Informatization management system based on AI agricultural environment data analysis

By introducing AI and Internet of Things technology in agricultural management, real-time monitoring and analysis of farmland environmental data, the problems of low efficiency and waste of resources in traditional agricultural management are solved, precise agricultural operations and risk warning are achieved, and the stability of agricultural production is improved. Sustainability.

CN119048028AInactive Publication Date: 2024-11-29FUZHOU CHANGYU AGRI DEV CO LTD
View PDF 4 Cites 0 Cited by

Patent Information

Application Number
CN202411529225.6
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-10-30
Publication Date
2024-11-29
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

The backward traditional agricultural management technology leads to low production efficiency, insufficient resource utilization, weak risk resistance, and easy to cause waste of water resources, soil pollution and ecological damage.

Method used

Design an AI-based agricultural environment data analysis information management system, including cloud data center, user interaction and data analysis terminal, using mobile management terminal, device management module, environmental real-time monitoring module and crop growth analysis module, through AI and IoT technology monitors and analyzes farmland environmental data in real time to realize intelligent irrigation, fertilization and insect removal operations.

Benefits of technology

It improves agricultural production efficiency, achieves precise agricultural operations, reduces resource waste, improves crop yield and quality, enhances the stability and sustainable development capabilities of agricultural production, and can early warning of natural disasters and pest risks.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119048028A_ABST
    Figure CN119048028A_ABST
Patent Text Reader

Abstract

The invention discloses an informatization management system based on AI agricultural environment data analysis, and relates to the technical field of agricultural management, the informatization management system comprises a cloud data center, a user interaction end and a data analysis end, the user interaction end comprises a mobile management terminal and an equipment management module; the data analysis terminal comprises an environment real-time monitoring module and a crop growth analysis module, and the mobile management terminal is used for being carried by a farm manager and checking and processing various types of data in time when farm crops are managed in real time; the equipment management module is used for controlling and managing various informatization equipment in the farmland informatization management system, and the environment real-time monitoring module is used for collecting various data of a crop planting and cultivating area for informatization management and uploading the data to the cloud data center. According to the invention, the agricultural production efficiency can be improved, the resource distribution can be optimized, intelligent monitoring and early warning can be realized, the management efficiency can be improved, and the intelligent and sustainable development of agriculture can be promoted.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention mainly relates to the technical field of agricultural management, and specifically to an information management system based on AI agricultural environmental data analysis. Background Art

[0002] Informatization of agricultural management is an important trend in the development of modern agriculture. It uses information technology to comprehensively upgrade agricultural production, operation, and management in order to improve agricultural production efficiency, optimize resource allocation, ensure the quality of agricultural products, and promote sustainable agricultural development. With the rapid development of technologies such as the Internet of Things, remote sensing technology, and sensors, a large amount of environmental data is generated in the agricultural production process, including meteorological data, soil data, and plant growth data. These data provide a solid foundation for agricultural big data analysis. Agricultural big data analysis can reveal the inherent laws of the agricultural production process and provide scientific decision-making support for agricultural production by collecting, storing, processing, and analyzing these data.

[0003] The following problems exist in traditional agricultural management: 1. Backward technology: Traditional agricultural management often relies on manual labor and simple farm tools, and lacks the application of modern science and technology, which leads to low production efficiency, insufficient resource utilization, and increased labor intensity; 2. Low resource utilization efficiency: In terms of irrigation, fertilization, pesticide use, etc., traditional agricultural management often lacks scientific planning, which easily leads to waste of water resources, soil pollution and ecological damage; excessive use of fertilizers and pesticides not only increases production costs, but may also have a negative impact on the environment and human health; 3. Weak risk resistance: Traditional agricultural management often lacks effective prevention and response measures when facing risks such as natural disasters, pests and diseases, resulting in poor stability in agricultural production and large fluctuations in farmers' income. Summary of the invention

[0004] Based on this, the purpose of the present invention is to provide an information management system based on AI agricultural environmental data analysis to solve the technical problems raised in the above background technology.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] An information management system based on AI agricultural environmental data analysis, comprising a cloud data center, a user interaction terminal connected to the cloud data center network, and a data analysis terminal, wherein the user interaction terminal comprises a mobile management terminal and an equipment management module; the data analysis terminal comprises an environmental real-time monitoring module and a crop growth analysis module;

[0007] The mobile management terminal is used by farm managers to carry and view and process various types of data in a timely manner when managing farm crops in real time;

[0008] The equipment management module is used to control and manage various types of information equipment in the farmland information management system, including irrigation system equipment, automatic fertilizer spreaders, various sensor equipment, and remote-controlled drones;

[0009] The real-time environmental monitoring module is used to collect various data of the crop planting and cultivation area for information management, and upload the data to the cloud data center; the data include meteorological information, soil data, and pest distribution;

[0010] The crop growth analysis module is used to analyze the various data of the crop planting and cultivation area obtained by the real-time environmental monitoring module, and send the analysis results to the cloud data center in real time;

[0011] The real-time environmental monitoring module includes a climate detection unit, a soil detection unit, and a pest detection unit;

[0012] The pest detection unit is used to obtain the pest distribution of the entire crop planting and cultivation area in real time through a remote-controlled drone and a YOLO algorithm, and upload the pest distribution to the cloud data center;

[0013] The YOLO algorithm can be simplified into a function that combines coordinate loss, confidence loss and category probability loss in the detection of crop pest quantity. The coordinate loss uses square error to calculate the difference between the center coordinates and size of the predicted bounding box and the true bounding box. The confidence loss is indirectly calculated through the error of the bounding box coordinates. The category probability loss uses multi-category cross entropy loss to calculate the difference between the predicted category probability distribution and the true category.

