Crop management method and device based on large model, and medium

By applying big model AI technology in agriculture, pest monitoring, yield prediction, intelligent irrigation and automated operations have been achieved, the problems of agricultural production efficiency and quality have been solved, the degree of intelligence and precision has been improved, and the modernization of agriculture has been promoted. Sustainable development.

CN119027059APending Publication Date: 2024-11-26SHANDONG LANGCHAO YUNTOU INFORMATION TECH CO LTD
View PDF 0 Cites 14 Cited by

Patent Information

Application Number
CN202411156196.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-22
Publication Date
2024-11-26

AI Technical Summary

Technical Problem

Agricultural production efficiency and quality are low, the degree of intelligence, precision and informatization is insufficient, and innovative applications are lacking.

Method used

Using AI technology based on big models, through convolutional neural networks, decision tree models, cluster analysis and deep learning algorithms, pest monitoring and diagnosis, yield prediction, intelligent irrigation and automated operations are realized, and the intelligence and precision of agricultural management are improved. level.

Benefits of technology

It has improved agricultural production efficiency and crop yields, reduced resource waste, reduced environmental pollution, improved farmers' quality of life, and promoted the modernization and sustainable development of agriculture.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119027059A_ABST
    Figure CN119027059A_ABST
Patent Text Reader

Abstract

The invention discloses a crop management method and device based on a large model, and a medium, belongs to the technical field of large models and AI, and aims to solve the technical problem of how to improve the agricultural generation efficiency and quality by using the large model and improve the intelligence, precision and informatization degree of agricultural production at the same time. According to the technical scheme, disease and insect pest monitoring and diagnosis are carried out, wherein disease and insect pests are rapidly recognized through the convolutional neural network image recognition technology; yield prediction: adopting an AI technology to predict crop yield by analyzing historical and real-time data, assisting planting decision, and optimizing crop layout and harvesting time; intelligent irrigation: analyzing soil and crop data through a decision tree model to realize accurate application of water and fertilizer resources; crop growth monitoring: analyzing the crop growth data according to clustering analysis and principal component analysis so as to obtain a crop growth change trend, and predicting a crop production condition; and automatic operation is realized.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of big models and AI technology, and in particular to a big model-based crop management method, equipment and medium. Background Art

[0002] The application background technologies of AI technology based on large models in agriculture cover artificial intelligence and machine learning, data science and big data technology, cloud computing and edge computing, Internet of Things (IoT) technology, geographic information system (GIS), agricultural science knowledge, software engineering and system integration, etc. These technologies support each other and jointly promote the process of agricultural modernization and realize efficient and environmentally friendly production methods. Through the integration and development of these background technologies, the agricultural field will usher in more advanced technological applications, improve agricultural production efficiency and crop yields, and promote the sustainable development of agriculture.

[0003] With the continuous development of agricultural production and the continuous advancement of technology, agricultural management is also facing some new challenges, as follows:

[0004] ① There is a lack of innovation and application in agricultural technology, and the efficiency and quality of agricultural production need to be improved.

[0005] ② The level of intelligence, precision and informatization in agricultural production is low.

[0006] The application background technologies of big model-based AI technology in agriculture cover artificial intelligence and machine learning, data science and big data technology, cloud computing and edge computing, Internet of Things (IoT) technology, geographic information system (GIS), agricultural science knowledge, software engineering and system integration, etc.

[0007] Therefore, how to use large models to improve agricultural production efficiency and quality, while at the same time improving the intelligence, precision and informatization of agricultural production is a technical problem that needs to be solved urgently. Summary of the invention

[0008] The technical task of the present invention is to provide a crop management method, equipment and medium based on a large model to solve the problem of how to use large models to improve agricultural production efficiency and quality, while improving the intelligence, precision and informatization of agricultural production.

[0009] The technical task of the present invention is achieved in the following way: a crop management method based on a large model, the method is as follows:

[0010] Pest and disease monitoring and diagnosis: Using convolutional neural network (CNN) image recognition technology to quickly identify pests and diseases, reduce the abuse of pesticides, and improve crop health and yield;

[0011] Yield forecasting: AI technology is used to predict crop yields by analyzing historical and real-time data, assist in planting decisions, and optimize crop layout and harvest time;

[0012] Smart irrigation: Analyze soil and crop data through decision tree models to achieve accurate application of water and fertilizer resources, reduce waste, and improve resource utilization efficiency;

[0013] Crop growth monitoring: Analyze crop growth data based on cluster analysis and principal component analysis to obtain crop growth trends and predict crop production conditions;

[0014] Automated operations: Use deep learning algorithms to detect and track the growth of crops and adjust agricultural machinery operation strategies.

[0015] As a preferred method, pest and disease monitoring and diagnosis are as follows:

[0016] Crop image data collection: Use sensors, cameras or drones installed in farmland to collect crop image data, and then obtain crop growth environment data; the crop growth environment data includes temperature, humidity and soil conditions;

[0017] Image preprocessing: preprocess the collected images; preprocessing includes resizing, cropping and noise removal to facilitate subsequent analysis;

[0018] Image feature extraction: Use convolutional neural networks (CNNs) to automatically extract features from images to distinguish healthy crops from crops affected by pests and diseases; the extracted features include texture, color, and shape;

[0019] Large model training: Use convolutional neural network (CNN) or support vector machine (SVM) supervised learning algorithms to train labeled data sets; the labeled data sets include images of healthy crops and images of crops affected by different pests and diseases;

[0020] Pest and disease identification and classification: Identify and classify pests and diseases in images through a trained large model; and use the large model to distinguish different fungal diseases, viral diseases, or insect pests.

