Phytoseiulus mite field breeding regulation and control method and system in combination with block pest data

Abstract

The invention discloses a phytoseiulus mite field breeding regulation and control method and system in combination with block pest data, and relates to the technical field of breeding management and control. The method comprises the following steps: for a target vegetation area, carrying out pest data calling and pest mite feature mining in a vegetation growth cycle, and determining distribution mode features; carrying out clustering analysis and trend mining on the distribution mode characteristics in a space-time dimension, and converting the distribution mode characteristics into a characteristic array; the breeding information of the phytoseiulus mites is determined; putting decision making and breeding optimization decision making are carried out, and a pre-regulation and control strategy is determined; and according to the pre-regulation strategy, putting control and breeding control of the phytoseiulus mites are carried out. The technical problems that in the prior art, the phytoseiulus mite putting opportunity selection is not accurate, the breeding regulation and control process lacks systematicness and data support, and consequently the prevention and control effect is unstable are solved, the dynamic optimization and accurate regulation and control of phytoseiulus mite putting and field breeding are achieved by fusing the block pest data, and the breeding efficiency is improved. Therefore, the insect pest prevention and control efficiency and the breeding resource utilization rate are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of aquaculture control, and particularly relates to a method and system for regulating the field aquaculture of Phytoseiulus persimilis by combining block pest data. Background Art

[0002] In the process of biological control of agricultural pests and diseases, Phytoseiulus persimilis is widely used in the green control of cash crops such as orchards and tea gardens because of its significant predation on various pest mites. However, the existing methods of field aquaculture and release of Phytoseiulus persimilis generally rely on empirical judgment, lacking dynamic perception of the vegetation growth status, pest occurrence, and distribution characteristics of pest mites, making it difficult to achieve precise matching of release timing and aquaculture regulation, resulting in low resource utilization efficiency and large fluctuations in pest control effects. Summary of the Invention

[0003] The present application provides a method and system for regulating the field aquaculture of Phytoseiulus persimilis by combining block pest data, which solves the technical problems in the prior art that the release timing of Phytoseiulus persimilis is not precise, and the aquaculture regulation process lacks systematicness and data support, resulting in unstable control effects.

[0004] In the first aspect of the present application, a method for regulating the field aquaculture of Phytoseiulus persimilis by combining block pest data is provided. The method includes: For a target vegetation area, call pest data and mine the characteristics of pest mites under the vegetation growth cycle to determine the distribution pattern characteristics; perform cluster analysis and trend mining on the distribution pattern characteristics in the spatio-temporal dimension and convert them into a feature matrix; determine the aquaculture information of Phytoseiulus persimilis, where the aquaculture information includes aquaculture environment - growth cycle; based on the feature matrix as the demand orientation and the aquaculture information as the basis, make a release decision and optimize the aquaculture decision for the target vegetation area to determine a pre-regulation strategy; according to the pre-regulation strategy, conduct release control and aquaculture control of Phytoseiulus persimilis.

[0005] In the second aspect of the present application, a system for regulating the field aquaculture of Phytoseiulus persimilis by combining block pest data is provided. The system includes: A feature mining module for calling pest data and mining the characteristics of pest mites under the vegetation growth cycle for a target vegetation area to determine the distribution pattern characteristics; an analysis module for performing cluster analysis and trend mining on the distribution pattern characteristics in the spatio-temporal dimension and converting them into a feature matrix; an information acquisition module for determining the aquaculture information of Phytoseiulus persimilis, where the aquaculture information includes aquaculture environment - growth cycle; a decision module for making a release decision and optimizing the aquaculture decision for the target vegetation area based on the feature matrix as the demand orientation and the aquaculture information as the basis to determine a pre-regulation strategy; a control module for conducting release control and aquaculture control of Phytoseiulus persimilis according to the pre-regulation strategy.

