Farmland management digital twinborn decision support system and control method thereof

Abstract

The invention relates to the technical field of smart agriculture, and provides a digital twinborn decision support system for farmland management and a control method thereof, and the system comprises a digital farmland module which is used for collecting a plurality of farmland parameters; the crop growth phenotype monitoring module is used for collecting and analyzing growth state phenotype data of crops to obtain simulation values of crop growth parameters; the crop production management model decision module comprises a crop knowledge model which is used for fitting a growth change curve of a compound according to a plurality of farmland parameters; the crop three-dimensional structure function model is used for performing three-dimensional reconstruction on a crop population and calculating true values of crop growth parameters; the production management decision calculation model is used for deciding a water and fertilizer regulation prescription according to the simulation value and the true value, determining disease information of crops according to the growth state phenotypic data and determining a pesticide application decision prescription in combination with pesticide application requirements, so that an agricultural implement performs water, fertilizer and pesticide comprehensive operation; according to the system, the crop planting management efficiency and the intelligent level are improved.

Description

Technical Field

[0001] The present invention relates to the field of smart agricultural technology, and in particular to a farmland management digital twin decision support system and a control method thereof. Background Art

[0002] With the development of agricultural Internet of Things technology, traditional agricultural planting management methods have been highly informationized. Planting managers usually use mobile phones to check field weather, soil moisture conditions, crop growth and other information, and perform production operations through water-fertilizer integration and agricultural machinery to achieve management of planting, fertilization, irrigation and spraying.

[0003] Among the related technologies, traditional crop production management technologies rely on experts or experienced technicians to make decisions on various aspects of planting management, which is prone to information missing or information islands. At the same time, there is a lack of real-time monitoring of crop growth data and environmental data, resulting in a large deviation between the generated planting management decision plan and the actual growth conditions of the crops, which in turn leads to low water and fertilizer utilization efficiency in crop planting and low farmland management efficiency. Summary of the invention

[0004] The present invention provides a farmland management digital twin decision support system and a control method thereof, which are used to solve the problem that when the prior art relies on experts or experienced technicians to make decisions on various links of planting management, information missing or information islands are prone to occur, and there is a lack of real-time monitoring of crop growth data and environmental data, resulting in large deviations in planting management decision plans, thereby improving the water and fertilizer utilization efficiency of crop planting and the efficiency of farmland management.

[0005] The present invention provides a farmland management digital twin decision support system, comprising: Digital farmland module, which is used to collect multiple farmland parameters, including farmland GIS map data, soil fertility data, soil temperature and humidity environment data, meteorological data and crop variety data; A crop growth phenotype monitoring module, which is used to collect crop growth status phenotype data and perform pipeline phenotype analysis to generate a phenotypic indicator GIS layer; wherein the growth status phenotype data includes RGB images, spectral images and point cloud data; The crop production management model decision module includes a crop knowledge model, a crop three-dimensional structure and function model, and a production management decision calculation model; wherein the crop knowledge model is used to fit the growth change curve of the crop according to the multiple farmland parameters to obtain the simulated value of the crop growth parameter; the crop three-dimensional structure and function model is used to perform three-dimensional reconstruction of the crop group according to the phenotypic index GIS layer to obtain a crop three-dimensional model, and calculate the real value of the crop growth parameter according to the crop three-dimensional model; the production management decision calculation model is used to make decisions on water and fertilizer regulation according to the simulated value and the real value, obtain a water and fertilizer decision prescription, determine the disease information of the crop according to the growth state phenotypic data, and determine the drug application decision prescription according to the pre-drug application requirements and the disease information, so that agricultural machinery can perform water, fertilizer and drug comprehensive operations; the production management decision calculation model is determined based on the coupling of the crop knowledge model and the crop three-dimensional structure and function model; the crop growth parameter includes at least one of the values ​​of the daily dynamic changes of leaf age, leaf area index and NDVI.

[0006] According to a farmland management digital twin decision support system provided by the present invention, the system also includes: A crop growth control digital twin module, which is connected to the digital farmland module, the crop growth phenotype monitoring module and the crop production management model decision module through an agent mechanism; the crop growth control digital twin module is used to construct crop twins, environment twins, agricultural machinery twins and interactive calculations between twins, and based on the interactive calculation results between the twins, comprehensive management of crop water, fertilizer and pesticides is performed in a virtual-real way.

[0007] According to a farmland management digital twin decision support system provided by the present invention, the crop growth control digital twin module includes: A five-dimensional digital twin model of crop growth, the five-dimensional digital twin model of crop growth includes physical entities, virtual entities, services, twin data and connections, the physical entities include planting managers, agricultural machinery, crops and environment; the virtual entities include geometric models, physical models, behavioral models and rule models corresponding to the physical entities; the services are used to encapsulate various types of data information and algorithm models; the twin data includes physical entity data, virtual entity data, service data, knowledge data and fusion derived data; the connection is used to provide data interaction for any two of the physical entity, the virtual entity, the service and the twin data; A plurality of Agent units, the plurality of Agent units include a sensor Agent, a phenotypic parameter parsing Agent, a prescription decision calculation Agent, a prescription push Agent and a human-computer interaction Agent; the digital farmland module and the crop growth phenotype monitoring module are respectively connected to the sensor Agent, the phenotypic parameter parsing Agent is connected to the crop growth phenotype monitoring module; the basic database, the sensor Agent, and the crop three-dimensional structure function model are respectively connected to the prescription decision calculation Agent; the physical entity, the virtual entity and the prescription decision calculation Agent are respectively connected to the prescription push Agent; the human-computer interaction Agent is connected to the prescription decision calculation Agent; A three-dimensional visualization unit is connected to the human-computer interaction Agent.

[0008] According to a farmland management digital twin decision support system provided by the present invention, the three-dimensional visualization unit includes: a farmland plot digital display panel, a soil moisture data display panel, a meteorological data display panel, a crop variety data display panel, a soil basic data display panel, a growth status phenotype data display panel, a crop virtual growth data display panel, a prescription GIS map data display panel, an intelligent agricultural machinery data display panel and a water-fertilizer integrated agricultural machinery data panel.

[0009] According to a farmland management digital twin decision support system provided by the present invention, the production management decision calculation model is also used to generate a plurality of prescription node data according to the farmland GIS map data, the water and fertilizer decision prescription and the pesticide decision prescription, and generate a prescription GIS map according to the prescription node data; and send the prescription GIS map to the agricultural machinery; The prescription node data include longitude and latitude, fertilizer application amount, water application amount, pesticide application amount and pesticide application type.

[0010] According to a farmland management digital twin decision support system provided by the present invention, the crop growth phenotype monitoring module includes: A drone airport device, the drone airport device is installed in a farmland area, and the drone airport device includes a variety of sensors; A flight path generation unit, the flight path generation unit is used to generate different flight paths according to different growth stages of the crops; A phenotypic data collection unit, the phenotypic data collection unit is used to control the drone airport equipment to collect the growth status phenotypic data from the farmland area according to the target flight route and the target sensor sampling frequency; A parsing unit, the parsing unit is used to extract a first phenotypic indicator from the RGB image, a second phenotypic indicator from the spectral image, and a third phenotypic indicator from the point cloud data, and generate the phenotypic indicator GIS layer based on the first phenotypic indicator, the second phenotypic indicator, the third phenotypic indicator and the coordinates of the target flight route; Among them, the first phenotypic indicator includes at least one of the crop leaf color, texture, spot distribution and crop canopy coverage; the second phenotypic indicator includes at least one of the crop canopy vegetation index NDVI, nitrogen content and water content; the third phenotypic indicator includes at least one of the crop plant height, leaf area, biomass and plant structure.

