Agricultural disaster risk prediction method and device for agricultural insurance policy, electronic equipment and medium

Abstract

The embodiment of the invention provides an agricultural disaster risk prediction method and device for agricultural insurance policies, electronic equipment and a medium, belongs to the technical field of artificial intelligence, and is suitable for the field of financial science and technology. The method comprises the following steps: acquiring original multi-source environment data; obtaining an agricultural insurance policy resource pool; the agricultural insurance policy resource pool comprises an original agricultural insurance policy, and the original agricultural insurance policy is used for underwriting a current crop; performing information extraction on the original agricultural insurance policy to obtain underwriting area information and current growth information of a current crop; based on the original multi-source environment data, the insured area information and the current growth information, performing knowledge retrieval on a preset agricultural disaster information base to obtain reference agricultural disaster information; performing risk prediction based on the reference agricultural disaster information to obtain agricultural disaster risk prediction data; and performing insurance policy screening on the original agricultural insurance policies based on the agricultural disaster risk prediction data to obtain disaster-affected agricultural insurance policies. According to the embodiment of the invention, the accuracy of agricultural disaster risk prediction can be improved.

Description

Agricultural insurance policy disaster risk prediction methods and devices, electronic equipment and media Technical Field

[0001] This application relates to the field of artificial intelligence technology, and is applicable to the financial technology field, particularly to a method and device, electronic device and medium for predicting agricultural disaster risks for agricultural insurance policies. Background Technology

[0002] Agricultural disaster risk prediction refers to the dynamic assessment of potential risks to agricultural production in a specific region and time period based on monitoring data of agricultural meteorology and biological disasters, combined with future weather trends and crop growth status. In fintech scenarios, agricultural disaster risk prediction methods can be used to predict the disaster risks faced by insured agricultural products.

[0003] Currently, agricultural disaster risk prediction is mainly achieved using statistical regression methods or machine learning models. However, in practical applications, statistical regression methods or machine learning models can only achieve local risk prediction for specific crops and regions, and cannot be applied to more regions and crops, resulting in poor applicability and affecting the accuracy of agricultural disaster risk prediction.

[0004] Therefore, improving the accuracy of agricultural disaster risk prediction has become an urgent technical problem to be solved. Summary of the Invention

[0005] The main objective of this application is to propose a method, device, electronic device, and medium for predicting agricultural disaster risks in agricultural insurance policies. This aims to solve the technical problem that agricultural disaster risk prediction is not applicable to more crops and regions due to its poor applicability, thereby improving the accuracy of agricultural disaster risk prediction.

[0006] To achieve the above objectives, a first aspect of this application proposes a method for predicting agricultural disaster risks based on agricultural insurance policies. The method includes: acquiring raw multi-source environmental data; acquiring an agricultural insurance policy resource pool; wherein the agricultural insurance policy resource pool includes raw agricultural insurance policies used to insure the current crop; extracting information from the raw agricultural insurance policies to obtain insured area information and current growth information of the current crop; performing knowledge retrieval on a preset agricultural disaster information database based on the raw multi-source environmental data, the insured area information, and the current growth information to obtain reference agricultural disaster information; performing risk prediction based on the reference agricultural disaster information to obtain agricultural disaster risk prediction data; and filtering the raw agricultural insurance policies based on the agricultural disaster risk prediction data to obtain disaster-affected agricultural insurance policies.

[0007] In some embodiments, the agricultural disaster information database includes multiple original agricultural disaster information sets. The step of performing knowledge retrieval on the preset agricultural disaster information database based on the original multi-source environmental data, the insured area information, and the current growth information to obtain reference agricultural disaster information includes: performing feature filtering on the original multi-source environmental data based on the insured area information to obtain initial multi-source environmental data; extracting features from the initial multi-source environmental data and the current growth information to obtain target multi-source risk information; matching the target multi-source risk information with each of the original agricultural disaster information sets to obtain information matching data; filtering the multiple original agricultural disaster information sets based on the information matching data to obtain historical agricultural disaster knowledge; and enhancing the target multi-source risk information based on the historical agricultural disaster knowledge to obtain the reference agricultural disaster information.

[0008] In some embodiments, the initial multi-source environmental data includes: current remote sensing image, current meteorological data, and real-time monitoring data of crop diseases and pests; the step of extracting features from the initial multi-source environmental data and the current growth information to obtain target multi-source risk information includes: obtaining current soil data based on the insured area information; identifying cultivated land based on the current remote sensing image to obtain cultivated land identification data; performing weather forecasting based on the current meteorological data to obtain weather forecast data; identifying the status of crop diseases and pests based on the real-time monitoring data of crop diseases and pests to obtain crop disease and pest status; predicting yield based on the current soil data and the current growth information to obtain crop yield prediction data; and fusing features from the cultivated land identification data, the weather forecast data, the crop disease and pest status, and the crop yield prediction data to obtain the target multi-source risk information.

[0009] In some embodiments, the step of fusing features from the farmland identification data, the weather forecast data, the crop pest and disease status data, and the crop yield forecast data to obtain the target multi-source risk information includes: performing text conversion on the farmland identification data to obtain farmland identification text; performing text conversion on the weather forecast data to obtain weather forecast text; performing text conversion on the crop pest and disease status to obtain pest and disease status text; performing text conversion on the crop yield forecast data to obtain crop yield forecast text; and performing structured processing on the farmland identification text, the weather forecast text, the pest and disease status text, and the crop yield forecast text to obtain the target multi-source risk information.

[0010] In some embodiments, the agricultural disaster information database is constructed in the following ways: acquiring historical disaster case data, agricultural insurance compensation data, historical crop growth data, historical meteorological data, and regional soil data; extracting features based on the historical disaster case data to obtain historical disaster features; analyzing the historical crop growth data and the historical meteorological data to obtain crop-meteorological relationship coefficients; analyzing the historical crop growth data and the regional soil data to obtain yield-soil relationship coefficients; extracting features based on the agricultural insurance compensation data to obtain agricultural insurance compensation features; confirming the historical disaster features, the crop-meteorological relationship coefficients, the yield-soil relationship coefficients, and the agricultural insurance compensation features as the original agricultural disaster information, and constructing the agricultural disaster information database.

[0011] In some embodiments, the step of performing risk prediction based on the reference agricultural disaster information to obtain agricultural disaster risk prediction data includes: classifying the disaster based on the reference agricultural disaster information to obtain a disaster risk level; generating disaster-affected area information based on the reference agricultural disaster information; estimating losses based on the reference agricultural disaster information to obtain estimated loss data; and generating the agricultural disaster risk prediction data based on the disaster risk level, the disaster-affected area information, and the estimated loss data.

