Plant pesticide application management method and system based on intelligent decision-making system

Abstract

The invention discloses a plant pesticide application management method and system based on an intelligent decision-making system. The method comprises the following steps: dividing a target farmland into a plurality of sub-regions, constructing a plant disease and insect pest recognition model by using an image processing algorithm, recognizing disease and insect pests of each sub-region, and generating a recognition result. And then analyzing an identification result based on a clustering algorithm, and determining a pest distribution area. A pesticide application decision-making system is established, a pesticide application scheme of the unmanned aerial vehicle is formulated, influences under different moving speeds and environment wind conditions are analyzed in combination with historical pesticide drift data, and finally the pesticide application scheme is optimized according to real-time environment data. The method can effectively reduce the pesticide consumption, reduce the environmental influence and improve the health level of crops.

Description

Technical Field

[0001] The present invention relates to the technical field of pesticide application, and in particular to a plant pesticide application management method and system based on an intelligent decision-making system. Background Art

[0002] With the rapid development of modern agriculture, pest control has become an important part of ensuring high yield and high quality of crops. Traditional pesticide application methods have many shortcomings, such as uneven application, pesticide waste, environmental pollution and potential threats to human health. Therefore, how to achieve accurate and efficient pesticide application has become the focus of agricultural research.

[0003] At present, the application of intelligent decision-making systems in the agricultural field is gradually increasing, especially in the identification of pests and diseases, pesticide application decisions and operation optimization. Using image processing technology, the growth of plants and the occurrence of pests and diseases in farmland can be monitored in real time, so as to achieve accurate identification. Through data analysis and clustering algorithms, the distribution areas of pests and diseases can be effectively identified, providing a basis for subsequent pesticide application plans. In addition, combined with drone technology, rapid pesticide application in the target area can be achieved, greatly improving the efficiency of pesticide application.

[0004] However, most of the existing intelligent pesticide application management systems focus on a single aspect, such as pest and disease identification or pesticide application path planning, lack a comprehensive decision support system, and cannot be dynamically adjusted according to real-time environmental conditions and historical data. This limits the accuracy and effectiveness of pesticide application, and the problem of pesticide drift has not been effectively solved. Therefore, it is of great practical significance to develop a plant pesticide application management method based on an intelligent decision-making system that can comprehensively consider pest and disease identification, optimization of application plans, and the influence of environmental factors.

[0005] The present invention aims to provide a systematic method for the management of plant pesticide application. By means of technical means such as image processing, cluster analysis, drone application and dynamic optimization, an intelligent decision-making system is constructed to achieve precise and intelligent application of pesticides, ultimately improving the yield and quality of crops and reducing negative impacts on the environment. Summary of the invention

[0006] In order to solve at least one of the above technical problems, the present invention proposes a plant pesticide application management method and system based on an intelligent decision-making system.

[0007] The first aspect of the present invention provides a plant pesticide application management method based on an intelligent decision-making system, comprising:

[0008] The target farmland is divided into N sub-areas, a plant disease and insect pest recognition model is constructed based on an image processing algorithm, and plant diseases and insect pests in each sub-area are recognized according to the plant disease and insect pest recognition model to obtain a plant disease and insect pest recognition result;

[0009] Analyzing the plant disease and insect pest identification results based on a clustering algorithm to determine the plant disease and insect pest distribution area of ​​the target farmland;

[0010] Constructing a pesticide application decision system for the target farmland, and determining a pesticide application plan for the drone in the plant pest distribution area based on the pesticide application decision system;

[0011] Acquire historical pesticide drift data of the drone applying pesticides at different application speeds and under different environmental wind conditions, determine the impact of the drone applying pesticides at different application speeds and under different environmental wind conditions on pesticide drift based on the historical pesticide drift data, and obtain impact data;

[0012] The pesticide application program is optimized according to the impact data and the real-time environmental wind conditions of the target farmland.

[0013] In this solution, the target farmland is divided into N sub-areas, a plant disease and insect pest recognition model is constructed based on an image processing algorithm, and plant diseases and insect pests in each sub-area are identified according to the plant disease and insect pest recognition model to obtain plant disease and insect pest recognition results, specifically:

[0014] Obtain the names of plants planted in the target farmland, and obtain plant diseased leaf image data of the plants planted in the farmland under the attack of different types of pests and diseases according to the plant names;

[0015] Extracting diseased leaf features from the image data of plant diseased leaves under the attack of different types of pests and diseases based on an image recognition algorithm, wherein the diseased leaf features include color features, spot features, leaf curling features, hole features, and texture features;

[0016] Constructing a plant diseased leaf feature database, importing the diseased leaf features into the plant diseased leaf feature database, and forming a one-to-one mapping relationship between image features and pest types;

[0017] Divide the target farmland into N sub-areas, obtain plant image data at a preset position in each sub-area, and extract plant leaf features at the preset position in each sub-area based on the plant image data;

[0018] Compare and match the plant leaf features with the plant diseased leaf feature database to determine the similarity between the plant leaf features and the plant diseased leaf features of each type of pest and disease, and determine the type of pest and disease at a preset position in each sub-area according to the similarity;

[0019] According to the types of pests and diseases at preset positions in each sub-area, the proportion of areas with pests and diseases in each sub-area is determined, and the severity of pests and diseases in each sub-area is determined according to the proportion data, so as to obtain the plant pest and disease identification results of each sub-area, and the plant pest and disease identification results include the types of pests and diseases and the severity of pests and diseases.