[0021] Real-time monitoring and early warning: Deploy the trained large model to farmland and conduct real-time monitoring through sensors and cameras. When potential pests and diseases are detected, an early warning is automatically issued to notify farmers to take measures.

[0022] Diagnosis and recommendations: Identify pests and diseases through large models, provide diagnostic results and prevention recommendations; at the same time, based on the type and severity of pests and diseases, recommend appropriate pesticides or biological control measures.

[0023] As a preferred method, the yield forecast is as follows:

[0024] Climate, soil and crop growth data collection: collect historical climate data, soil data and crop growth data; historical climate data includes temperature, rainfall and light; soil data includes soil type, pH value and organic matter content; crop growth data includes plant height, leaf area and biomass;

[0025] Climate, soil and crop growth data preprocessing: Clean and preprocess the collected climate, soil and crop growth data, including removing missing values, outliers and standardizing or normalizing the data;

[0026] Growth-related feature engineering: Extract relevant features based on the characteristics and needs of crop growth; relevant features include the duration of the growth stage and the climatic conditions during the key growth period;

[0027] Time series analysis: Use recurrent neural networks (RNN) or long short-term memory networks (LSTM) to model and analyze time series data, capture long-term dependencies in time series data, and use them to predict seasonal growth trends of crops;

[0028] Regression analysis: Use linear regression and support vector regression (SVR) regression algorithms in machine learning to build a yield prediction model and predict crop yields based on input features, such as predicting crop yields based on climate and soil conditions;

[0029] Model training and validation: Use historical data to train the yield prediction model and evaluate its performance through cross-validation. At the same time, adjust the network structure and learning rate-related parameters of the yield prediction model to improve its prediction accuracy and generalization ability.

[0030] Yield forecasting and decision support: Use the trained yield forecasting model to predict future climate and soil conditions, thereby predicting crop yields; and make more informed planting decisions based on the yield forecast results; planting decisions include selecting appropriate crop varieties and adjusting planting times.

[0031] As a preferred embodiment, the smart irrigation is as follows:

[0032] Soil moisture, climate conditions and crop growth data collection: collect soil moisture data, climate conditions data and crop growth data; soil moisture data includes real-time soil moisture content; climate conditions data includes temperature, rainfall and evaporation rate; crop growth data includes crop type and growth stage;

[0033] Soil moisture, climate conditions and crop growth data preprocessing: Clean and preprocess the collected soil moisture, climate conditions and crop growth data, including removing outliers and standardizing or normalizing;

[0034] Growth-related feature engineering: Extract relevant features based on irrigation requirements and crop growth characteristics; relevant features include soil moisture thresholds and climate condition thresholds;

[0035] Model training: Use decision tree or random forest algorithm to train relevant feature data; the decision tree algorithm recursively divides the data set into subsets to build a decision tree model for irrigation decision-making; the random forest algorithm improves the accuracy and stability of the decision tree model by building multiple decision trees and taking the average or majority vote;

[0036] Irrigation strategy formulation: formulate intelligent irrigation strategies based on soil moisture and climate conditions; the irrigation strategy includes irrigation time, irrigation amount and irrigation frequency;

[0037] Model validation and optimization: Use the cross-validation method to evaluate the performance of the decision tree model, and adjust the decision tree model parameters and irrigation strategies based on the prediction results of the decision tree model and the actual irrigation effect to improve the accuracy and practicality of the decision tree model;

[0038] Deployment and application: Deploy the trained decision tree model to farmland, realize intelligent irrigation through sensors and control systems, monitor soil moisture and climate conditions in real time through the decision tree model, and automatically control the irrigation system based on the prediction results.

[0039] As a preferred embodiment, the crop growth monitoring is as follows:

[0040] Growth data collection: collect crop growth data and growth environment data; crop growth data includes crop height, leaf area and biomass; crop growth environment data includes temperature, humidity and light;

[0041] Growth data preprocessing: cleaning and preprocessing the collected growth data; cleaning and preprocessing include removing outliers, standardization or normalization;

[0042] Cluster analysis of growth data: Use K-means or DBSCAN clustering algorithms in unsupervised learning to cluster crop growth data and obtain different growth patterns of crops; growth patterns include normal growth and those affected by pests and diseases;

[0043] Principal component analysis (PCA) of growth data: PCA is used to reduce the dimension of growth data and extract the main features. PCA is also used to reveal the changing trends in growth data, helping to understand and predict crop growth.

[0044] Growth pattern recognition: Based on the results of cluster analysis, different growth patterns are identified, and the characteristics and influencing factors of each growth pattern are analyzed to provide a basis for crop management;

[0045] Abnormal growth detection: Detect abnormalities in crop growth using features extracted from principal component analysis (PCA), discover abnormal growth in a timely manner, and take corresponding management measures such as adjusting irrigation or fertilization;

[0046] Growth assessment and optimization: Based on the results of cluster analysis and principal component analysis (PCA), crop growth is assessed and crop management strategies are optimized to improve crop yield and quality.

