Sweet potato planting pest analysis method and system based on multiple environmental characteristics

Abstract

The invention discloses a sweet potato planting pest analysis method and system based on various environmental characteristics, and the method comprises the steps: setting a plurality of monitoring points in a sweet potato planting region, obtaining multi-dimensional environmental characteristics and pest characteristic data in a historical time period, and converting the data into data items; performing association analysis on the data items by using an Apriori association algorithm, mapping the correlation between the environment and the insect pest characteristics, and screening associated monitoring points; constructing a classification model based on a decision tree, obtaining associated monitoring point data in real time, converting the associated monitoring point data into decision nodes, and performing classification model construction and prediction training; and finally, acquiring non-associated monitoring point data, importing the non-associated monitoring point data into a classification model for real-time classification, performing insect pest associated prediction in combination with an associated monitoring point analysis result, and generating a plurality of efficient insect pest prevention and control schemes. The method effectively integrates the multi-source environment data and the pest data, improves the pest prediction accuracy, achieves the multi-region correlation prediction, and provides a scientific prevention and control basis for sweet potato planting.

Description

Technical Field

[0001] The present invention relates to the field of intelligent analysis of agricultural planting, and more specifically, to a method and system for analyzing sweet potato planting pests based on multiple environmental characteristics. Background Art

[0002] During the sweet potato planting process, it is often threatened by pests and diseases such as soft rot, viral diseases, sweet potato borers, and sweet potato hornworms. Traditional prevention and control relies on manual experience, and there are problems of lag, inefficiency, and pesticide abuse. In the analysis process of sweet potato planting and pests in the existing technology, the monitoring resources for multiple planting areas are greatly consumed, and it is difficult to conduct a comprehensive analysis of multi-dimensional environmental data and pest data. Planting analysis and pest prediction are often based on a single type of data. There is a lack of effective sweet potato regional planting correlation analysis methods. At the same time, there is a lack of multi-region pest correlation prediction, resulting in low efficiency of its prevention and control strategies in large-scale planting environments, making it difficult to achieve holistic planting comprehensive evaluation and pest impact control. Moreover, for large-scale sweet potato planting, the environmental data and pest data in different monitoring points are complex and changeable, and it is difficult for existing technologies to conduct comprehensive evaluation analysis and correlation prediction. Therefore, there is an urgent need for an efficient sweet potato planting pest analysis method. Summary of the Invention

[0003] The present invention overcomes the defects of the prior art and proposes a sweet potato planting pest analysis method and system based on multiple environmental characteristics.

[0004] A first aspect of the present invention provides a method for analyzing sweet potato plant pests based on multiple environmental characteristics, comprising:

[0005] S102: In the sweet potato planting area, multiple monitoring points are set, and multi-dimensional environmental characteristic data and pest characteristic data of the multiple monitoring points are obtained in a historical period, and the environmental characteristic data and pest characteristic data are converted into multiple data items;

[0006] S104: Taking each monitoring point as an analysis unit, performing association analysis on the data items based on the Apriori association algorithm, and mapping them to the correlation analysis of environmental characteristics and pest characteristics, and screening out associated monitoring points;

[0007] S106: Constructing a classification model based on a decision tree. Within a planting cycle, real-time environmental characteristic data and pest characteristic data of associated monitoring points are obtained and converted into decision nodes. Nodes of the classification model are constructed based on the decision nodes, and prediction training is performed on the classification model.

[0008] S108: During a planting cycle, environmental characteristic data and pest characteristic data of non-associated monitoring points are obtained and imported into a classification model to perform real-time classification of the monitoring points to obtain classification results. Pest analysis is performed based on associated monitoring points, and pest association prediction is performed on multiple groups of monitoring points based on the classification results to generate N types of pest control plans.

[0009] In this solution, the S102 is specifically as follows:

[0010] Construct a map model based on sweet potato planting areas;

[0011] Based on the distribution of each sweet potato planting site, multiple monitoring points are set from the map model to ensure that the planting density of each monitoring point is consistent within the preset range;

[0012] A monitoring unit is set up based on each monitoring point. During a historical period, multi-dimensional environmental characteristic data and pest characteristic data of each monitoring point are obtained through the monitoring unit. Information is extracted for each environmental characteristic and each pest characteristic to form multiple data items.

[0013] In this solution, the S104 is specifically as follows:

[0014] For a historical period, multiple time nodes are divided according to the preset interval T;

[0015] Taking a monitoring point as the analysis unit, for each time node, the data items of the corresponding node are obtained from the multi-dimensional environmental characteristic data and pest characteristic data;

[0016] Taking the data items of the corresponding nodes as data item sets, and filtering out frequent item sets with minimum support based on the Apriori association algorithm;

[0017] Calculate the confidence in the frequent item set, filter the rules according to the minimum confidence, and obtain the association rules;

[0018] Based on the association rules, the association between data items is determined. If the data items corresponding to the environmental characteristics are associated with the data items corresponding to the pest characteristics, and the number of associations is greater than the preset association quantity, the current time node is marked as an associated node.