[0014] Preferably, the cloud data center is used to record and retain various categories of data from the user interaction end and the data analysis end; the various categories of data include equipment information, meteorological information, soil data, pest distribution, crop maintenance tasks, and analysis results.

[0015] Preferably, the mobile management terminal is specifically a movable terminal management device with an image display function, the mobile management terminal is connected to the cloud data center through a network, and receives and downloads various types of data from the cloud data center in real time;

[0016] The mobile management terminal includes a data download unit and a device status display unit;

[0017] The data download unit is used to download the latest data of each category at the cloud data center in real time, and overwrite the previous data of each category of the same set, so as to ensure that the storage status of each category of data of the device is optimal and up-to-date;

[0018] The equipment status display unit is used to display the equipment information of the irrigation system equipment, automatic fertilizer spreader, various sensor devices, and remote-controlled drones, so as to facilitate the administrator to view and manage; the equipment information includes working time, standby time, machine battery status, fault conditions, and location distribution coordinates; various sensor devices specifically include soil temperature sensors, soil PH sensors, soil moisture sensors, and soil nitrogen, phosphorus and potassium sensors.

[0019] Preferably, the device management module includes a device control unit and a device monitoring unit;

[0020] The equipment control unit has the highest management authority and is used to remotely control and manage the operation and standby of the irrigation system equipment, the automatic fertilizer applicator, various sensor devices, and the remote-controlled drone;

[0021] The equipment monitoring unit is used to obtain equipment information of the irrigation system equipment, automatic fertilizer spreader, various sensor devices, and remote-controlled drones; and send the equipment information to the cloud data center in real time.

[0022] Preferably, the climate detection unit is used to obtain the latest regional weather information through a regional weather station network, the weather information including temperature, humidity, wind speed, wind direction, and precipitation; and upload the weather information to the cloud data center;

[0023] The soil detection unit is used to monitor and obtain soil data through soil temperature sensors, soil pH sensors, soil moisture sensors, and soil nitrogen, phosphorus and potassium sensors set at different locations in the crop planting and cultivation area. The soil data includes soil temperature, soil moisture, fertility, and soil acidity and alkalinity; and the soil data is uploaded to the cloud data center.

[0024] Preferably, the YOLO algorithm assumes that each pest has a bounding box, a confidence score, and a series of category probabilities; the specific simplified expression is as follows:

[0025] in, : represents the sum of the loss values ​​of the comprehensive coordinates, confidence, and category probability; : represents the number of pests detected in the image; : is the coordinate loss of the i-th pest, measuring the position difference between the predicted bounding box and the true bounding box; : is the confidence loss of the i-th pest, including the confidence of the existence of the target in the bounding box and the confidence of the intersection and union ratio of the bounding box and the true bounding box; : is the category probability loss of the i-th pest, measuring the difference between the predicted category probability distribution and the true category probability distribution.

[0026] Preferably, for the prediction of a single crop planting and cultivation area, the square error can be expressed as: :yes The predicted pest coordinate values, : is the true value of the pest coordinate data;

[0027] The multi-category cross entropy loss expression is as follows:

[0028] in, :express The difference between the predicted value of a class and the true value of the class; : is the i-th pest in the one-hot encoding representation of the true label. It is a binary variable that takes the value of 1 if and only if the sample belongs to the category of the i-th pest, otherwise it takes the value of 0. In the pest quantity detection, if the pest in a grid cell is a second-category pest, then in the corresponding true label vector, the second element will be 1, and the remaining elements will be 0. :express Predict the probability of the category of the i-th pest; :express Predict the total amount of pest samples.

[0029] Preferably, the crop growth analysis module includes a daily maintenance unit and an insect pest analysis unit;

[0030] The daily maintenance unit is used to formulate crop maintenance tasks for crops in the crop planting and cultivation area according to the meteorological information and soil data detected and acquired by the real-time environmental monitoring module; the crop maintenance tasks include quantitative irrigation, soil fertility adjustment, cultivation area temperature adjustment, and soil pH adjustment;

[0031] The pest analysis unit is used to process and analyze the pest distribution using a Logistic growth model, and upload the analysis results to the cloud data center.

[0032] Preferably, the Logistic growth model is used to analyze the change in the number of pests under the condition of limited resources in crop planting and cultivation areas. The specific equation expression of the Logistic growth model is as follows:

[0033] in, : represents the relative value of pest population at a certain moment; : represents time; : is the pest population growth rate parameter, which indicates the maximum growth rate when resources are not limited; : Indicates the rate of change of population size over time.

