through the interpretability weight visualization module, it enhances the transparency of the decision-making logic and helps agricultural managers quickly locate the main causes of risk.

[0061] (4) This invention constructs an efficient, reliable and interpretable intelligent decision support system for agricultural meteorology through multi-source data fusion, spatiotemporal dynamic modeling, privacy-preserving collaborative learning and closed-loop feedback optimization, providing an innovative technical solution for addressing the challenges of climate change and achieving sustainable agricultural development. Attached Figure Description

[0062] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0063] Figure 1 This is a schematic diagram of the method flow for the agricultural meteorological intelligent decision support method and system embodiment that combines multi-source heterogeneous data according to the present invention.

[0064] Figure 2 This is a schematic diagram of step S2 in the embodiment of the agricultural meteorological intelligent decision support method and system combining multi-source heterogeneous data of the present invention.

[0065] Figure 3 This is a schematic diagram of the risk calculation process provided by the present invention for an agricultural meteorological intelligent decision support method and system embodiment that combines multi-source heterogeneous data.

[0066] Figure 4 This is a schematic diagram of the risk severity analysis process provided by the present invention for an agricultural meteorological intelligent decision support method and system embodiment that combines multi-source heterogeneous data.

[0067] Figure 5 This is a schematic diagram of an agricultural meteorological intelligent decision support system provided by an embodiment of the agricultural meteorological intelligent decision support method and system combining multi-source heterogeneous data according to the present invention.

[0068] Figure 6 This is a schematic diagram of the data acquisition and processing module provided in the embodiment of the intelligent decision support method and system for agricultural meteorology that combines multi-source heterogeneous data of the present invention.

[0069] Figure 7 This is a schematic diagram of the data fusion module provided in the embodiment of the agricultural meteorological intelligent decision support method and system combining multi-source heterogeneous data of the present invention.

[0070] Figure 8 This is a schematic diagram of the risk calculation and analysis module provided in the embodiment of the agricultural meteorological intelligent decision support method and system combining multi-source heterogeneous data of the present invention.

[0071] Figure 9This is a schematic diagram of the feedback optimization module provided in the embodiment of the agricultural meteorological intelligent decision support method and system combining multi-source heterogeneous data of the present invention. Detailed Implementation

[0072] To enable those skilled in the art to better understand the technical solution of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings.

[0073] like Figure 1-4 As shown in the figure, the agricultural meteorological intelligent decision support method combining multi-source heterogeneous data provided by the embodiments of the present invention is characterized by including the following steps:

[0074] Step S1. Acquisition and processing of multi-source heterogeneous data: Acquire agricultural multi-source heterogeneous data through official channels and monitoring equipment, and preprocess the multi-source heterogeneous data;

[0075] In step S1:

[0076] Official data includes historical temperature / precipitation / wind speed data released by the meteorological bureau, crop growth cycle data and soil moisture reports provided by the agricultural sector;

[0077] Monitoring equipment data is acquired through UAV remote sensing images, ground meteorological sensor arrays, and IoT soil probes;

[0078] Preprocessing includes spatiotemporal interpolation compensation for missing values, wavelet transform filtering of sensor noise data, and normalization of data with different dimensions through min-max standardization.

[0079] Step S2. Data extraction and fusion: Extract multi-source heterogeneous data based on correlation, then construct a spatiotemporal graph neural network, fuse the extracted multi-source heterogeneous data based on spatiotemporal correlation, and introduce an attention mechanism to dynamically weight the contribution of different data sources to the decision objective;

[0080] In step S2, the spatiotemporal graph neural network includes a bidirectional spatiotemporal convolution module. In the spatial dimension, a graph convolutional network is used to model the topological relationship of the fields, and in the temporal dimension, a gated recurrent unit is used to capture the evolution law of meteorological elements. The spatiotemporal graph neural network introduces a transfer learning mechanism to transfer the parameters of the meteorological early warning model established in the old main production area to the new planting area. The feature shift caused by soil type differences is adjusted through a domain adaptive algorithm.

[0081] The attention mechanism simultaneously calculates the overlap between the meteorological probability distribution and the crop harvest time window, and generates a disaster avoidance priority ranking. The attention mechanism includes the following:

[0082] (1) Source Domain Attention Layer: Calculate the feature similarity matrix between satellite remote sensing spectral data and ground meteorological station data.

[0083] (2) Spatiotemporal attention layer: Dynamically capture the spatiotemporal correlation pattern between typhoon path and crop planting distribution through a multi-head self-attention mechanism.