[0012] In some embodiments, after filtering the original agricultural insurance policies based on the agricultural disaster risk prediction data to obtain disaster-affected agricultural insurance policies, the method further includes: obtaining the policyholder information of the disaster-affected agricultural insurance policies; and issuing agricultural disaster risk warnings based on the agricultural disaster risk prediction data and the policyholder information.

[0013] To achieve the above objectives, a second aspect of this application proposes an agricultural disaster risk prediction device for agricultural insurance policies. The device includes: an environmental data acquisition module for acquiring raw multi-source environmental data; a policy acquisition module for acquiring an agricultural insurance policy resource pool, wherein the agricultural insurance policy resource pool includes raw agricultural insurance policies used to insure the current crop; a policy information extraction module for extracting information from the raw agricultural insurance policies to obtain insured area information and current growth information of the current crop; a knowledge retrieval module for performing knowledge retrieval on a preset agricultural disaster information database based on the raw multi-source environmental data, the insured area information, and the current growth information to obtain reference agricultural disaster information; an agricultural disaster risk prediction module for performing risk prediction based on the reference agricultural disaster information to obtain agricultural disaster risk prediction data; and a disaster-affected policy screening module for screening the raw agricultural insurance policies based on the agricultural disaster risk prediction data to obtain disaster-affected agricultural insurance policies.

[0014] To achieve the above objectives, a third aspect of this application provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the method described in the first aspect.

[0015] To achieve the above objectives, a fourth aspect of the present application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method described in the first aspect.

[0016] The agricultural disaster risk prediction method, device, electronic equipment, and medium proposed in this application, based on agricultural insurance policies, comprehensively grasp the external environmental conditions of crop growth by acquiring raw, multi-source environmental data, providing a foundation for subsequent analysis. It acquires an agricultural insurance policy resource pool, including original agricultural insurance policies used to insure the current crop, thus clarifying the scope of analysis. Next, information is extracted from the original agricultural insurance policies to obtain the insured area information and the current crop's growth information, accurately locating the current crop's location and growth status. Furthermore, based on the raw, multi-source environmental data, insured area information, and current growth information, a knowledge retrieval is performed on a pre-set agricultural disaster information database to quickly obtain reference agricultural disaster information matching the current situation, improving information acquisition efficiency. Finally, risk prediction is conducted based on reference agricultural disaster information to obtain agricultural disaster risk prediction data. Based on the agricultural disaster risk prediction data and the information on the insured area, the original agricultural insurance policies are screened to accurately identify the agricultural insurance policies affected by disasters. This helps to assess and process claims in a timely and accurate manner, solves the technical problem that agricultural disaster risk prediction cannot be applied to more crops and regions due to its poor applicability, improves the accuracy of agricultural disaster risk prediction, and enhances the quality and efficiency of agricultural insurance services. Attached Figure Description

[0017] Figure 1 is a flowchart of the agricultural disaster risk prediction method for agricultural insurance policies provided in this application embodiment; Figure 2 is a flowchart of step S104 in Figure 1; Figure 3 is a flowchart of step S202 in Figure 2; Figure 4 is a flowchart of step S306 in Figure 3; Figure 5 is a flowchart of step S105 in Figure 1; Figure 6 is a structural schematic diagram of the agricultural disaster risk prediction device for agricultural insurance policies provided in this application embodiment; Figure 7 is a hardware structure schematic diagram of the electronic device provided in this application embodiment. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0019] It should be noted that although functional modules are divided in the device schematic diagram and a logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than the module division in the device or the order in the flowchart. The terms "first," "second," etc., in the specification, claims, and the aforementioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence.

[0020] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only and is not intended to limit this application.

[0021] First, let's clarify some terms used in this application: Artificial intelligence (AI) is a new technical science that studies and develops theories, methods, technologies, and application systems for simulating, extending, and expanding human intelligence. AI is a branch of computer science that attempts to understand the essence of intelligence and produce a new type of intelligent machine that can react in a way similar to human intelligence. Research in this field includes robotics, speech recognition, image recognition, natural language processing, and expert systems. AI can simulate the information processes of human consciousness and thought. AI also refers to the theories, methods, technologies, and application systems that use digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0022] Natural Language Processing (NLP): NLP uses computers to process, understand, and utilize human language (such as Chinese and English). It is a branch of artificial intelligence and an interdisciplinary field of computer science and linguistics, often referred to as computational linguistics. NLP includes syntactic analysis, semantic analysis, and discourse understanding. It is commonly used in machine translation, handwritten and printed character recognition, speech recognition and text-to-speech conversion, intent recognition, information extraction and filtering, text classification and clustering, sentiment analysis, and opinion mining. It involves data mining, machine learning, knowledge acquisition, knowledge engineering, artificial intelligence research, and linguistic research related to language computation.

[0023] Agricultural disaster risk prediction refers to the dynamic assessment of potential risks to agricultural production in a specific region and time period based on monitoring data of agricultural meteorology and biological disasters, combined with future weather trends and crop growth status. The core of agricultural disaster risk prediction lies in quantifying the probability and severity of disasters, such as the threats to crop yield or quality from frost, drought, and pests and diseases. This provides a scientific basis for disaster prevention and mitigation deployments and production adjustments, thereby reducing disaster losses. In fintech scenarios, agricultural disaster risk prediction methods can be combined to predict the disaster risks faced by insured agricultural products.

[0024] Currently, agricultural disaster risk prediction is mainly achieved using statistical regression methods or machine learning models. However, in practical applications, statistical regression methods or machine learning models can only achieve local risk prediction under specific crops and specific regions, and cannot be applied to more regions and crops. Their applicability and generalization performance are poor, which affects the accuracy of agricultural disaster risk prediction.

[0025] Based on this, the embodiments of this application provide a method, device, electronic device and medium for predicting agricultural disaster risks for agricultural insurance policies, which aims to solve the technical problem that agricultural disaster risk prediction cannot be applied to more crops and regions due to its poor applicability, and improve the accuracy of agricultural disaster risk prediction.

[0026] The agricultural disaster risk prediction method, device, electronic equipment, and medium for agricultural insurance policies provided in this application are specifically illustrated through the following embodiments. First, the agricultural disaster risk prediction method for agricultural insurance policies in this application embodiment is described.

