Agricultural multi-source information fusion method and system and storage medium

Abstract

The invention relates to the technical field of agricultural information fusion, in particular to an agricultural multi-source information fusion method and system and a storage medium. The method comprises the following steps: collecting and fusing multi-source agricultural data, generating a three-dimensional reconstruction agricultural model, then analyzing a cross-layer coupling structure of the model, obtaining a ground-air segmentation agricultural model, deriving a regional supply-demand relationship, analyzing and compensating nutrient demands based on growth demand parameters, mapping a fertilization prescription, generating a fertilization prescription map, and finally obtaining a fertilization result. And performing supply development simulation by utilizing the fertilization prescription map to obtain a simulated supply response model, determining irrigation opportunity according to the simulated supply response model, generating an irrigation prescription map, interactively integrating the fertilization prescription map and the irrigation prescription map, and generating agricultural fusion feedback information. According to the invention, a more accurate and more reliable agriculture-oriented multi-source information fusion method is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of agricultural information fusion, and particularly to a method, a system and a storage medium for agricultural multi-source information fusion. Background Art

[0002] Existing agricultural management methods often struggle to adapt to the rapidly changing market demands and complex ecological environments. Inefficiency and resource waste have become common phenomena. There is an urgent need for an innovative solution to address these issues. The agricultural multi-source information fusion method has emerged, aiming to improve the intelligence and scientific level of agricultural production by integrating multiple data sources. However, during the process of agricultural data collection and processing, the phenomenon of data islands is widespread, and there is a lack of effective integration and sharing among various types of data, resulting in a reduction in the accuracy and reliability of data analysis results. Traditional decision support systems cannot respond in a timely manner to complex agricultural production demands. The fragmentation and inconsistency of information pose greater challenges to agricultural production management, thereby affecting agricultural production efficiency and sustainable development. Summary of the Invention

[0003] Based on this, it is necessary to provide a method, a system and a storage medium for agricultural multi-source information fusion to solve at least one of the above technical problems.

[0004] To achieve the above object, a method for agricultural multi-source information fusion includes the following steps:

[0005] Step S1: Collect multi-source agricultural data; perform multi-source fusion on the multi-source agricultural data to obtain fused agricultural data; perform three-dimensional agricultural reconstruction on the fused agricultural data to generate a reconstructed agricultural model;

[0006] Step S2: Dissect the cross-layer coupling structure of the reconstructed agricultural model to obtain a ground-air segmented agricultural model; perform agricultural supply-demand derivation based on the segmented agricultural model to generate a regional supply-demand relationship;

[0007] Step S3: Analyze the compensated nutrient requirements of the regional supply-demand relationship based on preset growth requirement parameters; perform fertilization prescription mapping on the compensated nutrient requirements to generate a fertilization prescription map;

[0008] Step S4: Perform supply and development simulation on the reconstructed agricultural model based on the fertilization prescription map to obtain a simulated supply response model; determine the irrigation timing based on the simulated supply response model to generate an irrigation prescription map;

[0009] Step S5: Interactively integrate the fertilization prescription map and the irrigation prescription map to generate agricultural fusion feedback information, so as to achieve agricultural multi-source information fusion.

[0010] The present invention realizes comprehensive coverage of information through the collection of multi-source agricultural data, provides data support from different sources, including information in multiple aspects such as meteorology, soil, and crop growth, helps to form a more accurate basis for agricultural decision-making. The process of integrating agricultural data uniformly processes various types of data, eliminates the island effect between data, improves the correlation and consistency of data, making subsequent analysis more reliable. The implementation of three-dimensional agricultural reconstruction introduces a spatial dimension to the agricultural model, enhances the expressiveness of the crop growth environment, and helps to better understand the interaction between crops and the environment. The process of dissecting the cross-layer coupling structure provides a clear perspective for the analysis of agricultural characteristics at different levels. The formation of the ground-air segmentation agricultural model enables a more detailed study of the relationship between soil and crop growth. The derivation of regional supply-demand relationships provides a scientific basis for agricultural managers, clarifies the demand and supply status of various resources. The analysis of compensatory nutrient requirements ensures that crops can obtain sufficient nutrient support during the growth process. The implementation of fertilization prescription mapping effectively docks the actual demand and fertilization plan, improves the accuracy and efficiency of fertilization. The generation of fertilization prescription maps provides intuitive fertilization guidance for farmers, reducing resource waste caused by improper fertilization. The simulation of supply development based on fertilization prescription maps can reflect the changing demand of crops for nutrients in real time. The establishment of the simulation supply response model provides a scientific basis for determining the irrigation timing. The generation of irrigation prescription maps ensures the scientific management of farmland water. The process of interactively integrating fertilization prescription maps and irrigation prescription maps provides a comprehensive perspective for agricultural management. The generation of agricultural integration feedback information improves the transparency of farming activities, promotes the circulation and sharing of information among different agricultural links, overall enhances the intelligent level and management efficiency of agricultural production, promotes the optimal allocation and sustainable utilization of resources, and provides strong technical support and guarantee for realizing the high efficiency and precision of modern agriculture.

[0011] Preferably, step S1 includes the following steps:

[0012] Step S11: Deploy a sensor network and drones in the agricultural area, and monitor crop growth through the sensor network and drones to obtain multi-source agricultural data;

[0013] Step S12: Perform cross-modal alignment on the multi-source agricultural data to generate aligned agricultural data; associate heterogeneous agricultural information according to the aligned agricultural data;

[0014] Step S13: Based on the associated heterogeneous agricultural information, fuse the multi-source agricultural data features into fused agricultural data;

[0015] Step S14: Map the fused agricultural data into agricultural point cloud data; perform voxelization processing on the agricultural point cloud data to obtain an agricultural voxel model;

[0016] Step S15: Perform reverse three-dimensional reconstruction based on the agricultural voxel model to generate a reconstructed agricultural model.

[0017] The present invention realizes efficient and real-time crop growth monitoring through the deployment of sensor networks and drones, ensuring the comprehensiveness and accuracy of data collection. Cross-modal alignment improves the compatibility and consistency of multi-source agricultural data. The association of heterogeneous agricultural information enhances the integration and application value of data. Feature fusion technology provides a richer view of agricultural data. The mapping of agricultural point cloud data provides a basis for three-dimensional models. Voxelization processing improves the computational efficiency and operability of the model. Reverse three-dimensional reconstruction technology realizes the intuitive expression and visualization of agricultural states. Overall, it forms an efficient and intelligent agricultural information fusion system, promoting the scientific and intelligent process of agricultural management and enhancing the sustainable development ability of agricultural production.

[0018] Preferably, the cross-layer coupling structure of the dissected and reconstructed agricultural model includes:

[0019] Perform horizontal layer division on the reconstructed agricultural model to distinguish surface, underground, and spatial layers, obtaining layered agricultural data;

[0020] Extract vertical profiles from the layered agricultural data to generate profile agricultural data;

[0021] Identify cross-layer interaction features from the profile agricultural data to obtain cross-layer coupling features;

[0022] Based on the cross-layer coupling features, dissect the reconstructed agricultural model into layers to obtain a ground-air segmented agricultural model.

[0023] The present invention improves the structural degree of the agricultural model through horizontal layer division, clearly distinguishing surface, underground, and spatial layers. The generation of layered agricultural data enhances the pertinence and accuracy of data analysis. Vertical profile extraction provides a more detailed view of agricultural states. The formation of profile agricultural data lays a foundation for subsequent analysis. Cross-layer interaction feature identification realizes the association and interaction between different layers, enhancing the comprehensiveness and applicability of the model. Layered dissection based on cross-layer coupling features optimizes the analytical ability of the agricultural model. The formation of the ground-air segmented agricultural model provides a scientific basis and decision support for precision agricultural management. Overall, it improves the intelligence and efficiency of agricultural production and promotes the sustainable development of modern agriculture.

[0024] Preferably, the agricultural supply and demand derivation based on the segmented agricultural model includes:

[0025] Extract surface action features from the ground-air segmented agricultural model to obtain surface crop growth features;

[0026] Conduct underground soil supply analysis on the ground-air segmented agricultural model to generate underground supply capacity;

[0027] Perform a spatial adaptability assessment on the ground-air segmentation agricultural model to obtain the spatial meteorological adaptability;

[0028] Based on the surface crop growth characteristics, underground supply capacity, and spatial meteorological adaptability, simulate the crop growth supply and demand to obtain the simulated supply and demand process;

[0029] Derive the regional supply and demand relationship of the segmentation agricultural model according to the simulated supply and demand process.