[0084] (3) An interpretable weight visualization module is set at the output end to display the cross-influence factors of soil pH and precipitation;

[0085] The dynamic calculation of edge weights between vertices uses the following formula, as shown below:

[0086]

[0087] Among them, W ij Let the weight be the edge weight between vertices i and j. The square of the Euclidean spatial distance between vertices i and j, α is the weighting coefficient of the spatial distance term α = 0.6, σ is the bandwidth of the Gaussian kernel function σ = 150m, β is the weighting coefficient of the feature correlation term β = 0.4, Corr(X i ,X j X represents the feature vectors X of nodes i and j. i ,X j The Pearson correlation coefficient;

[0088] The data fusion process uses a federated learning framework to achieve multi-subject data collaboration, and aggregates the local models of each farm node through a parameter server while protecting the privacy of agricultural enterprise data. The fused multi-dimensional feature vector is stored in the blockchain traceability system, with each feature dimension accompanied by a data source hash value and timestamp.

[0089] The specific steps of step S2 are as follows:

[0090] Step S1-1. Multi-source data extraction and preprocessing: Based on the Pearson correlation coefficient, the spatiotemporal correlation features of meteorological, soil and crop data are screened, and then a convolutional autoencoder is used to reduce the dimensionality of remote sensing images and sensor data to extract key features;

[0091] It should be noted that the meteorological data includes temperature (24-hour average), precipitation (6-hour cumulative), wind speed (10-meter height), and relative humidity (daytime minimum); the soil data includes electrical conductivity (20cm depth), volumetric water content (0-40cm average), soil pH (top layer), and available nitrogen content (0-20cm); and the crop data includes NDVI (Normalized Difference Vegetation Index), chlorophyll content (SPAD value), canopy cover, and crop growth stage coding.

[0092] Step S1-2. Spatiotemporal graph neural network construction: The farmland area is divided into N×N grid nodes, and a dynamic adjacency matrix is ​​generated based on spatial distance and data correlation. Then, a graph attention network is used to aggregate the features of adjacent 3 layers of nodes to capture local spatial correlation. Bidirectional LSTM is used to extract time series features.

[0093] Step S1-3. Multi-source contribution weighting: The semantic association weights of meteorological, soil, and crop data are calculated using a multi-head attention mechanism, and the allocation rules are determined based on the dynamic weights, as shown below:

[0094] When the meteorological error rate is less than 5%, a weight of 0.4-0.6 is assigned.

[0095] When the correlation of crop data is greater than 0.8, assign a weight of 0.3-0.5;

[0096] When the soil data weight update frequency is > 1 time / h, assign a weight of 0.2-0.4;

[0097] Step S1-4. Spatiotemporal attention optimization: Strengthen the feature weights of key time steps, attenuate the influence of historical redundant data, identify disaster-sensitive areas, and increase local weights.

[0098] Step S3. Data Calculation and Analysis: Calculate the risk of agricultural meteorology based on the contribution level, and then conduct a severity analysis of the risk;

[0099] The risk calculation in step S3 includes the following steps:

[0100] Step S2-1. Establish a multi-hazard coupling assessment model, and weight and fuse drought index, accumulated temperature deviation, and rainstorm disaster-causing factors;

[0101] Step S2-2. Generate 1000 sets of meteorological disturbance scenarios using Monte Carlo simulation;

[0102] Step S2-3. Output a five-level risk warning using the fuzzy comprehensive evaluation method.

[0103] In another embodiment provided by the present invention, the specific steps of the risk severity analysis in step S3 are as follows:

[0104] Step S3-1. Disaster probability classification and spatial location: Based on the risk level distribution map, identify high-risk areas (such as drought areas and low-lying flood-prone areas), overlay farmland boundary data to locate affected fields, and combine with historical disaster databases (10-year / 50-year / 100-year disaster records) to assess the historical recurrence period and extreme nature of the current risk;

[0105] Step S3-2. Crop Loss Prediction: Input the crop growth model to simulate yield loss rates under different risk scenarios:

[0106] Arid scenario: Soil moisture content < 0.15m 3 / m 3 At that time, crop yield loss can reach 25%-40%.