[0027] The embodiments of this application can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence (AI) refers to the theories, methods, technologies, and application systems that use digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0028] Foundational technologies for artificial intelligence generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing, operating / interactive systems, and mechatronics. AI software technologies mainly encompass computer vision, robotics, biometrics, speech processing, natural language processing, and machine learning / deep learning.

[0029] The method for predicting agricultural disaster risks for agricultural insurance policies provided in this application relates to the field of artificial intelligence technology. This method can be applied to a terminal, a server, or software running on either a terminal or a server. In some embodiments, the terminal can be a smartphone, tablet, laptop, desktop computer, etc.; the server can be configured as an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms; the software can be an application implementing the method for predicting agricultural disaster risks for agricultural insurance policies, but is not limited to the above forms.

[0030] This application can be used in a wide variety of general-purpose or special-purpose computer system environments or configurations. Examples include: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, and distributed computing environments including any of the above systems or devices. This application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform specific tasks or implement specific abstract data types. This application can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0031] It should be noted that in all specific embodiments of this application, when processing data related to user identity or characteristics, such as user information, user behavior data, user historical data, and user location information, user permission or consent is obtained first. Furthermore, the collection, use, and processing of this data comply with relevant laws, regulations, and standards. In addition, when embodiments of this application require access to sensitive personal information of users, separate permission or consent from the user is obtained through pop-ups or redirection to confirmation pages. Only after obtaining the user's separate permission or consent is the necessary user-related data required for the proper functioning of these embodiments acquired.

[0032] Figure 1 is an optional flowchart of the agricultural disaster risk prediction method for agricultural insurance policies provided in the embodiments of this application. The method in Figure 1 may include, but is not limited to, steps S101 to S106.

[0033] Step S101: Obtain raw multi-source environmental data; Step S102: Obtain an agricultural insurance policy resource pool; wherein, the agricultural insurance policy resource pool includes raw agricultural insurance policies, which are used to insure the current crop; Step S103: Extract information from the raw agricultural insurance policies to obtain the insured area information and the current growth information of the current crop; Step S104: Based on the raw multi-source environmental data, the insured area information, and the current growth information, perform knowledge retrieval on the preset agricultural disaster information database to obtain reference agricultural disaster information; Step S105: Based on the reference agricultural disaster information, perform risk prediction to obtain agricultural disaster risk prediction data; Step S106: Based on the agricultural disaster risk prediction data, perform policy screening on the raw agricultural insurance policies to obtain disaster-affected agricultural insurance policies.

[0034] Steps S101 to S106, as illustrated in this embodiment, comprehensively assess the external environmental conditions of crop growth by acquiring raw multi-source environmental data, providing a foundation for subsequent analysis. An agricultural insurance policy resource pool is obtained, including original agricultural insurance policies used to insure the current crop, thus defining the scope of analysis. Next, information is extracted from the original agricultural insurance policies to obtain the insured area information and the current crop's growth information, accurately locating the crop's current location and growth status. Furthermore, based on the raw multi-source environmental data, insured area information, and current growth information, a knowledge retrieval is performed on a pre-set agricultural disaster information database to quickly obtain reference agricultural disaster information matching the current situation, improving information acquisition efficiency. Finally, risk prediction is conducted based on reference agricultural disaster information to obtain agricultural disaster risk prediction data. Based on the agricultural disaster risk prediction data and the information on the insured area, the original agricultural insurance policies are screened to accurately identify the agricultural insurance policies affected by disasters. This helps to assess and process claims in a timely and accurate manner, solves the technical problem that agricultural disaster risk prediction cannot be applied to more crops and regions due to its poor applicability, improves the accuracy of agricultural disaster risk prediction, and enhances the quality and efficiency of agricultural insurance services.

[0035] In step S101 of some embodiments, the raw multi-source environmental data is a collection of unprocessed data related to agriculture and the environment from different channels and of different types. It can comprehensively and holistically grasp the external environmental conditions of crop growth and provide rich and detailed basic information for subsequent accurate analysis of the disaster risks that crops may face.

[0036] Specifically, the original multi-source environmental data includes: remote sensing images, meteorological data, crop pest and disease monitoring data, agricultural sensor sampling data, etc.; remote sensing images are image data that reflect the crop growth status, topography and other information of the preset area, which are obtained through remote sensing platforms such as satellites and aircraft. They are usually collected from a large area (such as various prefecture-level cities / districts and counties).

[0037] Meteorological data is collected by meteorological agencies and includes data on meteorological elements such as temperature, humidity, precipitation, wind speed, and sunshine in the current time period, as well as meteorological forecasts for future periods.

[0038] Crop pest and disease monitoring data is information about crops that may be affected by pests and diseases, which is detected and released by agricultural monitoring departments or related institutions. This includes the types of pests and diseases, the time of occurrence, the scope of occurrence, and the types of crops affected.

[0039] Agricultural sensor sampling data is data related to crop growth collected by sensors installed in cultivated land.

[0040] In step S102 of some embodiments, the agricultural policy resource pool is a collection of all agricultural insurance-related policies centrally stored and managed by the insurance company. The agricultural policy resource pool includes multiple original agricultural policies, which are used to insure the current crop, which is the insured object. The current crop can be corn, wheat, rice, fruit trees, etc., and is not limited to these. Each original agricultural policy contains information such as the policyholder's name, contact information, insured object, insured amount, and coverage area address.

[0041] Therefore, by analyzing the original agricultural insurance policy, information on the insured area and the current growth information of the crop can be extracted, accurately locating the crop's location and understanding its growth status. This provides crucial information for subsequent assessment of disaster risks in conjunction with environmental data, making risk assessment more targeted and accurate.

[0042] Specifically, the coverage area information refers to the specific geographical location of the insured object (current crop) as clearly specified in the original agricultural insurance policy. The current growth information of the current crop is reported regularly by farmers, including the current growth stage of the crop (such as sowing period, growing period, maturity period, etc.) and growth status (such as good growth, slow growth, etc.).

[0043] Before step S104 in some embodiments, the agricultural disaster risk prediction method for the agricultural insurance policy further includes constructing an agricultural disaster information database. Specifically, the construction process of the agricultural disaster information database includes, but is not limited to, the following steps: acquiring historical disaster case data, agricultural insurance claim data, historical crop growth data, historical meteorological data, and regional soil data; extracting features based on historical disaster case data to obtain historical disaster features; analyzing historical crop growth data and historical meteorological data to obtain crop-meteorological relationship coefficients; analyzing historical crop growth data and regional soil data to obtain yield-soil relationship coefficients; extracting features based on agricultural insurance claim data to obtain agricultural insurance claim features; confirming the historical disaster features, crop-meteorological relationship coefficients, yield-soil relationship coefficients, and agricultural insurance claim features as original agricultural disaster information, and constructing the agricultural disaster information database.

