Data management and control and intelligent analysis system and method for whole-process of agricultural product production

Abstract

The application discloses an agricultural product production whole-process data management and control and intelligent analysis system and method, relates to the technical field of data management and control, and comprises a data acquisition module for deploying Internet of Things sensors and multispectral unmanned aerial vehicles to generate a crop growth digital archive; a storage and deduction module for simulating growth deduction in a mechanism growth digital model; a yield optimization module for generating a yield optimization scheme according to target yield influencing factors; a data history module for obtaining a digital fingerprint ID of crops and generating a query two-dimensional code; a data query module for constructing a crop knowledge graph and constructing a natural language question and answer interface on the front end; a crop adaptation module for generating crop / pattern plug-ins and extending the crop / pattern adaptation plug-ins; and an automatic grading module for performing computer vision recognition on crop fruits and automatically grading the crop fruits. The application has the effect of improving the efficiency and accuracy of agricultural product production whole-process data management and analysis.

Description

Technical Field

[0001] This application relates to the technical field of data management, and in particular to a data management and intelligent analysis system and method for the entire process of agricultural production. Background Technology

[0002] With consumers paying more attention to the quality and safety of agricultural products and the fresh food industry developing on a large scale, the need for full-chain management of agricultural products from planting to production is becoming increasingly urgent.

[0003] In current technologies, data management of the entire agricultural production process often relies on traditional manual monitoring for data collection. This data is then manually analyzed and managed to achieve the desired results. However, manual methods are prone to inaccurate and incomplete data collection, leading to problems in subsequent data analysis. Furthermore, manual data analysis and management can be subject to data processing biases and are extremely time-consuming and labor-intensive. Additionally, manual analysis requires personnel with a deep understanding of agricultural products, which vary widely, necessitating different data processing methods for different types, placing significant pressure on staff. Therefore, efficient and accurate data management and intelligent analysis of the entire agricultural production process is a pressing issue that needs to be addressed. Summary of the Invention

[0004] The purpose of this invention is to provide a data management and intelligent analysis system and method for the entire agricultural production process, in order to solve the problems mentioned in the background art.

[0005] Firstly, this application provides a data management and intelligent analysis system for the entire agricultural production process, the system comprising:

[0006] Data acquisition module: used to deploy IoT sensors and multispectral drones in the field to monitor crop growth, obtain crop growth status data and agricultural operation data, and generate digital archives of crop growth;

[0007] Storage and simulation module: used to store the crop growth digital archive, establish a mechanism growth digital model, substitute multiple preset plans into the mechanism growth digital model to simulate growth, and record the simulation results and yield-affecting factors generated during the simulation process;

[0008] Production optimization module: used to filter multiple simulation results to obtain the optimal target plan, extract the target production influencing factors corresponding to the target plan, and generate a production optimization scheme based on the target production influencing factors;

[0009] Data history module: used to extract digital fingerprints from the crop growth digital files of different batches of crops, obtain the digital fingerprint ID of each batch of crops, store the digital fingerprint ID in the blockchain, and generate a query QR code;

[0010] Data query module: used to extract the relationship structure between different data in the crop growth digital archive, and to construct a crop knowledge graph based on the relationship structure and the crop growth digital archive, and to construct a natural language question answering interface on the front end based on the crop knowledge graph;

[0011] Crop adaptation module: used to extract historical plant data and / or historical pattern data of new crops and / or new planting patterns introduced to the farm, generate crop / pattern plugins based on the historical plant data and / or historical pattern data, and extend the crop / pattern adaptation plugins.

[0012] Automatic grading module: used to obtain the current crop grading standards in the market, perform computer vision recognition on crop fruits according to the crop grading standards, and perform automatic grading.

[0013] Preferably, the data acquisition module includes:

[0014] Sensor units and drone units;

[0015] Sensor unit: used to acquire field maps and divide the field into multiple small plots based on the field maps;

[0016] In each of the aforementioned small plots, IoT sensors are deployed to monitor the roots of crops and record agricultural operations to obtain initial growth data and agricultural operation data.

[0017] Unmanned aerial vehicle (UAV) unit: used to deploy multispectral UAVs around the field according to the field map. The multispectral UAVs inspect the crops in the field at preset time intervals and collect data on the stems, leaves and fruits of the crops to obtain second growth data.

[0018] By combining the first growth data and the second growth data, crop growth status data is obtained. By combining the growth status data and the agricultural operation data, a crop growth digital profile is generated.

[0019] Preferably, the storage deduction module includes:

[0020] Data storage unit and model derivation unit;

[0021] Data storage unit: used to identify erroneous data in the crop growth digital archive, correct errors, and store the corrected crop growth digital archive stably for a long period of time;

[0022] Model extrapolation unit: used to acquire local historical crop yields, historical meteorological data and current soil data, and to construct a digital model framework based on the historical crop yields, historical meteorological data and current soil data;

[0023] Based on the crop growth digital archive, the data in the crop growth digital archive is substituted into the digital model framework to construct a crop mechanistic growth digital model.