[0006] One or more technical solutions provided in this application have at least the following technical effects or advantages: First, for the target vegetation area, call pest data during the vegetation growth cycle and mine the characteristics of pest mites to determine the distribution pattern characteristics. Further, perform cluster analysis and trend mining on the distribution pattern characteristics in the spatio-temporal dimension and convert them into a feature matrix. Then, determine the breeding information of phytoseiid mites, where the breeding information includes the breeding environment - growth cycle. Then, with the feature matrix as the demand orientation and the breeding information as the basis, make release decisions and optimize breeding decisions based on the target vegetation area to determine the pre-regulation strategy. Finally, according to the pre-regulation strategy, conduct release control and breeding control of phytoseiid mites. This solves the technical problem in the prior art that the timing of releasing phytoseiid mites is not accurate enough, and the breeding regulation process lacks systematicness and data support, resulting in unstable prevention and control effects, and achieves the technical effect of realizing the dynamic optimization and precise regulation of the release and field breeding of phytoseiid mites by integrating block pest data, thereby improving the pest prevention and control efficiency and the utilization rate of breeding resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0008] Figure 1 It is a schematic flowchart of a method for regulating the field breeding of phytoseiid mites in combination with block pest data provided by an embodiment of this application; Figure 2 It is a schematic structural diagram of a system for regulating the field breeding of phytoseiid mites in combination with block pest data provided by an embodiment of this application.

[0009] Description of the reference numerals: Feature mining module 11, analysis module 12, information acquisition module 13, decision module 14, control module 15. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0010] By providing a method and system for regulating the field breeding of phytoseiid mites in combination with block pest data, this application solves the technical problem in the prior art that the timing of releasing phytoseiid mites is not accurate enough, and the breeding regulation process lacks systematicness and data support, resulting in unstable prevention and control effects.

[0011] The following will clearly and completely describe the technical solutions in the embodiments of the present application with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts belong to the scope of protection of the present application.

[0012] It should be noted that the terms "include" and "have" are intended to cover non-exclusive inclusion. For example, a process, method, system, product or server that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or modules that are not clearly listed or are inherent to these processes, methods, products or devices.

[0013] Embodiment 1, as Figure 1 shown, the present application provides a method for regulating the breeding of Phytoseiulus persimilis in the field by combining block pest data, wherein the method includes: For the target vegetation area, call pest data during the vegetation growth cycle and mine the characteristics of pest mites to determine the distribution pattern characteristics.

[0014] For the target vegetation area, obtain its geographical location information and crop types, and combine the standard growth cycle model of the crop to determine the analysis period and key growth stages. During the analysis period, call the pest data collected by the plant protection monitoring platform or agricultural perception terminal. The pest data includes, but is not limited to, information such as pest types, occurrence frequencies, and spatial distribution coordinates. At the same time, based on the previously deployed pest monitoring devices, image recognition models or manually collected data, mine the characteristics of the pest mite population distribution, density change, activity frequency, etc. in the target area to form a pest mite feature set. On this basis, based on the vegetation space unit division principle (such as by grid, regional block or natural field block), spatially discretize the target vegetation area, and map and match the above pest data and pest mite characteristics in combination with the time dimension, and extract the pest and pest mite index values of each space unit at different time points to form the distribution pattern characteristics in the time-space dimension. This distribution pattern characteristic is used to describe the dynamic changes of plant pests and pest mites in the target area, providing basic data support for subsequent clustering analysis and regulation strategy formulation.

[0015] Conduct clustering analysis and trend mining on the distribution pattern characteristics in the time-space dimension and convert them into a feature matrix.

[0016] After obtaining the distribution pattern features corresponding to each spatio-temporal unit in the target vegetation area, first set a preset class difference threshold for controlling the clustering accuracy as the basis for similarity discrimination in the spatio-temporal clustering process. Subsequently, use a spatio-temporal joint clustering algorithm (such as an improved ST-DBSCAN or a variant of K-means), input the distribution pattern features of each unit synchronously according to the time dimension and the space dimension, automatically identify the area blocks with similar pest-mite feature distributions through the clustering algorithm, and output multiple clustering blocks with spatio-temporal consistency. For each clustering block, extract its representative block features based on the clustering results, including pest conditions, mite density, change trend indicators, etc. On this basis, traverse all block features, combine with the time series of the vegetation growth cycle, analyze the change trend of mite features over time at the same block position, and form a corresponding change curve. This change curve is used to describe the dynamic evolution of mite features in each spatio-temporal unit and is stored with the block number and time series index as identifiers. Finally, combine the above block features and change curves and convert them into a feature matrix, where the row dimension represents the spatial block number, the column dimension represents the time series nodes or feature items, and the matrix elements are the corresponding numerical indicators. This feature matrix can be used as the input for the subsequent phytoseiid release and breeding regulation model and has the characteristics of high structure and strong spatio-temporal dynamics.