[0019] For a monitoring point, if each time node is an associated node, the monitoring point is marked as an associated monitoring point.

[0020] In this solution, the S106 is specifically as follows:

[0021] Build a classification model based on decision trees;

[0022] During a planting cycle, the environmental characteristic data and pest characteristic data of a related monitoring point are obtained in real time, and each characteristic data is converted into a conditional node of a decision tree to generate multiple conditional nodes;

[0023] Generate conditional nodes for all associated monitoring points, and based on the heuristic algorithm, add the conditional nodes to the classification model, and determine the root node and child nodes to form a classification model based on the complete decision tree;

[0024] The classification model is trained and optimized based on the monitoring data collected during the historical period.

[0025] In this solution, the S108 is specifically:

[0026] During a planting cycle, environmental characteristic data and pest characteristic data of non-correlated monitoring points are obtained and imported into the classification model for real-time classification of monitoring points to obtain multiple groups of monitoring points;

[0027] For each associated monitoring point, environmental status prediction and pest status prediction are performed based on environmental characteristic data and pest characteristic data to obtain environmental and pest prediction data;

[0028] Mapping the environment and pest prediction data of each associated monitoring point to each group of monitoring points, and performing pest distribution prediction based on the pest prediction data of each monitoring point in combination with the map model to obtain a pest prediction distribution map;

[0029] Based on the pest prediction distribution map, environment and pest prediction data, N kinds of pest control plans are set for the associated monitoring points and applied to multiple groups of monitoring points.

[0030] A second aspect of the present invention further provides a sweet potato planting pest analysis system based on multiple environmental characteristics, the system comprising: a memory and a processor, wherein the memory includes a sweet potato planting pest analysis program based on multiple environmental characteristics, and when the sweet potato planting pest analysis program based on multiple environmental characteristics is executed by the processor, the following steps are implemented:

[0031] S102: In the sweet potato planting area, multiple monitoring points are set, and multi-dimensional environmental characteristic data and pest characteristic data of the multiple monitoring points are obtained in a historical period, and the environmental characteristic data and pest characteristic data are converted into multiple data items;

[0032] S104: Taking each monitoring point as an analysis unit, performing association analysis on the data items based on the Apriori association algorithm, and mapping them to the correlation analysis of environmental characteristics and pest characteristics, and screening out associated monitoring points;

[0033] S106: Constructing a classification model based on a decision tree. Within a planting cycle, real-time environmental characteristic data and pest characteristic data of associated monitoring points are obtained and converted into decision nodes. Nodes of the classification model are constructed based on the decision nodes, and prediction training is performed on the classification model.

[0034] S108: During a planting cycle, environmental characteristic data and pest characteristic data of non-associated monitoring points are obtained and imported into a classification model to perform real-time classification of the monitoring points to obtain classification results. Pest analysis is performed based on associated monitoring points, and pest association prediction is performed on multiple groups of monitoring points based on the classification results to generate N types of pest control plans.

[0035] The third aspect of the present invention also provides a computer-readable storage medium, which includes a sweet potato planting pest analysis program based on multiple environmental characteristics. When the sweet potato planting pest analysis program based on multiple environmental characteristics is executed by a processor, the steps of the sweet potato planting pest analysis method based on multiple environmental characteristics as described in any one of the above items are implemented.

[0036] The present invention discloses a sweet potato planting pest analysis method and system based on multiple environmental characteristics, comprising: setting multiple monitoring points in the sweet potato planting area, obtaining multi-dimensional environmental characteristics and pest characteristic data within a historical period and converting them into data items; using the Apriori association algorithm to perform association analysis on the data items, mapping the correlation between environmental and pest characteristics, and screening associated monitoring points; constructing a classification model based on a decision tree, obtaining associated monitoring point data in real time and converting them into decision nodes, and performing classification model construction and prediction training; finally, obtaining non-associated monitoring point data and importing it into the classification model for real-time classification, combining the associated monitoring point analysis results to perform pest association prediction, and generating multiple efficient pest control solutions. The present invention effectively integrates multi-source environmental data and pest data, improves pest prediction accuracy, realizes multi-region association prediction, and provides a scientific control basis for sweet potato planting. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 A flowchart of a sweet potato planting pest analysis method based on multiple environmental characteristics of the present invention is shown;

[0038] Figure 2 A block diagram of a sweet potato planting pest analysis system based on multiple environmental characteristics of the present invention is shown. DETAILED DESCRIPTION

[0039] In order to more clearly understand the above-mentioned objects, 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, in the absence of conflict, the embodiments of the present application and the features therein can be combined with each other.

[0040] 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 scope of protection of the present invention is not limited to the specific embodiments disclosed below.

[0041] Figure 1 The present invention shows a flowchart of a sweet potato planting pest analysis method based on multiple environmental characteristics.