[0030] The present invention provides key indicators for crop growth through the extraction of surface action characteristics, ensuring the accuracy and effectiveness of growth characteristics. The analysis of underground supply capacity provides a scientific basis for the rational utilization of soil resources. The assessment of spatial meteorological adaptability enhances the adaptability to environmental changes. The simulation of crop growth supply and demand realizes a comprehensive prediction of agricultural production dynamics. The generation of the simulated supply and demand process provides data support for regional agricultural management. The derivation of the regional supply and demand relationship promotes the effective allocation and rational utilization of resources, and overall improves the scientific nature of agricultural production decision-making and the intelligent level of management, and promotes the development and application of precision agriculture.

[0031] Preferably, step S3 includes the following steps:

[0032] Step S31: Analyze the nutrient supply content of the regional supply and demand relationship; calculate the difference between the nutrient supply content and the preset growth demand parameters to obtain the nutrient demand difference;

[0033] Step S32: Perform a compensation demand mapping according to the nutrient demand difference to generate a compensated nutrient demand;

[0034] Step S33: Decompose each nutrient demand of the compensated nutrient demand to obtain a specific nutrient demand set; perform a comparison and superposition on the specific nutrient demand set according to the preset fertilizer components to obtain the fertilizer demand;

[0035] Step S34: Sort the nutrient demand urgency of the compensated nutrient demand to obtain the nutrient demand order; plan the application plan for the fertilizer demand based on the nutrient demand order to generate a fertilization prescription map.

[0036] The present invention provides an accurate baseline of nutritional requirements through the analysis of nutrient supply content. The differential calculation reveals the gap between nutrient supply and growth requirements. The compensation demand mapping ensures the accurate identification and fulfillment of nutrient requirements. The breakdown of nutrient requirements divides complex requirements into specific nutrient types, promoting the precise allocation of resources. The comparison and superposition of fertilizer demand amounts provide a basis for scientific fertilization. The ranking of nutrient requirement urgency provides a basis for the rational arrangement of fertilization timing and application amount. The generation of fertilization prescription maps optimizes the implementation effect of fertilization plans, overall enhancing the resource utilization efficiency and economic benefits of agricultural production and promoting the development of sustainable agriculture.

[0037] Preferably, the simulation of the replenishment development of the reconstructed agricultural model based on the fertilization prescription map includes:

[0038] Fuse the data of the reconstructed agricultural model and the fertilization prescription map to obtain the initial fertilization response data;

[0039] Dynamically simulate the nutrient release of the initial fertilization response data to obtain nutrient release dynamic data;

[0040] Estimate the crop absorption efficiency of the nutrient release dynamic data to obtain nutrient absorption efficiency data;

[0041] Predict the crop growth response based on the nutrient absorption efficiency data to obtain growth response prediction data;

[0042] Conduct replenishment modeling on the growth response prediction data to obtain a simulated replenishment response model.

[0043] The present invention realizes the comprehensive integration of information through the data fusion of the reconstructed agricultural model and the fertilization prescription map. The generation of the initial fertilization response data provides a basis for subsequent analysis. The dynamic simulation of nutrient release enhances the real-time monitoring of fertilization effects. The estimation of nutrient release dynamic data improves the understanding of crop absorption efficiency. The nutrient absorption efficiency data provides a quantitative basis for crop growth optimization. The growth response prediction data provides a scientific prediction of crop growth. The realization of replenishment modeling provides effective decision-making support for agricultural management. The generation of the simulated replenishment response model promotes the implementation of precision fertilization and management, overall enhancing the intelligence and sustainability of agricultural production and promoting the efficient utilization of resources and the improvement of economic benefits.

[0044] Preferably, the determination of irrigation timing according to the simulated replenishment response model includes:

[0045] Calculate the crop transpiration of the simulated replenishment response model to obtain transpiration water demand data;

[0046] Estimate the soil evaporation of the simulated replenishment response model to obtain evaporation water consumption data;

[0047] Calculate the water demand based on the transpiration water demand data and the evaporation water consumption data to generate a water replenishment gap;

[0048] Determine the optimal irrigation timing based on the water replenishment gap to obtain the irrigation timing;

[0049] Carry out irrigation planning according to the irrigation timing to generate an irrigation prescription map.

[0050] The present invention provides an accurate water demand assessment through the calculation of crop transpiration. The generation of transpiration water demand data provides a scientific basis for irrigation decision-making. The estimation of soil evaporation enhances the understanding of water loss. The evaporation water consumption data provides necessary parameters for water demand calculation. The generation of the water replenishment gap reveals the urgency of irrigation demand. The determination of the optimal irrigation timing provides guidance for the rational utilization of water resources. The implementation of irrigation planning ensures the efficiency and pertinence of water replenishment. The generation of the irrigation prescription map optimizes the irrigation management plan. Overall, it improves the irrigation efficiency and the sustainability of agricultural production, and promotes the scientific management and utilization of resources.

[0051] Preferably, step S5 includes the following steps:

[0052] Step S51: Align the fertilization prescription map and the irrigation prescription map in time series to obtain a time series aligned prescription;

[0053] Step S52: Detect the conflict operation points of the time series aligned prescription; perform spatio-temporal scheduling on the time series aligned prescription based on the conflict operation points to obtain an agricultural activity schedule;

[0054] Step S53: Feedback the agricultural activity schedule in a visual form to generate agricultural integration feedback information to achieve agricultural multi-source information integration.

[0055] The present invention ensures the coordination of agricultural activities through the time series alignment of the fertilization prescription map and the irrigation prescription map. The generation of the time series aligned prescription provides a basis for subsequent scheduling. The detection of conflict operation points enhances the ability to identify potential problems. The implementation of spatio-temporal scheduling optimizes the resource allocation and utilization efficiency. The generation of the agricultural activity schedule provides a systematic plan for agricultural management. The visual form of the agricultural integration feedback information improves the intuitiveness and effectiveness of information transmission. Overall, it promotes the integration and application of agricultural multi-source information, and promotes the scientific management and decision-making support of agricultural production.

[0056] The present invention also provides a system for agricultural multi-source information integration, which is used to execute the method for agricultural multi-source information integration as described above. The system for agricultural multi-source information integration includes:

[0057] A data acquisition module, which is used to collect multi-source agricultural data; perform multi-source fusion on the multi-source agricultural data to obtain fused agricultural data; perform three-dimensional agricultural reconstruction on the fused agricultural data to generate a reconstructed agricultural model;

[0058] A supply and demand derivation module, which is used to analyze the cross-layer coupling structure of the reconstructed agricultural model to obtain a ground-air segmented agricultural model; perform agricultural supply and demand derivation based on the segmented agricultural model to generate a regional supply and demand relationship;

[0059] A nutrient analysis module, which is used to analyze the compensated nutrient demand of the regional supply and demand relationship based on preset growth demand parameters; perform fertilization prescription mapping on the compensated nutrient demand to generate a fertilization prescription map;

[0060] A supply replenishment simulation module, which is used to perform supply replenishment development simulation on the reconstructed agricultural model based on the fertilization prescription map to obtain a simulated supply replenishment response model; determine the irrigation timing according to the simulated supply replenishment response model to generate an irrigation prescription map;

[0061] An information integration module, which is used to interactively integrate the fertilization prescription map and the irrigation prescription map to generate agricultural fusion feedback information to achieve agricultural multi-source information fusion.