[0034] Preferably, the Logistic growth model equation is already in explicit form with respect to x and t, and the right side is a function of x, so there is no need to further separate the variables, and the solution is as follows:

[0035] in, is a constant; r: is the pest population growth rate parameter, which indicates the maximum growth rate when resources are not limited; when r>0, the pest population will increase; when r<0, the pest population will not increase, and the case of r<0 is not considered here; : describes the temporal trends in pest populations under conditions of limited resources; : represents a point in time before the number of pests begins to increase significantly. In the Logistic growth model, The position of the Logistic growth model curve on the time axis was determined so that the Logistic growth model curve could correctly reflect the time point when the number of pests changes from slow growth to rapid growth.

[0036] In summary, the present invention mainly has the following beneficial effects:

[0037] This system improves agricultural production efficiency and realizes precision agricultural operations. Through AI technology, the system can realize intelligent fertilization, irrigation and pest control precision agricultural operations based on environmental data analysis, reduce resource waste, and significantly improve crop yield and quality; data visualization, the system uses charts and maps to display farmland data and crop growth conditions, so that farmers can intuitively understand the farmland conditions and make scientific decisions and adjustments;

[0038] This system enables efficient use of resources. AI technology can accurately calculate soil nutrient and water requirements, achieve precise application of water resources and fertilizers, reduce resource waste, and promote sustainable agricultural development.

[0039] At the same time, it can achieve intelligent monitoring and early warning, real-time monitoring, and use AI and IoT technology to monitor the farmland environment in real time, including key indicators such as soil temperature, soil moisture, fertility, soil pH, etc., to ensure the stability of the agricultural production environment;

[0040] The early warning system can predict weather changes in temperature, humidity, wind speed, wind direction, precipitation and crop growth status, and issue early warning information in a timely manner to help managers respond to risks such as natural disasters and pests and diseases in advance and reduce losses. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 It is the overall structural framework diagram of the present invention;

[0042] Figure 2 This is a specific structural framework diagram of the mobile management terminal of the present invention;

[0043] Figure 3 This is a specific structural framework diagram of the device management module of the present invention;

[0044] Figure 4 This is a specific structural framework diagram of the environment real-time monitoring module of the present invention;

[0045] Figure 5 This is a specific structural framework diagram of the crop growth analysis module of the present invention;

[0046] Figure 6 It is a schematic diagram of the regional division and equipment distribution of the soybean planting and cultivation area of ​​the present invention;

[0047] Figure 7 This is a schematic diagram of pest distribution in the soybean planting and cultivation area of ​​the present invention.

[0048] Description of the drawings: 10. Cloud data center; 20. User interaction terminal; 30. Data analysis terminal; 21. Mobile management terminal; 22. Equipment management module; 31. Real-time environmental monitoring module; 32. Crop growth analysis module; 211. Data download unit; 212. Equipment status display unit; 221. Equipment control unit; 222. Equipment monitoring unit; 311. Climate detection unit; 312. Soil detection unit; 313. Pest detection unit; 321. Daily maintenance unit; 322. Pest analysis unit. DETAILED DESCRIPTION

[0049] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0050] Example

[0051] like Figure 1 , 4 , 5, 6, and 7, the pest detection unit 313 is used to obtain the overall pest distribution of the crop planting and cultivation area in real time through a remote-controlled drone and the YOLO algorithm, and upload the pest distribution to the cloud data center 10; the YOLO algorithm can be simplified into a function that combines coordinate loss, confidence loss, and category probability loss in the detection of crop pest quantity, assuming that each pest consists of a bounding box, a confidence score, and a series of category probabilities; the specific simplified expression is as follows:

[0052] in, : represents the sum of the loss values ​​of the comprehensive coordinates, confidence, and category probability; : represents the number of pests detected in the image; : is the coordinate loss of the i-th pest, measuring the position difference between the predicted bounding box and the true bounding box; : is the confidence loss of the i-th pest, including the confidence of the existence of the target in the bounding box and the confidence of the intersection and union ratio of the bounding box and the true bounding box; : is the category probability loss of the i-th pest, measuring the difference between the predicted category probability distribution and the true category probability distribution; the coordinate loss uses the squared error To calculate the difference between the center coordinates and size of the predicted bounding box and the true bounding box;

[0053] For the prediction of a single crop planting and cultivation area, the square error It can be expressed as: :yes The predicted pest coordinate values, : is the true value of the pest coordinate data;

[0054] The confidence loss is calculated indirectly through the error of the bounding box coordinates;

[0055] Class probability loss uses multi-class cross entropy loss To calculate the difference between the predicted category probability distribution and the true category; multi-category cross entropy loss The expression is as follows:

[0056] in, :express The difference between the predicted value of a class and the true value of the class; : is the i-th pest in the one-hot encoding representation of the true label. It is a binary variable that takes the value of 1 if and only if the sample belongs to the category of the i-th pest, otherwise it takes the value of 0. In the pest quantity detection, if the pest in a grid cell is a second-class pest, then the second element in the corresponding true label vector is will be 1, and the rest of the elements will be 0; :express Predict the probability of the i-th pest category; the crop growth analysis module 32 includes a daily maintenance unit 321 and a pest analysis unit 322; the pest analysis unit 322 is used to process and analyze the pest distribution using a Logistic growth model, and upload the analysis results to the cloud data center 10; :express Predict the total amount of pest samples.