[0044] Specifically, by acquiring historical disaster case data, agricultural insurance claim data, historical crop growth data, historical meteorological data, and regional soil data, and performing feature extraction and analysis on these data respectively, we can obtain historical disaster characteristics, crop-meteorological relationship coefficients, yield-soil relationship coefficients, and agricultural insurance claim characteristics. This allows us to construct an agricultural disaster information database, which can provide comprehensive and accurate data support for subsequent agricultural disaster risk assessments, helping to prevent disasters in advance and reduce agricultural losses and insurance risks.

[0045] Among them, historical disaster case data are relevant records of past agricultural disaster events, including disaster type, time of occurrence, location, affected area, and degree of loss.

[0046] Agricultural insurance claims data contains detailed information about insurance companies’ past compensation payments to affected crops, such as the amount paid, the time of payment, and the recipients of the compensation.

[0047] Historical crop growth data is a record of the growth status of crops in different time periods and regions in the past, such as growth stage, growth height, and yield.

[0048] Historical meteorological data is past meteorological information, covering temperature, humidity, precipitation, sunshine, wind speed, etc.

[0049] Regional soil data refers to various indicators of soil within a specific region, such as soil fertility, pH, and texture.

[0050] In some embodiments, data mining, statistical analysis and other methods can be used to identify key features from historical disaster case data to obtain historical disaster features. These historical disaster features are representative and regular information, such as the peak season for disasters, common disaster types, and the degree of damage to different crops.

[0051] Statistical methods such as correlation analysis and regression analysis can be used to analyze the relationship between historical crop growth data and historical meteorological data, calculate the crop-meteorological correlation coefficient, and clarify the degree of influence of meteorological factors on crop growth. Specifically, the crop-meteorological correlation coefficient can reflect the numerical value of the correlation between meteorological factors (such as temperature and precipitation) and crop growth status (such as yield and growth rate). For example, comparing wheat growth data with temperature and precipitation data in a certain region, it can be found that for every 1°C increase in temperature, wheat yield may increase by 5%, and for every 10 mm increase in precipitation, wheat yield may increase by 3%, thus obtaining the corresponding crop-meteorological correlation coefficient.

[0052] Statistical analysis methods can be used to analyze the relationship between yield data in historical crop growth data and regional soil data, calculating the yield-soil correlation coefficient to understand the impact of soil conditions on crop yield. Specifically, this yield-soil correlation coefficient represents the degree of correlation between soil conditions (such as soil fertility, pH, etc.) and crop yield. For example, analyzing corn yield data and soil fertility data in a certain region reveals that for every grade increase in soil fertility, corn yield may increase by 10%, thus deriving the yield-soil correlation coefficient.

[0053] Data mining techniques can be used to extract features from agricultural insurance claims data to obtain agricultural insurance claims characteristics, such as high-incidence periods of claims, common causes of claims, and distribution of claims amounts. These agricultural insurance claims characteristics are used to characterize agricultural insurance terms and claims standards.

[0054] Finally, all the obtained historical disaster characteristics, crop-meteorological relationship coefficients, yield-soil relationship coefficients, and agricultural insurance compensation characteristics are used as raw agricultural disaster information and entered into the database management system to build a complete agricultural disaster information database, which can be queried by keywords.

[0055] Understandably, raw agricultural disaster information is an element of the agricultural disaster information database, which includes multiple raw agricultural disaster information sets.

[0056] Please refer to Figure 2. In some embodiments, step S104 may include, but is not limited to, steps S201 to S205: Step S201, performing feature filtering on the original multi-source environmental data based on the insured area information to obtain initial multi-source environmental data; Step S202, extracting features from the initial multi-source environmental data and current growth information to obtain target multi-source risk information; Step S203, matching the target multi-source risk information with each original agricultural disaster information to obtain information matching data; Step S204, filtering multiple original agricultural disaster information based on the information matching data to obtain agricultural disaster historical knowledge; Step S205, enhancing the target multi-source risk information based on agricultural disaster historical knowledge to obtain reference agricultural disaster information.

[0057] Steps S201 to S205, as illustrated in this embodiment, involve feature filtering of the original multi-source environmental data based on the insured area information to obtain initial multi-source environmental data. This allows for focusing on data from specific areas, removing irrelevant information, and improving data relevance and processing efficiency. Next, feature extraction is performed on the initial multi-source environmental data and current growth information to obtain target multi-source risk information, which is used for subsequent knowledge retrieval from the agricultural disaster information database. Further, information matching is performed between the target multi-source risk information and each original agricultural disaster information to obtain matching data. Based on this matching data, multiple original agricultural disaster information items are filtered to select those related to the original agricultural insurance policy, thus obtaining historical agricultural disaster knowledge and avoiding information redundancy. Finally, information enhancement is performed on the target multi-source risk information based on historical agricultural disaster knowledge to obtain reference agricultural disaster information. This information integrates multiple factors, providing a more accurate and comprehensive reference for agricultural disaster assessment and helping to improve the accuracy of agricultural disaster risk prediction.

[0058] In step S201 of some embodiments, data related to the insured area is filtered from the original multi-source environmental data based on the insured area information. This avoids processing a large amount of data from irrelevant areas, reduces the complexity and computational load of data processing, improves data processing efficiency, and ensures that the data analyzed subsequently is closely related to the insured area, thereby enhancing the relevance and accuracy of the analysis.

[0059] For example: Suppose the insured area is County A in a province, and the original multi-source environmental data includes meteorological and soil data from multiple counties across the province. By filtering, only the temperature, precipitation, and soil moisture data of County A are retained to obtain the initial multi-source environmental data.

[0060] In some embodiments, the initial multi-source environmental data includes: current remote sensing images, current meteorological data, and real-time monitoring data of crop diseases and pests.

[0061] The current remote sensing images are those relevant to the insured area information selected from all remote sensing images; the current meteorological data are those relevant to the insured area information selected from all meteorological data; and the real-time crop pest and disease monitoring data are those relevant to the insured area information selected from all crop pest and disease monitoring data.