[0047] Preferably, the automated operations are as follows:

[0048] Target detection and tracking: Use YOLO (You Only Look Once) or SSD (Single Shot MultiBox Detector) deep learning algorithms to detect and track crops, weeds or insects in farmland, automatically identify objects in images, and locate and classify them; the detected target information is used to guide the automatic spraying of pesticides or automated weeding operations of agricultural machinery;

[0049] Automated control of agricultural machinery: Use Q-learning or DDPG (Deep Deterministic Policy Gradient) reinforcement learning algorithms to achieve automated control of agricultural machinery. The reinforcement learning algorithm learns the optimal operation strategy through interaction with the environment. During the farmland operation, the agricultural machinery automatically adjusts the operation path and operation according to the learned strategy to improve the operation efficiency and quality.

[0050] Real-time feedback and adjustment: During operation, the agricultural machinery receives feedback from the target detection and tracking system in real time. Based on the feedback, the agricultural machinery adjusts its operation strategy in a timely manner to adapt to the actual situation in the farmland.

[0051] Agricultural machinery operation data collection and analysis: During the automated operation process, collect agricultural machinery operation data on operation speed, path and operation quality, use machine learning algorithms to analyze agricultural machinery operation data, and optimize agricultural machinery operation strategies;

[0052] Operation efficiency and quality evaluation: Based on the collected operation data, the efficiency and quality of agricultural machinery operations are evaluated, and the agricultural machinery operation strategy is adjusted according to the evaluation results to improve the operation effect.

[0053] Preferably, a convolutional neural network (CNN) uses convolutional layers to automatically and hierarchically learn features in an image. The convolutional layers extract local features on the image through a sliding window (filter), which are then passed to the next layer of the network. Features from low-level to high-level are captured through multiple layers of convolution and pooling (downsampling) operations, and finally classification or regression is performed through a fully connected layer.

[0054] Preferably, the principal component analysis (PCA) method is a statistical method for data dimensionality reduction or feature extraction; the principal component analysis (PCA) method maps the original data into a new coordinate system through a linear transformation, and the new coordinate system is determined by the covariance structure of the original data;

[0055] In the new coordinate system, the first coordinate axis (first principal component) is the direction of the largest variance in the original data, the second coordinate axis (second principal component) is the direction of the second largest variance, and so on.

[0056] Among them, Artificial Intelligence (AI) refers to the technology of simulating human intelligence through computer programs or machines to solve complex problems.

[0057] Deep Learning: A method of implementing artificial intelligence that uses multi-layer neural networks to learn features and recognize patterns in data.

[0058] Machine Learning: A technology that enables computers to learn from data and improve their performance.

[0059] Big Data: refers to a collection of data that is huge in scale and complex in type, requiring special technologies for effective processing and analysis.

[0060] Internet of Things (IoT): refers to a system that connects various information sensing devices to the network through the Internet to achieve intelligent management and control.

[0061] Smart Agriculture: A modern agricultural method that uses technologies such as the Internet of Things, big data, and artificial intelligence to improve agricultural production efficiency, reduce costs, and protect the environment.

[0062] Precision Agriculture: Based on the actual conditions of farmland, refined management methods are adopted to achieve high efficiency, high quality and low consumption of crop production.

[0063] Agricultural Remote Sensing: Using remote sensing technologies such as satellites and aircraft to obtain farmland information and provide a scientific basis for crop production.

[0064] Meteorological Data: refers to meteorological information closely related to agricultural production, such as temperature, humidity, wind speed, and precipitation.

[0065] Soil Data: includes information on soil characteristics such as soil texture, fertility, and moisture that affect crop growth.

[0066] Growth Model: A mathematical model established based on the growth laws of crops, used to predict crop growth conditions and yields.

[0067] Crop Monitoring: Real-time monitoring and analysis of crop growth conditions through remote sensing, ground observation and other means.

[0068] Decision Support System (DSS): A computer system that provides data analysis and decision-making recommendations for agricultural production.

[0069] Automation Control: Use computer technology to realize the automated operation of agricultural machinery and equipment to improve production efficiency.

[0070] Intelligent Optimization: Use artificial intelligence algorithms to optimize resource allocation, operation scheduling, etc. in the agricultural production process to reduce costs.

[0071] An electronic device comprising: a memory and at least one processor;

[0072] Wherein, the memory stores computer-executable instructions;

[0073] The at least one processor executes the computer-executable instructions stored in the memory, so that the at least one processor performs the above-mentioned large-model-based crop management method.

[0074] A computer-readable storage medium stores computer-executable instructions. When a processor executes the computer-executable instructions, the above-mentioned large-model-based crop management method is implemented.

[0075] The crop management method, device and medium based on the large model of the present invention have the following advantages:

[0076] (I) The application of AI technology based on large models in agriculture involves the cross-integration of multiple fields such as computer vision, machine learning and deep learning, data science and big data analysis, the Internet of Things, precision agriculture, intelligent agricultural machinery, agricultural economics and management, and environmental science and ecology. The application of these technologies has brought innovative solutions to agriculture, improved agricultural production efficiency, and promoted the sustainable development of agriculture; AI technology based on large models plays an important role in agriculture by real-time monitoring of crop health, analyzing agricultural data, providing data support, collecting real-time farmland data, optimizing resource allocation, improving agricultural operation efficiency and accuracy, assisting farmers in planting decisions and risk management, and promoting environmental protection and ecological balance; with the continuous advancement and development of technology, agriculture will be more intelligent, precise and automated in the future;