[0042] like Figure 1 As shown, the first aspect of the present invention provides a sweet potato planting pest analysis method based on multiple environmental characteristics, comprising:

[0043] S102: In the sweet potato planting area, multiple monitoring points are set, and multi-dimensional environmental characteristic data and pest characteristic data of the multiple monitoring points are obtained in a historical period, and the environmental characteristic data and pest characteristic data are converted into multiple data items;

[0044] S104: Taking each monitoring point as an analysis unit, performing association analysis on the data items based on the Apriori association algorithm, and mapping them to the correlation analysis of environmental characteristics and pest characteristics, and screening out associated monitoring points;

[0045] S106: Constructing a classification model based on a decision tree. Within a planting cycle, real-time environmental characteristic data and pest characteristic data of associated monitoring points are obtained and converted into decision nodes. Nodes of the classification model are constructed based on the decision nodes, and prediction training is performed on the classification model.

[0046] S108: During a planting cycle, environmental characteristic data and pest characteristic data of non-associated monitoring points are obtained and imported into a classification model to perform real-time classification of the monitoring points to obtain classification results. Pest analysis is performed based on associated monitoring points, and pest association prediction is performed on multiple groups of monitoring points based on the classification results to generate N types of pest control plans.

[0047] According to an embodiment of the present invention, the S102 is specifically:

[0048] Construct a map model based on sweet potato planting areas;

[0049] Based on the distribution of each sweet potato planting site, multiple monitoring points are set from the map model to ensure that the planting density of each monitoring point is consistent within the preset range;

[0050] A monitoring unit is set up based on each monitoring point. During a historical period, multi-dimensional environmental characteristic data and pest characteristic data of each monitoring point are obtained through the monitoring unit. Information is extracted for each environmental characteristic and each pest characteristic to form multiple data items.

[0051] It should be noted that the map model is a 2D model used to visualize and analyze planting conditions and pest infestations across different regions. Monitoring units, including soil monitoring devices, meteorological monitoring devices, and plant image detection devices, are used to obtain planting monitoring data. The historical period is a longer period used to analyze a certain amount of monitoring data to characterize the environmental, planting, and pest characteristics of sweet potato planting areas.

[0052] Environmental characteristics include multi-dimensional information such as temperature, humidity, light, soil EC value, and rainfall. Pest characteristics include data such as the area of ​​sweet potato leaf lesions, pest species monitored, number of pests monitored, and number of trapped pests. Pest types include sweet potato leaf beetles, sweet potato hawk moths, sweet potato weevils, stem nematodes, and aphids.

[0053] According to an embodiment of the present invention, the S104 is specifically as follows:

[0054] For a historical period, multiple time nodes are divided according to the preset interval T;

[0055] Taking a monitoring point as the analysis unit, for each time node, the data items of the corresponding node are obtained from the multi-dimensional environmental characteristic data and pest characteristic data;

[0056] Taking the data items of the corresponding nodes as data item sets, and filtering out frequent item sets with minimum support based on the Apriori association algorithm;

[0057] Calculate the confidence in the frequent item set, filter the rules according to the minimum confidence, and obtain the association rules;

[0058] Based on the association rules, the association between data items is determined. If the data items corresponding to the environmental characteristics are associated with the data items corresponding to the pest characteristics, and the number of associations is greater than the preset association quantity, the current time node is marked as an associated node.

[0059] For a monitoring point, if each time node is an associated node, the monitoring point is marked as an associated monitoring point.

[0060] It should be noted that the preset interval T is an adjustable parameter. The larger T is, the smaller the amount of data divided into time nodes, which can be used to dynamically control the amount of data analyzed. Here, the present invention divides the historical period into multiple time nodes for refined analysis of the correlation of monitoring data under different nodes, thereby screening out related monitoring points, and can effectively improve data utilization and data correlation analysis accuracy. The time interval T can be dynamically adjusted based on computing power resources to dynamically adjust the data volume for correlation analysis of data items.

[0061] The data items for a corresponding node include multiple data items, each corresponding to a characteristic type (environmental or pest). A certain amount of monitoring data can be acquired at a time node and converted into multiple data items for correlation analysis. The preset correlation quantity can be set to 3, meaning that at least three groups of data items are correlated, with each group of correlated data items containing at least one environmental characteristic and at least one pest characteristic.

[0062] According to an embodiment of the present invention, the S106 is specifically as follows:

[0063] Build a classification model based on decision trees;

[0064] During a planting cycle, the environmental characteristic data and pest characteristic data of a related monitoring point are obtained in real time, and each characteristic data is converted into a conditional node of a decision tree to generate multiple conditional nodes;

[0065] Generate conditional nodes for all associated monitoring points, and based on the heuristic algorithm, add the conditional nodes to the classification model, and determine the root node and child nodes to form a classification model based on the complete decision tree;

[0066] The classification model is trained and optimized based on the monitoring data collected during the historical period.

[0067] It should be noted that the classification model initially starts with an empty node tree. Nodes are subsequently added gradually, and the heuristic algorithm ID3 is introduced to determine node positions and form a complete decision tree. The classification model is trained and optimized based on historical monitoring data. Multidimensional environmental and pest characteristic data from multiple monitoring points during this period are used as training data for classification and prediction training. The planting cycle is a user-defined period used for periodic pest control and crop ecological regulation. The cycle length can be adjusted based on the density and analysis requirements of the monitoring period.