[0062] Please refer to Figure 3. In some embodiments, step S202 may include, but is not limited to, steps S301 to S306: Step S301, obtaining current soil data based on the insured area information; Step S302, identifying cultivated land based on the current remote sensing image to obtain cultivated land identification data; Step S303, making weather predictions based on current meteorological data to obtain weather prediction data; Step S304, identifying the status of crop diseases and pests based on real-time monitoring data; Step S305, making yield predictions based on current soil data and current growth information to obtain crop yield prediction data; Step S306, performing feature fusion on the cultivated land identification data, weather prediction data, crop disease and pest status, and crop yield prediction data to obtain target multi-source risk information.

[0063] Steps S301 to S306, as illustrated in this embodiment, obtain current soil data based on the insured area information to accurately grasp the basic soil conditions of the insured area. Next, farmland identification is performed based on current remote sensing images to obtain farmland identification data, which clarifies the current crop type and planting area. Weather forecasting is performed based on current meteorological data to obtain weather forecast data, predicting future weather changes. Status identification is performed based on real-time crop pest and disease monitoring data to obtain the status of crop pests and diseases, enabling timely detection of potential pest and disease damage. Yield prediction is performed based on current soil data and current growth information to obtain crop yield prediction data, assessing crop yield risk. Finally, feature fusion is performed on the farmland identification data, weather forecast data, crop pest and disease status, and crop yield prediction data to obtain target multi-source risk information. This comprehensively covers multiple factors such as soil, farmland, weather, pests and diseases, and yield, providing a comprehensive and accurate assessment of agricultural risk and contributing to improving the accuracy of agricultural disaster risk prediction.

[0064] In step S301 of some embodiments, based on the given insured area information, various relevant data of the soil in the area are collected through methods such as field sampling, soil sensor monitoring, and querying existing soil databases. This allows for a precise understanding of the soil conditions within the insured area, providing basic data support for subsequent crop growth analysis and yield prediction. The current soil data covers various characteristics of the soil in the insured area, such as soil type (clay, loam, sandy soil, etc.), soil fertility (nutrient content such as nitrogen, phosphorus, and potassium), soil pH, and soil moisture.

[0065] In step S302 of some embodiments, farmland identification can be performed using the current remote sensing image with a ViT model or a U-Net model to extract information such as farmland distribution, vegetation index changes, and flooded areas, thereby obtaining farmland identification data. This allows for the rapid and accurate acquisition of distribution information for large areas of farmland, which helps in understanding the actual situation of farmland within the protected area. The farmland identification data is a collection of information about the location, area, and shape of farmland.

[0066] In step S303 of some embodiments, weather forecasting is performed on current meteorological data based on a deep learning model (such as the TemporalFusionTransformer model). This forecast predicts precipitation, temperature, humidity, and other data for a specific future time period (7-30 days), obtaining weather forecast data. This helps to take timely preventative measures before weather changes occur, reducing crop losses. The weather forecast data is a prediction of weather conditions (such as sunny days, rainy days, temperature changes, wind speed, etc.) for a future period.

[0067] In step S304 of some embodiments, the collected real-time monitoring data of crop diseases and pests are organized and analyzed to determine the types and extent of crop diseases and pests, thereby obtaining the status of crop diseases and pests, enabling timely detection of crop disease and pest problems, so as to take targeted prevention and control measures.

[0068] In some embodiments, if the real-time monitoring data of pests and diseases is in the form of images, a target detection model (such as YOLOv8) can be used to identify leaf lesions and pest density.

[0069] In step S305 of some embodiments, current soil data and current growth information can be input into the crop yield prediction model. The model comprehensively considers the impact of factors such as soil fertility, water conditions, and crop growth stage on yield to generate crop yield prediction data, which makes it easier for insurance companies to assess risks based on yield prediction and formulate reasonable insurance rates and compensation plans.

[0070] Please refer to Figure 4. In some embodiments, step S306 may include, but is not limited to, steps S401 to S405: Step S401, converting farmland identification data into text to obtain farmland identification text; Step S402, converting weather forecast data into text to obtain weather forecast text; Step S403, converting crop pest and disease status into text to obtain pest and disease status text; Step S404, converting crop yield forecast data into text to obtain crop yield forecast text; Step S405, performing structured processing on the farmland identification text, weather forecast text, pest and disease status text, and crop yield forecast text to obtain target multi-source risk information.

[0071] Steps S401 to S405, as illustrated in this embodiment, convert farmland identification data, weather forecast data, crop pest and disease status data, and crop yield forecast data into text, unifying different types of data into text format. This yields farmland identification text, weather forecast text, pest and disease status text, and crop yield forecast text, resolving the incompatibility issue of different modal feature spaces. Next, the farmland identification text, weather forecast text, pest and disease status text, and crop yield forecast text undergo structuring processing, integrating scattered and disordered text information into clearly structured and standardized multi-source risk information. This more comprehensively and accurately reflects various risk factors in agricultural production, providing reliable and comprehensive data support for subsequent risk assessment and improving the accuracy of agricultural disaster risk prediction.

[0072] In step S401 of some embodiments, the farmland identification data is converted into natural language text to obtain farmland identification text, which is easier for computers to understand and process. For example, farmland identification text may include phrases such as "The current area contains 5 plots of farmland, located at coordinates (X1, Y1), (X2, Y2), etc., and the soil type is mainly yellow soil" or "The chlorophyll index of rice in the current area has decreased by 15% compared to last month, and the area suspected to be affected by floods is about 120 hectares."

[0073] In step S402 of some embodiments, the weather forecast data is converted into natural language text to obtain weather forecast text, which is easier for computers to understand and process. For example, the weather forecast text may be "The probability of rainfall in the next 10 days is 80%, with three consecutive days of rainfall expected and an average precipitation of 95 mm" or "In the coming week, the daytime high temperature in this area will be between 25-30℃, and the nighttime low temperature will be between 15-20℃, with moderate to heavy rain from Wednesday to Friday."

[0074] In step S403 of some embodiments, the crop pest and disease status is converted into text describing the pest and disease status in natural language, resulting in pest and disease status text that is easier for a computer to understand and process. For example, the pest and disease status text could be something like, "The risk index of rice planthoppers detected is 0.82, which is a high-risk area," or "Aphids were found on the wheat leaves in this plot, with an average of 10-15 aphids per plant, indicating a moderate incidence, mainly concentrated in the central and eastern parts of the plot."

[0075] In step S404 of some embodiments, the crop yield prediction data is converted into natural language text to obtain crop yield prediction text, which is easier for the model to understand and process. For example, the crop yield prediction text could be: "According to the forecast, the average yield of wheat in this region this year is expected to be between 450-500 kg per mu, and the total yield will increase compared to last year."