[0107] Flood scenario: When rainfall exceeds 100mm for 3 consecutive days, crop yield loss rate increases by 50%;

[0108] It should be noted that the crop growth model uses a logistic regression equation to calculate the probability of yield reduction, as shown below:

[0109] P(y|x)=1 / [1+exp(-0.5·SPEI+1.2·ΔNDVI)]

[0110] Wherein, SPEI is the standardized precipitation evapotranspiration index, and ΔNDVI is the vegetation index change rate;

[0111] Step S3-2. Economic Loss and Ecological Impact Assessment: Calculate direct economic losses (such as reduced yield and increased irrigation costs) and indirect losses (such as market fluctuations and insurance payouts), assess the risks of secondary disasters such as soil salinization and the spread of pests and diseases, and generate an ecological impact index (0-1 range).

[0112] Step S3-3. Expert consultation and verification: Conduct multidisciplinary consultations with meteorological and agricultural departments, use a digital twin system to simulate the disaster evolution process, verify the accuracy of the risk model, then revise the early warning threshold and optimize the severity classification rules.

[0113] Step S4. Decision generation and verification: Based on the severity analysis results, the AI ​​generates multiple decision options, inputs these options into the virtual verification model for verification, and selects the optimal decision option based on the verification results.

[0114] The specific steps for operating the virtual verification model are as follows:

[0115] (1) First, the changes in yield and soil salinity after the implementation of the decision-making scheme were simulated by coupling the WOFOST crop model and the SWAT hydrological model;

[0116] (2) When the simulated yield reduction rate is >8% or the soil electrical conductivity increase is >0.5dS / m, the marking scheme is high risk;

[0117] In step S4, Pareto front analysis is used to select the optimal solution, and the objective function is:

[0118] max(economic benefits / (0.8·cost + 0.2·ecological loss)).

[0119] Step S5. Solution Verification and Optimization: Apply the decision-making solution to adjust agricultural planting, and collect actual data during the adjustment process. Optimize the decision-making solution based on the actual data.

[0120] In step S5, the scheme optimization adopts a closed-loop feedback mechanism. When collecting data in step S5, field sensors are deployed to monitor soil moisture content and canopy temperature data in real time after the decision is implemented. When the deviation between the monitored data and the predicted value is >15%, the Bayesian network is triggered to dynamically adjust the model parameters. When applying the decision scheme for planting in step S5, an incremental learning dataset is generated every quarter, and the disaster identification feature space is optimized through comparative learning.

[0121] Agricultural meteorological intelligent decision support systems that combine multi-source heterogeneous data, such as Figure 5-9 As shown, the system includes a data acquisition and processing module, a data fusion module, a risk calculation and analysis module, a decision generation and verification module, and a feedback optimization module. The data acquisition and processing module collects multi-source heterogeneous agricultural data through official data interfaces and monitoring equipment, and also performs data preprocessing. The data fusion module fuses multi-source heterogeneous data based on a spatiotemporal graph neural network, and also dynamically weights the contribution of different data sources through an attention mechanism. The risk calculation and analysis module calculates agricultural meteorological risks based on the data fusion results and assesses the severity of the risks. The decision generation and verification module generates multiple decision schemes and performs simulation verification through a virtual verification model to select the best scheme. The feedback optimization module executes the decision schemes and monitors the execution effect in real time, and dynamically optimizes the model parameters through incremental learning.

[0122] The data acquisition and processing module includes an official data interface unit, an equipment acquisition unit, and a preprocessing unit. The official data interface unit accesses historical temperature / precipitation / wind speed data from the meteorological bureau, crop growth cycle data from the agricultural department, and soil moisture reports. The equipment acquisition unit integrates a UAV remote sensing image acquisition module, a ground meteorological sensor array, and an IoT soil probe. The preprocessing unit is configured with a spatiotemporal interpolation algorithm to compensate for missing values, a wavelet transform filter to eliminate sensor noise, and a min-max standardization module to unify data dimensions.

[0123] The decision generation and verification module couples the WOFOST crop model and the SWAT hydrological model to simulate changes in yield and soil salinity after the implementation of decision schemes, and also selects the optimal scheme based on a multi-objective optimization algorithm.

[0124] The data fusion module includes a spatiotemporal graph neural network construction unit, a transfer learning unit, an attention mechanism unit, a dynamic edge weight calculation unit, a federated learning unit, and a blockchain storage unit. The spatiotemporal graph neural network construction unit is used to deploy bidirectional spatiotemporal convolution modules. The transfer learning unit is used to transfer the parameters of the meteorological early warning model established in the old main production area to the new planting area. The attention mechanism unit is used to simultaneously calculate the overlap between the meteorological probability distribution and the crop harvest time window and generate a disaster avoidance priority ranking. The dynamic edge weight calculation unit dynamically adjusts the edge weights between vertices according to the formula. The federated learning unit aggregates the local model gradients of each farm node through the parameter server and performs privacy data localization training. The blockchain storage unit is used to collect the fused data.