[0068] In the obtained associated monitoring points, for the classification model, each associated monitoring point is a category. The role of the classification model is to classify the monitoring data of a certain monitoring point (multi-dimensional environmental characteristic data and pest characteristic data), and determine whether a certain monitoring point is similar to the associated monitoring point, or determine it to be another monitoring point. If it is similar to the associated monitoring point, determine which monitoring data feature is closest to the associated monitoring point and classify it into one category.

[0069] It's worth mentioning that traditional sweet potato planting and pest analysis consumes significant monitoring resources across multiple planting areas, making comprehensive analysis of multi-dimensional environmental and pest data difficult. Planting analysis and pest prediction often rely on a single type of data, lacking effective methods for regional sweet potato planting correlation analysis and multi-region pest correlation prediction. This results in inefficient prevention and control strategies for large-scale planting environments, making it difficult to achieve comprehensive planting assessments and pest impact control. For large-scale sweet potato plantings, the complex and variable environmental and pest data from different monitoring points makes comprehensive assessment, analysis, and correlation prediction difficult with existing technologies.

[0070] Based on this, the present invention collects multiple environmental characteristics and multiple pest characteristics in multiple planting areas, evaluates the correlation between environmental factors and pest factors based on the data item form and the Apriori association algorithm, and uses a multi-node form for judgment, thereby screening out monitoring point areas where the environment and pest factors have a correlation status (i.e., associated monitoring points). The associated monitoring points have a higher monitoring and analysis significance for the overall planting area. Through the associated monitoring points, the environmental changes and pest predictions of the overall planting area can be correlated and analyzed, and an efficient monitoring plan can be set, thereby realizing efficient analysis and efficient monitoring of planting and pests based on multidimensional data. Furthermore, in a planting cycle, the present invention performs conditional transformation on the real-time data collected by the associated monitoring points, constructs a classification decision tree (i.e., a classification model), and based on the decision tree, can perform real-time data collection and classification of the monitoring points in the subsequent real-time planting cycle, set multiple groups of monitoring points, and based on the monitoring points in the same group, their environmental change status and pest prediction evaluation are highly consistent with the corresponding associated monitoring points. Therefore, the planting environment evaluation and pest prediction analysis are performed on the associated monitoring points, and the generated pest control plan can be applied to other monitoring points in the same group, thereby improving the efficiency of associated pest control and achieving cost reduction and efficiency improvement in sweet potato planting.

[0071] The decision tree model can realize real-time and rapid classification of monitoring points, use machine learning models to perform real-time classification and evaluation of monitoring data, dynamically adjust associated areas and maintain the accuracy of predictive analysis, and effectively optimize the setting of prevention and control strategies. A plan setting is performed once in each real-time cycle, which can be applied to multiple groups of regions for simultaneous governance, reducing the manpower and material resources consumed by separate analysis of each region, while improving application efficiency and prevention and control effects.

[0072] According to an embodiment of the present invention, the S108 is specifically as follows:

[0073] During a planting cycle, environmental characteristic data and pest characteristic data of non-correlated monitoring points are obtained and imported into the classification model for real-time classification of monitoring points to obtain multiple groups of monitoring points;

[0074] For each associated monitoring point, environmental status prediction and pest status prediction are performed based on environmental characteristic data and pest characteristic data to obtain environmental and pest prediction data;

[0075] Mapping the environment and pest prediction data of each associated monitoring point to each group of monitoring points, and performing pest distribution prediction based on the pest prediction data of each monitoring point in combination with the map model to obtain a pest prediction distribution map;

[0076] Based on the pest prediction distribution map, environment and pest prediction data, N kinds of pest control plans are set for the associated monitoring points and applied to multiple groups of monitoring points.

[0077] It should be noted that, among the multiple groups of monitoring points, each group of monitoring points corresponds to a type of associated monitoring points, and the real-time solution for an associated monitoring point can be applied to the remaining monitoring points in the group. One monitoring point represents a monitored planting area. The environment and pest prediction data mapping of an associated monitoring point is associated with a group of monitoring points. N is the number of associated monitoring points, which is also the number of monitoring point groups. In the prediction of associated monitoring points, since there is a certain correlation between the environment and pests for the monitoring points in the same monitoring group, the prediction based on the associated monitoring points can be applied to the pest prediction of the remaining monitoring points, and the pest distribution and environmental status distribution analysis can be carried out in combination with the map model, and then an efficient prevention and control plan can be set up. The prevention and control plan includes biological pesticide spraying, natural enemy release, planting density adjustment, organic nutrient supply, irrigation setting, etc.

