Sustained and controlled release pesticide application method based on environmental characteristics

Abstract

The invention relates to the technical field of pesticide application, in particular to a sustained and controlled release pesticide application method based on environmental characteristics. Obtaining a pre-pesticide-applying scheme of each sub-pesticide-applying area, and predicting according to the pre-pesticide-applying scheme of each sub-pesticide-applying area in combination with the multi-layer perceptron network model to obtain a predicted release rate of the to-be-applied sustained and controlled release pesticide of each sub-pesticide-applying area at each sub-time node within the pre-pesticide-applying time period; according to the predicted release rate of the to-be-applied sustained and controlled release pesticide in each sub-time node in each sub-pesticide application region in the pre-application pesticide application time period, carrying out feasibility analysis on the pre-application pesticide application scheme of each sub-pesticide application region; and if a certain pesticide application sub-region cannot perform pesticide application treatment on the pesticide application sub-region based on the corresponding pre-pesticide application scheme, adjusting the pre-pesticide application scheme of the pesticide application sub-region to obtain a recommended pesticide application scheme. The pesticide release rate is accurately predicted, the pesticide application scheme is reasonably adjusted, and it is ensured that pests and diseases are effectively controlled.

Description

Technical Field

[0001] The invention relates to the technical field of pesticide application, in particular to a method for applying a sustained-release pesticide based on environmental characteristics. Background Art

[0002] With the development of agriculture and the improvement of people's awareness of environmental protection, slow-release pesticides, as a new type of pesticide formulation, have gradually attracted people's attention. Slow-release pesticides refer to new pesticide formulations that use certain technical means to slowly release pesticides after they are applied to farmland, thereby extending the effective period of pesticides, reducing the amount of pesticides used, and reducing the pollution of pesticides to the environment. However, the current application method of slow-release pesticides is mainly based on time or dosage control, while ignoring the impact of environmental factors on pesticide release. In fact, environmental factors such as temperature, humidity, light, soil texture, etc. will affect the release rate of slow-release pesticides. For example, in a high temperature and high humidity environment, the release rate of pesticides will accelerate. Therefore, how to adjust the application method of slow-release pesticides according to environmental characteristics to improve the utilization rate and effect of pesticides and reduce the pollution of pesticides to the environment is a problem that needs to be solved urgently in the current agricultural field. Summary of the invention

[0003] The present invention overcomes the deficiencies of the prior art and provides a method for applying a slow-release pesticide based on environmental characteristics.

[0004] To achieve the above-mentioned purpose, the technical solution adopted by the present invention is:

[0005] The first aspect of the present invention discloses a method for applying a sustained-release pesticide based on environmental characteristics, comprising the following steps:

[0006] Obtain the historical release rates of various slow-release pesticides under various environmental characteristics, and build a multi-layer perceptron network model of slow-release pesticides based on the historical release rates of various slow-release pesticides under various environmental characteristics;

[0007] Obtain the target pesticide application area, divide the target pesticide application area into several sub-pesticide application areas, obtain the pre-pesticide application plan for each sub-pesticide application area, and predict the release rate of the slow-release pesticide to be applied in each sub-pesticide application area at each sub-time node within the pre-pesticide application time period according to the pre-pesticide application plan for each sub-pesticide application area and in combination with the multi-layer perceptron network model;

[0008] Conduct feasibility analysis on the pre-pesticide treatment scheme for each sub-pesticide treatment area based on the predicted release rate of the slow-release pesticide to be applied in each sub-pesticide treatment area at each sub-time node within the pre-pesticide treatment time period, and determine whether the pre-pesticide treatment can be carried out on each sub-pesticide treatment area based on the corresponding pre-pesticide treatment scheme;

[0009] If a sub-pesticide application area can be treated with pesticides based on the corresponding pre-pesticide installation plan, the pre-pesticide installation plan for the sub-pesticide application area will not be adjusted; if a sub-pesticide application area cannot be treated with pesticides based on the corresponding pre-pesticide installation plan, the pre-pesticide installation plan for the sub-pesticide application area will be adjusted to obtain a recommended pesticide application plan.