[0024] Multiple different pre-set scenarios are substituted into the mechanism growth digital model, and the mechanism growth digital model performs mechanism growth deduction based on each different scenario to obtain multiple deduction results;

[0025] Record the simulation process information of each of the contingency plans during the simulation process, compare the simulation results horizontally to obtain the order of superiority and inferiority of the simulation results, identify the simulation process information based on the order of superiority and inferiority of the simulation results, and obtain the factors affecting the final output.

[0026] Preferably, the production optimization module includes:

[0027] Contingency plan screening unit and data analysis unit;

[0028] Contingency plan screening unit: used to extract the output data from each of the simulation results, and to extract the number of factors and the difficulty of factor processing of the output influencing factors in each simulation process;

[0029] The production data, the number of factors, and the processing difficulty of the factors are weighted and assigned, and the optimal index is calculated to evaluate and obtain the optimal target plan corresponding to the highest optimal index.

[0030] Data analysis unit: used to extract the target output influencing factors corresponding to the target plan, and identify the time period, scope of influence, depth of influence and cause of occurrence of the target output influencing factors;

[0031] By combining the time period of occurrence, the cause of occurrence, the scope of impact, and the depth of impact, a problem-solving plan is generated. Based on the problem-solving plan, the target plan is optimized to generate a production optimization plan.

[0032] Preferably, the data history module includes:

[0033] Digital fingerprint unit and query unit;

[0034] Digital fingerprint unit: used to extract crop growth characteristic data from the crop growth digital archives of different batches of crops;

[0035] Based on the crop growth characteristic data, unique feature identification is performed on the crop growth characteristic data to obtain the target unique feature;

[0036] Based on the unique characteristics of the target, digital fingerprints are extracted to obtain the digital fingerprint ID of each batch of crops, and the digital fingerprint ID is stored in the blockchain;

[0037] Query unit: Used to generate a query QR code for each digital fingerprint ID based on the digital fingerprint ID, and link the query QR code to the crop growth digital file corresponding to each digital fingerprint ID.

[0038] Preferably, the data query module includes:

[0039] Knowledge graph units and front-end question-answering interface;

[0040] Knowledge graph unit: used to extract plant data, pest and disease data, pesticide data and agronomic data from the crop growth digital archive;

[0041] Identify the relationship structure between the plant data, the pest and disease data, the pesticide data, and the agronomic data; and construct a crop knowledge graph for each batch of crops based on the relationship structure and the crop growth digital archive.

[0042] Front-end question answering interface: used to build a natural language question answering interface, which is connected to the crop knowledge graph information. The natural language question answering interface receives external natural language questions and performs semantic analysis on the natural language questions to obtain the query target.

[0043] Based on the query target, a search is performed in the crop knowledge graph to obtain the target output result;

[0044] The natural language question-answering interface also includes a condition combination query component, which includes four conditions: agricultural product type, growth cycle, time period, and data type.

[0045] Preferably, the crop adaptation module includes:

[0046] Plugin generation unit and adaptation extension unit;

[0047] Plugin generation unit: used to extract historical plant data and / or historical pattern data of new crops and / or new planting patterns introduced in the farm;

[0048] Data extraction is performed on the historical plant data and / or the historical pattern data to obtain plant growth information and / or pattern operation information;

[0049] Based on the plant growth information and / or the mode operation information, construct a crop / mode plugin;

[0050] Adaptation and expansion unit: used to extract basic field environmental information based on the crop growth digital archive, evaluate the adaptability of the crop / mode plugin based on the basic environmental information, and obtain adaptation error information;

[0051] Based on the adaptation error information, the crop / pattern plugin is adjusted to obtain the target crop / pattern plugin, and the system is adapted and extended based on the target crop / pattern plugin.

[0052] Preferably, the automatic grading module includes:

[0053] Market monitoring unit, automatic grading unit, and real-time adjustment unit;

[0054] Market monitoring unit: used to monitor the sales information of similar crops in the market, obtain real-time sales information, and obtain crop grading standards based on the real-time sales information;

[0055] Automatic grading unit: used to call a preset camera to perform computer vision recognition on crop fruits and obtain appearance condition data of crop fruits;

[0056] The crop fruits are graded according to the appearance condition data and the crop grading standards.

[0057] Real-time adjustment unit: used to monitor the crop grading standards and unsold crop fruits, and to determine whether the crop grading standards have changed;

[0058] If it is determined that the crop grading standard has changed, the automatic grading unit is adjusted in real time, the grade label information of the currently unsold crop fruits is recorded, and the grade label information is corrected in real time.

[0059] Secondly, this application provides a method for data management and intelligent analysis of the entire agricultural production process, the method comprising:

[0060] Deploying IoT sensors and multispectral drones in the fields monitors crop growth, obtains crop growth status data and agricultural operation data, and generates digital archives of crop growth.