[0057] The Logistic growth model is used to analyze the changes in the number of pests under the conditions of limited resources in crop planting and cultivation areas. The specific equation expression of the Logistic growth model is as follows:

[0058] in, : represents the relative value of pest population at a certain moment; : represents time; : is the pest population growth rate parameter, which indicates the maximum growth rate when resources are not limited; : Indicates the rate of change of population size over time; The logistic growth model equation is already in explicit form with respect to x and t, and the right side is a function of x, so there is no need to further separate the variables, and the solution is as follows:

[0059] in, is a constant; r: is the pest population growth rate parameter, which indicates the maximum growth rate when resources are not limited; when r>0, the pest population will increase; when r<0, the pest population will not increase, and the case of r<0 is not considered here; : describes the temporal trends in pest populations under conditions of limited resources; : represents a point in time before the number of pests begins to increase significantly. In the Logistic growth model, The position of the Logistic growth model curve on the time axis was determined so that the Logistic growth model curve could correctly reflect the time point when the number of pests changes from slow growth to rapid growth.

[0060] It should be noted that, in this embodiment, the pest detection unit 313 detects the number of pests in soybean growth by using the YOLO algorithm. The underlying principle is to obtain image information of the soybean cultivation area by remote-controlled drones, and extract images of pests from the image information, thereby calculating the number of pests per mu in the soybean planting and cultivation area c.

[0061] When calculating, we set up a simple execution script, which is as follows: import torch from models.experimental import attempt_load from utils.general import non_max_suppression, scale_coords from utils.augmentations import letterbox from PIL import Image # Load the model model = attempt_load('yolov5s.pt', map_location='cuda:0') # or 'yolov5s.pt', 'yolov5m.pt', 'yolov5l.pt', 'yolov5x.pt' model.eval() # Read the image img_path = 'path_to_soybean_leaf_image.jpg' img = Image.open(img_path) img = letterbox(img, new_shape=640)[0] img = img[:, :, ::-1].transpose(2, 0, 1) # BGR to RGB, to 3x416x416 img = np.ascontiguousarray(img) # Reasoning pred = model(img[None])[0] pred = non_max_suppression(pred, 0.4, 0.5, classes=None, agnostic=False) # Count the number of pests num_pests = len(pred) # ... (subsequent processing) d is the area in area b where the pest distribution is more serious;

[0062] After extracting information on various types and quantities of pests, the system processes and analyzes the distribution of pests through the Logistic growth model in the pest analysis unit 322. In the Logistic growth model, The position of the Logistic growth model curve on the time axis was determined so that the Logistic growth model curve could correctly reflect the time point when the number of pests changes from slow growth to rapid growth;

[0063] When the analysis result is that the soybean aphid category has reached a warning of rapid growth, the screen of the mobile management terminal 21 displays "category one hazard, measures: use ethyl spinetoram for prevention and control";

[0064] When the analysis result is that the red spider category has reached a warning of rapid growth, the screen of the mobile management terminal 21 displays "Category II hazard, measures: use spinetoram for prevention and treatment";

[0065] When the analysis result is a warning that the leaf-eating pest category has reached a rapid growth, the mobile management terminal 21 screen displays "category three hazards, measures: use chlorantraniliprole for prevention and control";

[0066] When the analysis result is that the bean moth category has reached a warning of rapid growth, the screen of the mobile management terminal 21 displays "category 4 harm, measures: use 1000 times diluted 50% phoxim emulsion for prevention and control";

[0067] When the analysis result is that the soybean bridge-building insect category has reached a warning of rapid growth, the screen of the mobile management terminal 21 displays "Category 5 hazard, measures: use 20% cypermethrin emulsifiable concentrate for prevention and control."

[0068] like Figure 1 , 2 , 3, including a cloud data center 10, a user interaction terminal 20 connected to the cloud data center 10, and a data analysis terminal 30, the user interaction terminal 20 includes a mobile management terminal 21, and a device management module 22; the data analysis terminal 30 includes an environment real-time monitoring module 31, and a crop growth analysis module 32;

[0069] The mobile management terminal 21 is used by farm managers to carry around and view and process various types of data in real time when managing farm crops;

[0070] The equipment management module 22 is used to control and manage various types of information equipment in the farmland information management system, including irrigation system equipment, automatic fertilizer applicators, various sensor equipment, and remote-controlled drones;

[0071] The real-time environmental monitoring module 31 is used to collect various data of the crop planting and cultivation area for information management, and upload the data to the cloud data center 10; the data includes meteorological information, soil data, and pest distribution;