[0010] Furthermore, the historical release rates of various slow-release pesticides under various environmental characteristics are obtained, and a multi-layer perceptron network model of slow-release pesticides is constructed according to the historical release rates of various slow-release pesticides under various environmental characteristics, specifically:

[0011] Acquire characteristic component information of various slow-release and controlled-release pesticides, and retrieve historical release rates of various slow-release and controlled-release pesticides under various environmental characteristics in a big data network according to the characteristic component information;

[0012] A multi-layer perceptron network is introduced, and the environmental features are set as the input layer of the multi-layer perceptron network, and the historical release rates under various environmental features are set as the output layer of the multi-layer perceptron network;

[0013] Selecting an activation function of a multi-layer perceptron network, introducing a gradient descent optimization algorithm, and setting a gradient descent optimization algebra according to the gradient descent optimization algorithm;

[0014] The input layer and output layer of the multilayer perceptron network are trained and learned through back propagation through the selected activation function. After each training cycle is completed, the number of nodes in the hidden layer of the multilayer perceptron network is obtained.

[0015] When the number of nodes in the hidden layer of the multilayer perceptron network is greater than the preset number of nodes, the weights and biases of the input layer and output layer of the multilayer perceptron network are readjusted through gradient descent optimization algebra to minimize the loss function;

[0016] Until the number of nodes in the hidden layer of the multilayer perceptron network is no more than the preset number of nodes, the multilayer perceptron network model is output.

[0017] Furthermore, according to the pre-application scheme of each sub-application area and combined with the multi-layer perceptron network model, the predicted release rate of the slow-release pesticide to be applied in each sub-application area at each sub-time node within the pre-application period is obtained, specifically:

[0018] Obtain the pre-pesticide installation plan for each sub-pesticide application area in the target pesticide application area, and obtain the pre-pesticide installation time period for each sub-pesticide application area in the target pesticide application area;

[0019] The pre-pesticide application time period of each sub-pesticide application area is divided into several sub-time nodes, and the predicted environmental characteristic data of each sub-pesticide application area at each sub-time node is obtained based on the meteorological software;

[0020] Acquire characteristic component information of the slow-release pesticide to be applied in each sub-application area according to the pre-application plan of each sub-application area;

[0021] The characteristic component information of the slow-release pesticide to be applied in each sub-application area and the predicted environmental characteristic data of each sub-application area at each sub-time node are imported into the multi-layer perceptron network model for prediction, so as to obtain the predicted release rate of the slow-release pesticide to be applied in each sub-application area in the target application area at each sub-time node within the pre-application time period.

[0022] Furthermore, the feasibility of the pre-pesticide treatment scheme for each sub-pesticide treatment area is analyzed according to the predicted release rate of the slow-release pesticide to be applied in each sub-pesticide treatment area at each sub-time node within the pre-pesticide treatment time period, and it is determined whether the pre-pesticide treatment can be performed on each sub-pesticide treatment area based on the corresponding pre-pesticide treatment scheme, specifically:

[0023] Constructing a predicted release rate curve of the slow-release pesticide to be applied in each sub-application area according to the predicted release rate of the slow-release pesticide to be applied in each sub-application area at each sub-time node within the pre-application time period;

[0024] Obtaining a preset release rate curve graph for each sub-application area, and calculating the degree of overlap between the preset release rate curve graph for each sub-application area and the corresponding predicted release rate curve graph based on the Euclidean distance algorithm;

[0025] If the overlap between the preset release rate curve of a sub-drug application area and the corresponding predicted release rate curve is greater than the preset overlap, it means that the sub-drug application area can be treated with pesticides based on the preset pesticide application plan;

[0026] If the degree of overlap between the preset release rate curve of a sub-application area and the corresponding predicted release rate curve is greater than the preset overlap, it means that the sub-application area cannot be sprayed based on the preset spraying plan.