[0061] The crop growth digital archive is stored, and a mechanistic growth digital model is established. Multiple preset plans are substituted into the mechanistic growth digital model to simulate growth and extrapolate, and the extrapolation results and yield-affecting factors generated during the extrapolation process are recorded.

[0062] The optimal target plan is obtained by filtering multiple simulation results, the target output influencing factors corresponding to the target plan are extracted, and an output optimization plan is generated based on the target output influencing factors.

[0063] Digital fingerprints are extracted from the crop growth digital files of different batches of crops to obtain the digital fingerprint ID of each batch of crops. The digital fingerprint ID is stored in the blockchain and a query QR code is generated.

[0064] Extract the relationship structure between different data in the crop growth digital archive, construct a crop knowledge graph based on the relationship structure and the crop growth digital archive, and construct a natural language question answering interface on the front end based on the crop knowledge graph;

[0065] Extract historical plant data and / or historical pattern data of new crops and / or new planting patterns introduced to the farm, generate crop / pattern plugins based on the historical plant data and / or historical pattern data, and extend them based on the crop / pattern adaptation plugins;

[0066] Obtain the current crop grading standards in the market, perform computer vision recognition on crop fruits according to the crop grading standards, and automatically grade them.

[0067] In summary, this application includes at least one of the following beneficial technical effects:

[0068] By monitoring crop growth in the field using IoT sensors and multispectral drones, data on crop growth status and agricultural operations are obtained, constructing digital crop growth archives. These archives are stored, and a mechanistic growth digital model is established. Simulations are performed using this model to obtain results under different scenarios and factors influencing yield, generating yield optimization plans. Digital fingerprints are extracted from different batches of crop growth archives to obtain digital fingerprint IDs, which are stored in the blockchain and used to generate query QR codes. A crop knowledge graph is constructed using the relationship structure between different data within the crop growth archives, and a natural language question-and-answer interface is built on the front end to process query requests and output results. Crop / pattern plugins are then generated for new plants or new planting patterns for adaptation and expansion. Finally, market crop grading standards are obtained, and crop fruits are automatically graded using computer vision recognition. This improves the efficiency and accuracy of data management and analysis throughout the entire agricultural production process. Attached Figure Description

[0069] Figure 1 This is a block diagram of the data management and intelligent analysis system for the entire agricultural production process provided in this application.

[0070] Figure 2 This is a flowchart illustrating the steps of the data management and intelligent analysis method for the entire agricultural production process provided in this application.

[0071] Explanation of reference numerals in the attached diagram: 1. Data acquisition module; 2. Storage and simulation module; 3. Yield optimization module; 4. Data history module; 5. Data query module; 6. Crop adaptation module; 7. Automatic grading module. Detailed Implementation

[0072] The following combination Figures 1-2 This application will be described in further detail, but the embodiments of the present invention are not limited thereto.

[0073] This application discloses a data management and intelligent analysis system and method for the entire agricultural production process.

[0074] In this embodiment of the application, an agricultural product production process data management and intelligent analysis system is provided, comprising:

[0075] Data acquisition module 1: used to deploy IoT sensors and multispectral drones in the field to monitor crop growth, obtain crop growth status data and agricultural operation data, and generate digital archives of crop growth;

[0076] Storage and simulation module 2: It is used to store the digital archive of crop growth, establish a mechanistic growth digital model, substitute multiple different preset plans into the mechanistic growth digital model to simulate growth and simulation, and record the simulation results and yield-affecting factors generated during the simulation process.

[0077] Production optimization module 3: It is used to filter multiple simulation results to obtain the optimal target plan, extract the target production influencing factors corresponding to the target plan, and generate production optimization schemes based on the target production influencing factors.

[0078] Data history module 4: Used to extract digital fingerprints from the crop growth digital archives of different batches of crops, obtain the digital fingerprint ID of each batch of crops, store the digital fingerprint ID in the blockchain, and generate a query QR code;

[0079] Data query module 5: used to extract the relationship structure between different data in the crop growth digital archive, and to build a crop knowledge graph based on the relationship structure and the crop growth digital archive, and to build a natural language question answering interface on the front end based on the crop knowledge graph;

[0080] Crop Adaptation Module 6: Used to extract historical plant data and / or historical pattern data of new crops and / or new planting patterns introduced to the farm, generate crop / pattern plugins based on historical plant data and / or historical pattern data, and extend the crop / pattern adaptation plugins.

[0081] Automatic grading module 7: Used to obtain the current crop grading standards in the market, perform computer vision recognition on crop fruits according to the crop grading standards, and perform automatic grading.

[0082] The data acquisition module includes:

[0083] Sensor units and drone units;

[0084] Sensor unit: used to acquire field maps and divide the field into multiple small plots based on the field maps;

[0085] IoT sensors are deployed in each small plot of land to monitor the roots of crops and record agricultural operations, thus obtaining initial growth data and agricultural operation data.