[0076] In step S405 of some embodiments, the farmland identification text, weather forecast text, pest and disease status text and crop yield forecast text are structured based on a preset natural language template and automatically assembled to obtain target multi-source risk information, which can comprehensively reflect the comprehensive information of potential risks in agricultural production, and is easier for the model to understand and process.

[0077] For example, information such as farmland location and area from farmland identification text, temperature and precipitation from weather forecast text, pest and disease types and severity from pest and disease status text, and yield forecast from crop yield forecast text can be extracted and integrated into a structured table to obtain target multi-source risk information. Alternatively, farmland identification text, weather forecast text, pest and disease status text, and crop yield forecast text can be sequentially concatenated to obtain a complete text, i.e., target multi-source risk information.

[0078] In step S203 of some embodiments, the similarity between the target multi-source risk information and each original agricultural disaster information can be calculated to obtain information matching data; wherein, the scene matching data is used to represent the degree of matching between the target multi-source risk information and each original agricultural disaster information. Specifically, the similarity calculation can employ methods such as cosine similarity calculation, Pearson similarity calculation, Euclidean distance calculation, and Manhattan distance calculation, and is not limited to these.

[0079] For example, if multi-source risk information indicates that a certain region is currently experiencing high temperatures and low rainfall, posing a drought risk, and this information is matched with data in an agricultural disaster information database, and it is found that an original agricultural disaster information record shows a moderate drought event that occurred in the same region under similar climatic conditions, then the information matching data between this original agricultural disaster information and the target multi-source risk information will be relatively high.

[0080] In step S204 of some embodiments, a preset similarity threshold can be set. Original agricultural disaster information in the information matching data that is less than the similarity threshold is filtered out, and original agricultural disaster information in the information matching data that is greater than or equal to the similarity threshold is retained as historical agricultural disaster knowledge. Alternatively, the original agricultural disaster information with the largest similarity in the information matching data can be selected as historical agricultural disaster knowledge. Or, the top few original agricultural disaster information in the information matching data can be selected as historical agricultural disaster knowledge, and this is not limited to these methods. By filtering out original agricultural disaster information that is irrelevant or has a low correlation with the target multi-source risk information, redundant information is removed, making the historical agricultural disaster knowledge more refined and targeted.

[0081] In step S205 of some embodiments, the retrieved historical knowledge of agricultural disasters is spliced ​​with the target multi-source risk information to obtain more comprehensive, accurate and valuable reference agricultural disaster information. For example, based on the fact that the loss rate of flood events in the same region over the past five years is approximately 0.35, and combined with remote sensing and meteorological analysis results, the possible loss range of the current disaster is inferred.

[0082] Please refer to Figure 5. In some embodiments, step S105 may include, but is not limited to, steps S501 to S502: Step S501, classifying disaster risk levels based on reference agricultural disaster information to obtain disaster risk levels; Step S502, generating disaster-affected area information based on reference agricultural disaster information; Step S503, estimating losses based on reference agricultural disaster information to obtain estimated loss data; Step S504, generating agricultural disaster risk prediction data based on disaster risk levels, disaster-affected area information, and estimated loss data.

[0083] Steps S501 to S504, as illustrated in the embodiments of this application, involve classifying disaster risks based on reference agricultural disaster information to obtain disaster risk levels and assess the severity of the disaster. Disaster-affected area information is generated based on the reference agricultural disaster information to clarify the scope of the disaster's impact. Loss estimation is performed based on the reference agricultural disaster information to obtain estimated loss data, quantifying the potential losses caused by the disaster. Finally, agricultural disaster risk prediction data is generated based on the disaster risk level, disaster-affected area information, and estimated loss data, providing comprehensive and accurate data for agricultural disaster prevention and mitigation, and improving the accuracy of agricultural disaster risk prediction.

[0084] In some embodiments, a pre-trained large language reasoning model (such as the GPT model, Qwen model, DeepSeek model, etc.) is selected as the risk prediction model. The risk prediction model takes reference agricultural disaster information as input and performs the above steps S501 to S504.

[0085] In addition, risk prediction models can also generate corresponding disaster relief and recovery guidance suggestions based on reference agricultural disaster information.

[0086] In step S501 of some embodiments, key features, such as disaster intensity indicators (e.g., rainfall, temperature deviation, pest and disease density, etc.) and the size of the affected area, are extracted from reference agricultural disaster information. Then, based on pre-set classification standards, this information is mapped to different disaster risk levels. A disaster risk level is a quantitative or qualitative description of the potential harm of an agricultural disaster, typically categorized into low, medium, high, or more detailed levels. Through level classification, the severity of different agricultural disasters can be quickly and intuitively understood, providing a basis for subsequent risk assessment and response strategy development.

[0087] In step S502 of some embodiments, the geographical location-related data in the reference agricultural disaster information is organized and analyzed to determine the actual regional boundaries and scope affected by the disaster, and is transformed into clear and specific disaster-affected area information. The disaster-affected area information refers to the geographical area where the agricultural disaster occurred and caused its impact, which may include specific administrative divisions (such as townships or counties), geographical coordinates, etc.

[0088] In step S503 of some embodiments, based on factors such as disaster type, intensity, and affected area in the referenced agricultural disaster information, and combined with relevant agricultural economic data (such as crop unit price, agricultural facility cost, etc.), the various losses caused by the agricultural disaster are estimated to obtain estimated loss data, which helps to accurately assess the degree of impact of the disaster on the agricultural economy. The estimated loss data includes, but is not limited to, crop yield loss, agricultural facility damage loss, and livestock losses.

[0089] In step S504 of some embodiments, the obtained disaster risk level, disaster area information and estimated loss data are integrated to obtain agricultural disaster risk prediction data. The essence of this agricultural disaster risk prediction data is a disaster summary report, which integrates multiple factors such as disaster risk level, disaster area and loss situation, including the probability of different levels of disasters occurring in different areas in the future and the possible scope of losses.

[0090] In step S106 of some embodiments, based on the information of the disaster-affected area recorded in the agricultural disaster risk prediction data, original agricultural insurance policies that meet the criteria for possible disaster are selected from the original agricultural insurance policy resource pool and used as disaster-affected agricultural insurance policies. This enables insurance companies to assess and process claims in a timely and accurate manner, improve service efficiency and quality, and protect the interests of the insured.