[0077] (II) The application of AI technology based on large models in agriculture has far-reaching significance from the perspective of scientific and technological innovation, social impact and sustainable development, as follows:

[0078] (1) Technological innovation:

[0079] ①Technological breakthrough: The development and deployment of large models marks a major breakthrough in AI technology in dealing with complex problems, especially in the field of multivariable and nonlinear systems such as agriculture;

[0080] ② Interdisciplinary integration: The application of AI technology in agriculture has promoted the deep integration of multiple disciplines such as computer science, data science, environmental science and agricultural science, and promoted interdisciplinary innovation;

[0081] (2) Productivity improvement:

[0082] ① Improve production efficiency: The application of AI technology can help farmers manage farms more effectively, improve agricultural production efficiency and crop yields, and thus increase income;

[0083] ② Improve the quality of life: Reduce farmers’ physical labor through automation and intelligent technology, and improve working conditions and quality of life;

[0084] ③Education and training: The introduction of AI technology requires new skills and knowledge, which promotes agriculture-related education and training and improves farmers’ technical level;

[0085] (3) Optimize the environment:

[0086] ① Resource optimization: Precision agriculture technology improves resource utilization efficiency by reducing resource waste;

[0087] ② Environmental protection: reducing the use of fertilizers and pesticides, reducing environmental pollution, and contributing to ecological protection and climate change mitigation;

[0088] ③ Long-term development: By providing data-driven decision support, AI technology can help farmers better plan long-term production strategies and respond to market changes and climate challenges;

[0089] In summary, the application of AI technology based on large models in agriculture has not only made remarkable achievements in scientific and technological innovation, but also played an important role in social impact and sustainable development. Further development and application will provide agriculture with a smarter, more efficient and more environmentally friendly production method, and promote the modernization and sustainable development of agriculture.

[0090] (III) The invention improves agricultural production efficiency: through automated operations, precision agricultural technology and management optimization, AI technology can significantly improve the yield and quality of crops, reduce resource waste and lower production costs;

[0091] (iv) The invention reduces environmental impact: precision agriculture and automated operations reduce the use of fertilizers, pesticides and water, reduce greenhouse gas emissions, and contribute to environmental protection and ecological balance;

[0092] (V) The invention increases farmers' income: it improves crop yield and quality, optimizes market decisions, and helps farmers increase their income and improve their living standards;

[0093] (6) This invention promotes agricultural development: AI technology helps farmers better plan long-term production strategies, cope with climate change and market changes, and promote the long-term development of agriculture;

[0094] (VII) The invention improves farmers’ technical level: The introduction of AI technology requires new skills and knowledge, promotes agriculture-related education and training, and improves farmers’ technical level;

[0095] (VIII) The invention enhances decision support: AI technology provides farmers with data-driven decision support, helping them better understand market and crop conditions and make more informed decisions;

[0096] (IX) This invention promotes agricultural modernization: The application of AI technology promotes the modernization and intelligence of agricultural production and improves the overall competitiveness of agriculture;

[0097] (10) The invention promotes agricultural scientific and technological progress: The development and application of AI technology promotes agricultural scientific and technological progress and promotes innovation in the agricultural field;

[0098] (XI) Data integration: The present invention integrates data from different sources and types (such as meteorological data, soil data, crop growth data, etc.) into a unified platform for comprehensive analysis and decision-making;

[0099] (XII) High degree of automation: Real-time monitoring and data collection of farmland can be achieved through the use of sensors, drones, satellite remote sensing and other technologies, reducing manual intervention and improving efficiency;

[0100] (13) Intelligent analysis: Use machine learning and artificial intelligence algorithms to conduct in-depth analysis of collected data and extract useful information, such as pest and disease prediction, soil fertility assessment, crop growth models, etc.;

[0101] (14) Accurate decision-making: Based on the analysis results, provide farmers with accurate farming advice, such as sowing time, irrigation plan, fertilization plan, pest and disease management, etc.;

[0102] (15) Optimizing resource allocation: Through simulation and optimization algorithms, help farmers rationally allocate resources, such as water, fertilizers, pesticides, etc., to minimize costs and maximize yields;

[0103] (XVI) Adaptive management: The present invention can dynamically adjust management strategies based on real-time data and forecasting models to cope with uncertainties such as climate change and market fluctuations;

[0104] (17) User-friendly interface: Providing easy-to-use interfaces and reports to enable farmers to easily access and understand complex data analysis results and make better decisions;

[0105] (XVIII) Scalability and compatibility: The present invention is flexibly designed to be compatible with existing agricultural management systems and technical equipment, and can be expanded and upgraded as technology develops and new data emerges.

[0106] (19) Environmental protection: By optimizing resource use and reducing the application of chemical substances, it helps to achieve sustainable agricultural practices and reduce the impact on the environment;

[0107] (XX) Economic Benefits: The present invention can bring significant economic benefits to farmers by increasing yields, reducing costs and mitigating risks. BRIEF DESCRIPTION OF THE DRAWINGS

[0108] The present invention is further described below in conjunction with the accompanying drawings.

[0109] Attached Figure 1 It is a flowchart of the crop management method based on the large model;

[0110] Attached Figure 2 This is a diagram of the working principle of CNN. DETAILED DESCRIPTION

[0111] The crop management method, equipment and medium based on the large model of the present invention are described in detail below with reference to the drawings and specific embodiments of the specification.