[0011] The present invention also provides a control method for a farmland management digital twin decision support system, comprising: Based on the digital farmland module, multiple farmland parameters are collected, and the multiple farmland parameters include farmland GIS map data, soil fertility data, soil temperature and humidity environment data, meteorological data and crop variety data; based on the crop growth phenotype monitoring module, the crop growth status phenotype data is collected and pipelined phenotype analysis is performed to generate a phenotypic indicator GIS layer; wherein the growth status phenotype data includes RGB images, spectral images and point cloud data; The crop knowledge model based on the crop production management model decision module fits the crop growth change curve according to the multiple farmland parameters to obtain the simulated values ​​of the crop growth parameters; the crop three-dimensional structure function model based on the crop production management model decision module reconstructs the crop group in three dimensions according to the phenotypic index GIS layer to obtain a crop three-dimensional model, and calculates the real values ​​of the crop growth parameters according to the crop three-dimensional model; the production management decision calculation model based on the crop production management model decision module makes decisions on water and fertilizer regulation according to the simulated values ​​and the real values ​​to obtain a water and fertilizer decision prescription, determines the disease information of the crops according to the growth status phenotypic data, and determines the pesticide decision prescription according to the pre-drug application requirements and the disease information, so that agricultural machinery can perform comprehensive water, fertilizer and pesticide operations; the production management decision calculation model is determined based on the coupling of the crop knowledge model and the crop three-dimensional structure function model; the crop growth parameters include at least one of the daily dynamic changing values ​​of leaf age, leaf area index and NDVI.

[0012] According to a control method of a farmland management digital twin decision support system provided by the present invention, the method further includes: Based on the crop growth management digital twin module, crop twins, environment twins, agricultural machinery twins and interactive calculations between the twins are constructed, and according to the interactive calculation results between the twins, comprehensive management of crop water, fertilizer and pesticides is carried out in a virtual-to-real manner; the crop growth management digital twin module is communicated and connected with the digital farmland module, the crop growth phenotype monitoring module and the crop production management model decision module through an Agent mechanism.

[0013] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, a control method for a digital twin decision support system for farmland management as described above is implemented.

[0014] The present invention also provides a non-transitory computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, it implements a control method for a digital twin decision support system for farmland management as described in any of the above-mentioned methods.

[0015] The farmland management digital twin decision support system and control method provided by the present invention collect multiple farmland parameters through the digital farmland module, collect growth status phenotypic data and perform pipeline phenotypic analysis through the crop growth phenotype monitoring module to generate phenotypic indicators, fit the growth change curve of the crop according to the multiple farmland parameters through the crop knowledge model to obtain the crop growth parameters, reconstruct the crop population in three dimensions according to the phenotypic indicators through the crop three-dimensional structure and function model, and then calculate the crop growth indicators, and regulate water, fertilizer and medicine according to the simulated values ​​and real values ​​of the crop growth parameters through the production management decision calculation model, so that agricultural machinery can perform comprehensive water, fertilizer and medicine operations, thereby improving the efficiency and intelligence level of crop planting management. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0017] Figure 1 It is one of the structural schematic diagrams of the farmland management digital twin decision support system provided by the present invention.

[0018] Figure 2 This is the second structural schematic diagram of the farmland management digital twin decision support system provided by the present invention.

[0019] Figure 3It is one of the flow charts of the control method of the farmland management digital twin decision support system provided by the present invention.

[0020] Figure 4 This is the second flow chart of the control method of the farmland management digital twin decision support system provided by the present invention.

[0021] Figure 5 It is a structural schematic diagram of the electronic device provided by the present invention.

[0022] Reference numerals: 110: Digital farmland module; 120: Crop growth phenotype monitoring module; 121: UAV airport equipment; 122: route generation unit; 123: phenotypic data collection unit; 124: analysis unit; 130: Crop production management model decision module; 131: Crop knowledge model; 132: Crop three-dimensional structure and function model; 133: Production management decision-making computational model; 140: Digital twin module for crop growth control; 141: Five-dimensional digital twin model for crop growth; 142: Agent unit; 143: 3D visualization unit. DETAILED DESCRIPTION

[0023] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0024] Combine the following Figure 1-Figure 4 The digital twin decision support system for farmland management and the control method thereof are described in the present invention.

[0025] Figure 1 This is one of the structural diagrams of the farmland management digital twin decision support system provided by the present invention, such as Figure 1 As shown, the farmland management digital twin decision support system includes: a digital farmland module 110, a crop growth phenotype monitoring module 120, and a crop production management model decision module 130.

[0026] The digital farmland module 110 is used to collect multiple farmland parameters, which include farmland GIS map data, soil fertility data, soil temperature and humidity environment data, meteorological data and crop variety data.

[0027] In this embodiment, in the early stage of crop planting and growth, plot image data is obtained through remote sensing maps and image data collected by drones, and the plot image data is segmented using an image deep learning segmentation model to obtain farmland plot boundary information; then, a farmland plot grid map composed of several plots is obtained based on the farmland plot boundary information; finally, the farmland plot grid map is processed based on GIS technology to obtain map layers of each farmland plot, i.e., digital farmland GIS map data, which includes attribute information such as plot number, longitude and latitude of the center point of the plot, elevation, and boundary of the plot.

[0028] In this embodiment, based on the elevation data of each plot of farmland, soil fertility sampling points are generated according to the method of not sampling neighboring plots with elevation differences less than a threshold value Tf, and soil fertility data of the sampling points are collected; then, interpolation technology is used to construct a digital farmland plot grid fertility map, i.e., soil fertility data. The soil fertility data establishes a plot digital farmland fertility database according to field information such as the plot number, longitude and latitude of the center point of the plot, and fertility (nitrogen, phosphorus, and potassium content).

[0029] In this embodiment, the installation position of the soil temperature and humidity environmental sensor is determined according to the method of not sampling the soil environment of the neighboring fields whose elevation data difference of each field is less than the threshold value Tm; the soil environment sensor is installed at the selected installation position of the soil temperature and humidity environmental sensor, and the soil temperature, humidity and other environmental data are collected regularly through the wireless module; the interpolation technology is used to construct a digital farmland field grid soil temperature and humidity environmental map, and then according to the field information such as the field number, the longitude and latitude of the field center point, the soil temperature, and the soil humidity, a field digital farmland soil temperature and humidity environmental database is established.

[0030] In this embodiment, meteorological sensors are installed on the fields, and meteorological environment data of the fields (for example, weather temperature, humidity, solar radiation, rainfall, wind speed and other information) are collected regularly through wireless modules; a digital farmland field meteorological database is established according to field information such as field number, longitude and latitude of the field center, weather temperature, humidity, solar radiation, rainfall, wind speed and so on.

[0031] In this embodiment, the main crop varieties in the planting area where the farmland is located are collected and selected, and a crop variety database is constructed according to information such as the district or county where the crop is located, disease resistance, and accumulated growth temperature.

[0032] The crop growth phenotype monitoring module 120 is used to collect crop growth status phenotype data and perform pipeline phenotype analysis to generate a phenotypic indicator GIS layer; wherein the growth status phenotype data includes RGB images, spectral images and point cloud data.

[0033] In this embodiment, the crop growth phenotype monitoring module 120 includes the following modules: drone airport equipment 121 , a route generation unit 122 , a phenotype data collection unit 123 and a parsing unit 124 .

[0034] The drone airport equipment 121 is installed in the farmland area, and the drone airport equipment 121 includes a variety of sensors.

[0035] In this embodiment, the multiple sensors include RGB image sensors, spectral image sensors and lidar sensors; wherein the RGB image sensor is used to obtain crop RGB image data, and further analyze to obtain phenotypic indicators such as crop leaf color, texture, lesion distribution, crop canopy coverage, etc.; the spectral image sensor is used to obtain crop multispectral image data, and further analyze to obtain indicators such as NDVI vegetation index, nitrogen content, moisture content, etc.; the lidar sensor is used to obtain crop three-dimensional point cloud data, and further analyze to obtain phenotypic indicators related to plant height, leaf area, biomass and plant structure.