[0086] Unmanned aerial vehicle (UAV) unit: Used to deploy multispectral UAVs around the field based on the field map. The multispectral UAVs patrol the field crops at preset time intervals and collect data on the crop stems, leaves and fruits to obtain secondary growth data.

[0087] By combining the first growth data and the second growth data, crop growth status data is obtained. By combining the growth status data and agricultural operation data, a digital profile of crop growth is generated.

[0088] In this application, taking a 50-acre tomato farm as an example, after obtaining a field map of the farm, the entire field is evenly divided into one hundred small plots of similar size. Then, IoT sensors are deployed at the center of each small plot. These sensors include soil temperature and humidity sensors and nitrogen, phosphorus, and potassium content sensors. The IoT sensors continuously monitor the soil moisture and nutrients around the tomato plants' roots and record agricultural operations such as irrigation time and fertilizer type, obtaining primary growth data and agricultural operation data. Simultaneously, based on the field map, three multispectral drones are deployed around the field. These drones patrol the tomato plants at preset intervals of 10:00 AM daily. The drones maintain a flight altitude of ten meters and perform multispectral scanning of the tomato stem thickness, leaf color, and fruit size, collecting secondary growth data. Then, the primary and secondary growth data are combined, for example, by correlating root nutrient data with leaf spectral data, to obtain overall growth status data of the tomato plants. Finally, the growth status data and recorded agricultural operation data are integrated to form a crop growth digital archive that includes plant height, leaf index, fruit quantity, irrigation records, and fertilization records. This archive is stored in the form of digital tables and images.

[0089] The storage simulation module includes:

[0090] Data storage unit and model derivation unit;

[0091] Data storage unit: used to identify erroneous data in the crop growth digital archive, correct errors, and store the corrected crop growth digital archive stably for a long period of time;

[0092] Model extrapolation unit: used to acquire local historical crop yields, historical meteorological data and current soil data, and to construct a digital model framework based on historical crop yields, historical meteorological data and current soil data;

[0093] Based on the crop growth digital archive, the data in the crop growth digital archive is substituted into the digital model framework to construct a digital model of crop growth mechanism.

[0094] Multiple different pre-set scenarios are substituted into the mechanism growth digital model, and the mechanism growth digital model performs mechanism growth deduction based on each different scenario to obtain multiple deduction results;

[0095] Record the simulation process information of each plan, compare the results of multiple simulations horizontally to obtain the order of superiority and inferiority of the simulation results, identify the simulation process information based on the order of superiority and inferiority of the simulation results, and obtain the factors affecting the final output.

[0096] In practice, taking this tomato farm as an example, the data storage unit identifies errors in the generated crop growth digital archive. The system scan found an abnormally high water volume value in one irrigation record, exceeding the normal range by three times, which was identified as a sensor false alarm. The system corrected this data, adjusting the water volume value to the normal range, and then stored the corrected crop growth digital archive in a cloud database for long-term stable storage. The model extrapolation unit first acquires historical tomato yield data, historical meteorological data, and current soil testing data from the past five years. Based on this data, a digital model framework including parameters such as light, temperature, humidity, and soil fertility is constructed. Then, the actual growth data from the crop growth digital archive is substituted into this framework to construct a mechanistic growth digital model of tomatoes. Next, three preset planting plans are substituted into the model for simulated growth extrapolation. Plan one involves increasing irrigation frequency, plan two involves increasing potassium fertilizer application, and plan three involves mulching. The mechanistic growth digital model was run on the three scenarios, yielding three projected yields of 8,000 kg / mu, 8,500 kg / mu, and 7,800 kg / mu, respectively. The system recorded detailed information during the simulation process for each scenario, such as the timing of drought occurrence and nutrient consumption curves. Finally, the three projected results were compared, ranked by yield, and the analysis of the simulation process information based on the ranking identified water stress and potassium deficiency as the key factors affecting the final yield.

[0097] The production optimization module includes:

[0098] Contingency plan screening unit and data analysis unit;

[0099] Contingency plan screening unit: used to extract production data from each simulation result, and to extract the number of factors and the difficulty of factor processing of production influencing factors in each simulation process;

[0100] Weights are assigned to production data, the number of factors, and the difficulty of factor processing, and the optimal index is calculated to evaluate the optimal target plan corresponding to the highest optimal index.

[0101] Data analysis unit: used to extract the factors affecting the target output corresponding to the target plan, and to identify the time period, scope, depth of influence and causes of the factors affecting the target output.

[0102] By combining the time period, cause, scope of impact, and depth of impact, a problem-solving plan is generated. Based on the problem-solving plan, the target plan is optimized to generate a production optimization plan.