[0112] Embodiment 1:

[0113] As attached Figure 1 As shown, this embodiment provides a crop management method based on a large model, and the method is specifically as follows:

[0114] S1. Pest and disease monitoring and diagnosis: Use convolutional neural network (CNN) image recognition technology to quickly identify pests and diseases, reduce the abuse of pesticides, and improve crop health and yield;

[0115] S2. Yield prediction: AI technology is used to predict crop yields by analyzing historical and real-time data, assist in planting decisions, and optimize crop layout and harvest time;

[0116] S3. Smart irrigation: Analyze soil and crop data through decision tree models to achieve accurate application of water and fertilizer resources, reduce waste and improve resource utilization efficiency;

[0117] S4. Crop growth monitoring: Analyze crop growth data based on cluster analysis and principal component analysis to obtain crop growth trend and predict crop production;

[0118] S5. Automated operations: Use deep learning algorithms to detect and track the growth of crops and adjust agricultural machinery operation strategies.

[0119] The pest monitoring and diagnosis in step S1 of this embodiment are as follows:

[0120] S101. Crop image data collection: using sensors, cameras or drones installed in the farmland to collect crop image data, and then obtain crop growth environment data; wherein the crop growth environment data includes temperature, humidity and soil conditions;

[0121] S102, image preprocessing: preprocessing the collected images; wherein the preprocessing includes resizing, cropping and noise removal to facilitate subsequent analysis;

[0122] S103, image feature extraction: using a convolutional neural network (CNN) to automatically extract features from images to distinguish healthy crops from crops affected by pests and diseases; wherein the extracted features include texture, color and shape;

[0123] S104, large model training: using a convolutional neural network (CNN) or a support vector machine (SVM) supervised learning algorithm to train a labeled data set; wherein the labeled data set includes images of healthy crops and images of crops affected by different pests and diseases;

[0124] S105, Pest and disease identification and classification: Identify and classify pests and diseases in images through the trained large model; and use the large model to distinguish different fungal diseases, viral diseases or insect pests;

[0125] S106, Real-time monitoring and early warning: Deploy the trained large model to the farmland and conduct real-time monitoring through sensors and cameras; when potential pests and diseases are detected, an early warning is automatically issued to notify farmers to take measures;

[0126] S107. Diagnosis and recommendations: Identify pests and diseases through large models, provide diagnostic results and prevention recommendations; at the same time, based on the type and severity of pests and diseases, recommend appropriate pesticides or biological control measures.

[0127] The production prediction in step S2 of this embodiment is as follows:

[0128] S201. Climate, soil and crop growth data collection: collect historical climate data, soil data and crop growth data; among which, historical climate data include temperature, rainfall and light; soil data include soil type, pH value and organic matter content; crop growth data include plant height, leaf area and biomass;

[0129] S202. Preprocessing of climate, soil and crop growth data: cleaning and preprocessing the collected climate, soil and crop growth data, including removing missing values, outliers and standardizing or normalizing the data;

[0130] S203, growth-related feature engineering: extract relevant features based on the characteristics and needs of crop growth; relevant features include the duration of the growth stage and the climatic conditions during the key growth period;

[0131] S204, Time Series Analysis: Model and analyze time series data using recurrent neural networks (RNN) or long short-term memory networks (LSTM) to capture long-term dependencies in time series data for predicting seasonal growth trends of crops;

[0132] S205, Regression analysis: Use linear regression and support vector regression (SVR) regression algorithms in machine learning to establish a yield prediction model and predict crop yields based on input features, such as predicting crop yields based on climate and soil conditions;

[0133] S206, model training and validation: Use historical data to train the yield prediction model, and evaluate the performance of the yield prediction model through cross-validation; at the same time, adjust the network structure and learning rate related parameters of the yield prediction model to improve the prediction accuracy and generalization ability of the yield prediction model;

[0134] S207. Yield forecasting and decision support: Use the trained yield forecasting model to predict future climate and soil conditions, thereby predicting crop yields; and make more informed planting decisions based on the yield forecast results; planting decisions include selecting appropriate crop varieties and adjusting planting times.

[0135] The smart irrigation in step S3 of this embodiment is specifically as follows:

[0136] S301, soil moisture, climate conditions and crop growth data collection: collect soil moisture data, climate condition data and crop growth data; wherein soil moisture data includes real-time soil moisture content; climate condition data includes temperature, rainfall and evaporation rate; crop growth data includes crop type and growth stage;

[0137] S302, preprocessing of soil moisture, climate conditions and crop growth data: cleaning and preprocessing the collected soil moisture, climate conditions and crop growth data, the cleaning and preprocessing including removal of outliers and standardization or normalization;

[0138] S303, growth-related feature engineering: extracting relevant features according to irrigation requirements and crop growth characteristics; wherein the relevant features include soil moisture threshold and climate condition threshold;

[0139] S304, model training: using a decision tree or random forest algorithm to train relevant feature data; wherein the decision tree algorithm recursively divides the data set into subsets to establish a decision tree model for irrigation decision-making; the random forest algorithm improves the accuracy and stability of the decision tree model by constructing multiple decision trees and taking an average or majority vote;

[0140] S305, irrigation strategy formulation: formulate intelligent irrigation strategy according to soil moisture and climate conditions; wherein the irrigation strategy includes irrigation time, irrigation amount and irrigation frequency;

[0141] S306, model validation and optimization: Use the cross-validation method to evaluate the performance of the decision tree model, and adjust the decision tree model parameters and irrigation strategies according to the prediction results of the decision tree model and the actual irrigation effect, so as to improve the accuracy and practicality of the decision tree model;

[0142] S307, Deployment and application: Deploy the trained decision tree model to farmland, realize intelligent irrigation through sensors and control systems, monitor soil moisture and climate conditions in real time through the decision tree model, and automatically control the irrigation system according to the prediction results.