[0062] The present invention realizes the comprehensive acquisition of multi-source agricultural data through the introduction of a data acquisition module, covering various aspects of information such as meteorology, soil, and crop growth, providing a solid data foundation for subsequent analysis and decision-making. The process of data fusion integrates various types of information, eliminates information silos, enhances the relevance of data, and avoids decision-making errors caused by inconsistent data. The implementation of three-dimensional agricultural reconstruction provides an intuitive representation of the spatial characteristics of the farmland environment, enhancing the understanding of the internal relationships within the agricultural system. The supply and demand derivation module realizes a profound analysis of the interactions between different levels of the agricultural system by dissecting the cross-layer coupling structure of the reconstructed agricultural model. The formation of the ground-air segmentation agricultural model enables the detailed study of the relationship between the soil and the crops, providing a scientific basis for the generation of regional supply and demand relationships. The operation of the nutrient analysis module ensures the accurate analysis of the compensated nutrient requirements. The analysis based on the growth requirement parameters enables the crops to obtain the necessary nutrient support during the growth period. The implementation of the fertilization prescription mapping effectively connects the actual requirements with the fertilization plan, improving the fertilization efficiency. The formation of the fertilization prescription map provides intuitive fertilization guidance for farmers. The replenishment simulation module based on the fertilization prescription map can reflect the changing nutrient requirements of the crops in real time. The establishment of the simulation replenishment response model provides a scientific basis for determining the irrigation timing. The generation of the irrigation prescription map ensures the scientific management of water. The information integration module generates agricultural integration feedback information by interactively integrating the fertilization prescription map and the irrigation prescription map, enhancing the transparency of farming activities, promoting the circulation and sharing of information among different agricultural links, and overall enhancing the intelligent level and management efficiency of agricultural production, promoting the optimal allocation and sustainable utilization of resources, providing strong technical support and guarantee for achieving the high efficiency and precision of modern agriculture, forming an agricultural management system centered on data, capable of adapting to the changes in different agricultural production environments and requirements, enhancing the decision-making ability and production efficiency of farmers, promoting the transformation and upgrading of agricultural production methods, and driving the development of sustainable agriculture and the protection of the ecological environment.

[0063] The present invention also provides a computer-readable storage medium storing a computer program, and when the computer program is executed, it realizes the method for fusing multi-source agricultural information as described in any one of the above.

[0064] The present invention provides a flexible program storage and management method through a computer-readable storage medium. The execution of the program realizes the automation and intelligence of agricultural data processing, enhances the efficiency and accuracy of data analysis and decision-making, supports the efficient fusion and application of multi-source data, promotes information sharing and collaboration in the agricultural production process, provides scientific agricultural model reconstruction and optimization, ensures the accurate implementation of fertilization and irrigation, optimizes the allocation and use of agricultural resources, enhances the visualization and operability of agricultural management, and overall promotes the development of agricultural informatization and intelligence, laying a foundation for the sustainable development of modern agriculture. Brief Description of the Drawings

[0065] Figure 1 It is a schematic flow chart of the steps of a multi-source information fusion method for agriculture;

[0066] Figure 2 It is a schematic flow chart of the detailed implementation steps of step S3;

[0067] The realization, functional characteristics and advantages of the object of the present invention will be further described with reference to the embodiments and the accompanying drawings. Detailed Embodiment

[0068] The technical method of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts belong to the scope of protection of the present invention.

[0069] In addition, the accompanying drawings are only schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and thus their repeated description will be omitted. Some of the block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software form, or in one or more hardware modules or integrated circuits, or in different networks and / or processor methods and / or microcontroller methods.

[0070] It should be understood that although the terms "first", "second", etc. may be used here to describe various units, these units should not be limited by these terms. These terms are only used to distinguish one unit from another. For example, without departing from the scope of the exemplary embodiments, the first unit may be called the second unit, and similarly the second unit may be called the first unit. The term "and / or" used here includes any and all combinations of one or more of the listed related items.

[0071] To achieve the above object, please refer to Figures 1 to 2 , a multi-source information fusion method for agriculture, including the following steps:

[0072] Step S1: Collect multi-source agricultural data; perform multi-source fusion on the multi-source agricultural data to obtain fused agricultural data; perform three-dimensional agricultural reconstruction on the fused agricultural data to generate a reconstructed agricultural model;

[0073] Step S2: Dissect the cross-layer coupling structure of the reconstructed agricultural model to obtain a ground-air segmented agricultural model; perform agricultural supply and demand derivation based on the segmented agricultural model to generate a regional supply and demand relationship;

[0074] Step S3: Analyze the compensated nutrient requirements for the regional supply-demand relationship based on the preset growth requirement parameters; perform a fertilization prescription mapping on the compensated nutrient requirements to generate a fertilization prescription map.

[0075] Step S4: Based on the fertilization prescription map, conduct a supply and development simulation on the reconstructed agricultural model to obtain a simulated supply response model; determine the irrigation timing according to the simulated supply response model to generate an irrigation prescription map.

[0076] Step S5: Interactively integrate the fertilization prescription map and the irrigation prescription map to generate agricultural integration feedback information, so as to achieve the integration of multi-source agricultural information.

[0077] The present invention realizes the comprehensive coverage of information through the collection of multi-source agricultural data, provides data support from different sources, including information in multiple aspects such as meteorology, soil, and crop growth, helps to form a more accurate basis for agricultural decision-making. The process of integrating agricultural data uniformly processes various types of data, eliminates the data silo effect, improves the correlation and consistency of data, makes subsequent analysis more reliable. The implementation of three-dimensional agricultural reconstruction introduces a spatial dimension into the agricultural model, improves the expressiveness of the crop growth environment, and helps to better understand the interaction between crops and the environment. The process of dissecting the cross-layer coupling structure provides a clear perspective for the analysis of agricultural characteristics at different levels. The formation of the ground-air segmentation agricultural model enables a more detailed study of the relationship between soil and crop growth. The derivation of the regional supply-demand relationship provides a scientific basis for agricultural managers, clarifies the demand and supply status of various resources. The analysis of the compensated nutrient requirements ensures that crops can obtain sufficient nutrient support during the growth process. The implementation of the fertilization prescription mapping effectively connects the actual demand with the fertilization plan, improves the accuracy and efficiency of fertilization. The generation of the fertilization prescription map provides intuitive fertilization guidance for farmers and reduces the waste of resources caused by improper fertilization. The supply and development simulation based on the fertilization prescription map can reflect the changing demand of crops for nutrients in real time. The establishment of the simulated supply response model provides a scientific basis for determining the irrigation timing. The generation of the irrigation prescription map ensures the scientific management of farmland water. The process of interactively integrating the fertilization prescription map and the irrigation prescription map provides a comprehensive perspective for agricultural management. The generation of agricultural integration feedback information improves the transparency of farming activities, promotes the circulation and sharing of information among different agricultural links, overall improves the intelligent level and management efficiency of agricultural production, promotes the optimal allocation and sustainable utilization of resources, and provides strong technical support and guarantee for the realization of high-efficiency and precision of modern agriculture.

[0078] In an embodiment of the present invention, the method for integrating multi-source agricultural information includes the following steps:

[0079] Step S1: Collect multi-source agricultural data; perform multi-source fusion on the multi-source agricultural data to obtain fused agricultural data; perform three-dimensional agricultural reconstruction on the fused agricultural data to generate a reconstructed agricultural model;

[0080] In this embodiment, to collect multi-source agricultural data, first, agricultural data from different sources are obtained through ground monitoring devices and satellite remote sensing technology, including information such as soil moisture, meteorological conditions, crop growth status, and fertilization history. An automatic weather station is used to obtain soil temperature and humidity and precipitation data. A high-resolution sensor is carried by an unmanned aerial vehicle (UAV) to obtain crop canopy information. In addition, macro-climate data are collected through satellite images and meteorological stations. The ground data and remote sensing data are combined, and all information is summarized to a unified platform through a data acquisition system for data storage and preliminary preprocessing, including noise filtering, data synchronization, time alignment, etc. After data collection, multi-source fusion of agricultural data is performed. The Kalman Filter algorithm is used to fuse the data, fully considering the spatio-temporal distribution characteristics of different data sources. The collected data from different sources are integrated into a unified data model to generate fused agricultural data, and the quality of the data is evaluated and corrected to ensure that the fused data has high accuracy and consistency. On this basis, three-dimensional agricultural reconstruction is performed. The spatial interpolation algorithm is used to perform spatial reconstruction on the agricultural data to construct a three-dimensional agricultural model containing multi-dimensional information such as soil, crops, and meteorology. The process of reconstructing the agricultural model includes performing three-dimensional spatial and temporal fusion on data at different levels, and finally obtaining a three-dimensional agricultural data model with high accuracy.