[0020] In this solution, the plant pest identification results are analyzed based on the clustering algorithm to determine the plant pest distribution area of ​​the target farmland, specifically:

[0021] According to the plant pest and disease identification result, coordinate information of sub-areas in the target farmland where plant pests and diseases exist is obtained, each coordinate information is regarded as a data point, a K-means clustering algorithm is introduced, the data points are analyzed by the elbow rule, and the number K of clusters required to be formed by the K-means clustering algorithm is determined;

[0022] Construct K clusters, randomly select K data points as the initial cluster center of each cluster, calculate the Euclidean distance of each data point to each initial cluster center, and assign each data point to the cluster to which the initial cluster center with the closest Euclidean distance belongs;

[0023] For each cluster, calculate the average value of all data points in the cluster, use the average value as the new cluster center of each cluster, repeatedly calculate the Euclidean distance of each data point to the new cluster center, assign clusters and update cluster centers until the change of cluster centers is less than the preset value, stop clustering operation, and get the clustering result;

[0024] The clustering results are visualized in a target farmland map to determine the distribution area of ​​plant diseases and insect pests in the target farmland.

[0025] In this solution, the pesticide application decision system for the target farmland is constructed, and the pesticide application plan for the drone in the plant pest distribution area is determined based on the pesticide application decision system, specifically:

[0026] Constructing a pesticide database, obtaining functional data of different types of pesticides based on Internet searches, the functional data including applicable plants, types of pests and diseases to be treated, applicability at plant growth stages, and application dosages at different severity levels of pests and diseases, and importing the functional data into the pesticide data;

[0027] Building a UAV pesticide application path planning model based on the A* algorithm, integrating the pesticide application path planning model with the pesticide database, and building a pesticide application decision-making system for the target farmland;

[0028] Acquire growth stage data of plants in the target farmland, and determine the severity of the plant pests and diseases in each plant pest and disease distribution area according to the plant pest and disease identification results;

[0029] Importing the plant pest identification results, plant growth stage data, and the severity of pests in each plant pest distribution area into the pesticide application decision system and comparing them with the pesticide database, determining the application suitability of each pesticide type in each plant pest distribution area, selecting the pesticide type with the highest application suitability as the pesticide applied in the plant pest distribution area, and determining the application dosage according to the severity of pests in the plant pest distribution area;

[0030] Acquire the pesticide spraying diameter data of the drone and the boundary coordinate data of each plant pest distribution area, import the pesticide spraying diameter data and the boundary coordinate data into the pesticide application decision system, and construct the pesticide application path of the drone through the A* algorithm, wherein the pesticide application path includes the flight path of the drone, the spraying path segment of the drone in the flight path, and the spraying suspension path segment;

[0031] Acquire the area data of the plant pest distribution area in the target farmland and the pesticide spraying efficiency requirement data, and calculate the current pesticide application moving speed of the drone according to the area data and the pesticide spraying efficiency requirement data;

[0032] The type of pesticide applied, the application dosage, the application path of the drone, and the current movement speed of the pesticide application are integrated to obtain the pesticide application plan of the drone in the area where plant pests and diseases are distributed.

[0033] In this solution, the historical pesticide drift data of the drone applying pesticides at different application speeds and different environmental wind conditions is obtained, and the impact of the drone applying pesticides at different application speeds and different environmental wind conditions on the pesticide drift is determined according to the historical pesticide drift data to obtain the impact data, which is specifically:

[0034] Obtain historical pesticide drift data during each pesticide application process of the drone within a historical preset time period, and record the drone's pesticide application movement speed and environmental wind conditions during each pesticide application process, wherein the historical pesticide drift data includes drift direction and drift distance, and the environmental wind conditions include wind force and wind direction;

[0035] Aligning the historical pesticide drift data, the speed of the drone spraying and the environmental wind conditions in time series, and constructing a feature matrix with the data after the alignment operation;

[0036] A pesticide drift influence model is constructed based on a multivariate linear regression method, the feature matrix is ​​imported into the pesticide drift influence model for learning and training, the regression coefficient of the pesticide drift influence model is calculated using the least squares method, and the contribution of each feature to the pesticide drift is determined according to the regression coefficient;

[0037] The characteristic matrix is ​​analyzed by variance analysis method to determine the differences of historical pesticide drift data under different environmental wind conditions and application speeds;

[0038] Determining the influence of each feature on the drift of the pesticide according to the contribution of each feature to the drift of the pesticide and the differences in historical pesticide drift data under different environmental wind conditions and application speeds;

[0039] The drift paths of the pesticides under different pesticide application moving speeds and different environmental wind conditions of the UAV are plotted according to the characteristic matrix, and the influence of different pesticide application moving speeds and different environmental wind conditions on the drift position of the pesticides is determined according to the drift paths to obtain the influence data.

[0040] In this solution, the pesticide application scheme is optimized according to the impact data and the real-time environmental wind conditions of the target farmland, specifically:

[0041] Analyzing the current pesticide application movement speed of the drone in the pesticide application scheme based on the impact data, and determining the drift amount and the drift direction of the first pesticide action position caused by the current pesticide application movement speed of the drone;

[0042] Calculating a first spraying path correction coefficient of the UAV according to the first pesticide action position drift amount and the first pesticide action position drift direction, and optimizing the pesticide application path of the UAV in the pesticide application scheme according to the first spraying path correction coefficient;

[0043] Acquire real-time environmental wind condition data in the target farmland, analyze the real-time environmental wind condition data based on the impact data, and determine the effect of the real-time environmental wind condition on the drift amount and drift direction of the second pesticide action position in the current pesticide application process;

[0044] The second spraying path correction coefficient of the UAV is calculated according to the drift of the second pesticide action position, and the pesticide application path of the UAV in the pesticide application plan is secondarily optimized according to the second spraying path correction coefficient to obtain the pesticide application optimization plan for the target farmland.