[0103] In application, taking the results of the farm's simulation as an example, the contingency plan screening unit first extracts the yield data from the three simulation results: 8,000, 8,500, and 7,800. Then, it extracts the yield-influencing factors identified in each simulation: Contingency Plan 1 has one factor—water stress—with moderate handling difficulty; Contingency Plan 2 has one factor—potassium deficiency—with low handling difficulty; and Contingency Plan 3 has two factors—uneven soil temperature and weed competition—with high handling difficulty. The system assigns weights of 0.5, 0.3, and 0.2 to the yield data, the number of factors, and the handling difficulty of the factors, respectively, and calculates the optimal index. The calculation shows that Contingency Plan 2 has the highest index, therefore it is determined as the target contingency plan. The data analysis unit then extracts the target yield-influencing factor corresponding to the target contingency plan, namely, potassium deficiency. The system identifies that this factor occurs during the fruit enlargement period, affects 60% of the planting area, and has an impact depth leading to smaller fruits. The cause is insufficient potassium supply from the soil and untimely topdressing. Then, combining the occurrence time, cause, impact range, and impact depth, a problem-solving plan is generated. The plan recommends applying a quick-acting potassium fertilizer to all plots at the early stage of fruit enlargement, at a rate of 10 kg per acre, in conjunction with irrigation. Finally, based on this problem-solving plan, the target plan is optimized by incorporating this topdressing operation into the original planting plan, generating the final yield optimization plan, which details the timing, dosage, and method of fertilization.

[0104] The data history module includes:

[0105] Digital fingerprint unit and query unit;

[0106] Digital fingerprint unit: used to extract crop growth characteristic data from the crop growth digital archives of different batches of crops;

[0107] Based on crop growth characteristic data, unique feature identification is performed on the crop growth characteristic data to obtain the target unique feature;

[0108] Based on the unique characteristics of the target, digital fingerprints are extracted to obtain the digital fingerprint ID of each batch of crops, and the digital fingerprint ID is stored in the blockchain;

[0109] Query Unit: Used to generate a query QR code for each digital fingerprint ID based on the digital fingerprint ID, and link the query QR code to the crop growth digital file corresponding to each digital fingerprint ID.

[0110] In application, taking different batches of tomatoes from the farm as an example, the digital fingerprint unit first extracts growth characteristic data from the crop growth digital archives of the first and second batches of tomatoes, including average plant height, first inflorescence fruit setting time, and average fruit sugar content. Based on this crop growth characteristic data, the system performs unique feature identification. It finds that the first inflorescence fruit setting time of the first batch of tomatoes is generally three days later than that of the second batch, and the average fruit sugar content is 0.5 degrees higher. These differences constitute the target unique feature. Then, based on this target unique feature, the system extracts digital fingerprints, generating digital fingerprint IDs "TOMATO-2023A-7.5" for the first batch of tomatoes and "TOMATO-2023B-7.0" for the second batch. These two digital fingerprint IDs are uploaded and stored in the blockchain network to ensure their immutability. The query unit generates a corresponding query QR code for each ID based on the stored digital fingerprint ID. For example, a QR code image is generated for ID "TOMATO-2023A-7.5". This QR code links to the complete crop growth digital archive of the first batch of tomatoes stored on the cloud server. After scanning the QR code, consumers can view the complete data history of the batch of tomatoes from planting to harvest on the webpage, including fertilization records, pest and disease control records, and harvest date.

[0111] The data query module includes:

[0112] Knowledge graph units and front-end question-answering interface;

[0113] Knowledge graph unit: used to extract plant data, pest and disease data, pesticide data, and agronomic data from crop growth digital archives;

[0114] Identify the relationship structure between plant data, pest and disease data, pesticide data, and agronomic data. Based on the relationship structure and crop growth digital archives, construct a crop knowledge graph for each batch of crops.

[0115] Front-end question answering interface: used to build a natural language question answering interface. The natural language question answering interface is connected to the crop knowledge graph information. The natural language question answering interface receives external natural language questions and performs semantic analysis on the natural language questions to obtain the query target.

[0116] Based on the query target, a search is performed in the crop knowledge graph to obtain the target output results;

[0117] The natural language question answering interface also has a condition combination query component, which includes four conditions: agricultural product type, growth cycle, time period, and data type.

[0118] In practice, taking the crop growth digital archive of the first batch of tomatoes from this farm as an example, the knowledge graph unit first extracts plant data (such as plant height and stem diameter), pest and disease data (such as aphid occurrence records), pesticide data (such as imidacloprid usage information), and agronomic data (such as pruning methods) from the archive. The system identifies the relationship structure between these data, for example, "aphid occurrence" leads to "imidacloprid use," and "plants recover growth" after "imidacloprid use." Based on these relationship structures and the original archive data, the system constructs a visualized crop knowledge graph, where nodes represent entities and lines represent relationships. The front-end question-and-answer interface constructs a natural language question-and-answer box, which is connected to the knowledge graph. When a user enters "When was the first batch of tomatoes sprayed with pesticide?" in the question-and-answer box, the interface performs semantic analysis on the question, obtaining the query target as "the time of pesticide application for the first batch of tomatoes." Based on this target, the system searches in the knowledge graph, finds the associated nodes of "first batch of tomatoes," "pesticide application operation," and "date," and finally outputs the target result "Imidacloprid was used to control aphids on June 15th." The Q&A interface also features a conditional combination query component, allowing users to combine and select agricultural product type "tomato", growth cycle "flowering period", time period "May to June", and data type "pest and disease data" for more precise queries.