[0017] Furthermore, perform clustering analysis on the distribution pattern features in the spatio-temporal dimension, including: Set a preset class difference; according to the preset class difference, perform joint clustering on the distribution pattern features based on the time dimension and the space dimension to determine the clustering blocks; according to the clustering blocks, extract the block features.

[0018] First, set a preset class difference parameter for controlling clustering accuracy. The preset class difference is used to measure the degree of feature similarity between different spatio-temporal units and can be determined based on the results of historical data analysis. For example, the difference threshold between feature vectors can be defined using methods such as Euclidean distance, Manhattan distance, or Mahalanobis distance. Then, based on the preset class difference, a spatio-temporal joint clustering algorithm is used to cluster the distribution pattern features in the target vegetation area. Specifically, the pest index and mite characteristics of each spatio-temporal unit are used to construct a joint feature vector according to the "time dimension + space dimension" and input into a clustering model (such as based on ST-DBSCAN, Gaussian Mixture Model, or density-adjustable K-means algorithm). Automatic grouping is performed based on the similarity between feature vectors to determine multiple clustering blocks with both spatial continuity and feature similarity. After completing the clustering block division, statistical analysis is performed on the distribution pattern feature data in each clustering block to extract its representative block features, such as the pest occurrence situation, mite population density, and co-variation trend within the block. The block features are used to reflect the overall situation of each clustering block in the dynamic changes of pests and diseases, providing a basis for subsequent trend analysis and matrix construction.

[0019] Furthermore, trend mining is carried out, including: Traverse the block features, perform trend analysis on the mite characteristics under the time series of the vegetation growth cycle for the same-position mapping to determine the trend curves, where each trend curve is marked with the regional location; based on the block features and the trend curves, convert and determine the feature matrix.

[0020] First, traverse the block features, that is, process the block features extracted from each clustering block obtained from the previous clustering analysis in turn. Combining with the time series of the vegetation growth cycle corresponding to the target vegetation area, the pest and mite indexes at consecutive time nodes of each block are sorted and matched in the time dimension to form a time series data set under the same-position mapping. Subsequently, based on the time series data, dynamic analysis is performed on the mite characteristics at each spatial position to identify their change trends at different time nodes. This trend analysis can be achieved through fitting methods (such as polynomial fitting, moving average, sliding window difference, etc.) to generate the mite characteristic trend curves corresponding to each clustering block, which are used to reflect the temporal evolution law of the mite population in this area. Each trend curve is marked with its corresponding clustering block or spatial position to ensure position consistency in the subsequent processing. Finally, based on the block features and the trend curves, dimension integration is performed according to the spatial position and time node, and it is converted into a standardized feature matrix. Each row of this feature matrix corresponds to a spatial clustering block, each column corresponds to a time node or feature item, and the numerical unit is the corresponding pest-mite index value or its derived statistic (such as growth rate, volatility, etc.).

[0021] Determine the breeding information of phytoseiid mites, where the breeding information includes the breeding environment - growth cycle.

[0022] First, determine the breeding environment according to the geographical information, climate conditions, and environmental factors of the target vegetation area. The breeding environment includes, but is not limited to, factors such as temperature, humidity, soil type, light intensity, wind speed, etc. that affect the growth and reproduction of phytoseiid mites. Data can be collected in real - time by installing environmental monitoring equipment (such as temperature - humidity sensors, soil detectors, etc.), or predictive analysis can be carried out through historical meteorological data provided by meteorological stations to obtain the required breeding environment parameters. Then, combined with the breeding environment factors and the environmental adaptability requirements of different phytoseiid mite species, determine the range of environmental parameters most suitable for the growth of phytoseiid mites. These environmental parameters will provide a basis for subsequent phytoseiid mite breeding decisions and be used to adjust the actual breeding conditions in the target vegetation area (such as adjusting temperature, humidity, or light, etc.). Next, determine the growth cycle of phytoseiid mites, that is, based on the life - cycle characteristics of phytoseiid mites (such as the survival conditions in different growth stages like eggs, larvae, adult mites, etc.) and the crop growth cycle in the target vegetation area, reasonably divide the growth cycle of phytoseiid mites. This cycle usually includes the preparation period for releasing mites in the early stage, the mite reproduction and growth cycle, and the monitoring and feedback adjustment period of mite efficacy in the later stage. Finally, through the comprehensive analysis of the breeding environment and growth cycle, determine the optimal breeding information of phytoseiid mites, that is, including suitable environmental conditions and optimized cycle arrangements, to ensure that phytoseiid mites can reproduce efficiently and play the best pest control role in the target vegetation area.