[0045] The second aspect of the present invention also provides a plant pesticide application management system based on an intelligent decision-making system, the system comprising: a memory, a processor, the memory comprising a plant pesticide application management method program based on the intelligent decision-making system, and when the plant pesticide application management method program based on the intelligent decision-making system is executed by the processor, the following steps are implemented:

[0046] The target farmland is divided into N sub-areas, a plant disease and insect pest recognition model is constructed based on an image processing algorithm, and plant diseases and insect pests in each sub-area are recognized according to the plant disease and insect pest recognition model to obtain a plant disease and insect pest recognition result;

[0047] Analyzing the plant disease and insect pest identification results based on a clustering algorithm to determine the plant disease and insect pest distribution area of ​​the target farmland;

[0048] Constructing a pesticide application decision system for the target farmland, and determining a pesticide application plan for the drone in the plant pest distribution area based on the pesticide application decision system;

[0049] Acquire historical pesticide drift data of the drone applying pesticides at different application speeds and under different environmental wind conditions, determine the impact of the drone applying pesticides at different application speeds and under different environmental wind conditions on pesticide drift based on the historical pesticide drift data, and obtain impact data;

[0050] The pesticide application program is optimized according to the impact data and the real-time environmental wind conditions of the target farmland.

[0051] In this solution, the pesticide application scheme is optimized according to the impact data and the real-time environmental wind conditions of the target farmland, specifically:

[0052] Analyzing the current pesticide application movement speed of the drone in the pesticide application scheme based on the impact data, and determining the drift amount and the drift direction of the first pesticide action position caused by the current pesticide application movement speed of the drone;

[0053] Calculating a first spraying path correction coefficient of the UAV according to the first pesticide action position drift amount and the first pesticide action position drift direction, and optimizing the pesticide application path of the UAV in the pesticide application scheme according to the first spraying path correction coefficient;

[0054] Acquire real-time environmental wind condition data in the target farmland, analyze the real-time environmental wind condition data based on the impact data, and determine the effect of the real-time environmental wind condition on the drift amount and drift direction of the second pesticide action position in the current pesticide application process;

[0055] The second spraying path correction coefficient of the UAV is calculated according to the drift of the second pesticide action position, and the pesticide application path of the UAV in the pesticide application plan is secondarily optimized according to the second spraying path correction coefficient to obtain the pesticide application optimization plan for the target farmland.

[0056] The present invention discloses a plant pesticide application management method and system based on an intelligent decision-making system. The method includes dividing the target farmland into multiple sub-areas, and using an image processing algorithm to construct a plant pest and disease recognition model, identifying the pest and disease in each sub-area, and generating a recognition result. Subsequently, the recognition results are analyzed based on a clustering algorithm to determine the distribution area of ​​the pest and disease. A pesticide application decision-making system is established to formulate a pesticide application plan for drones, and the impact of different moving speeds and environmental wind conditions is analyzed in combination with historical pesticide drift data, and finally the pesticide application plan is optimized according to real-time environmental data. This method can effectively reduce the amount of pesticide used, reduce environmental impact, and improve the health level of crops. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 A flow chart of a method for managing plant pesticide application based on an intelligent decision-making system according to the present invention is shown;

[0058] Figure 2 A flow chart of determining the distribution area of ​​plant diseases and insect pests in target farmland according to the present invention is shown;

[0059] Figure 3 A flow chart showing the optimization of the pesticide application scheme according to the present invention is shown;

[0060] Figure 4 A block diagram of a plant pesticide application management system based on an intelligent decision-making system of the present invention is shown. DETAILED DESCRIPTION

[0061] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0062] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0063] Figure 1 A flow chart of a plant pesticide application management method based on an intelligent decision-making system of the present invention is shown.

[0064] like Figure 1 As shown, the first aspect of the present invention provides a plant pesticide application management method based on an intelligent decision-making system, comprising:

[0065] S102, dividing the target farmland into N sub-areas, building a plant disease and insect pest recognition model based on an image processing algorithm, and identifying plant diseases and insect pests in each sub-area according to the plant disease and insect pest recognition model to obtain a plant disease and insect pest recognition result;

[0066] S104, analyzing the plant disease and insect pest identification result based on a clustering algorithm to determine the plant disease and insect pest distribution area of ​​the target farmland;

[0067] S106, constructing a pesticide application decision system for the target farmland, and determining a pesticide application plan for the drone in the plant pest distribution area based on the pesticide application decision system;

[0068] S108, obtaining historical pesticide drift data when the drone applies pesticides at different application speeds and under different environmental wind conditions, and determining the impact of the drone applying pesticides at different application speeds and under different environmental wind conditions on pesticide drift based on the historical pesticide drift data to obtain impact data;

[0069] S110, optimizing the pesticide application scheme according to the impact data and the real-time environmental wind conditions of the target farmland.

[0070] It should be noted that by dividing the target farmland into N sub-areas, the pest and disease situation in each area can be accurately located. The recognition model constructed using the image processing algorithm can monitor the health of plants in real time, ensure timely identification of pests and diseases, and reduce potential losses; the application of clustering algorithms to analyze the results of pest and disease identification can reveal the spatial distribution pattern of pests and diseases; by building a pesticide application decision system, the scientific formulation of drone spraying plans for different pest and disease distribution areas can be achieved, the spraying path can be optimized, the drug can be evenly covered, and the spraying efficiency can be improved; by obtaining historical pesticide drift data under different conditions, the impact of spraying movement speed and environmental wind conditions on pesticide drift can be deeply analyzed, and the pesticide application plan can be optimized according to the impact data and real-time environmental wind conditions. The spraying strategy can be dynamically adjusted to ensure the best spraying effect in different environments and minimize the drift and loss of pesticides.