[0119] The crop adaptation module includes:

[0120] Plugin generation unit and adaptation extension unit;

[0121] Plugin generation unit: used to extract historical plant data and / or historical pattern data of new crops and / or new planting patterns introduced in the farm;

[0122] Extract data from historical plant data and / or historical model data to obtain plant growth information and / or model operation information;

[0123] Build crop / model plugins based on plant growth information and / or model operation information;

[0124] Adaptation and Extension Unit: Used to extract basic environmental information from the field based on the crop growth digital archive, evaluate the adaptability of crop / mode plugins based on the basic environmental information, and obtain adaptation error information;

[0125] Based on the adaptation error information, the crop / pattern plugin is adjusted to obtain the target crop / pattern plugin, and the system is adapted and extended based on the target crop / pattern plugin.

[0126] In practice, taking the farm's plan to introduce strawberries as a new crop as an example, the plugin generation unit first extracts historical plant data from other farms that grow strawberries. This data includes strawberry plant height growth curves, suitable temperature and humidity ranges, common disease types, and corresponding agricultural operations. The system extracts key plant growth information from this historical plant data, such as the optimal daytime temperature of 15 to 25 degrees Celsius, and operational model information, such as the need to build rain shelters for disease prevention. Based on this plant growth information and operational model information, the system constructs a crop plugin called "Strawberry Cultivation," which encapsulates a growth pattern model and an agricultural operation calendar. The adaptation and expansion unit extracts basic field environmental information based on the farm's tomato crop growth data, such as an average annual temperature of 18 degrees Celsius and sandy loam soil. Then, this environmental information is used to evaluate the compatibility of the "Strawberry Cultivation" plugin. The evaluation found that the winter insulation measures required by the plugin are not compatible with the farm's low-temperature winter conditions. Based on the adaptation error message, the system adjusted the plugin, adding the operation suggestion of "adding a double layer of film in winter," resulting in the target crop plugin "Strawberry Planting (Farm-Specific Version)." Finally, the system loaded this target plugin, successfully extending the strawberry planting management module into the original system.

[0127] The automatic grading module includes:

[0128] Market monitoring unit, automatic grading unit, and real-time adjustment unit;

[0129] Market monitoring unit: used to monitor the sales information of similar crops in the market, obtain real-time sales information, and obtain crop grading standards based on real-time sales information;

[0130] Automatic grading unit: used to call a preset camera to perform computer vision recognition on crop fruits and obtain appearance condition data of crop fruits;

[0131] The crop fruits are graded based on appearance data and crop grading standards.

[0132] Real-time adjustment unit: Used to monitor crop grading standards and unsold crop fruits, and to determine whether the crop grading standards have changed;

[0133] If it is determined that the crop grading standard has changed, the automatic grading unit will be adjusted in real time, the grade label information of the currently unsold crop fruits will be recorded, and the grade label information will be corrected in real time.

[0134] In practice, taking the tomato harvest from this farm as an example, the market monitoring unit continuously monitors tomato sales information on major fruit and vegetable wholesale markets and e-commerce platforms to obtain real-time sales data. System analysis revealed that tomatoes with a diameter greater than six centimeters, a fully red color, and no blemishes currently command the highest market price. Based on this, the latest crop grading standard is "Extra Grade: Diameter > 6cm, fully red color, no defects." The automatic grading unit utilizes high-definition cameras installed on the sorting line to perform computer vision recognition on each tomato on the conveyor belt, obtaining appearance data for each fruit. If the diameter is 5.8cm, the color is 90% red, and there are minor blemishes at the stem end, the system compares the appearance data with the latest grading standard and grades the tomato as "Grade 1." The real-time adjustment unit simultaneously monitors the market grading standard and graded but unsold tomatoes. One day later, the market monitoring unit discovers that the grading standard has changed to "Extra Grade: Diameter > 7cm." The system determines that the standard has changed and immediately adjusts the automatic grading unit, changing the diameter threshold from six centimeters to seven centimeters. At the same time, the grade label information of unsold tomatoes in the current warehouse that were originally labeled "premium" is recorded, and these labels are corrected to "grade 1" in real time according to the new standard.

[0135] This invention provides a method for data management and intelligent analysis of the entire agricultural production process, using any of the above-mentioned agricultural production data management and intelligent analysis systems. The method includes the following:

[0136] S100: Deploy IoT sensors and multispectral drones in the field to monitor crop growth, obtain crop growth status data and agricultural operation data, and generate digital archives of crop growth.

[0137] S200: Store the crop growth digital archives and establish a mechanistic growth digital model. Substitute multiple preset plans into the mechanistic growth digital model to simulate growth and record the simulation results and yield-affecting factors generated during the simulation process.