[0091] For example, based on the predicted data of drought disasters that may occur to corn in a certain region, agricultural insurance policies that cover corn in that region and are located in drought-affected areas are selected from the agricultural insurance policy resource pool for subsequent claims processing.

[0092] In some embodiments, after step S106, the agricultural disaster risk prediction method for the agricultural insurance policy may also include, but is not limited to, the following steps: obtaining policyholder information of the affected agricultural insurance policy; issuing agricultural disaster risk warnings based on agricultural disaster risk prediction data and policyholder information.

[0093] Specifically, by first obtaining the policyholder information of agricultural insurance policies affected by disasters and clearly informing the recipients, agricultural disaster risk prediction data is then sent to the policyholders corresponding to the policyholder information, thereby realizing agricultural disaster risk warnings and enabling policyholders to take precautions in advance and reduce disaster losses.

[0094] Policyholder information includes the policyholder's name, contact information, address, and other basic identity information. This information is used to accurately contact the policyholder and ensure that alarm information is delivered to the truly affected policyholders in a timely and accurate manner, avoiding miscommunication or omission. Specifically, policyholder information for affected agricultural insurance policies can be found through the insurance company's business systems, databases, and other channels.

[0095] After identifying the policyholder, agricultural disaster risk prediction data can be sent to the policyholder to achieve agricultural disaster risk warning; risk warning information can be sent to the policyholder according to preset warning methods and content templates to achieve agricultural disaster risk warning, and this is not limited to these.

[0096] The agricultural disaster risk prediction method for agricultural insurance policies provided in this application can solve the problem of incompatibility of multimodal feature spaces. Furthermore, by combining the reasoning ability of a large model with an agricultural disaster information database, it enables domain-specificity and verifiability in risk prediction, which is conducive to improving the accuracy of agricultural disaster risk prediction.

[0097] Please refer to Figure 6. This application embodiment also provides an agricultural disaster risk prediction device for agricultural insurance policies, which can realize the above-mentioned agricultural disaster risk prediction method for agricultural insurance policies. The device includes: an environmental data acquisition module 601, used to acquire raw multi-source environmental data; a policy acquisition module 602, used to acquire an agricultural insurance policy resource pool; wherein, the agricultural insurance policy resource pool includes raw agricultural insurance policies, which are used to insure the current crop; a policy information extraction module 603, used to extract information from the raw agricultural insurance policies to obtain the insured area information and the current growth information of the current crop; a knowledge retrieval module 604, used to perform knowledge retrieval on a preset agricultural disaster information database based on the raw multi-source environmental data, the insured area information, and the current growth information to obtain reference agricultural disaster information; an agricultural disaster risk prediction module 605, used to perform risk prediction based on the reference agricultural disaster information to obtain agricultural disaster risk prediction data; and a disaster-affected policy screening module 606, used to screen the raw agricultural insurance policies based on the agricultural disaster risk prediction data to obtain disaster-affected agricultural insurance policies.

[0098] The specific implementation method of the agricultural disaster risk prediction device for this agricultural insurance policy is basically the same as the specific implementation method of the agricultural disaster risk prediction method for the above-mentioned agricultural insurance policy, and will not be repeated here.

[0099] This application also provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the above-mentioned method for predicting agricultural disaster risks in agricultural insurance policies. This electronic device can be any smart terminal, including tablet computers, in-vehicle computers, etc.

[0100] Please refer to Figure 7, which illustrates the hardware structure of an electronic device according to another embodiment. The electronic device includes: a processor 701, which can be implemented using a general-purpose CPU (Central Processing Unit), microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits, for executing related programs to implement the technical solutions provided in the embodiments of this application; and a memory 702, which can be implemented using a read-only memory (ROM), static storage device, dynamic storage device, or random access memory (RAM), etc. The memory 702 can store the operating system and other applications. When the technical solutions provided in the embodiments of this specification are implemented by software or firmware, the relevant program code is stored in the memory 702 and is called by the processor 701 to execute the agricultural disaster risk prediction method of the agricultural insurance policy of this application embodiment. The input / output interface 703 is used to realize information input and output. The communication interface 704 is used to realize communication interaction between this device and other devices. Communication can be realized by wired means (such as USB, network cable, etc.) or by wireless means (such as mobile network, WIFI, Bluetooth, etc.). The bus 705 transmits information between the various components of the device (such as processor 701, memory 702, input / output interface 703 and communication interface 704). The processor 701, memory 702, input / output interface 703 and communication interface 704 realize communication connection between each other within the device through the bus 705.

[0101] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method for predicting agricultural disaster risks in agricultural insurance policies.

[0102] Memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer-executable programs. Furthermore, memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, memory may optionally include memory remotely located relative to the processor, and these remote memories can be connected to the processor via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0103] The agricultural disaster risk prediction method, device, electronic equipment, and medium provided in this application embodiment, through acquiring raw multi-source environmental data, can comprehensively grasp the external environmental conditions of crop growth, providing a foundation for subsequent analysis. An agricultural policy resource pool is acquired, including raw agricultural policies used to insure the current crop, thus clarifying the scope of analysis. Next, information is extracted from the raw agricultural policies to obtain the insured area information and the current growth information of the crop, accurately locating the current crop's location and growth status. Furthermore, based on the raw multi-source environmental data, insured area information, and current growth information, a knowledge retrieval is performed on a pre-set agricultural disaster information database to quickly obtain reference agricultural disaster information matching the current situation, improving information acquisition efficiency. Finally, risk prediction is conducted based on reference agricultural disaster information to obtain agricultural disaster risk prediction data. Based on the agricultural disaster risk prediction data and the information on the insured area, the original agricultural insurance policies are screened to accurately identify the agricultural insurance policies affected by disasters. This helps to assess and process claims in a timely and accurate manner, solves the technical problem that agricultural disaster risk prediction cannot be applied to more crops and regions due to its poor applicability, improves the accuracy of agricultural disaster risk prediction, and enhances the quality and efficiency of agricultural insurance services.

[0104] The embodiments described in this application are for the purpose of more clearly illustrating the technical solutions of the embodiments of this application, and do not constitute a limitation on the technical solutions provided by the embodiments of this application. As those skilled in the art will know, with the evolution of technology and the emergence of new application scenarios, the technical solutions provided by the embodiments of this application are also applicable to similar technical problems.

[0105] Those skilled in the art will understand that the technical solutions shown in the figures do not constitute a limitation on the embodiments of this application, and may include more or fewer steps than shown, or combine certain steps, or different steps.