[0036] The flight path generating unit 122 is used to generate different flight paths according to different growth stages of the crops.

[0037] In this embodiment, the flight route is set according to the requirements for plant height and phenotype monitoring during the crop growth period and in combination with the geographical location of the field, and different flight routes corresponding to data collection are set for different growth periods.

[0038] The phenotypic data collection unit 123 is used to control the drone airport equipment 121 to collect growth status phenotypic data from the farmland area according to the target flight route and the target sensor sampling frequency.

[0039] In this embodiment, during the specified crop growth period, the drone airport is controlled to automatically fly over the farmland area according to the preset sensor collection frequency and the set route to collect crop growth status phenotypic data (including crop RGB images, spectral images and point cloud data), and transmit these data back to the server for subsequent extraction of phenotypic indicators.

[0040] The parsing unit 124 is used to extract a first phenotypic indicator from the RGB image, a second phenotypic indicator from the spectral image, and a third phenotypic indicator from the point cloud data, and generate a phenotypic indicator GIS layer based on the first phenotypic indicator, the second phenotypic indicator, the third phenotypic indicator and the coordinates of the target flight route.

[0041] In this embodiment, the site data of the corresponding field is first sampled from the collected crop RGB images, spectral images and point cloud data (the point cloud data can be obtained by laser radar) according to the route node position of the drone airport, and a pipelined phenotyping method is constructed through the following steps: (1) RGB pixels are extracted from RGB images based on the image deep learning model, and then phenotypic indicators such as crop leaf color, texture, lesion distribution and crop canopy coverage are calculated based on the RGB pixels based on the image algorithm, i.e., the first phenotypic indicator.

[0042] (2) Crop spectral pixels are extracted from spectral image data based on the image deep learning model. Then, the spectral inversion model is used to calculate the crop canopy vegetation index NDVI, nitrogen content, water content and other indicators based on the crop spectral pixels, i.e., the second phenotypic indicator.

[0043] (3) Based on the point cloud deep learning semantic model, the point cloud data is segmented and extracted for crop point clouds. Further, based on the point cloud deep learning instance model, the point cloud of crop leaves is segmented and extracted. The crop population canopy point cloud model is reconstructed based on the segmented and extracted point clouds, and the phenotypic indicators related to plant height, leaf area, biomass and plant structure, i.e., the third phenotypic indicator, are calculated.

[0044] (4) Based on the first phenotypic index, the second phenotypic index, and the third phenotypic index, combined with the route coordinate information, a phenotypic index GIS layer is generated (including: field number, longitude and latitude of the field center point, RGB image phenotypic index, spectral phenotypic index, point cloud phenotypic index, and crop canopy point cloud model, etc.), and the phenotypic index GIS layer is stored in the field crop phenotypic monitoring database.

[0045] The crop production management model decision module 130 includes a crop knowledge model 131, a crop three-dimensional structure function model 132, and a production management decision calculation model 133; wherein the crop knowledge model 131 is used to fit the growth change curve of the crop according to multiple farmland parameters to obtain the simulated value of the crop growth parameter.

[0046] In this embodiment, the crop production management model decision module 130 is used to calculate and generate the optimal prescription of seeds, fertilizers, water, pesticides, etc. in the crop planting management process.

[0047] In this embodiment, the crop knowledge model 131 is used to simulate planting with preset planting varieties, planting periods of varieties, planting spacing, seeding amounts and other planting plans. It can simulate and calculate the suitable growth curve of the crop throughout the growing season. The growth curve is a curve formed by the daily dynamic changes in leaf age, leaf area index and NDVI calculated based on the yield target and combined with basic data such as meteorology and soil.

[0048] Specifically, the appropriate dynamic curve of the leaf area index is simulated by the following steps: (1) Calculate the seedling emergence date based on the sowing date, that is, the accumulated temperature from the beginning of sowing to the date when the accumulated temperature is greater than T1; (2) Calculate the date when the leaf area index reaches its maximum from sowing to the date when the leaf area index reaches its maximum, that is, the date when the accumulated temperature is greater than T2 from sowing; (3) The DS value (a dimensionless index used to reflect the degree of accumulated temperature of the crop at the current growth stage relative to the time when its maximum leaf area index appears) at the growth stage is calculated by the following formula: DS=SGDD3 / SGDD1; Among them, SGDD1 is the accumulated temperature from seedling emergence to the time when the leaf area index reaches the maximum, and SGDD3 is the accumulated temperature from the seedling emergence day to the current day; (4) Calculate the relative leaf area index RLAI according to the fitting model function of DS; Calculate the daily maximum suitable leaf area index LAIomax according to the atmospheric temperature, radiation, and variety parameters; Calculate the daily maximum suitable leaf area index LAIopt by the following formula: LAIopt =RLAI LAIomax; Finally, the appropriate dynamic curve of leaf area index of crop growth was calculated based on LAIopt.

[0049] In this embodiment, the appropriate dynamic curve of NDVI is simulated by the following steps: (1) Calculate the cumulative growing degree days (RAGDD) of crop growth, that is, the accumulation of effective accumulated temperature (GDD) every day; (2) Calculate the relative NDVI value RNDVI, which is a fitting function of RAGDD and variety parameters; (3) Calculate the daily suitable value of NDVI by the following formula: NDVI=RNDVI NDVImax; Among them, NDVImax is a variety parameter, which is the maximum NDVI value during the entire growth period of the crop.

[0050] In this embodiment, the leaf age suitable dynamic curve is simulated by the following steps: The calculation function of leaf age is the cumulative accumulated temperature from the seedling emergence date / leaf-to-leaf thermal distance, where the leaf-to-leaf thermal distance is the variety parameter of the crop.

[0051] The crop three-dimensional structural function model 132 is used to perform three-dimensional reconstruction of the crop population according to the phenotypic indicator GIS layer, obtain a crop three-dimensional model, and calculate the true value of the crop growth parameter according to the crop three-dimensional model.

[0052] In this embodiment, the crop three-dimensional structural function model 132 obtains phenotypic data (including image data, point cloud data and other phenotypic data) of the crop during the key growth period through a phenotypic sensor, reconstructs the crop three-dimensional model by combining it with three-dimensional modeling technology, and then uses the crop three-dimensional model to calculate the crop growth function parameters (i.e., the above-mentioned true values) such as leaf age, leaf area index, and NDVI of the crop, thereby forming a crop growth monitoring method driven by monitoring data.

[0053] Specifically, the method for constructing a three-dimensional crop model includes: using crop canopy point cloud data and multispectral data collected by drones, and fusing the point cloud and spectral data to generate a spectral point cloud, then using ground elevation data to remove the ground point cloud of the plot, and using the centroid shrinkage method to obtain the center point of each crop plant, and then determining the root base point of the crop plant, and finally using the point cloud erosion search clustering algorithm to segment and mark to obtain the single canopy point cloud of each crop, thereby realizing three-dimensional reconstruction of the crop group.

[0054] In this embodiment, the method for calculating crop functional parameters based on a three-dimensional model includes: calculating plant height through crop plant point cloud data; obtaining leaf point cloud through point cloud data combined with a stem-leaf segmentation model, obtaining the number of plant leaves using a clustering algorithm, and calculating the crop leaf age; obtaining plant leaf point cloud through point cloud data combined with a stem-leaf segmentation model, gridding the leaf point cloud, calculating the plant leaf area, and further calculating the leaf area index; calculating the NDVI value of the crop population based on the spectral point cloud data.

[0055] In this embodiment, for the single-point location data calculated above, a GIS difference method is used to calculate and generate a model parameter map, which is stored in a model parameter database.