[0071] According to an embodiment of the present invention, the target farmland is divided into N sub-areas, a plant disease and insect pest recognition model is constructed based on an image processing algorithm, and plant diseases and insect pests in each sub-area are recognized according to the plant disease and insect pest recognition model to obtain a plant disease and insect pest recognition result, which is specifically:

[0072] Obtain the names of plants planted in the target farmland, and obtain plant diseased leaf image data of the plants planted in the farmland under the attack of different types of pests and diseases according to the plant names;

[0073] Extracting diseased leaf features from the image data of plant diseased leaves under the attack of different types of pests and diseases based on an image recognition algorithm, wherein the diseased leaf features include color features, spot features, leaf curling features, hole features, and texture features;

[0074] Constructing a plant diseased leaf feature database, importing the diseased leaf features into the plant diseased leaf feature database, and forming a one-to-one mapping relationship between image features and pest types;

[0075] Divide the target farmland into N sub-areas, obtain plant image data at a preset position in each sub-area, and extract plant leaf features at the preset position in each sub-area based on the plant image data;

[0076] Compare and match the plant leaf features with the plant diseased leaf feature database to determine the similarity between the plant leaf features and the plant diseased leaf features of each type of pest and disease, and determine the type of pest and disease at a preset position in each sub-area according to the similarity;

[0077] According to the types of pests and diseases at preset positions in each sub-area, the proportion of areas with pests and diseases in each sub-area is determined, and the severity of pests and diseases in each sub-area is determined according to the proportion data, so as to obtain the plant pest and disease identification results of each sub-area, and the plant pest and disease identification results include the types of pests and diseases and the severity of pests and diseases.

[0078] It should be noted that by using image recognition algorithms to extract the color features, spot features, leaf curl features, hole features and texture features of diseased leaves, the different manifestations of pests and diseases can be accurately captured; by constructing a structured plant diseased leaf feature database; extracting the plant leaf features of each sub-region and comparing them with the diseased leaf features in the feature database, the health status of the plants in the region can be effectively judged. This matching process helps to quickly identify the types of pests and diseases and reduce the time and cost of manual inspections; by determining the types of pests and diseases in each sub-region through similarity analysis, and further calculating the proportion of regions with pests and diseases, the distribution of pests and diseases in each sub-region can be fully understood; based on the extracted and analyzed pest and disease identification results, farmers can monitor and evaluate the farmland conditions in real time, take timely prevention and control measures, and ensure that pest and disease control is more effective. This dynamic feedback mechanism improves the growth safety of crops and reduces potential losses.

[0079] Figure 2 A flow chart of determining the distribution area of ​​plant diseases and insect pests in target farmland according to the present invention is shown.

[0080] According to an embodiment of the present invention, the plant disease and insect pest identification result is analyzed based on a clustering algorithm to determine the plant disease and insect pest distribution area of ​​the target farmland, specifically:

[0081] S202, obtaining coordinate information of sub-regions in the target farmland where plant diseases and insect pests exist according to the plant disease and insect pest identification result, treating each coordinate information as a data point, introducing a K-means clustering algorithm, analyzing the data points by the elbow rule, and determining the number K of clusters required to be formed by the K-means clustering algorithm;

[0082] S204, constructing K clusters, randomly selecting K data points as the initial cluster center of each cluster, calculating the Euclidean distance of each data point to each initial cluster center, and assigning each data point to the cluster to which the initial cluster center with the shortest Euclidean distance belongs;

[0083] S206, for each cluster, calculating the average value of all data points in the cluster, taking the average value as the new cluster center of each cluster, repeatedly calculating the Euclidean distance from each data point to the new cluster center, assigning clusters and updating cluster centers until the change of the cluster center is less than a preset value, stopping the clustering operation, and obtaining a clustering result;

[0084] S208, visualizing the clustering result in a target farmland map to determine the distribution area of ​​plant diseases and insect pests in the target farmland.

[0085] It should be noted that by introducing the K-means clustering algorithm to analyze the results of plant pest and disease identification, the pest and disease data can be effectively classified, so as to clarify the spatial distribution of various pests and diseases in farmland; the K-means clustering algorithm aggregates based on the similarity of data points, and classifies data points with similar characteristics into the same cluster. This means that areas with similar pest and disease characteristics will be effectively identified and concentrated together, thereby revealing the distribution of pests and diseases, and using Euclidean distance to evaluate the similarity between data points and cluster centers. This measurement method can effectively reflect the differences between pest and disease characteristics, so that similar pest and disease samples are assigned to the same cluster; in each iteration, the cluster center is updated according to the data points it contains to ensure that it more accurately reflects the clustering results; by clarifying the distribution areas of different pests and diseases, farmers can allocate pesticide application resources more reasonably, give priority to areas with serious pests and diseases, and thus improve the efficiency and effectiveness of pesticide application.

[0086] According to an embodiment of the present invention, the pesticide application decision system for the target farmland is constructed, and the pesticide application plan for the drone in the plant pest distribution area is determined based on the pesticide application decision system, specifically:

[0087] Constructing a pesticide database, obtaining functional data of different types of pesticides based on Internet searches, the functional data including applicable plants, types of pests and diseases to be treated, applicability at plant growth stages, and application dosages at different severity levels of pests and diseases, and importing the functional data into the pesticide data;

[0088] Building a UAV pesticide application path planning model based on the A* algorithm, integrating the pesticide application path planning model with the pesticide database, and building a pesticide application decision-making system for the target farmland;

[0089] Acquire growth stage data of plants in the target farmland, and determine the severity of the plant pests and diseases in each plant pest and disease distribution area according to the plant pest and disease identification results;

[0090] Importing the plant pest identification results, plant growth stage data, and the severity of pests in each plant pest distribution area into the pesticide application decision system and comparing them with the pesticide database, determining the application suitability of each pesticide type in each plant pest distribution area, selecting the pesticide type with the highest application suitability as the pesticide applied in the plant pest distribution area, and determining the application dosage according to the severity of pests in the plant pest distribution area;

[0091] Acquire the pesticide spraying diameter data of the drone and the boundary coordinate data of each plant pest distribution area, import the pesticide spraying diameter data and the boundary coordinate data into the pesticide application decision system, and construct the pesticide application path of the drone through the A* algorithm, wherein the pesticide application path includes the flight path of the drone, the spraying path segment of the drone in the flight path, and the spraying suspension path segment;

[0092] Acquire the area data of the plant pest distribution area in the target farmland and the pesticide spraying efficiency requirement data, and calculate the current pesticide application moving speed of the drone according to the area data and the pesticide spraying efficiency requirement data;

[0093] The type of pesticide applied, the application dosage, the application path of the drone, and the current movement speed of the pesticide application are integrated to obtain the pesticide application plan of the drone in the area where plant pests and diseases are distributed.