[0106] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0107] Those skilled in the art will understand that all or some of the steps in the methods disclosed above, as well as the functional modules / units in the systems and devices, can be implemented as software, firmware, hardware, or suitable combinations thereof.

[0108] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “having,” and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0109] It should be understood that in this application, "at least one (item)" means one or more, and "more than one" means two or more. "And / or" is used to describe the relationship between related objects, indicating that three relationships can exist. For example, "A and / or B" can represent three cases: only A exists, only B exists, and both A and B exist simultaneously, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one (item) of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one (item) of a, b, or c can represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.

[0110] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of the units described above is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0111] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0112] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0113] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes multiple instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing programs, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0114] The software tools or components not belonging to our company that appear in the embodiments of this application are for illustrative purposes only and do not represent actual use.

[0115] The preferred embodiments of the present application have been described above with reference to the accompanying drawings, but this does not limit the scope of the claims of the present application. Any modifications, equivalent substitutions, and improvements made by those skilled in the art without departing from the scope and substance of the embodiments of the present application shall be within the scope of the claims of the present application.

Claims

1. A method for predicting agricultural disaster risk in agricultural insurance policies, characterized in that, The method includes: acquiring raw multi-source environmental data; acquiring an agricultural insurance policy resource pool; wherein the agricultural insurance policy resource pool includes raw agricultural insurance policies used to insure the current crop; extracting information from the raw agricultural insurance policies to obtain insured area information and current growth information of the current crop; performing knowledge retrieval on a preset agricultural disaster information database based on the raw multi-source environmental data, the insured area information, and the current growth information to obtain reference agricultural disaster information; performing risk prediction based on the reference agricultural disaster information to obtain agricultural disaster risk prediction data; and filtering the raw agricultural insurance policies based on the agricultural disaster risk prediction data to obtain disaster-affected agricultural insurance policies.

2. The method according to claim 1, characterized in that, The agricultural disaster information database includes multiple original agricultural disaster information sets. The step of performing knowledge retrieval on the preset agricultural disaster information database based on the original multi-source environmental data, the insured area information, and the current growth information to obtain reference agricultural disaster information includes: performing feature filtering on the original multi-source environmental data based on the insured area information to obtain initial multi-source environmental data; extracting features from the initial multi-source environmental data and the current growth information to obtain target multi-source risk information; matching the target multi-source risk information with each of the original agricultural disaster information sets to obtain information matching data; filtering the multiple original agricultural disaster information sets based on the information matching data to obtain historical agricultural disaster knowledge; and enhancing the target multi-source risk information based on the historical agricultural disaster knowledge to obtain the reference agricultural disaster information.

3. The method according to claim 2, characterized in that, The initial multi-source environmental data includes: current remote sensing images, current meteorological data, and real-time monitoring data of crop diseases and pests. The step of extracting features from the initial multi-source environmental data and the current growth information to obtain target multi-source risk information includes: acquiring current soil data based on the insured area information; identifying cultivated land based on the current remote sensing image to obtain cultivated land identification data; performing weather forecasting based on the current meteorological data to obtain weather forecast data; identifying the status of crop diseases and pests based on the real-time monitoring data of crop diseases and pests to obtain crop disease and pest status; predicting yield based on the current soil data and the current growth information to obtain crop yield prediction data; and fusing features from the cultivated land identification data, the weather forecast data, the crop disease and pest status, and the crop yield prediction data to obtain the target multi-source risk information.

4. The method according to claim 3, characterized in that, The step of fusing features from the farmland identification data, weather forecast data, crop pest and disease status data, and crop yield forecast data to obtain the target multi-source risk information includes: converting the farmland identification data into text to obtain farmland identification text; converting the weather forecast data into text to obtain weather forecast text; converting the crop pest and disease status data into text to obtain pest and disease status text; converting the crop yield forecast data into text to obtain crop yield forecast text; and performing structured processing on the farmland identification text, weather forecast text, pest and disease status text, and crop yield forecast text to obtain the target multi-source risk information.

5. The method according to claim 2, characterized in that, The agricultural disaster information database is constructed through the following methods: acquiring historical disaster case data, agricultural insurance compensation data, historical crop growth data, historical meteorological data, and regional soil data; extracting features from the historical disaster case data to obtain historical disaster features; analyzing the historical crop growth data and the historical meteorological data to obtain crop-meteorological relationship coefficients; analyzing the historical crop growth data and the regional soil data to obtain yield-soil relationship coefficients; extracting features from the agricultural insurance compensation data to obtain agricultural insurance compensation features; and confirming the historical disaster features, the crop-meteorological relationship coefficients, the yield-soil relationship coefficients, and the agricultural insurance compensation features as the original agricultural disaster information, thereby constructing the agricultural disaster information database.

6. The method according to claim 1, characterized in that, The step of performing risk prediction based on the reference agricultural disaster information to obtain agricultural disaster risk prediction data includes: classifying disasters based on the reference agricultural disaster information to obtain disaster risk levels; generating disaster-affected area information based on the reference agricultural disaster information; estimating losses based on the reference agricultural disaster information to obtain estimated loss data; and generating the agricultural disaster risk prediction data based on the disaster risk level, the disaster-affected area information, and the estimated loss data.

7. The method according to any one of claims 1 to 6, characterized in that, After filtering the original agricultural insurance policies based on the agricultural disaster risk prediction data to obtain disaster-affected agricultural insurance policies, the method further includes: obtaining the policyholder information of the disaster-affected agricultural insurance policies; and issuing agricultural disaster risk warnings based on the agricultural disaster risk prediction data and the policyholder information.

8. A device for predicting agricultural disaster risks for agricultural insurance policies, characterized in that, The device includes: an environmental data acquisition module for acquiring raw multi-source environmental data; a policy acquisition module for acquiring an agricultural policy resource pool, wherein the agricultural policy resource pool includes raw agricultural policies used to insure the current crop; a policy information extraction module for extracting information from the raw agricultural policies to obtain insured area information and current growth information of the current crop; a knowledge retrieval module for performing knowledge retrieval on a preset agricultural disaster information database based on the raw multi-source environmental data, the insured area information, and the current growth information to obtain reference agricultural disaster information; an agricultural disaster risk prediction module for performing risk prediction based on the reference agricultural disaster information to obtain agricultural disaster risk prediction data; and a disaster-affected policy screening module for screening the raw agricultural policies based on the agricultural disaster risk prediction data to obtain disaster-affected agricultural policies.

9. An electronic device, characterized in that, The electronic device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the method according to any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the method of any one of claims 1 to 7.

Images