[0081] Step S2: Analyze the cross-layer coupling structure of the reconstructed agricultural model to obtain a ground-air segmented agricultural model; perform agricultural supply and demand derivation based on the segmented agricultural model to generate regional supply and demand relationships;

[0082] In this embodiment, the cross-layer coupling structure of the dissected and reconstructed agricultural model is analyzed to obtain the ground-air segmented agricultural model. First, based on the reconstructed three-dimensional agricultural model, the spatial analysis of each layer of data is carried out to identify the coupling relationships of different layers such as the ground layer (soil layer), crop layer (plant canopy layer), and atmosphere layer. And based on the coupling relationships of these layers, a cross-layer coupling analysis model is constructed. The impacts between layers are quantitatively calculated through numerical analysis methods, and the coupling function is used to model the interaction between different layers. For example, the impact of water supply on crop growth is determined through the relationship model between soil moisture and crop transpiration. The interaction between different agricultural layers is analyzed according to this relationship, and on this basis, the ground-air segmented agricultural model is generated. The agricultural area is divided into several spatial units, each unit contains different soil, water, crop, and meteorological data. Then, by analyzing the cross-layer coupling relationships of these areas, detailed agricultural information of each partition is obtained, and a spatial distribution model of the agricultural area is generated. Relying on this model, agricultural supply and demand derivation is carried out, and further the agricultural supply and demand relationships of each area are calculated. Based on the demand for crop water and nutrients, combined with the climate and soil conditions, the supply and demand situations of each area are generated. According to the segmented agricultural model, agricultural supply and demand derivation is carried out to generate the regional supply and demand relationship. Combining with the previously obtained ground-air segmented agricultural model, based on the soil characteristics, meteorological data, and crop demand parameters of each agricultural unit, agricultural supply and demand derivation is carried out. First, the water and fertilizer requirements of the crops are accurately predicted. The algorithms based on crop growth models (such as the CERES model) are used to calculate the water and fertilizer requirements of the crops at different growth stages. The precipitation in the region is corrected according to the meteorological data, and this data is used to judge whether there are water and nutrient gaps. Then, according to the water retention capacity of the soil in the region and the water absorption capacity of the crops, combined with the meteorological forecast data, the supply and demand situations of water and nutrients in different regions are derived, and a regional supply and demand relationship chart is generated, showing the water and nutrient supply situations of each region and whether there are gaps.

[0083] Step S3: Analyze the compensated nutrient requirements of the regional supply and demand relationship based on the preset growth demand parameters; perform fertilization prescription mapping on the compensated nutrient requirements to generate a fertilization prescription map;

[0084] In this embodiment, the compensation nutrient requirements are analyzed based on the preset growth requirement parameters for the regional supply-demand relationship. First, according to the regional supply-demand relationship diagram, the nutrient gaps in each region are analyzed. If the soil nutrients in a certain area are lower than the growth requirements of the crops, then this region needs to supplement nutrients. Combining with the growth requirement parameters of the crops, the classical crop fertilizer requirement formula (such as the NPK model) is used to calculate the compensation amount of each nutrient. The precise fertilization algorithm is used to quantify the missing nutrients to ensure the accuracy of the fertilization amount. After that, the compensation nutrient requirements are converted into a fertilization prescription map, and the fertilization requirement information is visualized through an agricultural management platform to generate a fertilization prescription map. The map will clearly indicate the fertilization amount, fertilization type, and fertilization time for each region.

[0085] Step S4: Based on the fertilization prescription map, conduct a supply and development simulation on the reconstructed agricultural model to obtain a simulated supply response model; determine the irrigation timing according to the simulated supply response model to generate an irrigation prescription map.

[0086] In this embodiment, based on the fertilization prescription map, a supply and development simulation is conducted on the reconstructed agricultural model to obtain a simulated supply response model. First, based on the fertilization data in the fertilization prescription map, the reconstructed agricultural model is simulated. Considering the impact of fertilization operations on the soil, water, and crop growth, through the crop growth simulation algorithm, combined with the soil water model, the changes in the farmland after fertilization are simulated. Relying on the supply response model to analyze the specific impact of fertilization on crop growth, the model will calculate parameters such as the changes in soil water and crop growth status after fertilization according to the fertilization amount, fertilization time, and different growth stages of the crops. Then, based on the simulated supply response model, combined with the crop growth requirements and soil water conditions, the irrigation timing is determined. The system will accurately determine the best irrigation time based on factors such as crop growth stage, meteorological conditions, and soil water, and generate an irrigation prescription map. The irrigation prescription map will clearly mark the irrigation time, irrigation amount, and irrigation area.

[0087] Step S5: Interactively integrate the fertilization prescription map and the irrigation prescription map to generate agricultural integration feedback information to achieve multi-source information integration for agriculture.

[0088] In this embodiment, the fertilization prescription map and the irrigation prescription map are interactively integrated to generate agricultural integration feedback information. First, through a data fusion algorithm, the fertilization prescription map and the irrigation prescription map are aligned in time series to ensure that the time points of fertilization and irrigation operations are consistent. Then, the priorities and possible conflict points of fertilization and irrigation operations are compared and analyzed. If there are conflicting operations (such as overlapping fertilization and irrigation), the operation order is adjusted to generate a coordinated agricultural operation schedule. After that, a visualization tool (such as a GIS platform) is used to display the integrated agricultural operation schedule in the form of maps and tables to generate agricultural integration feedback information. The feedback information includes the time, location, and crop type of each operation, and specific operation guidelines are given to ensure that all operations can be coordinated and executed. Finally, the agricultural activities are fed back through the management platform to ensure the accuracy and timeliness of the data.

[0089] Preferably, step S1 includes the following steps:

[0090] Step S11: Deploy a sensor network and drones in the agricultural area, and monitor crop growth through the sensor network and drones to obtain multi-source agricultural data;

[0091] Step S12: Perform cross-modal alignment on the multi-source agricultural data to generate aligned agricultural data; associate heterogeneous agricultural information according to the aligned agricultural data;

[0092] Step S13: Based on the associated heterogeneous agricultural information, fuse the features of the multi-source agricultural data into fused agricultural data;

[0093] Step S14: Map the fused agricultural data into agricultural point cloud data; perform voxelization processing on the agricultural point cloud data to obtain an agricultural voxel model;

[0094] Step S15: Perform reverse three-dimensional reconstruction according to the agricultural voxel model to generate a reconstructed agricultural model.

[0095] In this embodiment, key planting plots in the agricultural area are selected, and multiple types of sensors are deployed on the ground, including soil moisture sensors, light intensity sensors, air temperature and humidity sensors, and leaf nitrogen content sensors. They are evenly arranged in a grid pattern at intervals of every 10 meters, and a low-power sensor network is constructed using the ZigBee (short-range wireless communication protocol) networking method, enabling various sensors to upload data to the edge computing gateway through wireless communication nodes. At the same time, the flight routes of drones are planned over the agricultural area, and a multi-spectral camera and a LiDAR (Light Detection and Ranging) sensor are used to scan the crop canopy structure, vegetation index, and terrain data, generating a farmland remote sensing dataset containing spectral, point cloud, and image data. A spatio-temporal synchronization strategy is adopted to align the sensor data and the drone remote sensing data in terms of time and space. The interpolation method is used to perform time interpolation on the sensor data to make the data timestamps consistent with the shooting times of the drone remote sensing data. The GPS (Global Positioning System) information is used to correct the coordinates of the drone remote sensing data to match the measurement points of the ground sensors. Then, based on the statistical characteristics of the data distribution, the feature leveling scale of each modal data is calculated, and the normalization transformation and normal distribution reshaping methods are used to normalize different modal data to ensure that the data value distributions are consistent. Finally, the agricultural meteorological data, crop management information, and historical operation records are associated with the leveled agricultural data according to the plot code to form a complete agricultural information fusion dataset. The convolutional neural network (CNN, Convolutional Neural Network) is called to extract the deep features of the crop canopy spectral data, and the principal component analysis (PCA, Principal Component Analysis) is used to reduce the dimension to obtain the main components of the spectral features. Combining the soil moisture, nutrient concentration, and air humidity data collected by the ground sensors, the optimal feature combination is selected through the Bayesian Information Criterion (BIC), and the attention mechanism is used to allocate feature weights. Finally, the multi-modal feature fusion network aggregates the features of the crop growth environment data to generate fusion agricultural data containing crop growth status, environmental factors, and management information. According to the spatial information in the fusion agricultural data, the inverse distance weighting interpolation method (IDW, Inverse Distance Weighting) is used to perform spatial interpolation on the discrete sensor measurement point data, enabling the data to form a continuous distribution in space and matching the coordinates with the drone LiDAR point cloud data to generate fusion agricultural point cloud data. Subsequently, the agricultural point cloud data is divided into grids, and the octree partitioning method is used for hierarchical processing to divide the point cloud data into spatial units at different levels, and the crop canopy density, soil moisture gradient, and light uniformity parameters are calculated within each unit.Finally, an agricultural voxel model with hierarchical information is generated. Based on the hierarchical structure of the agricultural voxel model, the Marching Cubes algorithm is used to extract the isosurface to construct a three-dimensional grid of the crop canopy, and the Poisson Surface Reconstruction method is adopted for detail optimization to make the crop surface smoother. At the same time, the VoxelFusion algorithm is called to calculate the three-dimensional morphology of the soil profile and root distribution, and the terrain undulation is adjusted by combining the gravity field simulation to obtain a complete reconstructed agricultural model.