[0143] The crop growth monitoring in step S4 of this embodiment is specifically as follows:

[0144] S401, growth data collection: collecting crop growth data and growth environment data; wherein the crop growth data includes crop height, leaf area and biomass; the crop growth environment data includes temperature, humidity and light;

[0145] S402, growth data preprocessing: cleaning and preprocessing the collected growth data; wherein the cleaning and preprocessing include removing outliers, standardization or normalization;

[0146] S403. Cluster analysis of growth data: Use K-means or DBSCAN clustering algorithm in unsupervised learning to perform cluster analysis on crop growth data to obtain different growth patterns of crops; wherein the growth patterns include normal growth and the effects of pests and diseases;

[0147] S404. Principal component analysis (PCA) of growth data: PCA is used to reduce the dimension of growth data and extract main features. PCA is also used to reveal the changing trend in growth data, which helps understand and predict crop growth.

[0148] S405, growth pattern identification: Based on the results of cluster analysis, different growth patterns are identified, and the characteristics and influencing factors of each growth pattern are analyzed to provide a basis for crop management;

[0149] S406, abnormal growth detection: using the features extracted by principal component analysis (PCA) to detect abnormal conditions in the growth process of crops, timely discover abnormal growth, and take corresponding management measures such as adjusting irrigation or fertilization;

[0150] S407, Growth Assessment and Optimization: Based on the results of cluster analysis and principal component analysis (PCA), crop growth is assessed and crop management strategies are optimized to improve crop yield and quality.

[0151] The automation operation in step S5 of this embodiment is specifically as follows:

[0152] S501, Target Detection and Tracking: Use YOLO (You Only Look Once) or SSD (Single Shot MultiBox Detector) deep learning algorithms to detect and track crops, weeds or insects in farmland, automatically identify objects in images, and locate and classify objects; the detected target information is used to guide the automatic spraying of pesticides or automated weeding operations of agricultural machinery;

[0153] S502, automatic control of agricultural machinery: using Q-learning or DDPG (Deep Deterministic Policy Gradient) reinforcement learning algorithm to realize automatic control of agricultural machinery; wherein, the reinforcement learning algorithm learns the optimal operation strategy through interaction with the environment; during the farmland operation, the agricultural machinery automatically adjusts the operation path and operation according to the learned strategy to improve the operation efficiency and quality;

[0154] S503, real-time feedback and adjustment: During the operation, the agricultural machine receives feedback information from the target detection and tracking system in real time; and according to the feedback information, the agricultural machine adjusts the operation strategy in time to adapt to the actual situation in the farmland;

[0155] S504, agricultural machinery operation data collection and analysis: in the process of automated operation, collect agricultural machinery operation data on operation speed, path and operation quality, use machine learning algorithms to analyze agricultural machinery operation data, and optimize agricultural machinery operation strategies;

[0156] S505. Operation efficiency and quality assessment: Based on the collected operation data, the efficiency and quality of agricultural machinery operations are assessed, and the agricultural machinery operation strategy is adjusted according to the assessment results to improve the operation effect.

[0157] As attached Figure 2 As shown, the convolutional neural network (CNN) in this embodiment uses a convolutional layer to automatically and hierarchically learn features in an image. The convolutional layer extracts local features on the image through a sliding window (filter), and the local features are then passed to the next layer of the network; multiple layers of convolution and pooling (downsampling) operations are used to capture features from low-level to high-level, and finally classification or regression is performed through a fully connected layer.

[0158] The principal component analysis (PCA) method in this embodiment is a statistical method used for data dimension reduction or feature extraction; the principal component analysis (PCA) method maps the original data to a new coordinate system through linear transformation, and the new coordinate system is determined by the covariance structure of the original data;

[0159] In the new coordinate system, the first coordinate axis (first principal component) is the direction of the largest variance in the original data, the second coordinate axis (second principal component) is the direction of the second largest variance, and so on.

[0160] Embodiment 2:

[0161] This embodiment also provides an electronic device, including: a memory and at least one processor;

[0162] Wherein, the memory stores computer-executable instructions;

[0163] The at least one processor executes the computer-executable instructions stored in the memory, so that the at least one processor performs the large model-based crop management method described in any one of the present inventions.

[0164] The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The processor may be a microprocessor or any conventional processor, etc.

[0165] The memory can be used to store computer programs and / or modules. The processor realizes various functions of the electronic device by running or executing the computer programs and / or modules stored in the memory, and calling the data stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, at least one application required for a function, etc.; the data storage area can store data created according to the use of the terminal, etc. In addition, the memory can also include a high-speed random access memory, and can also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (SMC), a secure digital (SD) card, a flash memory card, at least one disk storage period, a flash memory device, or other volatile solid-state storage devices.