[0138] S300: Filter multiple simulation results to obtain the optimal target plan, extract the target output influencing factors corresponding to the target plan, and generate an output optimization plan based on the target output influencing factors;

[0139] S400: Extract digital fingerprints from the crop growth digital archives of different batches of crops to obtain the digital fingerprint ID of each batch of crops, store the digital fingerprint ID in the blockchain, and generate a query QR code.

[0140] S500: Extract the relationship structure between different data in the crop growth digital archive, construct a crop knowledge graph based on the relationship structure and the crop growth digital archive, and build a natural language question answering interface on the front end based on the crop knowledge graph;

[0141] S600: Extract historical plant data and / or historical pattern data of new crops and / or new planting patterns introduced to the farm, generate crop / pattern plugins based on historical plant data and / or historical pattern data, and extend them according to crop / pattern adaptation plugins;

[0142] S700: Acquires current crop grading standards in the market, performs computer vision recognition on crop fruits according to the crop grading standards, and automatically grades them.

[0143] The above are all preferred embodiments of this application and are not intended to limit the scope of protection of this application. Therefore, all equivalent changes made in accordance with the structure, shape and principle of this application should be covered within the scope of protection of this application.

Claims

1. A data management and intelligent analysis system for the entire agricultural production process, characterized in that: include: Data acquisition module: used to deploy IoT sensors and multispectral drones in the field to monitor crop growth, obtain crop growth status data and agricultural operation data, and generate digital archives of crop growth; Storage and simulation module: used to store the crop growth digital archive, establish a mechanism growth digital model, substitute multiple preset plans into the mechanism growth digital model to simulate growth, and record the simulation results and yield-affecting factors generated during the simulation process; Production optimization module: used to filter multiple simulation results to obtain the optimal target plan, extract the target production influencing factors corresponding to the target plan, and generate a production optimization scheme based on the target production influencing factors; Data history module: used to extract digital fingerprints from the crop growth digital files of different batches of crops, obtain the digital fingerprint ID of each batch of crops, store the digital fingerprint ID in the blockchain, and generate a query QR code; Data query module: used to extract the relationship structure between different data in the crop growth digital archive, and to construct a crop knowledge graph based on the relationship structure and the crop growth digital archive, and to construct a natural language question answering interface on the front end based on the crop knowledge graph; Crop adaptation module: used to extract historical plant data and / or historical pattern data of new crops and / or new planting patterns introduced to the farm, generate crop / pattern plugins based on the historical plant data and / or historical pattern data, and extend the crop / pattern adaptation plugins. Automatic grading module: used to acquire the current crop grading standards in the market, perform computer vision recognition on crop fruits according to the crop grading standards, and perform automatic grading; The storage simulation module includes: Data storage unit and model derivation unit; Data storage unit: used to identify erroneous data in the crop growth digital archive, correct errors, and store the corrected crop growth digital archive stably for a long period of time; Model extrapolation unit: used to acquire local historical crop yields, historical meteorological data and current soil data, and to construct a digital model framework based on the historical crop yields, historical meteorological data and current soil data; Based on the crop growth digital archive, the data in the crop growth digital archive is substituted into the digital model framework to construct a crop mechanistic growth digital model. Multiple different pre-set scenarios are substituted into the mechanism growth digital model, and the mechanism growth digital model performs mechanism growth deduction based on each different scenario to obtain multiple deduction results; Record the simulation process information of each of the contingency plans during the simulation process, compare the simulation results horizontally to obtain the order of superiority and inferiority of the simulation results, identify the simulation process information based on the order of superiority and inferiority of the simulation results, and obtain the factors affecting the final output.

2. The agricultural product production process data management and intelligent analysis system according to claim 1, characterized in that, The data acquisition module includes: Sensor units and drone units; Sensor unit: used to acquire field maps and divide the field into multiple small plots based on the field maps; In each of the aforementioned small plots, IoT sensors are deployed to monitor the roots of crops and record agricultural operations to obtain initial growth data and agricultural operation data. Unmanned aerial vehicle (UAV) unit: used to deploy multispectral UAVs around the field according to the field map. The multispectral UAVs inspect the crops in the field at preset time intervals and collect data on the stems, leaves and fruits of the crops to obtain second growth data. By combining the first growth data and the second growth data, crop growth status data is obtained. By combining the growth status data and the agricultural operation data, a crop growth digital profile is generated.

3. The agricultural product production process data management and intelligent analysis system according to claim 2, characterized in that, The production optimization module includes: Contingency plan screening unit and data analysis unit; Contingency plan screening unit: used to extract the output data from each of the simulation results, and to extract the number of factors and the difficulty of factor processing of the output influencing factors in each simulation process; The production data, the number of factors, and the processing difficulty of the factors are weighted and assigned, and the optimal index is calculated to evaluate and obtain the optimal target plan corresponding to the highest optimal index. Data analysis unit: used to extract the target output influencing factors corresponding to the target plan, and identify the time period, scope of influence, depth of influence and cause of occurrence of the target output influencing factors; By combining the time period of occurrence, the cause of occurrence, the scope of impact, and the depth of impact, a problem-solving plan is generated. Based on the problem-solving plan, the target plan is optimized to generate a production optimization plan.