[0072] The crop growth analysis module 32 is used to analyze various data of the crop planting and cultivation area obtained by the real-time environment monitoring module 31, and send the analysis results to the cloud data center 10 in real time;

[0073] The cloud data center 10 is used to record and store various types of data from the user interaction terminal 20 and the data analysis terminal 30; the various types of data include equipment information, weather information, soil data, pest distribution, crop maintenance tasks, and analysis results;

[0074] The mobile management terminal 21 is specifically a mobile terminal management device with an image display function. The mobile management terminal 21 is connected to the cloud data center 10 through a network, and receives and downloads various types of data from the cloud data center 10 in real time;

[0075] The mobile management terminal 21 includes a data download unit 211 and a device status display unit 212;

[0076] The data download unit 211 is used to download the latest time point of each category of data in the cloud data center 10 in real time, and overwrite the previous same set of each category of data to ensure that the storage status of each category of data of the device is optimal and up-to-date;

[0077] The equipment status display unit 212 is used to display the equipment information of the irrigation system equipment, the automatic fertilizer spreader, various sensor equipment, and the remote-controlled drone, so as to facilitate the administrator to view and manage; the equipment information includes working time, standby time, machine battery status, fault conditions, and location distribution coordinates;

[0078] The device management module 22 includes a device control unit 221 and a device monitoring unit 222;

[0079] The equipment control unit 221 has the highest management authority and is used to remotely control and manage the irrigation system equipment, the automatic fertilizer spreader, various sensor devices, and the operation and standby of the remote-controlled drone; the various sensor devices specifically include soil temperature sensors, soil pH sensors, soil moisture sensors, and soil nitrogen, phosphorus, and potassium sensors;

[0080] The equipment monitoring unit 222 is used to obtain equipment information of irrigation system equipment, automatic fertilizer applicators, various sensor devices, and remote-controlled drones; and send the equipment information to the cloud data center 10 in real time.

[0081] It should be noted that, in this embodiment, in daily maintenance and management, when the weather temperature is high, the soil moisture content is low, and the temperature of the soybean cultivation area exceeds the threshold, the mobile management terminal 21 will display a red warning message, and a pop-up window on the screen will display "The temperature in the cultivation area is too high. It is recommended to enable the sprinkler irrigation equipment. The start time: 1h, the interval time: 1h, and the number of executions: 3". The administrator checks the warning information and adjusts or directly executes the current suggestion according to the actual situation;

[0082] Furthermore, when the administrator judges based on his own experience that the red warning information needs to be adjusted, he can issue a task to the remote-controlled drone, and the remote-controlled drone will fly out to confirm the weather conditions and temperature status of the soybean planting and cultivation area c, thereby ensuring that the administrator's daily maintenance adjustments are more accurate and effective.

[0083] like Figure 1 , 2 As shown in FIG. 3 , the real-time environment monitoring module 31 includes a climate detection unit 311 , a soil detection unit 312 , and a pest detection unit 313 ;

[0084] The climate detection unit 311 is used to obtain the latest regional weather information through the regional weather station network, the weather information including temperature, humidity, wind speed, wind direction, and precipitation; and upload the weather information to the cloud data center 10;

[0085] The soil detection unit 312 is used to monitor and obtain soil data through soil temperature sensors, soil pH sensors, soil moisture sensors, and soil nitrogen, phosphorus, and potassium sensors set at different locations in the crop planting and cultivation area. The soil data includes soil temperature, soil moisture, fertility, and soil acidity and alkalinity; and upload the soil data to the cloud data center 10;

[0086] The daily maintenance unit 321 is used to formulate crop maintenance tasks for crops in the crop planting and cultivation area according to the meteorological information and soil data detected and acquired by the real-time environmental monitoring module 31; the crop maintenance tasks include quantitative irrigation, soil fertility adjustment, cultivation area temperature adjustment, and soil pH adjustment.

[0087] It should be noted that in this embodiment, in the early stage of information management, various equipment needs to be placed in the 500-mu soybean planting and cultivation area c. The main equipment to be controlled includes irrigation system equipment, automatic fertilizer applicators, various sensor equipment, and remote-controlled drones;

[0088] Among them, after the irrigation system equipment and the automatic fertilizer spreader are installed, the automatic fertilizer spreader and the irrigation system are fully debugged to check whether each part is operating normally and whether the irrigation system has problems such as leakage and blockage; in this soybean planting and cultivation area c, the sprinkler irrigation method is used to evenly irrigate soybean crops, improve irrigation efficiency, and increase air humidity, which is conducive to the growth of soybeans. After the arrangement is completed, the irrigation system equipment and the automatic fertilizer spreader are connected to the network through the mobile management terminal 21, and the equipment information including working time, standby time, machine battery status, fault conditions, and location distribution coordinates are recorded;

[0089] The soil detection unit 312 arranges various sensor devices, including soil temperature sensors, soil pH sensors, soil moisture sensors, and soil nitrogen, phosphorus, and potassium sensors, in a 500-mu soybean planting and cultivation area c to obtain information such as soil temperature, soil moisture, fertility, and soil acidity and alkalinity in the soybean planting and cultivation area c; the mobile management terminal 21 is also used to connect and record various sensor devices, especially coordinate information;

[0090] The remote-controlled drone can be placed indoors in the cultivation area without fixed location requirements. The remote-controlled drone can be registered and connected to the Internet through the mobile management terminal 21 to achieve real-time control and maintain a state where maintenance tasks can be performed at any time.