[0023] Taking the characteristic matrix as the demand orientation and the breeding information as the basis, make release decisions and optimized breeding decisions based on the target vegetation area, and determine the pre - regulation strategy.

[0024] First, taking the feature matrix as the demand orientation and combining the spatio-temporal characteristics of the target vegetation area, a demand model for release decision-making and breeding regulation is constructed. Each column in the feature matrix represents the dynamic change data of pests and pest mites at different time nodes, and the rows represent the pest characteristics of different spatial blocks. On this basis, combined with the biological characteristics of phytoseiid mites, the required mite release amount, reproduction cycle, and subsequent breeding management requirements at each block and different time nodes are calculated. Then, according to the above demand model, combined with breeding information (including breeding environment parameters and growth cycle), using decision optimization algorithms (such as linear programming, integer programming, or multi-objective optimization algorithms), the release decision-making and optimized breeding decision-making for the target vegetation area are calculated. Specifically, based on the regional characteristic data in the feature matrix, the severity of pests and the population density of pest mites in each area are analyzed, and according to the release effect of phytoseiid mites and the prevention and control requirements of crops, the optimal release amount at each time node and each spatial area is determined; according to the breeding information of phytoseiid mites (including environmental temperature and humidity, growth cycle, etc.) and the release strategy, combined with the requirements of the growth cycle, the growth and reproduction cycle of phytoseiid mites are optimized to ensure the maximization of mite growth and prevention and control effects during the breeding process. Finally, according to the results of the release decision-making and breeding decision-making, a pre-regulation strategy is determined. This pre-regulation strategy is based on the output of the aforementioned decision model, comprehensively considering the pest prevention and control requirements, the specific situation of phytoseiid mite breeding and release, and a holistic pre-regulation plan that adapts to the needs of the target vegetation area is proposed.

[0025] Furthermore, making the release decision-making and optimized breeding decision-making based on the target vegetation area includes: Taking the feature matrix as the orientation, combining with the release decision-making unit to make the block release amount decision and determine the pre-release strategy of phytoseiid mites; taking the pre-release strategy as the orientation, combining with the breeding decision-making unit to make the breeding regulation decision and determine the pre-breeding strategy; adding the pre-release strategy and the pre-breeding strategy into the pre-regulation strategy.

[0026] First, guided by the feature matrix and combined with the release decision-making unit, make a decision on the release amount of each block. According to the pest data and mite characteristics in the feature matrix, analyze the pest occurrence situation and mite population distribution characteristics in different spatial regions, and determine the release requirements of phytoseiid mites at different time nodes in each region. The release decision-making unit uses this data and calculates the most appropriate release amount based on a specific release model (such as a regression model or a machine learning model) to ensure the best pest control effect and resource utilization rate in the target vegetation area. Through this decision-making process, a pre-release strategy for phytoseiid mites is generated, that is, the release amount of phytoseiid mites is accurately set according to the needs in each region. Next, guided by the pre-release strategy and combined with the breeding decision-making unit, make a breeding regulation decision to ensure that the growth and reproduction of phytoseiid mites can match the release strategy. The breeding decision-making unit will estimate its growth cycle according to the life cycle of phytoseiid mites (such as eggs, larvae, adult mites, etc.) and combined with natural conditions such as temperature and humidity in the target area, and optimize the prediction model through historical data to predict the reproduction effect of phytoseiid mites under different seasons and environmental conditions. Furthermore, according to the life cycle of phytoseiid mites and the influence of the natural environment in the field, determine the appropriate breeding density to ensure that the reproduction quantity of phytoseiid mites is sufficient to meet the pest control requirements. Based on the growth cycle and breeding density, formulate a pre-breeding strategy adapted to the actual field conditions to ensure that phytoseiid mites achieve the best control effect at the appropriate time, appropriate density and environment. Finally, integrate the pre-release strategy with the pre-breeding strategy to form an overall pre-regulation strategy, which comprehensively considers the pest control requirements, resource utilization efficiency, biological characteristics of phytoseiid mites and breeding environment, and formulates a complete release and breeding regulation plan for phytoseiid mites.