4. The agricultural product production process data management and intelligent analysis system according to claim 3, characterized in that, The data history module includes: Digital fingerprint unit and query unit; Digital fingerprint unit: used to extract crop growth characteristic data from the crop growth digital archives of different batches of crops; Based on the crop growth characteristic data, unique feature identification is performed on the crop growth characteristic data to obtain the target unique feature; Based on the unique characteristics of the target, digital fingerprints are extracted to obtain the digital fingerprint ID of each batch of crops, and the digital fingerprint ID is stored in the blockchain; Query unit: Used to generate a query QR code for each digital fingerprint ID based on the digital fingerprint ID, and link the query QR code to the crop growth digital file corresponding to each digital fingerprint ID.

5. The agricultural product production process data management and intelligent analysis system according to claim 4, characterized in that, The data query module includes: Knowledge graph units and front-end question-answering interface; Knowledge graph unit: used to extract plant data, pest and disease data, pesticide data and agronomic data from the crop growth digital archive; Identify the relationship structure between the plant data, the pest and disease data, the pesticide data, and the agronomic data; and construct a crop knowledge graph for each batch of crops based on the relationship structure and the crop growth digital archive. Front-end question answering interface: used to build a natural language question answering interface, which is connected to the crop knowledge graph information. The natural language question answering interface receives external natural language questions and performs semantic analysis on the natural language questions to obtain the query target. Based on the query target, a search is performed in the crop knowledge graph to obtain the target output result; The natural language question-answering interface also includes a condition combination query component, which includes four conditions: agricultural product type, growth cycle, time period, and data type.

6. The agricultural product production process data management and intelligent analysis system according to claim 5, characterized in that, The crop adaptation module includes: Plugin generation unit and adaptation extension unit; Plugin generation unit: used to extract historical plant data and / or historical pattern data of new crops and / or new planting patterns introduced in the farm; Data extraction is performed on the historical plant data and / or the historical pattern data to obtain plant growth information and / or pattern operation information; Based on the plant growth information and / or the mode operation information, construct a crop / mode plugin; Adaptation and expansion unit: used to extract basic field environmental information based on the crop growth digital archive, evaluate the adaptability of the crop / mode plugin based on the basic environmental information, and obtain adaptation error information; Based on the adaptation error information, the crop / pattern plugin is adjusted to obtain the target crop / pattern plugin, and the system is adapted and extended based on the target crop / pattern plugin.

7. The agricultural product production process data management and intelligent analysis system according to claim 6, characterized in that, The automatic grading module includes: Market monitoring unit, automatic grading unit, and real-time adjustment unit; Market monitoring unit: used to monitor the sales information of similar crops in the market, obtain real-time sales information, and obtain crop grading standards based on the real-time sales information; Automatic grading unit: used to call a preset camera to perform computer vision recognition on crop fruits and obtain appearance condition data of crop fruits; The crop fruits are graded according to the appearance condition data and the crop grading standards. Real-time adjustment unit: used to monitor the crop grading standards and unsold crop fruits, and to determine whether the crop grading standards have changed; If it is determined that the crop grading standard has changed, the automatic grading unit is adjusted in real time, the grade label information of the currently unsold crop fruits is recorded, and the grade label information is corrected in real time.

8. A method for data management and intelligent analysis of the entire agricultural production process, wherein the method uses any one of the data management and intelligent analysis systems for the entire agricultural production process as described in claims 1-7, characterized in that, The method includes: Deploying IoT sensors and multispectral drones in the fields monitors crop growth, obtains crop growth status data and agricultural operation data, and generates digital archives of crop growth. The crop growth digital archive is stored, and a mechanistic growth digital model is established. Multiple preset plans are substituted into the mechanistic growth digital model to simulate growth and extrapolate, and the extrapolation results and yield-affecting factors generated during the extrapolation process are recorded. The optimal target plan is obtained by filtering multiple simulation results, the target output influencing factors corresponding to the target plan are extracted, and an output optimization plan is generated based on the target output influencing factors. Digital fingerprints are extracted from the crop growth digital files of different batches of crops to obtain the digital fingerprint ID of each batch of crops. The digital fingerprint ID is stored in the blockchain and a query QR code is generated. Extract the relationship structure between different data in the crop growth digital archive, construct a crop knowledge graph based on the relationship structure and the crop growth digital archive, and construct a natural language question answering interface on the front end based on the crop knowledge graph; Extract historical plant data and / or historical pattern data of new crops and / or new planting patterns introduced to the farm, generate crop / pattern plugins based on the historical plant data and / or historical pattern data, and extend them based on the crop / pattern adaptation plugins; Obtain the current crop grading standards in the market, perform computer vision recognition on crop fruits according to the crop grading standards, and automatically grade them.

Images