[0078] According to an embodiment of the present invention, the further embodiment includes:

[0079] In a planting cycle, the multi-dimensional environmental characteristic data and pest characteristic data of the associated monitoring points are serialized, and a data sequence is formed based on each characteristic to obtain the environmental sequence and pest sequence;

[0080] Using the LSTM prediction model, we set the prediction interval to one planting cycle, predicted the environmental sequence and the pest sequence, and obtained the environmental prediction sequence and the pest prediction sequence.

[0081] Based on the classification results, the classification probability of non-associated monitoring points is obtained, and the prediction deviation rate is set according to the classification probability;

[0082] The prediction sequence is numerically adjusted based on the deviation rate to obtain the environmental prediction sequence and pest prediction sequence of the non-correlated monitoring points, and the deviation rate of the prediction sequence of the non-correlated monitoring points and the related monitoring points is ensured to be equal to the prediction deviation rate;

[0083] The environment and pest status are assessed and control plans are analyzed through the prediction sequence of all monitoring points.

[0084] It should be noted that the environmental sequence and the pest sequence both include multiple sequences, which are simplified here by feature type. The prediction sequence includes two categories: environmental sequence and pest sequence. Each monitoring point for which there is a classification result corresponds to a prediction deviation rate. The environmental and pest prediction data (i.e., the prediction sequence) can be used for prediction evaluation and program analysis through the above-mentioned sequence prediction form. The classification probability is the probability value of the non-associated monitoring point being finally classified into a certain category through the decision tree. The larger the value, the more similar the monitoring characteristics of the monitoring point are to the associated monitoring point of this category, and the prediction deviation rate is inversely proportional to the classification probability. The prediction deviation rate is used to set the degree of deviation between the pest and environmental monitoring conditions of non-monitoring points and the prediction data of associated monitoring points. The deviation rate is set based on the classification probability and is subsequently adjusted based on the prediction data of the associated monitoring points to improve the overall prediction accuracy of the planting area. At the same time, the prevention and control plan can be adjusted to a certain extent.

[0085] Figure 2 A block diagram of a sweet potato planting pest analysis system based on multiple environmental characteristics of the present invention is shown.

[0086] A second aspect of the present invention further provides a sweet potato planting pest analysis system 2 based on multiple environmental characteristics, the system comprising: a memory 21 and a processor 22, wherein the memory 21 includes a sweet potato planting pest analysis program based on multiple environmental characteristics, and when the sweet potato planting pest analysis program based on multiple environmental characteristics is executed by the processor 22, the following steps are implemented:

[0087] S102: In the sweet potato planting area, multiple monitoring points are set, and multi-dimensional environmental characteristic data and pest characteristic data of the multiple monitoring points are obtained in a historical period, and the environmental characteristic data and pest characteristic data are converted into multiple data items;

[0088] S104: Taking each monitoring point as an analysis unit, performing association analysis on the data items based on the Apriori association algorithm, and mapping them to the correlation analysis of environmental characteristics and pest characteristics, and screening out associated monitoring points;

[0089] S106: Constructing a classification model based on a decision tree. Within a planting cycle, real-time environmental characteristic data and pest characteristic data of associated monitoring points are obtained and converted into decision nodes. Nodes of the classification model are constructed based on the decision nodes, and prediction training is performed on the classification model.

[0090] S108: During a planting cycle, environmental characteristic data and pest characteristic data of non-associated monitoring points are obtained and imported into a classification model to perform real-time classification of the monitoring points to obtain classification results. Pest analysis is performed based on associated monitoring points, and pest association prediction is performed on multiple groups of monitoring points based on the classification results to generate N types of pest control plans.

[0091] According to an embodiment of the present invention, the S102 is specifically:

[0092] Construct a map model based on sweet potato planting areas;

[0093] Based on the distribution of each sweet potato planting site, multiple monitoring points are set from the map model to ensure that the planting density of each monitoring point is consistent within the preset range;

[0094] A monitoring unit is set up based on each monitoring point. During a historical period, multi-dimensional environmental characteristic data and pest characteristic data of each monitoring point are obtained through the monitoring unit. Information is extracted for each environmental characteristic and each pest characteristic to form multiple data items.

[0095] It should be noted that the map model is a 2D model used to visualize and analyze planting conditions and pest infestations across different regions. Monitoring units, including soil monitoring devices, meteorological monitoring devices, and plant image detection devices, are used to obtain planting monitoring data. The historical period is a longer period used to analyze a certain amount of monitoring data to characterize the environmental, planting, and pest characteristics of sweet potato planting areas.

[0096] Environmental characteristics include multi-dimensional information such as temperature, humidity, light, soil EC value, and rainfall. Pest characteristics include data such as the area of ​​sweet potato leaf lesions, pest species monitored, number of pests monitored, and number of trapped pests. Pest types include sweet potato leaf beetles, sweet potato hawk moths, sweet potato weevils, stem nematodes, and aphids.