[0091] The climate detection unit 311 obtains the latest meteorological information from the local area meteorological station network at three time periods every day, including temperature, humidity, wind speed, wind direction, and precipitation. The mobile management terminal 21 obtains meteorological information and soil information in the soybean planting and cultivation area c, and based on the soybean growth conditions, performs real-time control to maintain the best growth environment for soybean growth.

[0092] The working principle of the present invention is:

[0093] In the application of the information management system of the present invention, a 500-mu soybean planting and cultivation area C is used as a management case; Figure 6 b shown is a uniformly divided area;

[0094] In the early stage of information management, it is necessary to place various equipment in the 500-acre soybean planting and cultivation area C. The main equipment to be controlled includes irrigation system equipment, automatic fertilizer applicators, various sensor equipment, and remote-controlled drones;

[0095] Among them, after the irrigation system equipment and the automatic fertilizer spreader are installed, the automatic fertilizer spreader and the irrigation system are fully debugged to check whether each part is operating normally and whether the irrigation system has problems such as leakage and blockage; in this soybean planting and cultivation area c, the sprinkler irrigation method is used to evenly irrigate soybean crops, improve irrigation efficiency, and increase air humidity, which is conducive to the growth of soybeans. After the arrangement is completed, the irrigation system equipment and the automatic fertilizer spreader are connected to the network through the mobile management terminal 21, and the equipment information including working time, standby time, machine battery status, fault conditions, and location distribution coordinates are recorded;

[0096] In the arrangement of various sensor devices in the soil detection unit 312, soil temperature sensors, soil PH sensors, soil moisture sensors, and soil nitrogen, phosphorus and potassium sensors are evenly arranged in the 500-mu soybean planting and cultivation area c to obtain information such as soil temperature, soil moisture, fertility, and soil acidity and alkalinity in the soybean planting and cultivation area c; the mobile management terminal 21 is also used to connect and record various sensor devices, especially coordinate information; Figure 6 The a shown is the arrangement point;

[0097] The remote-controlled drone can be placed indoors in the cultivation area without fixed location requirements. The remote-controlled drone can be registered and connected to the Internet through the mobile management terminal 21 to achieve real-time control and maintain a state where maintenance tasks can be performed at any time.

[0098] The climate detection unit 311 obtains the latest meteorological information from the local area meteorological station network at three time periods every day, including temperature, humidity, wind speed, wind direction, and precipitation. The mobile management terminal 21 obtains meteorological information and soil information in the soybean planting and cultivation area c, and performs real-time control based on the soybean growth conditions to maintain the best growth environment for soybean growth; in daily maintenance and management, when the weather temperature is high, the soil moisture content is low, and the temperature in the soybean cultivation area exceeds the threshold, the mobile management terminal 21 will display a red warning message, and the screen pop-up window will display "The temperature in the cultivation area is too high. It is recommended to enable the sprinkler equipment. The start time: 1h, the interval time: 1h, the number of executions: 3" The administrator checks the warning information and adjusts or directly executes the current suggestion according to the actual situation;

[0099] Furthermore, when the administrator determines based on his own experience that the red warning information needs to be adjusted, he can issue a task to the remote-controlled drone, and the remote-controlled drone will fly out to the soybean planting and cultivation area C to confirm the weather conditions and the temperature status of the soybean planting and cultivation area C, thereby ensuring that the administrator's daily maintenance and adjustment are more accurate and effective;

[0100] In the process of soybean growth, the most difficult to control is insect pests. There are many kinds of pests that affect soybean growth. Among them, the pests that have the greatest impact on soybean plant leaves include soybean aphids, red spiders, leaf-feeding pests, bean hawk moths, and soybean bridge-building insects.

[0101] In this system, the pest detection unit 313 detects the number of pests in soybean growth by using the YOLO algorithm. The underlying principle is to obtain image information of the soybean cultivation area by remote-controlled drones, and extract images of pests from the image information, thereby calculating the number of pests per mu in the soybean cultivation area c.

[0102] When calculating, we set up a simple execution script, which is as follows: import torch from models.experimental import attempt_load from utils.general import non_max_suppression, scale_coords from utils.augmentations import letterbox from PIL import Image # Load the model model = attempt_load('yolov5s.pt', map_location='cuda:0') # or 'yolov5s.pt', 'yolov5m.pt', 'yolov5l.pt', 'yolov5x.pt' model.eval() # Read the image img_path = 'path_to_soybean_leaf_image.jpg' img = Image.open(img_path) img = letterbox(img, new_shape=640)[0] img = img[:, :, ::-1].transpose(2, 0, 1) # BGR to RGB, to 3x416x416 img = np.ascontiguousarray(img) # Reasoning pred = model(img[None])[0] pred = non_max_suppression(pred, 0.4, 0.5, classes=None, agnostic=False) # Count the number of pests num_pests = len(pred) # ... (subsequent processing)