[0094] It should be noted that the pesticide database is constructed and the UAV pesticide application path planning model is constructed based on the A* algorithm to integrate into the pesticide application decision system. By comparing the plant pest identification results, plant growth stage data, and the severity of pests and diseases in each plant pest distribution area with the pesticide database in the pesticide application decision system, the most suitable pesticide type and application dosage in the plant pest distribution area are determined; the A* algorithm can effectively construct the UAV pesticide application path, mainly because it is efficient and accurate in path planning. The algorithm selects the optimal path through heuristic search, combining the cost from the current node to the target node with the estimated cost. It can take obstacles, terrain and spraying range into consideration, and adjust the path in real time to ensure that the drone can apply pesticides efficiently and accurately in the farmland, while minimizing flight time and resource consumption; by identifying the distribution area of ​​plant pests and diseases and evaluating their severity, it can select the most suitable type of pesticide and application dosage in a targeted manner, reduce unnecessary drug waste, and improve application efficiency; the A* algorithm is used for drone application path planning, which can achieve the shortest flight distance and the selection of the optimal spraying route, reduce flight time and energy consumption, and improve the overall efficiency of pesticide application; the application path includes the drone flight path, the drone's pesticide spraying status or suspended spraying status at each flight position, and the drone will only spray in the plant pest and disease distribution area, and spray when it flies to the spraying path section, otherwise it will suspend spraying when it flies to the suspended spraying path section.

[0095] According to an embodiment of the present invention, the historical pesticide drift data of the drone applying pesticides under different application moving speeds and different environmental wind conditions is obtained, and the influence of the drone applying pesticides under different application moving speeds and different environmental wind conditions on the pesticide drift is determined according to the historical pesticide drift data to obtain the influence data, specifically:

[0096] Obtain historical pesticide drift data during each pesticide application process of the drone within a historical preset time period, and record the drone's pesticide application movement speed and environmental wind conditions during each pesticide application process, wherein the historical pesticide drift data includes drift direction and drift distance, and the environmental wind conditions include wind force and wind direction;

[0097] Aligning the historical pesticide drift data, the speed of the drone spraying and the environmental wind conditions in time series, and constructing a feature matrix with the data after the alignment operation;

[0098] A pesticide drift influence model is constructed based on a multivariate linear regression method, the feature matrix is ​​imported into the pesticide drift influence model for learning and training, the regression coefficient of the pesticide drift influence model is calculated using the least squares method, and the contribution of each feature to the pesticide drift is determined according to the regression coefficient;

[0099] The characteristic matrix is ​​analyzed by variance analysis method to determine the differences of historical pesticide drift data under different environmental wind conditions and application speeds;

[0100] Determining the influence of each feature on the drift of the pesticide according to the contribution of each feature to the drift of the pesticide and the differences in historical pesticide drift data under different environmental wind conditions and application speeds;

[0101] The drift paths of the pesticides under different pesticide application moving speeds and different environmental wind conditions of the UAV are plotted according to the characteristic matrix, and the influence of different pesticide application moving speeds and different environmental wind conditions on the drift position of the pesticides is determined according to the drift paths to obtain the influence data.

[0102] It should be noted that in the actual application process of drones, pesticide drift will affect the effect of application, resulting in resource waste and environmental pollution. Different application speeds and environmental wind conditions (such as wind force and wind direction) will significantly affect the drift behavior of pesticides, resulting in the inability to accurately apply pesticides to the target area; therefore, a pesticide drift impact model is constructed through the multivariate linear regression method to identify the impact of drone application at different application speeds and different environmental wind conditions on pesticide drift, thereby optimizing the application plan and improving the safety and effectiveness of pesticide use; multivariate linear regression allows multiple independent variables (such as application speed, wind speed, wind direction, etc.) to be considered at the same time, so that the model can capture the complex relationship between these variables and reflect the multiple influences in the actual application process. Through the least squares method, the model can calculate the regression coefficient of each independent variable and quantify the specific contribution of each factor to pesticide drift; the model is trained based on historical data, so it can reflect the real situation and improve the reliability of prediction. By analyzing historical drift data, key factors that significantly affect drift can be identified; each feature is an influencing feature on pesticide drift, namely, the drone application speed and environmental wind conditions; the pesticide drift position is the position where the actual application position of the pesticide deviates from the target application position due to the influence of environmental wind or drone movement speed; the influencing data includes the degree of influence of different application speeds and different environmental wind conditions on pesticide drift, and the degree of influence indicates at what speed the drone will have a higher degree of pesticide drift or under what conditions the environmental wind is, and also includes the influence of different application speeds and different environmental wind conditions on the pesticide drift position, and the position influence includes the drift direction influence and the drift distance influence.

[0103] Figure 3 A flow chart showing the optimization of the pesticide application scheme according to the present invention is shown.