[0027] Furthermore, construct a release decision-making unit, including: Taking mite characteristics as independent variables and release amount as the dependent variable, mine the linear release relationship according to the pest data; construct a release decision-making unit according to the linear release relationship; integrate the pest data to determine the first training sample, and train the release decision-making unit until convergence.

[0028] First, taking the characteristics of pest mites as independent variables and the release amount as the dependent variable, a linear release relationship is established using the collected pest data. Specifically, first, various characteristics of pest mites (such as pest mite population density, activity frequency, damage degree, etc.) are extracted from the pest data as independent variables; at the same time, according to the requirements of the control effect of phytoseiid mites, the release amount of phytoseiid mites is selected as the dependent variable; then, based on these data, regression analysis methods (such as the least squares method, ridge regression, etc.) are used to explore the linear relationship between the characteristics of pest mites and the release amount. Next, according to the linear release relationship, a release decision-making unit is constructed. The core function of the release decision-making unit is to calculate the required release amount of phytoseiid mites based on the input pest mite characteristic data. This unit transforms the aforementioned linear release relationship into a decision-making model, inputs the real-time monitoring data of pest mites into the model, and automatically calculates the optimal release amount for each region and each time node according to the linear relationship. Finally, the pest data is integrated to determine the first training sample. By combining historical pest data with the corresponding actual release amounts, the first training sample is constructed. The first training sample includes the pairing of pest mite characteristic data and actual release amount data, forming a basic data set for training the release decision-making unit. Using the first training sample, the release decision-making unit is trained through machine learning algorithms (such as the gradient descent method, support vector machine, etc.) to optimize the model parameters until the model converges, that is, to achieve the optimal effect of the release decision-making unit in predicting the release amount of phytoseiid mites.

[0029] Furthermore, a breeding decision-making unit is constructed, including: Taking the breeding environment factors as independent variables and the growth cycle as the dependent variable, a linear breeding relationship is mined according to historical breeding information; according to the linear breeding relationship, a breeding decision-making unit is constructed; the historical breeding information is integrated to determine the second training sample, and the breeding decision-making unit is trained until it converges.

[0030] First, taking the aquaculture environmental factors as independent variables and the growth cycle as the dependent variable, linear aquaculture relationships are mined using historical aquaculture information. Specifically, the historical aquaculture information includes the growth data of phytoseiid mites under different aquaculture environments (such as temperature, humidity, light intensity, etc.) and the corresponding growth cycles (such as the time required from release to the mite reproduction stage). Through regression analysis methods (such as the least squares method, ridge regression, etc.), the linear relationship between aquaculture environmental factors and the growth cycle is analyzed, so as to determine the influence law of environmental factor changes on the growth cycle of phytoseiid mites. This linear relationship provides a mathematical model for subsequent aquaculture decisions and can predict the growth cycle based on environmental data. Next, according to the linear aquaculture relationship, an aquaculture decision-making unit is constructed. The function of this unit is to predict the growth cycle of phytoseiid mites based on the input environmental data. By converting the linear relationship extracted from historical aquaculture data into a decision-making model, the aquaculture decision-making unit can automatically predict the growth cycle of phytoseiid mites according to the current aquaculture environment (such as temperature and humidity changes, etc.), and then guide the regulation measures in the aquaculture process (such as cycle extension or shortening, etc.). This aquaculture decision-making unit can adopt a regression model, neural network or other machine learning models to automatically predict the appropriate growth cycle based on the input environmental data. Finally, historical aquaculture information is integrated to determine the second training sample. According to the environmental factors and actual growth cycle data in historical aquaculture records, the second training sample is constructed. The second training sample includes the corresponding data of environmental factors and actual growth cycles and is used to train the aquaculture decision-making unit. Through the second training sample, machine learning algorithms (such as gradient descent method, support vector machine, etc.) are used to train the aquaculture decision-making unit to optimize the parameters of the model until the model can accurately predict the growth cycle under different environmental conditions. When the training process reaches convergence, the model can provide support for decisions in the actual aquaculture process.