[0103] like Figure 7 As shown, d is the area in area b where the pest distribution is more serious;

[0104] After extracting information on various types and quantities of pests, the system processes and analyzes the distribution of pests through the Logistic growth model in the pest analysis unit 322. In the Logistic growth model, The position of the Logistic growth model curve on the time axis was determined so that the Logistic growth model curve could correctly reflect the time point when the number of pests changes from slow growth to rapid growth;

[0105] When the analysis result is that the soybean aphid category has reached a warning of rapid growth, the screen of the mobile management terminal 21 displays "category one hazard, measures: use ethyl spinetoram for prevention and control";

[0106] When the analysis result is that the red spider category has reached a warning of rapid growth, the screen of the mobile management terminal 21 displays "Category II hazard, measures: use spinetoram for prevention and treatment";

[0107] When the analysis result is a warning that the leaf-eating pest category has reached a rapid growth, the mobile management terminal 21 screen displays "category three hazards, measures: use chlorantraniliprole for prevention and control";

[0108] When the analysis result is that the bean moth category has reached a warning of rapid growth, the screen of the mobile management terminal 21 displays "category 4 damage, measures: use 1000 times diluted 50% phoxim emulsion for prevention and control";

[0109] When the analysis result is that the soybean bridge-building insect category has reached a warning of rapid growth, the screen of the mobile management terminal 21 displays "category five hazards, measures: use 20% cypermethrin emulsifiable concentrate for prevention and control";

[0110] It should be further explained that this system only uses insect pests as an example. The YOLO algorithm and Logistic growth model of this system can also analyze the leaf color of the cultivated plants in the cultivation area to determine the growth status of the plants, which will not be elaborated here.

[0111] The above embodiments are only for illustrating the technical idea of ​​the present invention, and cannot be used to limit the protection scope of the present invention. Any changes made on the basis of the technical solution in accordance with the technical idea proposed by the present invention shall fall within the protection scope of the present invention.

Claims

1. An information management system based on AI agricultural environmental data analysis, comprising a cloud data center (10), a user interaction terminal (20) connected to the cloud data center (10) through a network, and a data analysis terminal (30), characterized in that: The user interaction end (20) comprises a mobile management terminal (21) and a device management module (22); the data analysis end (30) comprises an environmental real-time monitoring module (31) and a crop growth analysis module (32); The mobile management terminal (21) is used to be carried by farm managers to timely check and process various types of data when managing farm crops in real time; The equipment management module (22) is used to control and manage various types of information equipment in the farmland information management system, and the various types of information equipment include irrigation system equipment, automatic fertilizer spreaders, various types of sensor equipment, and remote-controlled drones; The real-time environmental monitoring module (31) is used to collect various data of the crop planting and cultivation area for information management, and upload the various data to the cloud data center (10); the various data include meteorological information, soil data, and pest distribution; The crop growth analysis module (32) is used to analyze various data of the crop planting and cultivation area acquired by the real-time environmental monitoring module (31), and send the analysis results to the cloud data center (10) in real time; The real-time environmental monitoring module (31) comprises a climate detection unit (311), a soil detection unit (312), and a pest detection unit (313); The pest detection unit (313) is used to obtain the pest distribution situation of the entire crop planting and cultivation area in real time through a remote-controlled drone and a YOLO algorithm, and upload the pest distribution situation to the cloud data center (10); The YOLO algorithm can be simplified into a function that combines coordinate loss, confidence loss, and category probability loss in the detection of crop pest quantity; the coordinate loss uses square error to calculate the difference between the center coordinates and size of the predicted bounding box and the true bounding box; the confidence loss is indirectly calculated through the error of the bounding box coordinates; the category probability loss uses multi-category cross entropy loss To calculate the difference between the predicted category probability distribution and the true category.

2. According to claim 1, an information management system based on AI agricultural environmental data analysis is characterized in that: The cloud data center (10) is used to record and store various types of data from the user interaction terminal (20) and the data analysis terminal (30); the various types of data include equipment information, meteorological information, soil data, pest distribution, crop maintenance tasks, and analysis results.

3. The information management system based on AI agricultural environmental data analysis according to claim 2 is characterized in that: The mobile management terminal (21) is specifically a movable terminal management device with an image display function. The mobile management terminal (21) is connected to the cloud data center (10) via a network, and receives and downloads various types of data from the cloud data center (10) in real time; The mobile management terminal (21) comprises a data downloading unit (211) and a device status display unit (212); The data download unit (211) is used to download the data of each category at the latest time point in real time in the cloud data center (10), and overwrite the data of each category of the previous same set, so as to ensure that the storage state of each category of data of the device is optimal and up-to-date; The device status display unit (212) is used to display the device information of the irrigation system equipment, the automatic fertilizer spreader, various sensor devices, and the remote-controlled drone, so as to facilitate the administrator to view and manage; the device information includes working time, standby time, machine battery status, fault conditions, and location distribution coordinates; the various sensor devices specifically include soil temperature sensors, soil pH sensors, soil moisture sensors, and soil nitrogen, phosphorus, and potassium sensors.