[0125] The risk calculation and analysis module includes a multi-hazard coupling assessment unit, a Monte Carlo simulation unit, and a fuzzy comprehensive evaluation unit. The multi-hazard coupling assessment unit is used to weight and fuse drought index, accumulated temperature deviation, and rainstorm disaster factors. The Monte Carlo simulation unit is used to generate 1,000 sets of meteorological disturbance scenarios. The fuzzy comprehensive evaluation unit outputs a five-level risk warning based on the results generated by the Monte Carlo simulation unit.

[0126] The feedback optimization module includes a closed-loop feedback unit, an incremental learning unit, and a storage unit. The closed-loop feedback unit monitors soil moisture content and canopy temperature data in real time by deploying field sensors. The incremental learning unit generates an incremental dataset every quarter and optimizes the disaster identification feature space through comparative learning.

[0127] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. An intelligent decision support method for agricultural meteorology that combines multi-source heterogeneous data, characterized in that: Includes the following steps: Step S1. Acquisition and processing of multi-source heterogeneous data: Acquire agricultural multi-source heterogeneous data through official channels and monitoring equipment, and preprocess the multi-source heterogeneous data; Step S2. Data extraction and fusion: Based on the Pearson correlation coefficient, the spatiotemporal correlation features of meteorological, soil and crop data are screened, and a convolutional autoencoder is used to reduce the dimensionality of the data to extract key features; Farmland areas are divided into N×N grid nodes, and a dynamic adjacency matrix is ​​generated based on spatial distance and data correlation to construct a spatiotemporal graph neural network. The features of adjacent three-layer nodes are aggregated through a graph attention network to capture local spatial correlations, and bidirectional LSTM is used to extract time series features to achieve the fusion of multi-source heterogeneous data. Furthermore, a multi-head attention mechanism is introduced, which dynamically weights the contribution of different data sources to the decision-making objective by calculating the semantic association weights of meteorological, soil, and crop data; Step S3. Data Calculation and Analysis: Calculate the risk of agricultural meteorology based on the contribution level, and then conduct a severity analysis of the risk; Step S4. Decision generation and verification: Based on the severity analysis results, the AI ​​generates multiple decision options, inputs these options into the virtual verification model for verification, and selects the optimal decision option based on the verification results. Step S5. Solution Verification and Optimization: Apply the decision-making solution to adjust agricultural planting, and collect actual data during the adjustment process. Optimize the decision-making solution based on the actual data.

2. The agricultural meteorological intelligent decision support method combining multi-source heterogeneous data according to claim 1, characterized in that, In step S1: Official data includes historical temperature / precipitation / wind speed data released by the meteorological bureau, crop growth cycle data and soil moisture reports provided by the agricultural sector; Monitoring equipment data is acquired through UAV remote sensing images, ground meteorological sensor arrays, and IoT soil probes; Preprocessing includes spatiotemporal interpolation compensation for missing values, wavelet transform filtering of sensor noise data, and normalization of data with different dimensions through min-max standardization.

3. The agricultural meteorological intelligent decision support method combining multi-source heterogeneous data according to claim 1, characterized in that, In step S2, the spatiotemporal graph neural network includes a bidirectional spatiotemporal convolution module. In the spatial dimension, a graph convolutional network is used to model the topological relationship of the fields, and in the temporal dimension, a gated recurrent unit is used to capture the evolution law of meteorological elements. The spatiotemporal graph neural network introduces a transfer learning mechanism to transfer the meteorological early warning model parameters established in the old main production area to the new planting area. The feature shift caused by soil type differences is adjusted through a domain adaptive algorithm. The attention mechanism simultaneously calculates the overlap between the meteorological probability distribution and the crop harvest time window to generate a disaster avoidance priority ranking. The attention mechanism includes the following: (1) Source Domain Attention Layer: Calculate the feature similarity matrix between satellite remote sensing spectral data and ground meteorological station data; (2) Spatiotemporal attention layer: dynamically capture the spatiotemporal correlation pattern between typhoon path and crop planting distribution through a multi-head self-attention mechanism; (3) An interpretable weight visualization module is set at the output end to display the cross-influence factors of soil pH and precipitation; The dynamic calculation of edge weights between vertices uses the following formula, which is shown below: ； in, Let the weight be the edge weight between vertices i and j. The square of the Euclidean spatial distance between vertices i and j, where α is the weighting coefficient of the spatial distance term (α=0.6), σ is the bandwidth of the Gaussian kernel function (σ=150m), and β is the weighting coefficient of the feature correlation term (β=0.4). The feature vectors of nodes i and j The Pearson correlation coefficient; The data fusion process employs a federated learning framework to achieve multi-agent data collaboration, and aggregates the local models of each farm node through a parameter server while protecting the privacy of agricultural enterprise data. The fused multidimensional feature vector is stored in a blockchain traceability system, with each feature dimension accompanied by a data source hash value and a timestamp.