[0104] According to an embodiment of the present invention, the pesticide application scheme is optimized according to the impact data and the real-time environmental wind conditions of the target farmland, specifically:

[0105] S302, analyzing the current pesticide application moving speed of the drone in the pesticide application scheme based on the impact data, and determining the drift amount and the drift direction of the first pesticide action position caused by the current pesticide application moving speed of the drone;

[0106] S304, calculating a first spraying path correction coefficient of the UAV according to the first pesticide action position drift amount and the first pesticide action position drift direction, and optimizing the pesticide application path of the UAV in the pesticide application scheme according to the first spraying path correction coefficient;

[0107] It should be noted that the drone applies pesticides according to the pesticide application plan within the plant pest distribution area. However, since the drone has a certain speed and altitude, the spraying of pesticides will drift with the inertia of the drone's flight speed, causing the drone to fly to the boundary of the plant pest distribution area. At this time, the pesticide will be sprayed outside the plant pest distribution area due to inertia. Therefore, by analyzing the current pesticide application movement speed of the drone through the impact data, the drift amount of the first pesticide action position caused by the drone's flight inertia and the drift direction of the first pesticide action position can be determined to calculate the first spraying path correction coefficient of the drone. The first spraying path correction coefficient can enable the drone to suspend spraying when it is close to the boundary of the plant pest distribution area through the analysis of the drift amount of the first pesticide action position and the drift direction of the first pesticide action position, so that the pesticide in the area close to the boundary of the plant pest distribution area can be sprayed by inertia, avoiding the impact of pesticides on plants in non-plant pest distribution areas and the waste of pesticides. The first spraying path correction coefficient is how far away from the boundary of the plant pest distribution area the pesticide application is suspended in advance.

[0108] S306, acquiring real-time environmental wind condition data in the target farmland, analyzing the real-time environmental wind condition data based on the impact data, and determining the effect of the real-time environmental wind condition on the drift amount and drift direction of the second pesticide action position in the current pesticide application process;

[0109] S308, calculating a second spraying path correction coefficient of the UAV according to the drift of the second pesticide action position, performing secondary optimization on the pesticide application path of the UAV in the pesticide application plan according to the second spraying path correction coefficient, and obtaining a pesticide application optimization plan for the target farmland.

[0110] It should be noted that the environmental wind in the target farmland will also cause the drift of pesticides, that is, the actual spraying position of the pesticide drifts from the target position; therefore, the real-time environmental wind condition data in the target farmland is analyzed by influencing data, the drift amount and drift direction of the second pesticide action position are determined, and the second spraying path correction coefficient is calculated. The second spraying path correction coefficient is the correction distance coefficient of the UAV flight path. For example, the environmental wind causes the pesticide position in the farmland to drift to the left, and the position drift is 2 meters. The flight path of the UAV is adjusted to the right by 2 meters to correct the pesticide position operation, so that the pesticide can be accurately applied to the target position after correction, avoiding waste of pesticides or failure to apply pesticides to the target position, resulting in useless work in pest control, improving the accuracy of pesticide application and reducing the environmental impact of pesticides on non-plant pest distribution areas.

[0111] Figure 4 A block diagram of a plant pesticide application management system based on an intelligent decision-making system of the present invention is shown.

[0112] The second aspect of the present invention further provides a plant pesticide application management system 4 based on an intelligent decision-making system, the system comprising: a memory 41, a processor 42, the memory comprising a plant pesticide application management method program based on an intelligent decision-making system, and when the plant pesticide application management method program based on an intelligent decision-making system is executed by the processor, the following steps are implemented:

[0113] The target farmland is divided into N sub-areas, a plant disease and insect pest recognition model is constructed based on an image processing algorithm, and plant diseases and insect pests in each sub-area are recognized according to the plant disease and insect pest recognition model to obtain a plant disease and insect pest recognition result;

[0114] Analyzing the plant disease and insect pest identification results based on a clustering algorithm to determine the plant disease and insect pest distribution area of ​​the target farmland;

[0115] Constructing a pesticide application decision system for the target farmland, and determining a pesticide application plan for the drone in the plant pest distribution area based on the pesticide application decision system;

[0116] Acquire historical pesticide drift data of the drone applying pesticides at different application speeds and under different environmental wind conditions, determine the impact of the drone applying pesticides at different application speeds and under different environmental wind conditions on pesticide drift based on the historical pesticide drift data, and obtain impact data;

[0117] The pesticide application program is optimized according to the impact data and the real-time environmental wind conditions of the target farmland.

[0118] According to an embodiment of the present invention, the pesticide application scheme is optimized according to the impact data and the real-time environmental wind conditions of the target farmland, specifically:

[0119] Analyzing the current pesticide application movement speed of the drone in the pesticide application scheme based on the impact data, and determining the drift amount and the drift direction of the first pesticide action position caused by the current pesticide application movement speed of the drone;

[0120] Calculating a first spraying path correction coefficient of the UAV according to the first pesticide action position drift amount and the first pesticide action position drift direction, and optimizing the pesticide application path of the UAV in the pesticide application scheme according to the first spraying path correction coefficient;

[0121] Acquire real-time environmental wind condition data in the target farmland, analyze the real-time environmental wind condition data based on the impact data, and determine the effect of the real-time environmental wind condition on the drift amount and drift direction of the second pesticide action position in the current pesticide application process;

[0122] The second spraying path correction coefficient of the UAV is calculated according to the drift of the second pesticide action position, and the pesticide application path of the UAV in the pesticide application plan is secondarily optimized according to the second spraying path correction coefficient to obtain the pesticide application optimization plan for the target farmland.

[0123] The present invention discloses a plant pesticide application management method and system based on an intelligent decision-making system. The method includes dividing the target farmland into multiple sub-areas, and using an image processing algorithm to construct a plant pest and disease recognition model, identifying the pest and disease in each sub-area, and generating a recognition result. Subsequently, the recognition results are analyzed based on a clustering algorithm to determine the distribution area of ​​the pest and disease. A pesticide application decision-making system is established to formulate a pesticide application plan for drones, and the impact of different moving speeds and environmental wind conditions is analyzed in combination with historical pesticide drift data, and finally the pesticide application plan is optimized according to real-time environmental data. This method can effectively reduce the amount of pesticide used, reduce environmental impact, and improve the health level of crops.