[0096] Preferably, the cross-layer coupling structure of the dissected and reconstructed agricultural model includes:

[0097] Perform horizontal stratification on the reconstructed agricultural model to distinguish the surface, underground, and spatial levels, and obtain stratified agricultural data;

[0098] Extract vertical profiles from the stratified agricultural data to generate profile agricultural data;

[0099] Identify cross-layer interaction features from the profile agricultural data to obtain cross-layer coupling features;

[0100] Based on the cross-layer coupling features, dissect the reconstructed agricultural model into layers to obtain a ground-air segmented agricultural model.

[0101] In this embodiment, based on the three-dimensional point cloud data of the reconstructed agricultural model, the K-means (K-means clustering) method is used to group the data points according to the height range, and the data points are divided into a surface layer, a subsurface layer, and a spatial layer. The surface layer includes vegetation, farmland soil, and surface water bodies. The subsurface layer covers root distribution, soil profile, and groundwater dynamics. The spatial layer involves canopy structure, air humidity, and light distribution. Subsequently, the Delaunay triangulation method is used to grid the data points of each layer, and a continuous layered surface is generated through the B-spline interpolation method. Finally, a layered agricultural data including surface, subsurface, and spatial levels is constructed. Based on the data structures of the surface, subsurface, and spatial layers, a regular grid sampling method is used to perform profile sampling on the agricultural area, generating vertical slices at intervals of every 5 meters, and the Marching Squares algorithm is used to construct a two-dimensional profile contour. The distributions of soil moisture, root density, and air humidity within the profile are calculated through linear interpolation. At the same time, the geographic information system (GIS) tool is used to perform geographic registration on the profile data to ensure that each profile can be aligned with the geographic coordinate system. Finally, a series of vertical profile agricultural data is obtained. According to the characteristic distributions of each layer in the profile agricultural data, the mutual information analysis method is used to calculate the characteristic correlations between the surface, subsurface, and spatial layers. For example, the correlation between the surface vegetation density and the subsurface root distribution is calculated, and the influence of the canopy structure on soil transpiration is identified. At the same time, a long short-term memory network (LSTM) is used to establish a temporal feature analysis model to predict the coupling relationship of different-level features over time, and the principal component analysis (PCA) is combined to reduce the dimensionality and optimize the feature data, extracting the most representative cross-layer interaction features, including the influence coefficient of surface vegetation transpiration on soil moisture, the relationship model between subsurface root absorption efficiency and canopy photosynthetic rate, etc. Finally, a cross-layer coupling feature dataset is obtained. The adaptive segmentation algorithm based on graph segmentation is called to cluster the cross-layer coupling feature data, dividing the agricultural model into ground-air interaction regions, independent surface regions, and subsurface independent regions, and attribute assignment is performed on each region through the Bayesian classification method to ensure that the data at different levels are correctly classified. Subsequently, the voxel fusion method (VoxelFusion) is used to refine the boundary region between the surface and the subsurface to ensure the fineness of the ground-air segmentation model, and the multi-scale grid reconstruction algorithm is used to optimize the segmentation boundary so that it can accurately reflect the hierarchical relationship of the agricultural ecosystem. Finally, a complete ground-air segmentation agricultural model is obtained.

[0102] Preferably, the agricultural supply and demand derivation according to the segmented agricultural model includes:

[0103] Extract the surface action characteristics of the ground-air segmentation agricultural model to obtain the surface crop growth characteristics;

[0104] Conduct an underground soil supply analysis on the ground-air segmentation agricultural model to generate the underground supply capacity;

[0105] Conduct a spatial adaptability assessment on the ground-air segmentation agricultural model to obtain the spatial meteorological adaptability;

[0106] Conduct a crop growth supply-demand simulation based on the surface crop growth characteristics, underground supply capacity, and spatial meteorological adaptability to obtain the simulated supply-demand process;

[0107] Deduce the regional supply-demand relationship of the segmented agricultural model according to the simulated supply-demand process.

[0108] In this embodiment, based on the surface layer data in the ground-air segmentation agricultural model, the normalized difference vegetation index (NDVI) calculation method is used to quantitatively analyze the growth status of the crop coverage area. At the same time, spectral analysis technology is used to evaluate the pigment content of crop leaves to determine the crop growth health status. Combining multi-temporal remote sensing image data, the crop growth rate is calculated by the time series analysis method, and the support vector regression (SVR) model is used to predict the future growth trend of the crop. At the same time, the crop root-shoot ratio model is called to preliminarily estimate the crop root distribution to form a complete surface crop growth characteristic data set. Based on the underground layer data in the ground-air segmentation agricultural model, the time domain reflectometry (TDR) sensor data is called to accurately measure the soil water content. At the same time, the soil profile monitoring data is used to analyze the content of main nutrients such as nitrogen, phosphorus and potassium, and the effective soil water supply capacity is calculated by combining the field water holding capacity and the transpiration coefficient. The permeability coefficient analysis method is used to evaluate the soil's ability to supply different crop roots, and the root distribution model is combined to calculate the absorption potential of different crops for water and nutrients. At the same time, the underground microbial activity data is called, and the soil nutrient conversion efficiency is analyzed based on the microbial community metabolic network, and finally an underground supply capacity data set is formed. Based on the spatial layer data in the ground-air segmentation agricultural model, the weather station data and the unmanned aerial vehicle meteorological sensor data are called to calculate the air temperature, humidity, wind speed and light intensity, and the Kriging Interpolation method is used to spatially process the regional meteorological variables. At the same time, combined with the crop growth model, the adaptability of the current meteorological conditions to different crop varieties is analyzed. The analytic hierarchy process (AHP) is used to calculate the meteorological factor weights, and the Mahalanobis Distance method is used to classify and evaluate the meteorological adaptability, and finally a spatial meteorological fitness data set is generated. The crop growth process model (Crop Growth Model) is called. Based on the surface crop growth characteristic data, the growth rate of different growth stages of the crop is simulated. At the same time, combined with the underground supply capacity data, the water and nutrient supply amounts that the crop can obtain are calculated through the crop root absorption model, and the meteorological impact factors are used to dynamically adjust the crop growth rate. The discrete event simulation (DES) method is used to calculate the crop supply-demand relationship at different time steps, and finally a simulated supply-demand process data set including the crop supply-demand change trend is generated. The supply-demand flow analysis method based on graph theory is called, and based on the simulated supply-demand process data, a supply-demand network diagram of the agricultural area is constructed.Use the Max-Flow Min-Cut Algorithm to analyze the supply-demand equilibrium state of crops in different regions, and based on the local clustering algorithm, stratify the supply-demand data by region to form the supply-demand distribution pattern of different agricultural regions. At the same time, call the prediction method based on Bayesian inference to calculate the supply-demand gap that will occur in the future period and form a regional supply-demand relationship dataset.

[0109] Preferably, step S3 includes the following steps:

[0110] Step S31: Analyze the nutrient supply content of the regional supply-demand relationship; calculate the difference between the nutrient supply content and the preset growth demand parameters to obtain the nutrient demand difference;

[0111] Step S32: Perform a compensation demand mapping based on the nutrient demand difference to generate a compensated nutrient demand;

[0112] Step S33: Decompose each nutrient demand of the compensated nutrient demand to obtain a specific nutrient demand set; compare and superimpose the specific nutrient demand set according to the preset fertilizer components to obtain the fertilizer demand;

[0113] Step S34: Sort the nutrient demand urgency of the compensated nutrient demand to obtain the nutrient demand order; plan the application plan for the fertilizer demand based on the nutrient demand order to generate a fertilization prescription map.