[0031] According to the pre-regulation strategy, the release control and aquaculture control of phytoseiid mites are carried out.

[0032] Under the guidance of the pre-regulation strategy, according to the previously generated pre-release strategy, the pest development situation and the change of pest mite population in the target vegetation area are monitored in real time. Through plant protection monitoring equipment, remote sensing technology or image recognition technology, data such as pest density and pest mite activity in the target area are collected; based on these data, combined with the release quantity decision in the pre-release strategy, the release quantity and release time of phytoseiid mites are adjusted in real time to ensure that the quantity and distribution of phytoseiid mites can effectively respond to the spread of pests and pest mites. Secondly, according to the pre-aquaculture strategy, the aquaculture environment is regulated in real time to ensure that the growth and reproduction conditions of phytoseiid mites in the target area reach the best state. Through an environmental monitoring system (such as temperature and humidity sensors, etc.), aquaculture environmental data are continuously obtained and compared and analyzed with the requirements in the pre-aquaculture strategy; according to the actual environmental changes, environmental parameters are dynamically adjusted, such as adjusting factors such as temperature, humidity, and light through automated equipment to ensure that phytoseiid mites can grow in the best environment.

[0033] Furthermore, after the release control and breeding control of phytoseiid mites are carried out, including: With the implementation of the pre-regulation strategy, the release tracking and breeding supervision are carried out synchronously to determine the first release response and the second breeding response; the response off-axis evaluation is carried out on the first release response to determine the first feedback information; the response off-axis evaluation is carried out on the second breeding response to determine the second feedback information; according to the first feedback information and the second feedback information, the feedback regulation management of the release and breeding of phytoseiid mites is carried out.

[0034] During the implementation of the pre-regulation strategy, the release situation of phytoseiid mites and the breeding environment are tracked and supervised in real time through a monitoring system (such as remote sensing, unmanned aerial vehicle, sensor and other technologies) to determine the first release response and the second breeding response; the first release response, that is, the release effect of phytoseiid mites in the target vegetation area, is evaluated according to the pest control level and the change of the population of harmful mites; the second breeding response, that is, the growth state and the progress of the growth cycle of phytoseiid mites, is evaluated through environmental monitoring and breeding data to understand whether the growth of phytoseiid mites in the breeding process meets the expectations. Next, the response off-axis evaluation is carried out on the first release response. The response off-axis evaluation refers to the deviation analysis of the release effect to judge the difference between the actual release effect and the predetermined target. By comparing the pest control data with the expected prevention and control effect, it is determined whether there is too much or too little release, and the first feedback information is calculated according to the deviation analysis. This information reflects the effect and existing problems of the current release strategy. Similarly, the response off-axis evaluation is carried out on the second breeding response. By comparing the differences between the data such as the growth cycle and reproduction situation of phytoseiid mites and the expected breeding target, the second feedback information is calculated, which can reveal whether there are problems such as improper environmental regulation, too long or too short growth cycle in the breeding process, and thus provide a basis for subsequent control measures to be adjusted. According to the first feedback information and the second feedback information, the feedback regulation management of the release and breeding of phytoseiid mites is carried out. Combining the first feedback information and the second feedback information, the release and breeding strategies are dynamically adjusted by using a feedback control model. For example, if the excessive release amount leads to resource waste or too strong prevention and control effect, the system can adjust the release amount; if the growth cycle is too long or the environmental parameters do not meet the growth requirements of phytoseiid mites, the system will adjust environmental factors such as temperature and humidity or optimize the growth cycle. Through the feedback regulation management, it is ensured that the release and breeding of phytoseiid mites are always in the best state, so as to improve the pest control efficiency, resource utilization rate and overall prevention and control effect.