[0056] The production management decision-making calculation model 133 is used to make decisions on water and fertilizer regulation based on crop growth parameters and crop growth indicators, obtain water and fertilizer decision prescriptions, determine crop disease information based on growth status phenotypic data, and determine pesticide application decision prescriptions according to pre-pesticide requirements and disease information, so that agricultural machinery can perform integrated water, fertilizer and pesticide operations; the production management decision-making calculation model 133 is determined based on the coupling of the crop knowledge model 131 and the crop three-dimensional structure and function model 132; the crop growth parameters include at least one of the daily dynamic changes in the values ​​of leaf age, leaf area index and NDVI.

[0057] In this embodiment, the crop growth parameters, i.e., the change curve of the key indicators of crop growth (including the leaf area index LAI and the normalized vegetation index NDVI, which are the daily dynamic changes of the values) are obtained by simulating and calculating the crop growth parameters through the above-mentioned crop knowledge model 131 according to the basic geographical data, the real-time meteorological soil moisture data and the variety characteristic data; the obtained crop population is three-dimensionally modeled through the crop three-dimensional structure function model 132, and then the crop growth indicators (including the leaf area index LAI and the normalized vegetation index NDVI) of the current period are calculated through the crop population three-dimensional model; finally, the leaf area index LAI1 and the normalized vegetation index NDVI1 (crop growth indicators) measured by the production management decision calculation model 133 are compared with the leaf area index LAI2 and the normalized vegetation index NDVI2 (crop growth parameters) of the period calculated by the crop knowledge model 131, and the current fertilization control method is calculated by comprehensive decision; specifically, the difference TLAI of the leaf area index LAI is calculated by the following formula: TLAI = LAI1 - LAI2.

[0058] The difference TNDVI of the normalized difference vegetation index NDVI1 is calculated by the following formula: TNDVI = NDVI1 - NDVI2; Among them, in the current time, the amount of fertilizer per unit area N is a linear function of TLAI and TNDVI; the amount of nitrogen applied per unit area of ​​crops N is calculated according to the above method. If N is less than 0, no fertilization operation is performed.

[0059] In this embodiment, in order to regulate crop water demand, the soil moisture content data of the field moisture monitoring station is obtained in real time, and the variety parameters of the crop variety's water requirements are obtained to calculate the irrigation amount; based on the crop growth data monitored by the phenotypic drone, the current disease type and disease severity are identified, and the type and amount of pesticide application are calculated according to the pesticide application requirements of the disease database to form a pesticide application plan.

[0060] The digital twin decision support system for farmland management provided by the embodiment of the present invention collects multiple farmland parameters through the digital farmland module, collects growth status phenotypic data and performs pipeline phenotypic analysis through the crop growth phenotype monitoring module to generate phenotypic indicators, fits the growth change curve of crops according to multiple farmland parameters through the crop knowledge model to obtain crop growth parameters, reconstructs the crop population in three dimensions according to the phenotypic indicators through the crop three-dimensional structure and function model, and then calculates the crop growth indicators, and regulates water, fertilizer and medicine according to the simulated values ​​and real values ​​of the crop growth parameters through the production management decision calculation model, so that agricultural machinery can perform comprehensive water, fertilizer and medicine operations, thereby improving the efficiency and intelligence level of crop planting management.

[0061] In some embodiments, the farmland management digital twin decision support system also includes: a crop growth control digital twin module 140.

[0062] The crop growth control digital twin module 140 communicates with the models in the digital farmland module, the crop growth phenotype monitoring module and the crop production management model decision module 130 through the Agent mechanism; the crop growth control digital twin module is used to construct crop twins, environment twins, agricultural machinery twins and interactive calculations between each twin, and based on the interactive calculation results between each twin, comprehensive management of crop water, fertilizer and pesticide is performed in a virtual-real way.

[0063] In this embodiment, the crop growth control digital twin module 140 is composed of a five-dimensional digital twin model 141 of crop growth, a crop twin information communication mechanism and prescription decision calculation, and a three-dimensional visualization unit 143. The twin module provides water and fertilizer management for crop planting by controlling the real in a virtual way.

[0064] Specifically, the crop growth management and control digital twin module 140 includes: a crop growth digital twin five-dimensional model 141, multiple Agent units 142 and a three-dimensional visualization unit 143.

[0065] Among them, the five-dimensional digital twin model of crop growth 141 includes physical entities, virtual entities, services, twin data and connections. Physical entities include planting managers, agricultural machinery, crops and environment; virtual entities include geometric models, physical models, behavioral models and rule models corresponding to physical entities; services are used to encapsulate various types of data information and algorithm models; twin data includes physical entity data, virtual entity data, service data, knowledge data and fused derivative data; connections are used to provide data interaction for any two of physical entities, virtual entities, services and twin data.

[0066] In this embodiment, physical entities correspond to people (planting managers), machines (intelligent agricultural machinery and equipment), objects (crops), and environments (environment); virtual entities correspond to geometric models (three-dimensional models of humans, machines, objects, and environments), physical models (physical properties of humans, machines, and environments), behavioral models (real-time responses and behaviors of humans, machines, and environments), and rule models (association rules of knowledge graphs of humans, machines, and environments); services are service-oriented encapsulation of various types of data information and algorithm models in the system, realizing the functional and business services of the twin platform; connections include the interaction between physical entities and twin data, the interaction between physical entities and virtual entities, the interaction between physical entities and services, the interaction between virtual entities and twin data, the interaction between virtual entities and services, and the interaction between services and twin data.

[0067] In this embodiment, the crop growth management digital twin module 140 uses an agent mechanism to establish a communication connection between the crop twin and the human-machine interface.

[0068] exist Figure 2 In the illustrated embodiment, multiple Agent units include a sensor Agent (a crop growth environment monitoring Agent and a crop growth phenotype monitoring Agent), a phenotypic parameter parsing Agent, a prescription decision calculation Agent, a prescription push Agent, and a human-computer interaction Agent; the digital farmland module and the crop growth phenotype monitoring module are respectively connected to the sensor Agent (corresponding to the connection between the soil moisture and meteorological monitoring Internet of Things platform and the crop growth environment monitoring Agent, and the connection between the drone phenotyping platform and the crop growth phenotype monitoring Agent), and the phenotypic parameter parsing Agent is connected to the crop growth phenotype monitoring Agent, which is used to perform phenotypic analysis on the crop growth phenotype data through graphics images and artificial intelligence algorithms; the basic database, the human-computer interaction Agent, the crop growth environment monitoring Agent, and the crop growth model are respectively connected to the prescription decision calculation Agent; the physical entity (corresponding to the water-fertilizer integration, the plant protection and agricultural machinery physical entity), the virtual entity (corresponding to the water-fertilizer integration, the plant protection and agricultural machinery virtual twin) and the prescription decision calculation Agent are respectively connected to the prescription push Agent; the human-computer interaction Agent is connected to the three-dimensional visualization panel.

[0069] Figure 2 This is the second structural diagram of the farmland management digital twin decision support system provided by the present invention. Figure 2 In the illustrated embodiment, the crop growth environment monitoring Agent is used to obtain environmental data in real time by monitoring the physical entity of the field environment, and the crop growth phenotype monitoring Agent is used to obtain crop growth data by monitoring the physical entity of crop growth. These two Agents obtain virtual values ​​of growth environment data through the field environment virtual twin, the field planting scene twin and the crop virtual twin respectively; the phenotypic parameter analysis Agent is used to receive multimodal data collected by the drone phenotyping platform, and calculate the crop growth phenotypic indicators; the prescription decision calculation Agent is used to use the virtual value of the growth environment data obtained by the environmental monitoring Internet of Things sensor and the crop growth phenotypic data obtained by the phenotypic analysis as input, and optimize and calculate to obtain the current water and fertilizer prescription; the prescription push Agent is used to convert the water and fertilizer prescription into an intelligent agricultural machinery operation signal and send it to the agricultural machinery for operation execution; the human-computer interaction Agent has a built-in three-dimensional visualization modeling engine, which performs three-dimensional visualization of crop growth, Internet of Things devices, phenotyping drones and agricultural machinery operations, and receives user interaction information to provide a visual display; a crop twin information interaction mechanism is established through the above-mentioned multiple agents to form a decision-making calculation mutual feedback mechanism.