[0027] Furthermore, if a sub-pesticide application area cannot be treated with pesticides based on the corresponding pre-pesticide application plan, the pre-pesticide application plan for the sub-pesticide application area is adjusted to obtain a recommended pesticide application plan, specifically:

[0028] If a sub-pesticide application area cannot be treated with pesticides based on the pre-application plan, then the pesticide application record of the sub-pesticide application area is obtained;

[0029] According to the pesticide application record, a historical release rate curve diagram of the sub-pesticide application area after various historical slow-release and controlled-release pesticide applications under various historical environmental characteristic data is obtained;

[0030] and obtaining the predicted environmental characteristic data of the sub-application area and a preset release rate curve, and calculating the Pearson correlation coefficient between the predicted environmental characteristic data of the sub-application area and various historical environmental characteristic data based on the Pearson correlation coefficient algorithm;

[0031] The historical release rate curve graphs corresponding to the historical environmental characteristic data whose Pearson correlation coefficient is not greater than the preset coefficient threshold are eliminated, and the historical release rate curve graphs corresponding to the historical environmental characteristic data whose Pearson correlation coefficient is greater than the preset coefficient threshold are retained, so as to obtain a class of historical release rate curve graphs;

[0032] Calculate the overlap between the preset release rate curve and each type of historical release rate curve, remove the type of historical release rate curve corresponding to the preset overlap, and retain the type of historical release rate curve corresponding to the preset overlap, to obtain the type II historical release rate curve;

[0033] Acquire characteristic component information of historical slow-release pesticides corresponding to the two types of historical release rate curves, and calculate component similarity of characteristic component information between the slow-release pesticide to be applied and the historical slow-release pesticides corresponding to the two types of historical release rate curves;

[0034] The historical controlled-release pesticides with the highest similarity in composition are taken as the recommended controlled-release pesticides for application in the sub-application area, and a recommended application plan for the sub-application area is generated.

[0035] The following steps are also included:

[0036] Obtain the predicted release rate of the slow-release pesticide to be applied in each sub-application area at each sub-time node within the pre-application time period;

[0037] The dissolution completion time node of the controlled-release pesticide to be applied in each sub-application area is calculated according to the predicted release rate of each sub-time node of the controlled-release pesticide to be applied in each sub-application area within the pre-application time period;

[0038] Determine whether the dissolution completion time node of the slow-release pesticide to be applied in each sub-application area is within the pre-application time period;

[0039] If the time node of completion of dissolution of the controlled-release pesticide to be applied in a certain sub-application area is within the pre-application time period, the time node of completion of dissolution of the controlled-release pesticide to be applied in the application area is generated, and early warning information is generated based on the time node of completion of dissolution of the controlled-release pesticide to be applied in the application area.

[0040] The second aspect of the present invention discloses a controlled-release pesticide application system based on environmental characteristics, the controlled-release pesticide application system comprising a memory and a processor, the memory storing a controlled-release pesticide application method program, and when the controlled-release pesticide application method program is executed by the processor, the steps of any one of the controlled-release pesticide application methods are implemented.

[0041] The present invention solves the technical defects existing in the background technology, and has the following beneficial effects: obtaining the pre-drugation scheme of each sub-drug application area, and predicting the predicted release rate of each sub-time node of the slow-release pesticide to be applied in each sub-drug application area according to the pre-drugation scheme of each sub-drug application area and in combination with the multi-layer perceptron network model; feasibility analysis of the pre-drugation scheme of each sub-drug application area is performed according to the predicted release rate of each sub-time node of the slow-release pesticide to be applied in each sub-drug application area in the pre-drug application time period; if a sub-drug application area can be treated with pesticides based on the corresponding pre-drugation scheme, the pre-drugation scheme of the sub-drug application area is not adjusted; if a sub-drug application area cannot be treated with pesticides based on the corresponding pre-drugation scheme, the pre-drugation scheme of the sub-drug application area is adjusted to obtain a recommended pesticide application scheme. By accurately predicting the pesticide release rate and reasonably adjusting the pesticide application scheme to ensure that pests and diseases are effectively controlled, pesticide residues and negative impacts on the ecological environment can be reduced, and intelligent adjustment of the application of slow-release pesticides can be achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, drawings of other embodiments can be obtained based on these drawings without paying creative work.