[0035] In summary, the embodiments of the present application at least have the following technical effects: First, for the target vegetation area, call the pest data during the vegetation growth cycle and mine the characteristics of pest mites to determine the distribution pattern characteristics. Further, perform cluster analysis and trend mining on the distribution pattern characteristics in the spatio-temporal dimension and convert them into a feature matrix. Then, determine the breeding information of phytoseiid mites, where the breeding information includes the breeding environment - growth cycle. Next, with the feature matrix as the demand orientation and the breeding information as the basis, make release decisions and optimize breeding decisions based on the target vegetation area to determine the pre-regulation strategy. Finally, according to the pre-regulation strategy, conduct release control and breeding control of phytoseiid mites. This solves the technical problem in the prior art that the release timing of phytoseiid mites is not accurate enough, and the breeding regulation process lacks systematicness and data support, resulting in unstable prevention and control effects, and achieves the technical effect of realizing the dynamic optimization and precise regulation of the release and field breeding of phytoseiid mites by integrating pest data in blocks, thereby improving the pest prevention and control efficiency and the utilization rate of breeding resources.

[0036] Embodiment 2, based on the same inventive concept as the method for regulating the field breeding of phytoseiid mites by combining block pest data in the foregoing embodiment, as Figure 2 shown, the present application provides a system for regulating the field breeding of phytoseiid mites by combining block pest data, wherein the system includes: A feature mining module 11, configured to call pest data during the vegetation growth cycle and mine the characteristics of pest mites for the target vegetation area to determine the distribution pattern characteristics; an analysis module 12, configured to perform cluster analysis and trend mining on the distribution pattern characteristics in the spatio-temporal dimension and convert them into a feature matrix; an information acquisition module 13, configured to determine the breeding information of phytoseiid mites, where the breeding information includes the breeding environment - growth cycle; a decision module 14, configured to make release decisions and optimize breeding decisions based on the target vegetation area with the feature matrix as the demand orientation and the breeding information as the basis to determine the pre-regulation strategy; a control module 15, configured to perform release control and breeding control of phytoseiid mites according to the pre-regulation strategy.

[0037] Further, the analysis module 12 is configured to execute the following method: Set a preset class difference; according to the preset class difference, perform joint clustering on the distribution pattern characteristics based on the time dimension and the space dimension to determine the clustering blocks; extract block characteristics according to the clustering blocks.

[0038] Further, the analysis module 12 is configured to execute the following method: Traverse the block characteristics, perform an analysis of the trend change of the characteristics of pest mites under the same position mapping with the time series during the vegetation growth cycle to determine the trend change curves, where each trend change curve is marked with the regional position; based on the block characteristics and the trend change curves, convert and determine the feature matrix.

[0039] Further, the decision-making module 14 is configured to execute the following method: Guided by the feature matrix, combine with the placement decision-making unit to make a decision on the block placement amount, and determine the pre-placement strategy of the phytoseiid mite; guided by the pre-placement strategy, combine with the breeding decision-making unit to make a breeding regulation decision, and determine the pre-breeding strategy; add the pre-placement strategy and the pre-breeding strategy to the pre-regulation strategy.

[0040] Further, the decision-making module 14 is configured to execute the following method: Taking the pest mite characteristics as the independent variable and the placement amount as the dependent variable, mine the linear placement relationship according to the pest data; construct a placement decision-making unit according to the linear placement relationship; integrate the pest data to determine the first training sample, and train the placement decision-making unit until convergence.

[0041] Further, the decision-making module 14 is configured to execute the following method: Taking the breeding environmental factors as the independent variable and the growth cycle as the dependent variable, mine the linear breeding relationship according to the historical breeding information; construct a breeding decision-making unit according to the linear breeding relationship; integrate the historical breeding information to determine the second training sample, and train the breeding decision-making unit until convergence.

[0042] Further, the control module 15 is configured to execute the following method: Synchronously conduct placement tracking and breeding supervision along with the execution of the pre-regulation strategy to determine the first placement response and the second breeding response; conduct response off-axis evaluation on the first placement response to determine the first feedback information; conduct response off-axis evaluation on the second breeding response to determine the second feedback information; according to the first feedback information and the second feedback information, conduct feedback regulation management on the placement and breeding of the phytoseiid mite.

[0043] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description and does not represent the superiority or inferiority of the embodiments. And the above description of specific embodiments of this specification has been made. The processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0044] The above are only the preferred embodiments of the present application and are not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

[0045] This specification and the accompanying drawings are merely illustrative of the present application and are considered to cover any and all modifications, variations, combinations, or equivalents within the scope of the present application. Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalent technologies, the present application is intended to include these changes and modifications.