[0097] According to an embodiment of the present invention, the S104 is specifically as follows:

[0098] For a historical period, multiple time nodes are divided according to the preset interval T;

[0099] Taking a monitoring point as the analysis unit, for each time node, the data items of the corresponding node are obtained from the multi-dimensional environmental characteristic data and pest characteristic data;

[0100] Taking the data items of the corresponding nodes as data item sets, and filtering out frequent item sets with minimum support based on the Apriori association algorithm;

[0101] Calculate the confidence in the frequent item set, filter the rules according to the minimum confidence, and obtain the association rules;

[0102] Based on the association rules, the association between data items is determined. If the data items corresponding to the environmental characteristics are associated with the data items corresponding to the pest characteristics, and the number of associations is greater than the preset association quantity, the current time node is marked as an associated node.

[0103] For a monitoring point, if each time node is an associated node, the monitoring point is marked as an associated monitoring point.

[0104] It should be noted that the preset interval T is an adjustable parameter. The larger T is, the smaller the amount of data divided into time nodes, which can be used to dynamically control the amount of data analyzed. Here, the present invention divides the historical period into multiple time nodes for refined analysis of the correlation of monitoring data under different nodes, thereby screening out related monitoring points, and can effectively improve data utilization and data correlation analysis accuracy. The time interval T can be dynamically adjusted based on computing power resources to dynamically adjust the data volume for correlation analysis of data items.

[0105] The data items for a corresponding node include multiple data items, each corresponding to a characteristic type (environmental or pest). A certain amount of monitoring data can be acquired at a time node and converted into multiple data items for correlation analysis. The preset correlation quantity can be set to 3, meaning that at least three groups of data items are correlated, with each group of correlated data items containing at least one environmental characteristic and at least one pest characteristic.

[0106] According to an embodiment of the present invention, the S106 is specifically as follows:

[0107] Build a classification model based on decision trees;

[0108] During a planting cycle, the environmental characteristic data and pest characteristic data of a related monitoring point are obtained in real time, and each characteristic data is converted into a conditional node of a decision tree to generate multiple conditional nodes;

[0109] Generate conditional nodes for all associated monitoring points, and based on the heuristic algorithm, add the conditional nodes to the classification model, and determine the root node and child nodes to form a classification model based on the complete decision tree;

[0110] The classification model is trained and optimized based on the monitoring data collected during the historical period.

[0111] It should be noted that the classification model initially starts with an empty node tree. Nodes are subsequently added gradually, and the heuristic algorithm ID3 is introduced to determine node positions and form a complete decision tree. The classification model is trained and optimized based on historical monitoring data. Multidimensional environmental and pest characteristic data from multiple monitoring points during this period are used as training data for classification and prediction training. The planting cycle is a user-defined period used for periodic pest control and crop ecological regulation. The cycle length can be adjusted based on the density and analysis requirements of the monitoring period.

[0112] In the obtained associated monitoring points, for the classification model, each associated monitoring point is a category. The role of the classification model is to classify the monitoring data of a certain monitoring point (multi-dimensional environmental characteristic data and pest characteristic data), and determine whether a certain monitoring point is similar to the associated monitoring point, or determine it to be another monitoring point. If it is similar to the associated monitoring point, determine which monitoring data feature is closest to the associated monitoring point and classify it into one category.

[0113] It's worth mentioning that traditional sweet potato planting and pest analysis consumes significant monitoring resources across multiple planting areas, making comprehensive analysis of multi-dimensional environmental and pest data difficult. Planting analysis and pest prediction often rely on a single type of data, lacking effective methods for regional sweet potato planting correlation analysis and multi-region pest correlation prediction. This results in inefficient prevention and control strategies for large-scale planting environments, making it difficult to achieve comprehensive planting assessments and pest impact control. For large-scale sweet potato plantings, the complex and variable environmental and pest data from different monitoring points makes comprehensive assessment, analysis, and correlation prediction difficult with existing technologies.

[0114] Based on this, the present invention collects multiple environmental characteristics and multiple pest characteristics in multiple planting areas, evaluates the correlation between environmental factors and pest factors based on the data item form and the Apriori association algorithm, and uses a multi-node form for judgment, thereby screening out monitoring point areas where the environment and pest factors have a correlation status (i.e., associated monitoring points). The associated monitoring points have a higher monitoring and analysis significance for the overall planting area. Through the associated monitoring points, the environmental changes and pest predictions of the overall planting area can be correlated and analyzed, and an efficient monitoring plan can be set, thereby realizing efficient analysis and efficient monitoring of planting and pests based on multidimensional data. Furthermore, in a planting cycle, the present invention performs conditional transformation on the real-time data collected by the associated monitoring points, constructs a classification decision tree (i.e., a classification model), and based on the decision tree, can perform real-time data collection and classification of the monitoring points in the subsequent real-time planting cycle, set multiple groups of monitoring points, and based on the monitoring points in the same group, their environmental change status and pest prediction evaluation are highly consistent with the corresponding associated monitoring points. Therefore, the planting environment evaluation and pest prediction analysis are performed on the associated monitoring points, and the generated pest control plan can be applied to other monitoring points in the same group, thereby improving the efficiency of associated pest control and achieving cost reduction and efficiency improvement in sweet potato planting.