[0166] Embodiment 3:

[0167] This embodiment also provides a computer-readable storage medium, in which a plurality of instructions are stored, and the instructions are loaded by a processor, so that the processor executes the crop management method based on a large model in any embodiment of the present invention. Specifically, a system or device equipped with a storage medium can be provided, on which a software program code that implements the functions of any of the above embodiments is stored, and a computer (or CPU or MPU) of the system or device reads and executes the program code stored in the storage medium.

[0168] In this case, the program code itself read from the storage medium can realize the function of any one of the above-mentioned embodiments, and thus the program code and the storage medium storing the program code constitute a part of the present invention.

[0169] The storage medium embodiments for providing the program code include a floppy disk, a hard disk, a magneto-optical disk, an optical disk (such as CD-ROM, CD-R, CD-RW, DVD-ROM, DVD-RAM, DVD-RW, DVD+RW), a magnetic tape, a non-volatile memory card, and a ROM. Alternatively, the program code can be downloaded from a server computer by a communication network.

[0170] In addition, it should be clear that the functions of any of the above embodiments can be implemented not only by executing the program code read by the computer, but also by enabling an operating system operating on the computer to complete part or all of the actual operations based on instructions from the program code.

[0171] In addition, it can be understood that the program code read from the storage medium is written to a memory provided in an expansion board inserted into the computer or written to a memory provided in an expansion unit connected to the computer, and then based on the instructions of the program code, a CPU installed on the expansion board or the expansion unit is enabled to perform part or all of the actual operations, thereby realizing the functions of any of the above-mentioned embodiments.

[0172] 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 replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A crop management method based on a large model, characterized in that: The method is as follows: Pest and disease monitoring and diagnosis: Use convolutional neural network image recognition technology to quickly identify pests and diseases; Yield forecasting: AI technology is used to predict crop yields by analyzing historical and real-time data, assist in planting decisions, and optimize crop layout and harvest time; Smart irrigation: Analyze soil and crop data through decision tree models to achieve precise application of water and fertilizer resources; Crop growth monitoring: Analyze crop growth data based on cluster analysis and principal component analysis to obtain crop growth trends and predict crop production conditions; Automated operations: Use deep learning algorithms to detect and track the growth of crops and adjust agricultural machinery operation strategies.

2. The crop management method based on a large model according to claim 1, characterized in that: The details of pest and disease monitoring and diagnosis are as follows: Crop image data collection: Use sensors, cameras or drones installed in farmland to collect crop image data, and then obtain crop growth environment data; the crop growth environment data includes temperature, humidity and soil conditions; Image preprocessing: preprocess the collected images; preprocessing includes resizing, cropping and noise removal; Image feature extraction: Use convolutional neural networks to automatically extract features from images to distinguish healthy crops from crops affected by pests and diseases; the extracted features include texture, color, and shape; Large model training: Use convolutional neural networks or support vector machine supervised learning algorithms to train labeled data sets; the labeled data sets include images of healthy crops and images of crops affected by different pests and diseases; Pest and disease identification and classification: Identify and classify pests and diseases in images through a trained large model; and use the large model to distinguish different fungal diseases, viral diseases, or insect pests. Real-time monitoring and early warning: Deploy the trained large model to farmland and conduct real-time monitoring through sensors and cameras. When potential pests and diseases are detected, an early warning is automatically issued to notify farmers to take measures. Diagnosis and recommendations: Identify pests and diseases through large models, provide diagnostic results and prevention recommendations; at the same time, based on the type and severity of pests and diseases, recommend appropriate pesticides or biological control measures.

3. The crop management method based on a large model according to claim 1, characterized in that: The production forecast is as follows: Climate, soil and crop growth data collection: collect historical climate data, soil data and crop growth data; historical climate data includes temperature, rainfall and light; soil data includes soil type, pH value and organic matter content; crop growth data includes plant height, leaf area and biomass; Climate, soil and crop growth data preprocessing: Clean and preprocess the collected climate, soil and crop growth data, including removing missing values, outliers and standardizing or normalizing the data; Growth-related feature engineering: Extract relevant features based on the characteristics and needs of crop growth; relevant features include the duration of the growth stage and the climatic conditions during the key growth period; Time series analysis: Use recurrent neural networks or long short-term memory networks to model and analyze time series data, capture long-term dependencies in time series data, and use them to predict seasonal growth trends of crops; Regression analysis: Use linear regression and support vector regression algorithms in machine learning to build a yield prediction model and predict crop yields based on input features; Model training and validation: Use historical data to train the yield prediction model and evaluate its performance through cross-validation. At the same time, adjust the network structure and learning rate-related parameters of the yield prediction model to improve its prediction accuracy and generalization ability. Yield forecasting and decision support: Use the trained yield forecasting model to predict future climate and soil conditions, thereby predicting crop yields; and make more informed planting decisions based on the yield forecast results; planting decisions include selecting appropriate crop varieties and adjusting planting times.