[0114] In this embodiment, the agricultural soil database is called to detect the contents of nitrogen (N), phosphorus (P), potassium (K) and trace elements in the soil samples of the target area. The atomic absorption spectrometry (AAS) is used to measure the soil nutrient content. Combining with the hyperspectral remote sensing data, the principal component analysis (PCA) method is adopted to calculate the spatial nutrient distribution of the regional soil. Based on the growth demand models of different crop types, the nutrient demand parameters at each crop growth stage are extracted. The two-way difference calculation method is used to compare the current soil nutrient supply capacity with the crop nutrient demand, and a demand difference dataset of different nutrients is generated. The compensation demand mapping model based on geographically weighted regression (GWR) is called. Taking the nutrient demand difference data as the input, the nutrient compensation demand equations for different plots are constructed. The inverse distance weighting (IDW) interpolation method is used to smooth the spatially uneven nutrient demand data. Combining with the crop root distribution model, the absorption efficiency of different crops for specific nutrients is calculated, and a plot-level nutrient compensation demand dataset is generated. At the same time, the yield potential index of the plot is used to weight and adjust the compensation demand data to optimize the spatial distribution pattern of the compensated nutrient demand. The agronomy expert database is called to extract the nutrient absorption coefficients of different crops, and the actual demand for compensated nutrients is calculated by combining with the soil buffering capacity model. The Lagrangian optimization-based nutrient ratio algorithm is used to decompose the nitrogen, phosphorus, potassium and trace element demands, and a specific nutrient demand collection is generated. The fertilizer composition database is called to extract the composition information of available fertilizers in the region, and the linear regression analysis method is used to match and calculate the specific nutrient demand collection and the fertilizer composition data. The least squares optimization method is used to adjust the fertilizer ratio, and the optimal fertilizer demand is calculated to form a plot-level fertilizer application plan. The urgency evaluation model based on the entropy weight method is called. Taking the soil nutrient supply rate, crop growth cycle, soil moisture condition and current meteorological conditions as input variables, the urgency scores of different nutrients are calculated. The dynamic weighted sorting method is used to generate a nutrient demand sequence dataset. The fertilization optimization module based on the geographic information system (GIS) is called to perform plot-level spatial allocation of the fertilizer demand data, and combined with geographical factors such as terrain slope, drainage condition and soil permeability, the fertilization path planning is optimized, and finally a fertilization prescription map is generated.

[0115] Preferably, the simulation of the replenishment and development of the reconstructed agricultural model based on the fertilization prescription map includes:

[0116] Fuse the data of the reconstructed agricultural model and the fertilization prescription map to obtain the initial fertilization response data;

[0117] Dynamically simulate the nutrient release of the initial fertilization response data to obtain the dynamic nutrient release data;

[0118] Estimate the crop absorption efficiency of the dynamic nutrient release data to obtain the nutrient absorption efficiency data;

[0119] Predict the crop growth response based on the nutrient absorption efficiency data to obtain the growth response prediction data;

[0120] Conduct a replenishment modeling on the growth response prediction data to obtain the simulated replenishment response model.

[0121] In this embodiment, data fusion is performed on the reconstructed agricultural model and the fertilization prescription map. First, multi-source agricultural data is collected, including information such as soil moisture, soil type, climate data, crop growth stage, and fertilization history. These data are sourced from sensors, weather stations, and agricultural management systems. Through data cleaning and preprocessing, noise and missing values are removed to ensure the accuracy of the data. Next, the weighted fusion method is used to fuse data from different sources with different weights, and the weights are set according to the reliability and relevance of the data. Through the standardization process of different data sources, the unity and comparability after data fusion are ensured. Finally, the fused data is used as input to generate initial fertilization response data, which contains comprehensive information on factors such as fertilization patterns, soil fertility, climate change, and crop requirements. Dynamically simulate the nutrient release of the initial fertilization response data. Using information such as the fertilization amount, fertilization time, and soil type in the initial fertilization response data, construct a dynamic nutrient release model, combine with actual environmental variables such as soil fertility, temperature, and humidity, and use the finite difference method to perform time series simulation of nutrient release. During the simulation process, the physical and chemical properties of the soil, such as soil water holding capacity, soil pH, and fertilizer solubility, will be considered to dynamically calculate the nutrient release rate under different fertilization patterns, and finally obtain dynamic nutrient release data, which presents the fertilizer release process and release rate under different environmental conditions. Specific meteorological data and soil moisture data are used to dynamically adjust the model parameters during the simulation process, and estimate the crop absorption efficiency of the dynamic nutrient release data. Based on the dynamic nutrient release data, combined with the crop growth stage, root distribution, and crop growth requirements, estimate the crop nutrient absorption efficiency through a crop physiological model. First, determine the root absorption capacity of the crop, and establish a root absorption efficiency function using the relationship between soil moisture and nutrient concentration and the distance of the crop roots. Then, according to the availability of nutrients under different fertilization patterns, combined with the nutritional requirements of the crop, use the backpropagation algorithm to estimate the crop nutrient absorption efficiency, and finally obtain nutrient absorption efficiency data, which accurately represents the nutrient utilization rate of the crop under different fertilization conditions. Predict the crop growth response based on the nutrient absorption efficiency data. First, combine the nutrient absorption efficiency data with the crop growth model to simulate the growth process of the crop in a specific environment. The growth model includes growth parameters such as photosynthetic rate, evapotranspiration, and leaf area index. Simulate the growth process of the crop under different fertilization conditions through a dynamic systems model, combine meteorological data and soil data, and use numerical simulation methods to predict the growth of the crop. During the simulation process, continuously adjust parameters such as fertilizer application rate, fertilization timing, and soil moisture to obtain growth prediction data of the crop at different growth stages after fertilization. Perform replenishment modeling on the growth response prediction data.Based on the growth response prediction data, a crop replenishment response model is established using the Support Vector Regression (SVR) method in machine learning. First, key features in the growth response prediction data, such as the plant height, leaf area index, dry matter accumulation of the crop, etc., are selected and combined with the nutrient absorption efficiency data as inputs. The replenishment requirements of crop growth are modeled through the support vector regression model, and the model is trained using the training data so that the model can predict the replenishment requirements of the crop according to different fertilization patterns and environmental changes, and finally an accurate replenishment response model is output.

[0122] Preferably, the determination of the irrigation timing according to the simulated replenishment response model includes:

[0123] Calculate the crop transpiration of the simulated replenishment response model to obtain the transpiration water demand data;

[0124] Estimate the soil evaporation of the simulated replenishment response model to obtain the evaporation water consumption data;

[0125] Calculate the water demand based on the transpiration water demand data and the evaporation water consumption data to generate a water replenishment gap;

[0126] Determine the optimal irrigation timing based on the water replenishment gap to obtain the irrigation timing;

[0127] Carry out irrigation planning according to the irrigation timing to generate an irrigation prescription map.