[0124] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

[0125] The units described above as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units; they may be located in one place or distributed on multiple network units; some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0126] In addition, all functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may be separately used as a unit, or two or more units may be integrated into one unit; the above-mentioned integrated units may be implemented in the form of hardware or in the form of hardware plus software functional units.

[0127] A person of ordinary skill in the art can understand that: all or part of the steps of implementing the above method embodiments can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above method embodiments; and the aforementioned storage medium includes: mobile storage devices, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks or optical disks, and other media that can store program codes.

[0128] Alternatively, if the above-mentioned integrated unit of the present invention is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiment of the present invention can be essentially or partly reflected in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as mobile storage devices, ROM, RAM, magnetic disks or optical disks.

[0129] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A plant pesticide application management method based on an intelligent decision-making system, characterized in that: The following steps are involved: The target farmland is divided into N sub-areas, a plant disease and insect pest recognition model is constructed based on an image processing algorithm, and plant diseases and insect pests in each sub-area are recognized according to the plant disease and insect pest recognition model to obtain a plant disease and insect pest recognition result; Analyzing the plant disease and insect pest identification results based on a clustering algorithm to determine the plant disease and insect pest distribution area of ​​the target farmland; Constructing a pesticide application decision system for the target farmland, and determining a pesticide application plan for the drone in the plant pest distribution area based on the pesticide application decision system; Acquire historical pesticide drift data of the drone applying pesticides at different application speeds and under different environmental wind conditions, determine the impact of the drone applying pesticides at different application speeds and under different environmental wind conditions on pesticide drift based on the historical pesticide drift data, and obtain impact data; The pesticide application program is optimized according to the impact data and the real-time environmental wind conditions of the target farmland.

2. A method for managing the application of plant pesticides based on an intelligent decision-making system according to claim 1, characterized in that: The target farmland is divided into N sub-areas, a plant disease and insect pest recognition model is constructed based on an image processing algorithm, and plant disease and insect pest recognition results are obtained by identifying the plant disease and insect pest in each sub-area according to the plant disease and insect pest recognition model, specifically: Obtain the names of plants planted in the target farmland, and obtain plant diseased leaf image data of the plants planted in the farmland under the attack of different types of pests and diseases according to the plant names; Extracting diseased leaf features from the image data of plant diseased leaves under the attack of different types of pests and diseases based on an image recognition algorithm, wherein the diseased leaf features include color features, spot features, leaf curling features, hole features, and texture features; Constructing a plant diseased leaf feature database, importing the diseased leaf features into the plant diseased leaf feature database, and forming a one-to-one mapping relationship between image features and pest types; Divide the target farmland into N sub-areas, obtain plant image data at a preset position in each sub-area, and extract plant leaf features at the preset position in each sub-area based on the plant image data; Compare and match the plant leaf features with the plant diseased leaf feature database to determine the similarity between the plant leaf features and the plant diseased leaf features of each type of pest and disease, and determine the type of pest and disease at a preset position in each sub-area according to the similarity; According to the types of pests and diseases at preset positions in each sub-area, the proportion of areas with pests and diseases in each sub-area is determined, and the severity of pests and diseases in each sub-area is determined according to the proportion data, so as to obtain the plant pest and disease identification results of each sub-area, and the plant pest and disease identification results include the types of pests and diseases and the severity of pests and diseases.

3. The method for managing the application of plant pesticides based on an intelligent decision-making system according to claim 1, characterized in that: The plant pest and disease identification results are analyzed based on the clustering algorithm to determine the plant pest and disease distribution area of ​​the target farmland, specifically: According to the plant pest and disease identification result, coordinate information of sub-areas in the target farmland where plant pests and diseases exist is obtained, each coordinate information is regarded as a data point, a K-means clustering algorithm is introduced, the data points are analyzed by the elbow rule, and the number K of clusters required to be formed by the K-means clustering algorithm is determined; Construct K clusters, randomly select K data points as the initial cluster center of each cluster, calculate the Euclidean distance of each data point to each initial cluster center, and assign each data point to the cluster to which the initial cluster center with the closest Euclidean distance belongs; For each cluster, calculate the average value of all data points in the cluster, use the average value as the new cluster center of each cluster, repeatedly calculate the Euclidean distance of each data point to the new cluster center, assign clusters and update cluster centers until the change of cluster centers is less than the preset value, stop clustering operation, and get the clustering result; The clustering results are visualized in a target farmland map to determine the distribution area of ​​plant diseases and insect pests in the target farmland.

4. The method for managing the application of plant pesticides based on an intelligent decision-making system according to claim 1, characterized in that: The pesticide application decision system for the target farmland is constructed, and the pesticide application plan for the drone in the plant pest distribution area is determined based on the pesticide application decision system, specifically: Constructing a pesticide database, obtaining functional data of different types of pesticides based on Internet searches, the functional data including applicable plants, types of pests and diseases to be treated, applicability at plant growth stages, and application dosages at different severity levels of pests and diseases, and importing the functional data into the pesticide data; Building a UAV pesticide application path planning model based on the A* algorithm, integrating the pesticide application path planning model with the pesticide database, and building a pesticide application decision-making system for the target farmland; Acquire growth stage data of plants in the target farmland, and determine the severity of the plant pests and diseases in each plant pest and disease distribution area according to the plant pest and disease identification results; Importing the plant pest identification results, plant growth stage data, and the severity of pests in each plant pest distribution area into the pesticide application decision system and comparing them with the pesticide database, determining the application suitability of each pesticide type in each plant pest distribution area, selecting the pesticide type with the highest application suitability as the pesticide applied in the plant pest distribution area, and determining the application dosage according to the severity of pests in the plant pest distribution area; Acquire the pesticide spraying diameter data of the drone and the boundary coordinate data of each plant pest distribution area, import the pesticide spraying diameter data and the boundary coordinate data into the pesticide application decision system, and construct the pesticide application path of the drone through the A* algorithm, wherein the pesticide application path includes the flight path of the drone, the spraying path segment of the drone in the flight path, and the spraying suspension path segment; Acquire the area data of the plant pest distribution area in the target farmland and the pesticide spraying efficiency requirement data, and calculate the current pesticide application moving speed of the drone according to the area data and the pesticide spraying efficiency requirement data; The type of pesticide applied, the application dosage, the application path of the drone, and the current movement speed of the pesticide application are integrated to obtain the pesticide application plan of the drone in the area where plant pests and diseases are distributed.