[0070] The three-dimensional visualization unit 143 is connected to the human-computer interaction Agent.

[0071] In this embodiment, the three-dimensional visualization unit 143 includes: a farmland plot digital display panel, a soil moisture data display panel, a meteorological data display panel, a crop variety data display panel, a soil basic data display panel, a growth status phenotype data display panel, a crop virtual growth data display panel, a prescription GIS map data display panel, an intelligent agricultural machinery data display panel and a water-fertilizer integrated agricultural machinery data panel; each data panel and the corresponding twin are connected through an Agent to provide users with data query, crop growth simulation, water and fertilizer management decision prescription calculation, and online and visualized interactive operations for virtual control of real operations.

[0072] The digital twin decision support system for farmland management provided by the embodiment of the present invention, by setting a crop growth control digital twin module, constructs crop twins, environment twins, agricultural machinery twins and interactive calculations between each twin through an agent mechanism, and performs crop planting water and fertilizer management in a virtual-to-real manner according to the interactive calculation results between each twin. It can independently form effective management plans and prescriptions, open up unmanned application scenarios of agricultural generation management including data monitoring, decision-making calculations, and autonomous operations, effectively save labor costs, and further improve crop yields and water and fertilizer utilization efficiency.

[0073] In some embodiments, the production management decision calculation model is also used to generate multiple prescription node data based on farmland GIS map data, water and fertilizer decision prescriptions, and pesticide decision prescriptions, and generate a prescription GIS map based on the prescription node data; the prescription GIS map is sent to agricultural machinery; wherein the prescription node data includes longitude and latitude, fertilizer amount, water amount, pesticide amount and pesticide type.

[0074] In this embodiment, the production management decision-making calculation model generates prescriptions for each node according to the field GIS location map and forms a prescription GIS map. The prescription node data is recorded and stored in the GIS node data format.

[0075] In this embodiment, after obtaining the prescription GIS map, the prescription GIS map is sent to the agricultural machinery in accordance with the specified protocol format; for water and fertilizer operations, the prescription GIS map is sent to the controller unit of the water and fertilizer integrated equipment, and the water and fertilizer integrated equipment performs automated variable water and fertilizer irrigation operations according to the operation prescription; for pesticide application operations, the pesticide application GIS prescription is sent to the pesticide application plant protection drone, and the pesticide application plant protection drone performs automated variable pesticide application operations according to the GIS prescription and route.

[0076] The digital twin decision support system for farmland management provided by the embodiment of the present invention generates multiple prescription node data according to farmland GIS map data, water and fertilizer regulation and pesticide application decision prescription results through a production management decision calculation model, and generates a prescription GIS map according to the prescription node data, and sends it to agricultural machinery to perform different operations, thereby improving the comprehensive operation efficiency of water, fertilizer and pesticide.

[0077] Figure 3 This is one of the flow charts of the control method of the farmland management digital twin decision support system provided by the present invention, such as Figure 3 As shown, the control method of the farmland management digital twin decision support system includes the following steps: Step 310: Based on the digital farmland module, multiple farmland parameters are collected, and the multiple farmland parameters include farmland GIS map data, soil fertility data, soil temperature and humidity environmental data, meteorological data, and crop variety data; based on the crop growth phenotype monitoring module, the crop growth status phenotype data is collected and pipelined phenotype analysis is performed to generate a phenotypic indicator GIS layer; wherein the growth status phenotype data includes RGB images, spectral images, and point cloud data.

[0078] In this step, in the early stage of crop planting and growth, the plot image data is obtained through remote sensing maps and image data collected by drones, and the image deep learning segmentation model is used to segment the plot image data to obtain the farmland plot boundary information; then, based on the farmland plot boundary information, a farmland plot grid map composed of several plots is obtained; finally, the farmland plot grid map is processed based on GIS technology to obtain the map layers of each farmland plot, that is, digital farmland GIS map data, which includes attribute information such as plot number, longitude and latitude of the center point of the plot, elevation, and boundary of the plot.

[0079] In this step, soil fertility sampling points are generated according to the elevation data of each farmland plot, and the soil fertility data of the sampling points are collected according to the method of not sampling the neighboring plots with elevation difference less than the threshold Tf; then the interpolation technology is used to construct a digital farmland plot grid fertility map, that is, soil fertility data. The soil fertility data establishes a digital farmland fertility database according to the field information such as the plot number, the longitude and latitude of the center point of the plot, and the fertility (nitrogen, phosphorus, potassium content) of the plot.

[0080] In this step, the installation location of the soil temperature and humidity environmental sensor is determined according to the method of not sampling the soil environment of the neighboring fields whose elevation data difference of each field is less than the threshold value Tm; the soil environment sensor is installed at the selected installation location of the soil temperature and humidity environmental sensor, and the soil temperature, humidity and other environmental data are collected regularly through the wireless module; the interpolation technology is used to construct a digital farmland field grid soil temperature and humidity environmental map, and then according to the field information such as the field number, the longitude and latitude of the field center, the soil temperature, and the soil humidity, a field digital farmland soil temperature and humidity environmental database is established.

[0081] In this step, meteorological sensors are installed on the fields, and the meteorological environment data of the fields (for example, weather temperature, humidity, solar radiation, rainfall, wind speed and other information) are collected regularly through wireless modules; a digital farmland field meteorological database is established according to the field number, longitude and latitude of the field center, weather temperature, humidity, solar radiation, rainfall, wind speed and other field information.

[0082] In this step, the main crop varieties in the planting area where the farmland is located are collected and selected, and a crop variety database is constructed according to information such as the district or county where it is located, disease resistance, and accumulated growth temperature.

[0083] In this embodiment, the crop growth phenotype monitoring module includes the following modules: drone airport equipment, route generation unit, phenotype data collection unit and analysis unit.

[0084] Among them, the drone airport equipment is installed in the farmland area, and the drone airport equipment includes a variety of sensors; for example, RGB image sensors, spectral image sensors and lidar sensors; among them, the RGB image sensor is used to obtain crop RGB image data, and further analyze to obtain crop leaf color, texture, spot distribution, crop canopy coverage and other phenotypic indicators; the spectral image sensor is used to obtain crop multispectral image data, and further analyze to obtain NDVI vegetation index, nitrogen content, moisture content and other indicators; the lidar sensor is used to obtain crop three-dimensional point cloud data, and further analyze to obtain plant height, leaf area, biomass and plant structure related phenotypic indicators.

[0085] In this embodiment, the route generating unit is used to generate different flight routes according to different growth stages of the crops.

[0086] In this embodiment, the phenotypic data collection unit is used to control the drone airport equipment to collect growth status phenotypic data from the farmland area according to the target flight route and the target sensor sampling frequency.

[0087] In this embodiment, the parsing unit is used to extract a first phenotypic indicator from the RGB image, extract a second phenotypic indicator from the spectral image, extract a third phenotypic indicator from the point cloud data of the phenotypic indicator GIS layer, and generate a phenotypic indicator GIS layer based on the first phenotypic indicator, the second phenotypic indicator, the third phenotypic indicator and the coordinates of the target flight route.

[0088] In this embodiment, the site data of the corresponding field is first sampled from the collected crop RGB images, spectral images and point cloud data according to the route node position of the drone airport, and a pipelined phenotyping method is constructed through the following steps: (1) RGB pixels are extracted from RGB images based on the image deep learning model, and then phenotypic indicators such as crop leaf color, texture, lesion distribution and crop canopy coverage are calculated based on the RGB pixels based on the image algorithm, i.e., the first phenotypic indicator.