4. The information management system based on AI agricultural environmental data analysis according to claim 1 is characterized in that: The device management module (22) comprises a device control unit (221) and a device monitoring unit (222); The equipment control unit (221) has the highest management authority and is used to remotely control and manage the operation and standby of the irrigation system equipment, the automatic fertilizer applicator, various sensor devices, and the remote-controlled drone; The equipment monitoring unit (222) is used to obtain equipment information of the irrigation system equipment, the automatic fertilizer applicator, various sensor devices, and the remote-controlled drone; and to send the equipment information to the cloud data center (10) in real time.

5. The information management system based on AI agricultural environmental data analysis according to claim 1 is characterized in that: The climate detection unit (311) is used to obtain the latest regional weather information through a regional weather station network, the weather information including temperature, humidity, wind speed, wind direction, and precipitation; and upload the weather information to the cloud data center (10); The soil detection unit (312) is used to monitor and obtain soil data through soil temperature sensors, soil pH sensors, soil moisture sensors, and soil nitrogen, phosphorus, and potassium sensors arranged at different locations in the crop planting and cultivation area, wherein the soil data includes soil temperature, soil moisture, fertility, and soil acidity and alkalinity; and upload the soil data to the cloud data center (10).

6. The information management system based on AI agricultural environmental data analysis according to claim 1 is characterized in that: The YOLO algorithm assumes that each pest has a bounding box, a confidence score, and a series of category probabilities; the specific simplified expression is as follows: ; in, : represents the sum of the loss values ​​of the comprehensive coordinates, confidence, and category probability; : represents the number of pests detected in the image; : is the coordinate loss of the i-th pest, measuring the position difference between the predicted bounding box and the true bounding box; : is the confidence loss of the i-th pest, including the confidence of the existence of the target in the bounding box and the confidence of the intersection and union ratio of the bounding box and the true bounding box; : is the category probability loss of the i-th pest, measuring the difference between the predicted category probability distribution and the true category probability distribution.

7. The information management system based on AI agricultural environmental data analysis according to claim 1 is characterized in that: For the prediction of a single crop planting and cultivation area, the square error It can be expressed as: ; :yes The predicted pest coordinate values, : is the true value of the pest coordinate data; The multi-class cross entropy loss The expression is as follows: ; in, :express The difference between the predicted value of the category and the true value of the category; : is the i-th pest in the one-hot encoding representation of the true label. It is a binary variable that takes the value of 1 if and only if the sample belongs to the category of the i-th pest, otherwise it takes the value of 0. In the pest quantity detection, if the pest in a grid cell is a second-class pest, then in the corresponding true label vector, the second element will be 1, and the rest of the elements will be 0; :express Predict the probability of the category of the i-th pest; :express Predict the total amount of pest samples.

8. The information management system based on AI agricultural environmental data analysis according to claim 1 is characterized in that: The crop growth analysis module (32) comprises a daily maintenance unit (321) and an insect pest analysis unit (322); The daily maintenance unit (321) is used to formulate crop maintenance tasks for crops in the crop planting and cultivation area according to the meteorological information and soil data detected and acquired by the real-time environmental monitoring module (31); the crop maintenance tasks include quantitative irrigation, soil fertility adjustment, cultivation area temperature adjustment, and soil pH adjustment; The pest analysis unit (322) is used to process and analyze the pest distribution using a Logistic growth model, and upload the analysis results to the cloud data center (10).

9. The information management system based on AI agricultural environmental data analysis according to claim 8 is characterized in that: The Logistic growth model is used to analyze the change in the number of pests under the condition of limited resources in crop planting and cultivation areas. The specific equation expression of the Logistic growth model is as follows: ; in, : represents the relative value of pest population at a certain moment; : represents time; : is the pest population growth rate parameter, which indicates the maximum growth rate when resources are not limited; : Indicates the rate of change of population size over time.

10. The information management system based on AI agricultural environmental data analysis according to claim 9 is characterized in that: The logistic growth model equation is already in explicit form with respect to x and t, and the right side is a function of x, so there is no need to further separate the variables, and the solution is as follows: ; in, is a constant; r: is the pest population growth rate parameter, which indicates the maximum growth rate when resources are not limited; when r>0, the pest population will increase; when r<0, the pest population will not increase, and the case of r<0 is not considered here; : describes the temporal trends in pest populations under conditions of limited resources; : represents a point in time before the number of pests begins to increase significantly. In the Logistic growth model, The position of the Logistic growth model curve on the time axis was determined so that the Logistic growth model curve could correctly reflect the time point when the number of pests changes from slow growth to rapid growth.

Citation Information

Patent Citations

  • Oilseed rape pest identification method

    CN111199245A

  • AI intelligent agricultural harvesting management system based on Internet of Things and application thereof

    CN117893346A

  • Efficient artificial grassland comprehensive management method

    CN117933579A

  • Platform using cloud-based Big Data

    KR102271036B1