4. The agricultural meteorological intelligent decision support method combining multi-source heterogeneous data according to claim 1, characterized in that, The specific steps of step S2 are as follows: Step S1-1. Multi-source data extraction and preprocessing: Based on the Pearson correlation coefficient, the spatiotemporal correlation features of meteorological, soil and crop data are screened, and then a convolutional autoencoder is used to reduce the dimensionality of remote sensing images and sensor data to extract key features; Step S1-2. Spatiotemporal graph neural network construction: The farmland area is divided into N×N grid nodes, and a dynamic adjacency matrix is ​​generated based on spatial distance and data correlation. Then, a graph attention network is used to aggregate the features of adjacent 3 layers of nodes to capture local spatial correlation. Bidirectional LSTM is used to extract time series features. Step S1-3. Multi-source contribution weighting: The semantic association weights of meteorological, soil, and crop data are calculated using a multi-head attention mechanism, and the allocation rules are determined based on the dynamic weights, as shown below: When the meteorological error rate is less than 5%, a weight of 0.4-0.6 is assigned. When the correlation of crop data is greater than 0.8, assign a weight of 0.3-0.5; When the soil data weight update frequency is > 1 time / h, assign a weight of 0.2-0.4; Step S1-4. Spatiotemporal attention optimization: Strengthen the feature weights of key time steps, attenuate the influence of historical redundant data, identify disaster-sensitive areas, and increase local weights.

5. The agricultural meteorological intelligent decision support method combining multi-source heterogeneous data according to claim 1, characterized in that, The risk calculation in step S3 includes the following steps: Step S2-1. Establish a multi-hazard coupling assessment model, and weight and fuse drought index, accumulated temperature deviation, and rainstorm disaster-causing factors; Step S2-2. Generate 1000 sets of meteorological disturbance scenarios using Monte Carlo simulation; Step S2-3. Output a five-level risk warning using the fuzzy comprehensive evaluation method.

6. The agricultural meteorological intelligent decision support method combining multi-source heterogeneous data according to claim 1, characterized in that, The specific steps of the risk severity analysis in step S3 are as follows: Step S3-1. Disaster probability classification and spatial location: Based on the risk level distribution map, identify high-risk areas, overlay farmland boundary data to locate affected fields, and combine with the historical disaster database to assess the historical recurrence period and extreme nature of the current risk; Step S3-2. Crop Loss Prediction: Input the crop growth model to simulate yield loss rates under different risk scenarios: In drought conditions, when soil moisture content is less than 0.15 m³ / m³, crop yield loss can reach 25%-40%. Flood scenario: When rainfall exceeds 100mm for three consecutive days, crop yield loss rate increases by 50%; Step S3-2. Economic Loss and Ecological Impact Assessment: Calculate direct economic losses (such as reduced yield and increased irrigation costs) and indirect losses (such as market fluctuations and insurance payouts), assess the risks of secondary disasters such as soil salinization and the spread of pests and diseases, and generate an ecological impact index (0-1 range). Step S3-3. Expert consultation and verification: Conduct multidisciplinary consultations with meteorological and agricultural departments, use a digital twin system to simulate the disaster evolution process, verify the accuracy of the risk model, then revise the early warning threshold and optimize the severity classification rules.

7. The agricultural meteorological intelligent decision support method combining multi-source heterogeneous data according to claim 1, characterized in that, The virtual verification model in step S4 is an agricultural digital twin system, and the specific operation steps of the virtual verification model are as follows: (1) First, the changes in yield and soil salinity after the implementation of the decision-making scheme were simulated by coupling the WOFOST crop model and the SWAT hydrological model; (2) When the simulated yield reduction rate is >8% or the soil electrical conductivity increase is >0.5dS / m, the marking scheme is high risk; In step S4, Pareto front analysis is used to select the optimal solution.