[0089] (2) Crop spectral pixels are extracted from spectral image data based on the image deep learning model. Then, the spectral inversion model is used to calculate the crop canopy vegetation index NDVI, nitrogen content, water content and other indicators based on the crop spectral pixels, i.e., the second phenotypic indicator.

[0090] (3) Based on the point cloud deep learning semantic model, the point cloud data is segmented and extracted for crop point clouds. Further, based on the point cloud deep learning instance model, the point cloud of crop leaves is segmented and extracted. The crop population canopy point cloud model is reconstructed based on the segmented and extracted point clouds, and the phenotypic indicators related to plant height, leaf area, biomass and plant structure, i.e., the third phenotypic indicator, are calculated.

[0091] (4) Based on the first phenotypic index, the second phenotypic index, and the third phenotypic index, combined with the route coordinate information, a phenotypic index GIS layer is generated (including: field number, longitude and latitude of the field center point, RGB image phenotypic index, spectral phenotypic index, point cloud phenotypic index, and crop canopy point cloud model, etc.), and the phenotypic index GIS layer is stored in the field crop phenotypic monitoring database.

[0092] Step 320: the crop knowledge model based on the crop production management model decision module fits the crop growth change curve according to multiple farmland parameters to obtain simulated values ​​of the crop growth parameters; the crop three-dimensional structure function model based on the crop production management model decision module reconstructs the crop population in three dimensions according to the phenotypic index GIS layer to obtain a crop three-dimensional model, and calculates the true values ​​of the crop growth parameters according to the crop three-dimensional model; the production management decision calculation model based on the crop production management model decision module makes decisions on water and fertilizer regulation according to the simulated values ​​and the true values ​​to obtain a water and fertilizer decision prescription, determines the disease information of the crops according to the growth status phenotypic data, and determines the pesticide application decision prescription according to the pre-pesticide requirements and the disease information, so that agricultural machinery can perform integrated water, fertilizer and pesticide operations; the production management decision calculation model is determined based on the coupling of the crop knowledge model and the crop three-dimensional structure function model; the crop growth parameters include at least one of the values ​​of the daily dynamic changes of leaf age, leaf area index and NDVI.

[0093] In this step, the crop knowledge model is used to simulate planting with preset planting varieties, planting periods, planting spacing, seeding amounts and other planting plans. The suitable growth curve of the crop throughout the growing season can be simulated and calculated. The growth curve is a curve formed by the daily dynamic changes of leaf age, leaf area index and NDVI calculated based on the yield target and basic data such as meteorology and soil.

[0094] In this step, the crop three-dimensional structural function model obtains the phenotypic data of the crop in the key growth period (including image data, point cloud data and other phenotypic data) through the phenotypic sensor, reconstructs the crop three-dimensional model by combining the three-dimensional modeling technology, and then uses the crop three-dimensional model to calculate the crop growth functional parameters (i.e., crop growth indicators) such as leaf age, leaf area index, and NDVI, to form a crop growth monitoring method driven by monitoring data.

[0095] Specifically, the method for constructing a three-dimensional crop model includes: using crop canopy point cloud data and multispectral data collected by drones, and fusing the point cloud and spectral data to generate a spectral point cloud, then using ground elevation data to remove the ground point cloud of the plot, and using the centroid shrinkage method to obtain the center point of each crop plant, and then determining the root base point of the crop plant, and finally using the point cloud erosion search clustering algorithm to segment and mark to obtain the single canopy point cloud of each crop, thereby realizing three-dimensional reconstruction of the crop group.

[0096] In this embodiment, the method for calculating crop functional parameters based on a three-dimensional model includes: calculating plant height through crop plant point cloud data; obtaining leaf point cloud through point cloud data combined with a stem-leaf segmentation model, obtaining the number of plant leaves using a clustering algorithm, and calculating the crop leaf age; obtaining plant leaf point cloud through point cloud data combined with a stem-leaf segmentation model, gridding the leaf point cloud, calculating the plant leaf area, and further calculating the leaf area index; calculating the NDVI value of the crop population based on the spectral point cloud data.

[0097] In this embodiment, for the single-point location data calculated above, a GIS difference method is used to calculate and generate a model parameter map, which is stored in a model parameter database.

[0098] In this embodiment, the above-mentioned crop knowledge model is used to simulate the change curve of key indicators of crop growth (including the daily dynamic change values ​​of leaf area index LAI and normalized vegetation index NDVI) according to basic geographical data, real-time meteorological soil moisture data and variety characteristic data to determine the simulated values ​​of crop growth parameters; the obtained crop population is three-dimensionally modeled through the crop three-dimensional structure function model, and then the true values ​​of crop growth parameters (including leaf area index LAI and normalized vegetation index NDVI) of the current period are calculated through the crop population three-dimensional model; finally, the leaf area index LAI1 and normalized vegetation index NDVI1 (crop growth indicators) measured by the production management decision calculation model are compared with the leaf area index LAI2 and normalized vegetation index NDVI2 (crop growth parameters) of the period calculated by the crop knowledge model, and the current water and fertilizer regulation method is calculated by comprehensive decision.

[0099] In this embodiment, for crop water demand regulation, soil moisture content data from the field moisture monitoring station is obtained in real time, and the variety parameters of the crop variety's water requirements are obtained to calculate the irrigation amount; for crop drug demand regulation, crop growth data monitored by phenotypic drones are used to identify crop disease information (current disease type and disease severity), and the type and amount of drug application are calculated according to the drug application requirements of the disease database to form a drug application plan.

[0100] The control method of the digital twin decision support system for farmland management provided by the embodiment of the present invention collects multiple farmland parameters through the digital farmland module, collects growth status phenotypic data through the crop growth phenotype monitoring module and performs pipeline phenotypic analysis to generate phenotypic indicators, fits the growth change curve of the crop according to the multiple farmland parameters through the crop knowledge model to obtain the crop growth parameters, reconstructs the crop population in three dimensions according to the phenotypic indicators through the crop three-dimensional structure function model, and then calculates the crop growth indicators, and regulates water, fertilizer and medicine according to the simulated values ​​and real values ​​of the crop growth parameters through the production management decision calculation model, so that agricultural machinery can perform comprehensive water, fertilizer and medicine operations, thereby improving the efficiency and intelligence level of crop planting management.

[0101] In some embodiments, the control method of the digital twin decision support system for farmland management also includes: constructing crop twins, environment twins, agricultural machinery twins and interactive calculations between the twins based on the crop growth control digital twin module, and performing comprehensive management of crop water, fertilizer and pesticides in a virtual-to-real manner according to the interactive calculation results between the twins; the crop growth control digital twin module communicates with the digital farmland module, the crop growth phenotype monitoring module and the crop production management model decision module through the Agent mechanism.

[0102] In this embodiment, the crop growth control digital twin module consists of a five-dimensional digital twin model of crop growth, a crop twin information communication mechanism and prescription decision calculation, and a three-dimensional visualization unit. The twin module provides water and fertilizer management for crop planting by controlling the real in a virtual way.

[0103] Specifically, the crop growth management and control digital twin module includes: a five-dimensional crop growth digital twin model, multiple agent units and a three-dimensional visualization unit.

[0104] Among them, the five-dimensional digital twin model of crop growth includes physical entities, virtual entities, services, twin data and connections. Physical entities include planting managers, agricultural machinery, crops and environment; virtual entities include geometric models, physical models, behavioral models and rule models corresponding to physical entities; services are used to encapsulate various types of data information and algorithm models; twin data includes physical entity data, virtual entity data, service data, knowledge data and fused derivative data; connections are used to provide data interaction for any two of physical entities, virtual entities, services and twin data.