[0128] In this embodiment, to calculate the crop transpiration of the simulated recharge response model, first, based on the crop growth data provided by the simulated recharge response model, information such as the leaf area index of the crop, crop type, ambient temperature, humidity, etc. is selected. Combining the relationship between photosynthesis and transpiration of the crop, a transpiration calculation method based on the Penman-Monteith formula is used to determine the transpiration of the crop at different growth stages. In the specific steps, environmental data such as the crop leaf area index, relative air humidity, wind speed, light intensity, etc. are input. Using the correlation between the stomatal conductance of the crop and transpiration, the transpiration water requirement of the crop is calculated, and the transpiration water requirement data is accumulated day by day. During the processing, if meteorological data is missing, interpolation methods are used to supplement the missing data. To estimate the soil evaporation of the simulated recharge response model, based on parameters such as soil type, soil humidity, and temperature, a spatio-temporal transformation model is used for the estimation of soil evaporation. First, the humidity data of the soil surface layer is obtained. Combining the water permeability and evaporation rate of the soil, the Hargreaves-Samani method is used to estimate the soil evaporation. The model considers the influence of the change in soil surface temperature on the evaporation rate under different climate conditions and dynamically adjusts the calculation of soil evaporation according to the change in air temperature. Combining the change trend of soil humidity, evaporation water consumption data is obtained. During the simulation, if the humidity data of the soil surface layer is missing, the data of nearby soil humidity sensors are used for calculation. To calculate the water requirement based on the transpiration water requirement data and evaporation water consumption data, first, the transpiration water requirement data and evaporation water consumption data are added together to obtain the total water requirement of the crop and soil. Further, through the water holding capacity of the soil and the current soil humidity, the water supply gap is calculated. If the total water requirement is greater than the existing soil moisture, the supply gap is the amount of water to be supplemented. If the current soil humidity is sufficient, the supply gap is zero. During the calculation, if the soil humidity data is missing, it can be estimated by the weighted average method of multi-point soil humidity data to obtain accurate water supply gap data. To determine the optimal irrigation timing based on the water supply gap, first, a water supply gap threshold is set. When the water supply gap exceeds the threshold, it can be determined as the irrigation timing. In the specific process, an adaptive threshold algorithm is used to dynamically adjust the water supply gap threshold according to the correlation between historical irrigation records and meteorological data to ensure the accuracy of the irrigation timing determination. During the judgment, the growth stage of the crop and environmental changes are combined, and the dynamic changes of transpiration, evaporation, and water gap are comprehensively considered to output the optimal irrigation timing data. To make an irrigation plan according to the irrigation timing, using the irrigation timing data, combined with the growth requirements of the crop and soil type, an irrigation plan is formulated. First, according to the determination of the irrigation timing, the start time of irrigation is determined. Then, according to the soil type and water holding capacity, the irrigation amount and irrigation method are set, and precise irrigation methods such as drip irrigation or sprinkler irrigation are used for water supply. The calculation of the irrigation amount is adjusted according to the soil water gap and the water requirement of the crop. Finally, the irrigation plan data is mapped into the irrigation prescription map to generate a precise irrigation plan.

[0129] Preferably, step S5 includes the following steps:

[0130] Step S51: Align the fertilization prescription map and the irrigation prescription map in time series to obtain a time series aligned prescription;

[0131] Step S52: Detect the conflict operation points of the time series aligned prescription; perform spatio-temporal scheduling on the time series aligned prescription based on the conflict operation points to obtain the agricultural activity schedule;

[0132] Step S53: Feedback the agricultural activity schedule in a visual form to generate agricultural integration feedback information to achieve agricultural multi-source information integration.

[0133] In this embodiment, the time series alignment of the fertilization prescription map and the irrigation prescription map is carried out. First, based on the data points in the fertilization prescription map and the irrigation prescription map, the time dimension and time interval in the two charts are determined, and the data interpolation algorithm is used to unify the time axes of the two. Suppose the fertilization operation is carried out monthly, while the irrigation operation is carried out weekly. For the inconsistent time intervals, linear interpolation or spline interpolation methods are adopted to synchronize the data of the two at the same time points for further analysis. Next, combined with the seasonal characteristics of crop growth, the time series data is refined and adjusted to ensure that the operation time nodes of the two can be accurately reflected within a unified time frame. After processing, the obtained time series aligned prescription will include the time arrangements of fertilization and irrigation activities. Detect the conflict operation points in the time series aligned prescription. Adopt a rule-based algorithm. First, detect whether there is an overlap in the time points of fertilization and irrigation operations. If the time points of fertilization and irrigation operations conflict, it is determined as a conflict operation point. In the specific process, according to the farmland management regulations and crop requirements, conflict detection rules are set. For example, fertilization and irrigation cannot be carried out at the same time point. If a conflict occurs, based on the detection results of the conflict points, the conflicting operation points are further marked, and a priority is set for each conflict point. If multiple conflict points occur simultaneously, the operation with a higher priority will be adjusted according to the scheduling rules. Use a scheduling algorithm to perform spatio-temporal scheduling on the conflict points, and re-arrange the conflict points according to factors such as crop growth requirements, climate conditions, and soil moisture to ensure that the activities of the two can be carried out without interfering with each other, and obtain a scheduled agricultural activity schedule. The agricultural activity schedule is fed back in a visual form. First, use GIS (Geographic Information System) technology and graphic visualization tools to spatially display the agricultural activity schedule information. In the specific process, first obtain the geographical location information of the farmland, associate the schedules of fertilization and irrigation with the specific field locations, and use map annotation to show the agricultural activities at each time node. At the same time, for the convenience of agricultural operators to view, color coding is added. The fertilization operation is set to green, the irrigation operation is set to blue, and the conflict operation points are marked as red. The system outputs the schedule of agricultural activities according to the schedule table, and displays this information in the form of charts and graphs in the interface to ensure that the agricultural activity schedule is clear at a glance. The user can view the specific time, location, and priority of each agricultural operation in the visual interface. After the visual feedback of the agricultural activity schedule is completed, agricultural integration feedback information is generated. The feedback information will include all the spatio-temporally scheduled agricultural operation arrangements, as well as potential conflict solutions, the schedule of agricultural activities, crop varieties, specific requirements for fertilization and irrigation, etc. These data are transmitted to the agricultural operators and are updated in real time through an intelligent management system to ensure that the agricultural activities are carried out on schedule.

[0134] The present invention also provides an agricultural multi-source information fusion system for implementing the above-mentioned agricultural multi-source information fusion method. The agricultural multi-source information fusion system includes:

[0135] A data acquisition module, configured to collect multi-source agricultural data; perform multi-source fusion on the multi-source agricultural data to obtain fused agricultural data; perform three-dimensional agricultural reconstruction on the fused agricultural data to generate a reconstructed agricultural model;

[0136] A supply and demand derivation module, configured to dissect the cross-layer coupling structure of the reconstructed agricultural model to obtain a ground-air segmented agricultural model; perform agricultural supply and demand derivation based on the segmented agricultural model to generate a regional supply and demand relationship;

[0137] A nutrient analysis module, configured to analyze the compensated nutrient requirements of the regional supply and demand relationship based on preset growth requirement parameters; perform fertilization prescription mapping on the compensated nutrient requirements to generate a fertilization prescription map;

[0138] A supply replenishment simulation module, configured to perform supply replenishment development simulation on the reconstructed agricultural model based on the fertilization prescription map to obtain a simulated supply replenishment response model; determine the irrigation timing based on the simulated supply replenishment response model to generate an irrigation prescription map;

[0139] An information integration module, configured to interactively integrate the fertilization prescription map and the irrigation prescription map to generate agricultural fusion feedback information, so as to achieve agricultural multi-source information fusion.

[0140] The present invention realizes the comprehensive acquisition of multi-source agricultural data through the introduction of a data acquisition module, covering various aspects of information such as meteorology, soil, and crop growth, providing a solid data foundation for subsequent analysis and decision-making. The process of data fusion integrates various types of information, eliminates information silos, enhances the relevance of data, and avoids decision-making errors caused by inconsistent data. The implementation of three-dimensional agricultural reconstruction provides an intuitive representation of the spatial characteristics of the farmland environment, enhancing the understanding of the internal relationships within the agricultural system. The supply-demand derivation module realizes a profound analysis of the interactions between different levels of the agricultural system by dissecting the cross-layer coupling structure of the reconstructed agricultural model. The formation of the ground-air segmentation agricultural model enables the detailed study of the relationship between the soil and the crops, providing a scientific basis for the generation of regional supply-demand relationships. The operation of the nutrient analysis module ensures the accurate analysis of the compensatory nutrient requirements. The analysis based on the growth demand parameters enables the crops to obtain the necessary nutrient support during the growth period. The implementation of the fertilization prescription mapping effectively connects the actual demand with the fertilization plan, improving the fertilization efficiency. The formation of the fertilization prescription map provides intuitive fertilization guidance for farmers. The replenishment simulation module based on the fertilization prescription map can reflect the changing nutrient requirements of the crops in real time. The establishment of the simulation replenishment response model provides a scientific basis for determining the irrigation timing. The generation of the irrigation prescription map ensures the scientific management of water. The information integration module generates agricultural integration feedback information by interactively integrating the fertilization prescription map and the irrigation prescription map, enhancing the transparency of farming activities, promoting the circulation and sharing of information among different agricultural links, generally improving the intelligent level and management efficiency of agricultural production, promoting the optimal allocation and sustainable utilization of resources, providing strong technical support and guarantee for the realization of high-efficiency and precision in modern agriculture, forming an agricultural management system centered on data, capable of adapting to the changes in different agricultural production environments and requirements, enhancing the decision-making ability and production efficiency of farmers, promoting the transformation and upgrading of agricultural production methods, and driving the development of sustainable agriculture and the protection of the ecological environment.