4. The crop management method based on a large model according to claim 1, characterized in that: Smart irrigation is as follows: Soil moisture, climate conditions and crop growth data collection: collect soil moisture data, climate conditions data and crop growth data; soil moisture data includes real-time soil moisture content; climate conditions data includes temperature, rainfall and evaporation rate; crop growth data includes crop type and growth stage; Soil moisture, climate conditions and crop growth data preprocessing: Clean and preprocess the collected soil moisture, climate conditions and crop growth data, including removing outliers and standardizing or normalizing; Growth-related feature engineering: Extract relevant features based on irrigation requirements and crop growth characteristics; relevant features include soil moisture thresholds and climate condition thresholds; Model training: Use decision tree or random forest algorithm to train relevant feature data; the decision tree algorithm recursively divides the data set into subsets to build a decision tree model for irrigation decision-making; the random forest algorithm improves the accuracy and stability of the decision tree model by building multiple decision trees and taking the average or majority vote; Irrigation strategy formulation: formulate intelligent irrigation strategies based on soil moisture and climate conditions; the irrigation strategy includes irrigation time, irrigation amount and irrigation frequency; Model validation and optimization: Use the cross-validation method to evaluate the performance of the decision tree model, and adjust the decision tree model parameters and irrigation strategies based on the prediction results of the decision tree model and the actual irrigation effect to improve the accuracy and practicality of the decision tree model; Deployment and application: Deploy the trained decision tree model to farmland, realize intelligent irrigation through sensors and control systems, monitor soil moisture and climate conditions in real time through the decision tree model, and automatically control the irrigation system based on the prediction results.

5. The crop management method based on a large model according to claim 1, characterized in that: Crop growth monitoring is as follows: Growth data collection: collect crop growth data and growth environment data; crop growth data includes crop height, leaf area and biomass; crop growth environment data includes temperature, humidity and light; Growth data preprocessing: cleaning and preprocessing the collected growth data; cleaning and preprocessing include removing outliers, standardization or normalization; Cluster analysis of growth data: Use K-means or DBSCAN clustering algorithms in unsupervised learning to cluster crop growth data and obtain different growth patterns of crops; growth patterns include normal growth and those affected by pests and diseases; Principal component analysis of growth data: principal component analysis is used to reduce the dimension of growth data and extract the main features. At the same time, principal component analysis is used to reveal the changing trends in growth data, helping to understand and predict crop growth; Growth pattern recognition: Based on the results of cluster analysis, different growth patterns are identified, and the characteristics and influencing factors of each growth pattern are analyzed to provide a basis for crop management; Abnormal growth detection: Use the features extracted by principal component analysis to detect abnormalities in the growth process of crops, discover abnormal growth in time, and take corresponding management measures such as adjusting irrigation or fertilization; Growth assessment and optimization: Based on the results of cluster analysis and principal component analysis, crop growth is assessed and crop management strategies are optimized to improve crop yield and quality.

6. The crop management method based on a large model according to any one of claims 1 to 5, characterized in that: The automation operations are as follows: Target detection and tracking: Use YOLO or SSD deep learning algorithms to detect and track crops, weeds or insects in farmland, automatically identify objects in images, and locate and classify them; the detected target information is used to guide the automatic spraying of pesticides or automated weeding operations of agricultural machinery; Automatic control of agricultural machinery: Use Q-learning or DDPG reinforcement learning algorithms to realize automatic control of agricultural machinery. The reinforcement learning algorithm learns the optimal operation strategy through interaction with the environment. During the farmland operation, the agricultural machinery automatically adjusts the operation path and operation according to the learned strategy to improve the operation efficiency and quality. Real-time feedback and adjustment: During operation, the agricultural machinery receives feedback from the target detection and tracking system in real time. Based on the feedback, the agricultural machinery adjusts its operation strategy in a timely manner to adapt to the actual situation in the farmland. Agricultural machinery operation data collection and analysis: During the automated operation process, collect agricultural machinery operation data on operation speed, path and operation quality, use machine learning algorithms to analyze agricultural machinery operation data, and optimize agricultural machinery operation strategies; Operation efficiency and quality assessment: Based on the collected operation data, the efficiency and quality of agricultural machinery operations are assessed, and the agricultural machinery operation strategy is adjusted according to the assessment results to improve the operation effect.

7. The crop management method based on a large model according to claim 1, characterized in that: Convolutional neural networks use convolutional layers to automatically and hierarchically learn features in images. The convolutional layer extracts local features on the image through a sliding window, and the local features are then passed to the next layer of the network; multiple layers of convolution and pooling operations are used to capture features from low-level to high-level, and finally classification or regression is performed through a fully connected layer.

8. The crop management method based on a large model according to claim 5, characterized in that: The principal component analysis method is a statistical method used for data dimensionality reduction or feature extraction. The principal component analysis method maps the original data into a new coordinate system through linear transformation. The new coordinate system is determined by the covariance structure of the original data. In the new coordinate system, the first axis is the direction of the largest variance in the original data, the second axis is the direction of the second largest variance, and so on.

9. An electronic device, characterized in that: include: memory and at least one processor; Wherein, the memory stores computer-executable instructions; The at least one processor executes the computer-executable instructions stored in the memory, so that the at least one processor performs the large model-based crop management method according to any one of claims 1 to 8.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, and when the processor executes the computer-executable instructions, the large model-based crop management method as described in any one of claims 1 to 8 is implemented.

Citation Information

Cited By

  • Intelligent optimization management method for stereoscopic cultivation of vegetables

    CN119378760A

  • Accurate crop management method and system based on crop growth model

    CN119417639A

  • Crop yield prediction method and system based on multi-data fusion

    CN119558484A

  • A method and system for predicting crop yield based on multi-data fusion

    CN119558484B

  • Big data analysis platform based on smart agriculture

    CN119886871A