[0105] In this embodiment, physical entities correspond to people (planting managers), machines (intelligent agricultural machinery and equipment), objects (crops), and environments (environment); virtual entities correspond to geometric models (three-dimensional models of humans, machines, objects, and environments), physical models (physical properties of humans, machines, and environments), behavioral models (real-time responses and behaviors of humans, machines, and environments), and rule models (association rules of knowledge graphs of humans, machines, and environments); services are service-oriented encapsulation of various types of data information and algorithm models in the system, realizing the functional and business services of the twin platform; connections include the interaction between physical entities and twin data, the interaction between physical entities and virtual entities, the interaction between physical entities and services, the interaction between virtual entities and twin data, the interaction between virtual entities and services, and the interaction between services and twin data.

[0106] In this embodiment, the crop growth management digital twin module uses the Agent mechanism to establish a communication connection between the crop twin and the human-machine.

[0107] In this embodiment, multiple Agent units include a sensor agent, a phenotypic parameter analysis agent, a prescription decision calculation agent, a prescription push agent and a human-computer interaction agent; the digital farmland module and the crop growth phenotype monitoring module are respectively connected to the sensor agent, and the phenotypic parameter analysis agent is connected to the crop growth phenotype monitoring module; the basic database, the sensor agent, and the crop three-dimensional structure and function model are respectively connected to the prescription decision calculation agent; the physical entity, the virtual entity and the prescription decision calculation agent are respectively connected to the prescription push agent; the human-computer interaction agent is connected to the prescription decision calculation agent.

[0108] In this embodiment, the sensor agent is used to obtain environmental data and crop growth data in real time; the phenotypic parameter analysis agent is used to receive multimodal data collected by the drone phenotyping platform and calculate the crop growth phenotypic indicators; the prescription decision calculation agent is used to use the environmental data obtained by the environmental monitoring Internet of Things sensor and the crop growth phenotypic data obtained by phenotypic analysis as input, and optimize and calculate the current water and fertilizer prescription; the prescription push agent is used to convert the water and fertilizer prescription into intelligent agricultural machinery operation signals and send them to agricultural machinery for operation execution; the human-computer interaction agent has a built-in three-dimensional visualization modeling engine to perform three-dimensional visualization of crop growth, Internet of Things devices, phenotyping drones and agricultural machinery operations, and receive user interaction information to provide a visual display; a crop twin information interaction mechanism is established through the above multiple agents to form a decision-making calculation mutual feedback mechanism.

[0109] In this embodiment, the three-dimensional visualization unit is connected to the human-computer interaction Agent; the three-dimensional visualization unit includes: a farmland plot digital display panel, a soil moisture data display panel, a meteorological data display panel, a crop variety data display panel, a soil basic data display panel, a growth status phenotype data display panel, a crop virtual growth data display panel, a prescription GIS map data display panel, an intelligent agricultural machinery data display panel and a water-fertilizer integrated agricultural machinery data panel; each data panel and the corresponding twin are connected through the Agent to provide users with data query, crop growth simulation, water and fertilizer management decision prescription calculation, and online and visualized interactive operations of virtual control and real operation management.

[0110] The control method of the digital twin decision support system for farmland management provided by the embodiment of the present invention, by setting a crop growth control digital twin module, constructs crop twins, environment twins, agricultural machinery twins and interactive calculations between each twin through an agent mechanism, and performs crop planting water and fertilizer management in a virtual-to-real manner according to the interactive calculation results between each twin. It can independently form effective management plans and prescriptions, open up unmanned application scenarios of agricultural generation management of data monitoring, decision-making calculations, and autonomous operations, effectively save labor costs, and further improve crop yields and water and fertilizer utilization efficiency.

[0111] Figure 4 This is the second flow chart of the control method of the farmland management digital twin decision support system provided by the present invention. Figure 4 In the embodiment shown, the Agent mechanism is first used to establish communication connections between the field environment monitoring Internet of Things system (corresponding to the digital farmland module collection) and the UAV phenotyping monitoring system (corresponding to the crop growth phenotyping monitoring module) and the five-dimensional digital twin model of crop growth management (corresponding to the five-dimensional digital twin model of crop growth), and the prescription decision calculation Agent corresponding to the crop production management model decision module uses the virtual value of the crop growth environment data obtained by the field environment monitoring Internet of Things system and the crop growth phenotypic data obtained by the UAV phenotyping monitoring system as input, and calculates the water and fertilizer management prescription through the crop production management decision calculation method of the coupling knowledge model and the structure-function model, and sends the water and fertilizer management prescription to the agricultural machinery for operation according to the prescription map and push protocol through the prescription push Agent, and at the same time, the crop growth, Internet of Things devices, phenotyping drones and agricultural machinery operations are displayed on the three-dimensional visualization panel (corresponding to the three-dimensional visualization unit) of the crop growth control digital twin through the human-computer interaction agent.

[0112] Figure 5 is a schematic diagram of the structure of the electronic device provided by the present invention, such as Figure 5As shown, the electronic device may include: a processor 510, a communication interface 520, a memory 530 and a communication bus 540, wherein the processor 510, the communication interface 520 and the memory 530 communicate with each other through the communication bus 540. The processor 510 may call the logic instructions in the memory 530 to execute the control method of the farmland management digital twin decision support system, the method comprising: collecting multiple farmland parameters based on the digital farmland module, the multiple farmland parameters including farmland GIS map data, soil fertility data, soil temperature and humidity environmental data, meteorological data and crop variety data; collecting the growth status phenotype data of the crop based on the crop growth phenotype monitoring module and performing pipeline phenotype analysis to generate a phenotypic indicator GIS layer; wherein the growth status phenotype data includes RGB images, spectral images and point cloud data; the crop knowledge model based on the crop production management model decision module fits the growth change curve of the crop according to multiple farmland parameters to obtain the simulated value of the crop growth parameter; based on the crop production management The crop three-dimensional structure function model of the model decision module reconstructs the crop group in three dimensions according to the phenotypic index GIS layer to obtain the crop three-dimensional model, and calculates the real value of the crop growth parameter according to the crop three-dimensional model; the production management decision calculation model based on the crop production management model decision module makes decisions on water and fertilizer regulation according to the simulated value and the real value, obtains the water and fertilizer decision prescription, determines the disease information of the crop according to the growth status phenotypic data, and determines the pesticide decision prescription according to the pre-pesticide requirements and disease information, so as to provide agricultural machinery with water, fertilizer and pesticide integrated operations; the production management decision calculation model is determined based on the coupling of the crop knowledge model and the crop three-dimensional structure function model; the crop growth parameters include at least one of the daily dynamic changes in the values ​​of leaf age, leaf area index and NDVI.

[0113] In addition, the logic instructions in the above-mentioned memory 530 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0114] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to execute the control method of the digital twin decision support system for farmland management provided by the above-mentioned methods, the method comprising: collecting multiple farmland parameters based on a digital farmland module, the multiple farmland parameters including farmland GIS map data, soil fertility data, soil temperature and humidity environmental data, meteorological data and crop variety data; collecting crop growth status phenotypic data based on a crop growth phenotype monitoring module and performing pipelined phenotypic analysis to generate a phenotypic indicator GIS layer; wherein the growth status phenotypic data includes RGB images, spectral images and point cloud data; a crop knowledge model based on a crop production management model decision module fits the growth change curve of the crop according to multiple farmland parameters to obtain Simulated values ​​of crop growth parameters; the crop three-dimensional structure function model based on the crop production management model decision module reconstructs the crop group in three dimensions according to the phenotypic index GIS layer to obtain a crop three-dimensional model, and calculates the true values ​​of the crop growth parameters according to the crop three-dimensional model; the production management decision calculation model based on the crop production management model decision module makes decisions on water and fertilizer regulation according to the simulated values ​​and the true values, obtains water and fertilizer decision prescriptions, determines the disease information of crops according to the growth status phenotypic data, and determines the pesticide decision prescription according to the pre-pesticide requirements and disease information, so that agricultural machinery can perform comprehensive water, fertilizer and pesticide operations; the production management decision calculation model is determined based on the coupling of the crop knowledge model and the crop three-dimensional structure function model; the crop growth parameters include at least one of the daily dynamic changes in the values ​​of leaf age, leaf area index and NDVI.