5. The method for managing the application of plant pesticides based on an intelligent decision-making system according to claim 1, characterized in that: The historical pesticide drift data of the drone applying pesticides at different application speeds and different environmental wind conditions is obtained, and the influence of the drone applying pesticides at different application speeds and different environmental wind conditions on the pesticide drift is determined according to the historical pesticide drift data to obtain the influence data, specifically: Obtain historical pesticide drift data during each pesticide application process of the drone within a historical preset time period, and record the drone's pesticide application movement speed and environmental wind conditions during each pesticide application process, wherein the historical pesticide drift data includes drift direction and drift distance, and the environmental wind conditions include wind force and wind direction; Aligning the historical pesticide drift data, the speed of the drone spraying and the environmental wind conditions in time series, and constructing a feature matrix with the data after the alignment operation; A pesticide drift influence model is constructed based on a multivariate linear regression method, the feature matrix is ​​imported into the pesticide drift influence model for learning and training, the regression coefficient of the pesticide drift influence model is calculated using the least squares method, and the contribution of each feature to the pesticide drift is determined according to the regression coefficient; The characteristic matrix is ​​analyzed by variance analysis method to determine the differences of historical pesticide drift data under different environmental wind conditions and application speeds; Determining the influence of each feature on the drift of the pesticide according to the contribution of each feature to the drift of the pesticide and the differences in historical pesticide drift data under different environmental wind conditions and application speeds; The drift paths of the pesticides under different pesticide application moving speeds of the UAV and different environmental wind conditions are drawn according to the characteristic matrix, and the influence of different pesticide application moving speeds and different environmental wind conditions on the drift position of the pesticides is determined according to the drift paths to obtain the influence data.

6. The method for managing the application of plant pesticides based on an intelligent decision-making system according to claim 1, characterized in that: The pesticide application scheme is optimized according to the impact data and the real-time environmental wind conditions of the target farmland, specifically: Analyzing the current pesticide application movement speed of the drone in the pesticide application scheme based on the impact data, and determining the drift amount and the drift direction of the first pesticide action position caused by the current pesticide application movement speed of the drone; Calculating a first spraying path correction coefficient of the UAV according to the first pesticide action position drift amount and the first pesticide action position drift direction, and optimizing the pesticide application path of the UAV in the pesticide application scheme according to the first spraying path correction coefficient; Acquire real-time environmental wind condition data in the target farmland, analyze the real-time environmental wind condition data based on the impact data, and determine the effect of the real-time environmental wind condition on the drift amount and drift direction of the second pesticide action position in the current pesticide application process; The second spraying path correction coefficient of the UAV is calculated according to the drift of the second pesticide action position, and the pesticide application path of the UAV in the pesticide application plan is secondarily optimized according to the second spraying path correction coefficient to obtain the pesticide application optimization plan for the target farmland.

7. A plant pesticide application management system based on an intelligent decision-making system, characterized in that: The plant pesticide application management system based on the intelligent decision-making system includes a storage device and a processor, wherein the storage device includes a plant pesticide application management method program based on the intelligent decision-making system, and when the plant pesticide application management method program based on the intelligent decision-making system is executed by the processor, the following steps are implemented: The target farmland is divided into N sub-areas, a plant disease and insect pest recognition model is constructed based on an image processing algorithm, and plant diseases and insect pests in each sub-area are recognized according to the plant disease and insect pest recognition model to obtain a plant disease and insect pest recognition result; Analyzing the plant disease and insect pest identification results based on a clustering algorithm to determine the plant disease and insect pest distribution area of ​​the target farmland; Constructing a pesticide application decision system for the target farmland, and determining a pesticide application plan for the drone in the plant pest distribution area based on the pesticide application decision system; Acquire historical pesticide drift data of the drone applying pesticides at different application speeds and under different environmental wind conditions, determine the impact of the drone applying pesticides at different application speeds and under different environmental wind conditions on pesticide drift based on the historical pesticide drift data, and obtain impact data; The pesticide application program is optimized according to the impact data and the real-time environmental wind conditions of the target farmland.

8. The plant pesticide application management system based on the intelligent decision-making system according to claim 7, characterized in that: The pesticide application scheme is optimized according to the impact data and the real-time environmental wind conditions of the target farmland, specifically: Analyzing the current pesticide application movement speed of the drone in the pesticide application scheme based on the impact data, and determining the drift amount and the drift direction of the first pesticide action position caused by the current pesticide application movement speed of the drone; Calculating a first spraying path correction coefficient of the UAV according to the first pesticide action position drift amount and the first pesticide action position drift direction, and optimizing the pesticide application path of the UAV in the pesticide application scheme according to the first spraying path correction coefficient; Acquire real-time environmental wind condition data in the target farmland, analyze the real-time environmental wind condition data based on the impact data, and determine the effect of the real-time environmental wind condition on the drift amount and drift direction of the second pesticide action position in the current pesticide application process; The second spraying path correction coefficient of the UAV is calculated according to the drift of the second pesticide action position, and the pesticide application path of the UAV in the pesticide application plan is secondarily optimized according to the second spraying path correction coefficient to obtain the pesticide application optimization plan for the target farmland.

Images