8. The agricultural meteorological intelligent decision support method combining multi-source heterogeneous data according to claim 1, characterized in that, In step S5, the scheme optimization adopts a closed-loop feedback mechanism. When collecting data in step S5, field sensors are deployed to monitor soil moisture content and canopy temperature data in real time after the decision is implemented. When the deviation between the monitored data and the predicted value is >15%, the Bayesian network is triggered to dynamically adjust the model parameters. When applying the decision scheme for planting in step S5, an incremental learning dataset is generated every quarter, and the disaster identification feature space is optimized through comparative learning.

9. An intelligent decision support system for agricultural meteorology that integrates multi-source heterogeneous data, characterized in that: The system includes a data acquisition and processing module, a data fusion module, a risk calculation and analysis module, a decision generation and verification module, and a feedback optimization module. The data acquisition and processing module collects multi-source heterogeneous agricultural data through official data interfaces and monitoring equipment, and also preprocesses the data. The data fusion module includes a spatiotemporal graph neural network construction unit, which divides farmland areas into N×N grid nodes and generates a dynamic adjacency matrix based on spatial distance and data correlation. The data fusion module is also configured to aggregate features of adjacent three layers of nodes through a graph attention network. To capture local spatial correlations, time-series features are extracted using bidirectional LSTM to achieve the fusion of multi-source heterogeneous data. The data fusion module also calculates the semantic association weights of meteorological, soil, and crop data through a multi-head attention mechanism to dynamically weight the contribution of different data sources. The risk calculation and analysis module calculates agricultural meteorological risks based on the data fusion results and assesses the severity of the risks. The decision generation and verification module generates multiple decision schemes and performs simulation verification through a virtual verification model to select the best scheme. The feedback optimization module executes the decision schemes and monitors the execution effect in real time, dynamically optimizing the model parameters through incremental learning. The data acquisition and processing module includes an official data interface unit, an equipment acquisition unit, and a preprocessing unit. The official data interface unit accesses historical temperature / precipitation / wind speed data from the meteorological bureau, crop growth cycle data from the agricultural department, and soil moisture reports. The equipment acquisition unit integrates a UAV remote sensing image acquisition module, a ground meteorological sensor array, and an IoT soil probe. The preprocessing unit is configured with a spatiotemporal interpolation algorithm to compensate for missing values, a wavelet transform filter to eliminate sensor noise, and a min-max standardization module to unify data dimensions. The decision generation and verification module couples the WOFOST crop model and the SWAT hydrological model to simulate changes in yield and soil salinity after the implementation of the decision scheme, and also selects the optimal scheme based on a multi-objective optimization algorithm.

10. The agricultural meteorological intelligent decision support system combining multi-source heterogeneous data according to claim 9, characterized in that, The data fusion module includes a spatiotemporal graph neural network construction unit, a transfer learning unit, an attention mechanism unit, a dynamic edge weight calculation unit, a federated learning unit, and a blockchain storage unit. The spatiotemporal graph neural network construction unit is used to deploy a bidirectional spatiotemporal convolution module. The transfer learning unit is used to transfer the meteorological early warning model parameters established in the old main production area to the new planting area. The attention mechanism unit is used to simultaneously calculate the overlap between the meteorological probability distribution and the crop harvest time window, and generate a disaster avoidance priority ranking. The dynamic edge weight calculation unit dynamically adjusts the edge weights between vertices according to a formula. The federated learning unit aggregates the local model gradients of each farm node through a parameter server and performs privacy data localization training. The blockchain storage unit is used to collect the fused data. The risk calculation and analysis module includes a multi-hazard coupling assessment unit, a Monte Carlo simulation unit, and a fuzzy comprehensive evaluation unit. The multi-hazard coupling assessment unit is used to weight and fuse drought index, accumulated temperature deviation, and rainstorm disaster-causing factors. The Monte Carlo simulation unit is used to generate 1,000 sets of meteorological disturbance scenarios. The fuzzy comprehensive evaluation unit outputs a five-level risk warning based on the results generated by the Monte Carlo simulation unit. The feedback optimization module includes a closed-loop feedback unit, an incremental learning unit, and a storage unit. The closed-loop feedback unit monitors soil moisture content and canopy temperature data in real time by deploying field sensors. The incremental learning unit generates an incremental dataset every quarter and optimizes the disaster identification feature space through comparative learning. The incremental learning unit generates an incremental dataset every quarter and optimizes the disaster identification feature space through comparative learning.

Images