[0141] The present invention also provides a computer-readable storage medium storing a computer program, and when the computer program is executed, it realizes the method for fusing multi-source agricultural information as described in any one of the above.

[0142] The present invention provides a flexible program storage and management method through the computer-readable storage medium. The execution of the program realizes the automation and intelligence of agricultural data processing, enhances the efficiency and accuracy of data analysis and decision-making, supports the efficient fusion and application of multi-source data, promotes information sharing and collaboration in the agricultural production process, provides scientific agricultural model reconstruction and optimization, ensures the accurate implementation of fertilization and irrigation, optimizes the allocation and use of agricultural resources, enhances the visualization and operability of agricultural management, generally promotes the development of agricultural informatization and intelligence, and lays a foundation for the sustainable development of modern agriculture.

[0143] Therefore, from any perspective, the embodiments should be regarded as exemplary and non-limiting. The scope of the present invention is defined by the appended claims rather than the above description. Thus, all changes falling within the meaning and scope of the equivalent elements of the application documents are intended to be encompassed by the present invention.

[0144] The above are only specific embodiments of the present invention, enabling those skilled in the art to understand or implement the present invention. Various modifications to these embodiments will be obvious to those skilled in the art. The general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to these embodiments shown herein, but rather to the broadest scope consistent with the principles and novel features invented herein.

Claims

1. A method for agricultural multi-source information fusion, characterized in that: The following steps are involved: Step S1: Collect multi-source agricultural data; perform multi-source fusion on the multi-source agricultural data to obtain fused agricultural data; Carry out three-dimensional agricultural reconstruction on the fused agricultural data and generate a reconstructed agricultural model; Step S2: dissect and reconstruct the cross-layer coupling structure of the agricultural model to obtain a ground-space segmentation agricultural model; deduce agricultural supply and demand based on the segmentation agricultural model to generate a regional supply and demand relationship; Step S3: analyzing the compensatory nutrient requirements of the regional supply and demand relationship based on the preset growth requirement parameters; Map the fertilizer prescription to compensate for the nutrient demand and generate a fertilizer prescription map; Step S4: Based on the fertilization prescription map, the reconstructed agricultural model is simulated for replenishment development to obtain a simulated replenishment response model; the irrigation timing is determined according to the simulated replenishment response model to generate an irrigation prescription map; Step S5: interactively integrate the fertilization prescription map and the irrigation prescription map to generate agricultural fusion feedback information to achieve agricultural multi-source information fusion.

2. The agricultural multi-source information fusion method according to claim 1 is characterized in that: Step S1 includes the following steps: Step S11: deploying sensor networks and drones in agricultural areas, and monitoring crop growth through the sensor networks and drones to obtain multi-source agricultural data; Step S12: performing cross-modal alignment on multi-source agricultural data to generate leveled agricultural data; and associating heterogeneous agricultural information based on the leveled agricultural data; Step S13: fusing the multi-source agricultural data features into fused agricultural data based on the associated heterogeneous agricultural information; Step S14: mapping the fused agricultural data into agricultural point cloud data; performing voxel processing on the agricultural point cloud data to obtain an agricultural voxel model; Step S15: Perform reverse 3D reconstruction based on the agricultural voxel model to generate a reconstructed agricultural model.

3. The agricultural multi-source information fusion method according to claim 1 is characterized in that: The cross-layer coupling structure of the dissected and reconstructed agricultural model includes: The reconstructed agricultural model is horizontally layered to distinguish the surface, underground and spatial layers to obtain layered agricultural data; Extract vertical sections from stratified agricultural data to generate profile agricultural data; The cross-layer interaction features of the profile agricultural data are identified to obtain the cross-layer coupling features; The reconstructed agricultural model is layered and dissected based on the cross-layer coupling characteristics to obtain a ground-to-space segmentation agricultural model.

4. The agricultural multi-source information fusion method according to claim 1 is characterized in that: The derivation of agricultural supply and demand based on the segmented agricultural model includes: Extract the surface action characteristics of the ground-space segmentation agricultural model to obtain the surface crop growth characteristics; Conduct underground soil supply analysis on the ground-air segmentation agricultural model to generate underground supply capacity; The spatial adaptability of the ground-space segmentation agricultural model was evaluated to obtain the spatial meteorological adaptability; The crop growth supply and demand are simulated based on the surface crop growth characteristics, underground supply capacity and space meteorological adaptability to obtain the simulated supply and demand process; The regional supply and demand relationship of the segmented agricultural model is derived based on the simulated supply and demand process.

5. The agricultural multi-source information fusion method according to claim 1 is characterized in that: Step S3 includes the following steps: Step S31: analyzing the nutrient supply content of the regional supply and demand relationship; calculating the difference between the nutrient supply content and the preset growth requirement parameter to obtain the nutrient requirement difference; Step S32: performing compensation requirement mapping according to the nutrient requirement difference to generate compensation nutrient requirement; Step S33: decomposing the compensation nutrient demand into various nutrient demands to obtain a collection of specific nutrient demands; comparing and superimposing the collection of specific nutrient demands according to preset fertilizer components to obtain fertilizer demand; Step S34: sorting the compensatory nutrient needs according to the urgency of the nutrient needs to obtain a nutrient demand sequence; planning the fertilizer application plan based on the nutrient demand sequence to generate a fertilization prescription map.

6. The agricultural multi-source information fusion method according to claim 1 is characterized in that: The replenishment development simulation of the reconstructed agricultural model based on the fertilization prescription map includes: The reconstructed agricultural model and the fertilization prescription map are fused to obtain the initial data of fertilization response; Dynamically simulate the nutrient release of the initial data of fertilization response to obtain dynamic data of nutrient release; Estimate the crop absorption efficiency of nutrient release dynamic data to obtain nutrient absorption efficiency data; Predicting crop growth response based on nutrient absorption efficiency data to obtain growth response prediction data; Recharge modeling was performed on the growth response prediction data to obtain a simulated recharge response model.

7. The agricultural multi-source information fusion method according to claim 1 is characterized in that: The irrigation timing determined according to the simulated recharge response model includes: Calculate the crop transpiration of the simulated recharge response model and obtain the transpiration water demand data; Estimate soil evaporation from the simulated recharge response model and obtain evaporative water consumption data; Calculate water demand based on transpiration water demand data and evaporation water consumption data to generate water supply gap; Determine the best irrigation time based on the water supply gap and obtain the irrigation time; Irrigation planning is carried out according to the irrigation timing to generate an irrigation prescription map.

8. The agricultural multi-source information fusion method according to claim 1 is characterized in that: Step S5 includes the following steps: Step S51: performing time series alignment on the fertilization prescription map and the irrigation prescription map to obtain a time series alignment prescription; Step S52: Detect conflicting operation points of the timing alignment prescription; perform spatial and temporal scheduling on the timing alignment prescription based on the conflicting operation points to obtain a farming activity schedule; Step S53: Feedback the farming activity schedule in a visual form to generate agricultural integration feedback information to achieve agricultural multi-source information fusion.

9. An agricultural multi-source information fusion system, characterized in that: Used to execute the agricultural multi-source information fusion method as claimed in claim 1, the agricultural multi-source information fusion system comprises: The data acquisition module is used to collect multi-source agricultural data; to perform multi-source fusion on the multi-source agricultural data to obtain fused agricultural data; to perform three-dimensional agricultural reconstruction on the fused agricultural data to generate a reconstructed agricultural model; The supply and demand derivation module is used to dissect and reconstruct the cross-layer coupling structure of the agricultural model to obtain the ground-space segmentation agricultural model; the agricultural supply and demand are deduced based on the segmentation agricultural model to generate the regional supply and demand relationship; The nutrient analysis module is used to analyze the compensatory nutrient demand of the regional supply and demand relationship based on the preset growth demand parameters; to map the compensatory nutrient demand with the fertilization prescription and generate a fertilization prescription map; The replenishment simulation module is used to simulate the replenishment development of the reconstructed agricultural model based on the fertilization prescription map to obtain a simulated replenishment response model; the irrigation timing is determined according to the simulated replenishment response model to generate an irrigation prescription map; The information integration module is used to interactively integrate fertilization prescription maps and irrigation prescription maps to generate agricultural integration feedback information, so as to realize multi-source information fusion for agriculture.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed, the agricultural multi-source information fusion method as described in any one of claims 1 to 8 is implemented.

Images