[0115] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0116] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0117] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A farmland management digital twin decision support system, characterized in that: include: Digital farmland module, which is used to collect multiple farmland parameters, including farmland GIS map data, soil fertility data, soil temperature and humidity environment data, meteorological data and crop variety data; A crop growth phenotype monitoring module, which is used to collect crop growth status phenotype data and perform pipeline phenotype analysis to generate a phenotypic indicator GIS layer; wherein the growth status phenotype data includes RGB images, spectral images and point cloud data; The crop production management model decision module includes a crop knowledge model, a crop three-dimensional structure and function model, and a production management decision calculation model; wherein the crop knowledge model is used to fit the growth change curve of the crop according to the multiple farmland parameters to obtain the simulated value of the crop growth parameter; the crop three-dimensional structure and function model is used to perform three-dimensional reconstruction of the crop group according to the phenotypic index GIS layer to obtain a crop three-dimensional model, and calculate the real value of the crop growth parameter according to the crop three-dimensional model; the production management decision calculation model is used to make decisions on water and fertilizer regulation according to the simulated value and the real value, obtain a water and fertilizer decision prescription, determine the disease information of the crop according to the growth state phenotypic data, and determine the drug application decision prescription according to the pre-drug application requirements and the disease information, so that agricultural machinery can perform water, fertilizer and drug comprehensive operations; the production management decision calculation model is determined based on the coupling of the crop knowledge model and the crop three-dimensional structure and function model; the crop growth parameter includes at least one of the values ​​of the daily dynamic changes of leaf age, leaf area index and NDVI.

2. The farmland management digital twin decision support system according to claim 1, characterized in that: The system further comprises: A crop growth control digital twin module, which is connected to the digital farmland module, the crop growth phenotype monitoring module and the crop production management model decision module through an agent mechanism; the crop growth control digital twin module is used to construct crop twins, environment twins, agricultural machinery twins and interactive calculations between twins, and based on the interactive calculation results between the twins, comprehensive management of crop water, fertilizer and pesticides is performed in a virtual-real way.

3. The farmland management digital twin decision support system according to claim 2 is characterized in that: The crop growth control digital twin module includes: A five-dimensional digital twin model of crop growth, the five-dimensional digital twin model of crop growth includes physical entities, virtual entities, services, twin data and connections, the physical entities include planting managers, agricultural machinery, crops and environment; the virtual entities include geometric models, physical models, behavioral models and rule models corresponding to the physical entities; the services are used to encapsulate various types of data information and algorithm models; the twin data includes physical entity data, virtual entity data, service data, knowledge data and fusion derived data; the connection is used to provide data interaction for any two of the physical entity, the virtual entity, the service and the twin data; A plurality of Agent units, the plurality of Agent units include a sensor Agent, a phenotypic parameter parsing Agent, a prescription decision calculation Agent, a prescription push Agent and a human-computer interaction Agent; the digital farmland module and the crop growth phenotype monitoring module are respectively connected to the sensor Agent, the phenotypic parameter parsing Agent is connected to the crop growth phenotype monitoring module; the basic database, the sensor Agent, and the crop three-dimensional structure function model are respectively connected to the prescription decision calculation Agent; the physical entity, the virtual entity and the prescription decision calculation Agent are respectively connected to the prescription push Agent; the human-computer interaction Agent is connected to the prescription decision calculation Agent; A three-dimensional visualization unit is connected to the human-computer interaction Agent.

4. The farmland management digital twin decision support system according to claim 3 is characterized in that: The three-dimensional visualization unit includes: a farmland plot digital display panel, a soil moisture data display panel, a meteorological data display panel, a crop variety data display panel, a soil basic data display panel, a growth status phenotype data display panel, a crop virtual growth data display panel, a prescription GIS map data display panel, an intelligent agricultural machinery data display panel and a water-fertilizer integrated agricultural machinery data panel.

5. The farmland management digital twin decision support system according to claim 1, characterized in that: The production management decision calculation model is also used to generate a plurality of prescription node data according to the farmland GIS map data, the water and fertilizer decision prescription and the pesticide decision prescription, and generate a prescription GIS map according to the prescription node data; Sending the prescription GIS map to the agricultural machinery; The prescription node data include longitude and latitude, fertilizer application amount, water application amount, pesticide application amount and pesticide application type.

6. The farmland management digital twin decision support system according to claim 1, characterized in that: The crop growth phenotype monitoring module comprises: A drone airport device, the drone airport device is installed in a farmland area, and the drone airport device includes a variety of sensors; A flight path generation unit, the flight path generation unit is used to generate different flight paths according to different growth stages of the crops; A phenotypic data collection unit, the phenotypic data collection unit is used to control the drone airport equipment to collect the growth status phenotypic data from the farmland area according to the target flight route and the target sensor sampling frequency; A parsing unit, the parsing unit is used to extract a first phenotypic indicator from the RGB image, a second phenotypic indicator from the spectral image, and a third phenotypic indicator from the point cloud data, and generate the phenotypic indicator GIS layer based on the first phenotypic indicator, the second phenotypic indicator, the third phenotypic indicator and the coordinates of the target flight route; Among them, the first phenotypic indicator includes at least one of the crop leaf color, texture, spot distribution and crop canopy coverage; the second phenotypic indicator includes at least one of the crop canopy vegetation index NDVI, nitrogen content and water content; the third phenotypic indicator includes at least one of the crop plant height, leaf area, biomass and plant structure.

7. A control method for a farmland management digital twin decision support system, characterized in that: include: Based on the digital farmland module, multiple farmland parameters are collected, and the multiple farmland parameters include farmland GIS map data, soil fertility data, soil temperature and humidity environment data, meteorological data and crop variety data; based on the crop growth phenotype monitoring module, the crop growth status phenotype data is collected and pipelined phenotype analysis is performed to generate a phenotypic indicator GIS layer; wherein the growth status phenotype data includes RGB images, spectral images and point cloud data; The crop knowledge model based on the crop production management model decision module fits the crop growth change curve according to the multiple farmland parameters to obtain the simulated values ​​of the crop growth parameters; the crop three-dimensional structure function model based on the crop production management model decision module reconstructs the crop group in three dimensions according to the phenotypic index GIS layer to obtain a crop three-dimensional model, and calculates the real values ​​of the crop growth parameters according to the crop three-dimensional model; the production management decision calculation model based on the crop production management model decision module makes decisions on water and fertilizer regulation according to the simulated values ​​and the real values ​​to obtain a water and fertilizer decision prescription, determines the disease information of the crops according to the growth status phenotypic data, and determines the pesticide decision prescription according to the pre-drug application requirements and the disease information, so that agricultural machinery can perform comprehensive water, fertilizer and pesticide operations; the production management decision calculation model is determined based on the coupling of the crop knowledge model and the crop three-dimensional structure function model; the crop growth parameters include at least one of the daily dynamic changing values ​​of leaf age, leaf area index and NDVI.

8. The control method of the farmland management digital twin decision support system according to claim 7, characterized in that: The method further comprises: Based on the crop growth management digital twin module, crop twins, environment twins, agricultural machinery twins and interactive calculations between the twins are constructed, and according to the interactive calculation results between the twins, comprehensive management of crop water, fertilizer and pesticides is carried out in a virtual-to-real manner; the crop growth management digital twin module is communicated and connected with the digital farmland module, the crop growth phenotype monitoring module and the crop production management model decision module through an Agent mechanism.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, it implements the control method of the farmland management digital twin decision support system as described in any one of claims 7 to 8.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, it implements the control method of the farmland management digital twin decision support system as described in any one of claims 7 to 8.

Images