[0115] The decision tree model can realize real-time and rapid classification of monitoring points, use machine learning models to perform real-time classification and evaluation of monitoring data, dynamically adjust associated areas and maintain the accuracy of predictive analysis, and effectively optimize the setting of prevention and control strategies. A plan setting is performed once in each real-time cycle, which can be applied to multiple groups of regions for simultaneous governance, reducing the manpower and material resources consumed by separate analysis of each region, while improving application efficiency and prevention and control effects.

[0116] According to an embodiment of the present invention, the S108 is specifically as follows:

[0117] During a planting cycle, environmental characteristic data and pest characteristic data of non-correlated monitoring points are obtained and imported into the classification model for real-time classification of monitoring points to obtain multiple groups of monitoring points;

[0118] For each associated monitoring point, environmental status prediction and pest status prediction are performed based on environmental characteristic data and pest characteristic data to obtain environmental and pest prediction data;

[0119] Mapping the environment and pest prediction data of each associated monitoring point to each group of monitoring points, and performing pest distribution prediction based on the pest prediction data of each monitoring point in combination with the map model to obtain a pest prediction distribution map;

[0120] Based on the pest prediction distribution map, environment and pest prediction data, N kinds of pest control plans are set for the associated monitoring points and applied to multiple groups of monitoring points.

[0121] It should be noted that, among the multiple groups of monitoring points, each group of monitoring points corresponds to a type of associated monitoring points, and the real-time solution for an associated monitoring point can be applied to the remaining monitoring points in the group. One monitoring point represents a monitored planting area. The environment and pest prediction data mapping of an associated monitoring point is associated with a group of monitoring points. N is the number of associated monitoring points, which is also the number of monitoring point groups. In the prediction of associated monitoring points, since there is a certain correlation between the environment and pests for the monitoring points in the same monitoring group, the prediction based on the associated monitoring points can be applied to the pest prediction of the remaining monitoring points, and the pest distribution and environmental status distribution analysis can be carried out in combination with the map model, and then an efficient prevention and control plan can be set up. The prevention and control plan includes biological pesticide spraying, natural enemy release, planting density adjustment, organic nutrient supply, irrigation setting, etc.

[0122] The third aspect of the present invention also provides a computer-readable storage medium, which includes a sweet potato planting pest analysis program based on multiple environmental characteristics. When the sweet potato planting pest analysis program based on multiple environmental characteristics is executed by a processor, the steps of the sweet potato planting pest analysis method based on multiple environmental characteristics as described in any one of the above items are implemented.

[0123] The present invention discloses a sweet potato planting pest analysis method and system based on multiple environmental characteristics, comprising: setting multiple monitoring points in the sweet potato planting area, obtaining multi-dimensional environmental characteristics and pest characteristic data within a historical period and converting them into data items; using the Apriori association algorithm to perform association analysis on the data items, mapping the correlation between environmental and pest characteristics, and screening associated monitoring points; constructing a classification model based on a decision tree, obtaining associated monitoring point data in real time and converting them into decision nodes, and performing classification model construction and prediction training; finally, obtaining non-associated monitoring point data and importing it into the classification model for real-time classification, combining the associated monitoring point analysis results to perform pest association prediction, and generating multiple efficient pest control solutions. The present invention effectively integrates multi-source environmental data and pest data, improves pest prediction accuracy, realizes multi-region association prediction, and provides a scientific control basis for sweet potato planting.

[0124] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the units is merely a logical function division. In actual implementation, there may be other division methods, 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 across multiple network units; some or all of the units may be selected according to actual needs to achieve the purpose of the scheme of this 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] Those skilled in the art will appreciate that all or part of the steps of the above-mentioned method embodiments may be implemented by hardware associated with program instructions, and the aforementioned program may be stored in a computer-readable storage medium. When the program is executed, the program executes the steps of the above-mentioned method embodiments. The aforementioned storage medium includes various media that can store program codes, such as mobile storage devices, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical disks.

[0128] Alternatively, if the above-mentioned integrated unit of the present invention is implemented in the form of a software functional 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, or the part that contributes to the prior art, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a number of instructions for enabling a computer device (which can be a personal computer, server, or 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 description is only a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with this technical field can easily think of changes or replacements within the technical scope disclosed by the present invention, which should be covered by the scope of protection of the present invention.

Claims

1. A sweet potato planting pest analysis method based on multiple environmental characteristics, characterized in that: include: S102: In the sweet potato planting area, multiple monitoring points are set, and multi-dimensional environmental characteristic data and pest characteristic data of the multiple monitoring points are obtained in a historical period, and the environmental characteristic data and pest characteristic data are converted into multiple data items; S104: Taking each monitoring point as an analysis unit, performing association analysis on the data items based on the Apriori association algorithm, and mapping them to the correlation analysis of environmental characteristics and pest characteristics, and screening out associated monitoring points; S106: Constructing a classification model based on a decision tree. Within a planting cycle, real-time environmental characteristic data and pest characteristic data of associated monitoring points are obtained and converted into decision nodes. Nodes of the classification model are constructed based on the decision nodes, and prediction training is performed on the classification model. S108: During a planting cycle, environmental characteristic data and pest characteristic data of non-associated monitoring points are obtained and imported into a classification model to perform real-time classification of the monitoring points to obtain classification results. Pest analysis is performed based on associated monitoring points, and pest association prediction is performed on multiple groups of monitoring points based on the classification results to generate N types of pest control plans.