Claims

1. A method for regulating and controlling phytoseiid mites in field cultivation in combination with block pest data, characterized in that: The method comprises: For the target vegetation area, call the pest data and mine the characteristics of harmful mites during the vegetation growth cycle to determine the distribution pattern characteristics; Performing cluster analysis and trend mining on the distribution pattern characteristics in the spatiotemporal dimension and converting them into feature arrays; Determining the breeding information of the phytoseiid mite, wherein the breeding information includes the breeding environment-growth cycle; Taking the characteristic array as demand orientation and the breeding information as basis, making a placement decision and optimizing a breeding decision based on the target vegetation area, and determining a pre-regulation strategy; According to the pre-regulation strategy, the release and breeding of phytoseiid mites are controlled.

2. The method for controlling the cultivation of phytoseiid mites in the field in combination with block pest data according to claim 1, characterized in that: Performing cluster analysis on the distribution pattern characteristics in the spatiotemporal dimension includes: Set the preset class difference; According to the preset class difference, the distribution pattern features are jointly clustered based on the time dimension and the space dimension to determine the clustering block; According to the clustered blocks, block features are extracted.

3. The method for controlling the cultivation of phytoseiid mites in the field in combination with block pest data according to claim 2, characterized in that: Conduct trend mining, including: Traversing the block features, performing a trend change analysis of the harmful mite features under the same position mapping in a time series under the vegetation growth cycle, and determining a trend change curve, wherein each trend change curve is marked with a regional position; Based on the block features and the trend curve, a feature matrix is ​​determined by conversion.

4. The method for controlling phytoseiid mite field breeding in combination with block pest data according to claim 1, characterized in that: Make placement decisions and optimize breeding decisions based on the target vegetation area, including: Guided by the characteristic array, the block release amount decision is made in combination with the release decision unit to determine the pre-release strategy of phytoseiid mites; Guided by the pre-stocking strategy, the aquaculture regulation and control decision is made in combination with the aquaculture decision-making unit to determine the pre-aquaculture strategy; The pre-delivery strategy and the pre-breeding strategy are added to the pre-regulation strategy.

5. The method for controlling the cultivation of phytoseiid mites in the field in combination with block pest data according to claim 4, characterized in that: Construct a delivery decision unit, including: Taking the characteristics of harmful mites as independent variables and the release amount as dependent variable, mining the linear release relationship according to the pest data; According to the linear delivery relationship, a delivery decision unit is constructed; The pest data is integrated to determine a first training sample, and the placement decision unit is trained until convergence.

6. The method for controlling phytoseiid mite field breeding in combination with block pest data according to claim 4, characterized in that: Construct a breeding decision-making unit, including: Taking the aquaculture environment factors as independent variables and the growth cycle as dependent variables, the linear aquaculture relationship is mined based on historical aquaculture information; According to the linear breeding relationship, a breeding decision unit is constructed; The historical breeding information is integrated to determine a second training sample, and the breeding decision unit is trained until convergence.

7. The method for controlling phytoseiid mite field breeding in combination with block pest data according to claim 1, characterized in that: After the release and breeding control of phytoseiid mites, including: With the execution of the pre-regulation strategy, release tracking and breeding supervision are carried out simultaneously to determine a first release response and a second breeding response; Performing a response off-axis evaluation on the first delivery response to determine first feedback information; Performing off-axis response evaluation on the second breeding response to determine second feedback information; Feedback regulation and management of the release and breeding of phytoseiid mites is performed based on the first feedback information and the second feedback information.

8. A control system for phytoseiid field breeding combined with block pest data, characterized in that: A method for controlling the field breeding of phytoseiid mites in combination with block pest data according to any one of claims 1 to 7, the system comprising: The feature mining module is used to call pest data and mine pest mite features during the vegetation growth cycle in the target vegetation area to determine the distribution pattern characteristics; An analysis module, used to perform cluster analysis and trend mining on the distribution pattern characteristics in the spatiotemporal dimension and convert them into a feature array; An information acquisition module, used to determine the breeding information of the phytoseiid mite, wherein the breeding information includes the breeding environment-growth cycle; The decision-making module is used to make placement decisions and optimize breeding decisions based on the target vegetation area based on the characteristic array and the breeding information, and determine the pre-control strategy; the control module is used to perform placement control and breeding control of phytoseiid mites according to the pre-control strategy.

Images