2. A sweet potato planting pest analysis method based on multiple environmental characteristics according to claim 1, characterized in that, The S102 is specifically as follows: Construct a map model based on sweet potato planting areas; Based on the distribution of each sweet potato planting site, multiple monitoring points are set from the map model to ensure that the planting density of each monitoring point is consistent within the preset range; A monitoring unit is set up based on each monitoring point. During a historical period, multi-dimensional environmental characteristic data and pest characteristic data of each monitoring point are obtained through the monitoring unit. Information is extracted for each environmental characteristic and each pest characteristic to form multiple data items.

3. A sweet potato planting pest analysis method based on multiple environmental characteristics according to claim 1, characterized in that, The S104 is specifically as follows: For a historical period, multiple time nodes are divided according to the preset interval T; Taking a monitoring point as the analysis unit, for each time node, the data items of the corresponding node are obtained from the multi-dimensional environmental characteristic data and pest characteristic data; Taking the data items of the corresponding nodes as data item sets, and filtering out frequent item sets with minimum support based on the Apriori association algorithm; Calculate the confidence in the frequent item set, filter the rules according to the minimum confidence, and obtain the association rules; Based on the association rules, the association between data items is determined. If the data items corresponding to the environmental characteristics are associated with the data items corresponding to the pest characteristics, and the number of associations is greater than the preset association quantity, the current time node is marked as an associated node. For a monitoring point, if each time node is an associated node, the monitoring point is marked as an associated monitoring point.

4. A sweet potato planting pest analysis method based on multiple environmental characteristics according to claim 1, characterized in that, The S106 is specifically as follows: Build a classification model based on decision trees; During a planting cycle, the environmental characteristic data and pest characteristic data of a related monitoring point are obtained in real time, and each characteristic data is converted into a conditional node of a decision tree to generate multiple conditional nodes; Generate conditional nodes for all associated monitoring points, and based on the heuristic algorithm, add the conditional nodes to the classification model, and determine the root node and child nodes to form a classification model based on the complete decision tree; The classification model is trained and optimized based on the monitoring data collected during the historical period.

5. A sweet potato planting pest analysis method based on multiple environmental characteristics according to claim 1, characterized in that, The S108 is specifically as follows: During a planting cycle, environmental characteristic data and pest characteristic data of non-correlated monitoring points are obtained and imported into the classification model for real-time classification of monitoring points to obtain multiple groups of monitoring points; For each associated monitoring point, environmental status prediction and pest status prediction are performed based on environmental characteristic data and pest characteristic data to obtain environmental and pest prediction data; Mapping the environment and pest prediction data of each associated monitoring point to each group of monitoring points, and performing pest distribution prediction based on the pest prediction data of each monitoring point in combination with the map model to obtain a pest prediction distribution map; Based on the pest prediction distribution map, environment and pest prediction data, N kinds of pest control plans are set for the associated monitoring points and applied to multiple groups of monitoring points.

6. A sweet potato planting pest analysis system based on multiple environmental characteristics, characterized in that: The system includes: a memory and a processor. The memory includes a sweet potato planting pest analysis program based on multiple environmental characteristics. When the sweet potato planting pest analysis program based on multiple environmental characteristics is executed by the processor, the following steps are implemented: S102: In the sweet potato planting area, multiple monitoring points are set, and multi-dimensional environmental characteristic data and pest characteristic data of the multiple monitoring points are obtained in a historical period, and the environmental characteristic data and pest characteristic data are converted into multiple data items; S104: Taking each monitoring point as an analysis unit, performing association analysis on the data items based on the Apriori association algorithm, and mapping them to the correlation analysis of environmental characteristics and pest characteristics, and screening out associated monitoring points; S106: Constructing a classification model based on a decision tree. Within a planting cycle, real-time environmental characteristic data and pest characteristic data of associated monitoring points are obtained and converted into decision nodes. Nodes of the classification model are constructed based on the decision nodes, and prediction training is performed on the classification model. S108: During a planting cycle, environmental characteristic data and pest characteristic data of non-associated monitoring points are obtained and imported into a classification model to perform real-time classification of the monitoring points to obtain classification results. Pest analysis is performed based on associated monitoring points, and pest association prediction is performed on multiple groups of monitoring points based on the classification results to generate N types of pest control plans.

7. A computer-readable storage medium, characterized in that The computer-readable storage medium includes a sweet potato planting pest analysis program based on multiple environmental characteristics. When the sweet potato planting pest analysis program based on multiple environmental characteristics is executed by a processor, the steps of the sweet potato planting pest analysis method based on multiple environmental characteristics as described in any one of claims 1 to 5 are implemented.

Images