[0043] Figure 1 The first method flow chart of a method for applying a slow-release pesticide based on environmental characteristics;

[0044] Figure 2 A second method flow chart of a method for applying a slow-release pesticide based on environmental characteristics;

[0045] Figure 3 The system block diagram of a controlled-release pesticide application system based on environmental characteristics. DETAILED DESCRIPTION

[0046] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0047] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0048] like Figure 1 As shown, the first aspect of the present invention discloses a method for applying a sustained-release pesticide based on environmental characteristics, comprising the following steps:

[0049] S102: Obtaining historical release rates of various slow-release pesticides under various environmental characteristics, and constructing a multi-layer perceptron network model of slow-release pesticides according to the historical release rates of various slow-release pesticides under various environmental characteristics;

[0050] S104: Obtain a target pesticide application area, and divide the target pesticide application area into a plurality of sub-pesticide application areas, obtain a pre-pesticide application plan for each sub-pesticide application area, and predict the release rate of the slow-release pesticide to be applied in each sub-pesticide application area at each sub-time node within the pre-pesticide application time period according to the pre-pesticide application plan for each sub-pesticide application area in combination with a multi-layer perceptron network model;

[0051] S106: Performing a feasibility analysis on the pre-pesticide application scheme for each sub-pesticide application area according to the predicted release rate of the slow-release controlled-release pesticide to be applied in each sub-pesticide application area at each sub-time node within the pre-pesticide application time period, and determining whether the pre-pesticide application scheme can be used to apply pesticides to each sub-pesticide application area;

[0052] S108: If a sub-pesticide application area can be treated with pesticides based on the corresponding pre-pesticide installation plan, the pre-pesticide installation plan for the sub-pesticide application area will not be adjusted; if a sub-pesticide application area cannot be treated with pesticides based on the corresponding pre-pesticide installation plan, the pre-pesticide installation plan for the sub-pesticide application area will be adjusted to obtain a recommended pesticide application plan.

[0053] Among them, environmental characteristics include but are not limited to temperature, light intensity, humidity, wind speed and other environmental data.

[0054] It should be noted that, firstly, a multi-layer perceptron network model was constructed by collecting historical release rate data of various slow-release pesticides under different environmental conditions. This model can simulate and predict the release rate of pesticides under specific environmental conditions. Then, the target application area is subdivided into multiple sub-application areas, and a preliminary application plan is formulated for each sub-area. With the help of the above-constructed model, the system can predict the predicted release rate of each node in each sub-application area within a preset time period. Subsequently, the feasibility of the pre-application plan is evaluated by comparing the predicted release rate with the actual demand. If the predicted release rate of a sub-application area meets the requirements, there is no need to adjust its pre-application plan; on the contrary, if the prediction results show that the existing plan cannot meet the application needs, the pre-application plan needs to be adjusted to ensure the best application effect and environmental protection standards. By accurately predicting the release rate of pesticides, reasonably adjusting the application plan, and ensuring that pests and diseases are effectively controlled, pesticide residues and negative impacts on the ecological environment can be reduced, and the intelligent application of slow-release pesticides is realized, providing strong technical support for the realization of precision agriculture and green agriculture.

[0055] Furthermore, if Figure 2 As shown, the historical release rates of various slow-release pesticides under various environmental characteristics are obtained, and a multi-layer perceptron network model of slow-release pesticides is constructed according to the historical release rates of various slow-release pesticides under various environmental characteristics. Specifically,

[0056] S202: Acquire characteristic component information of various slow-release and controlled-release pesticides, and retrieve historical release rates of various slow-release and controlled-release pesticides under various environmental characteristics in a big data network according to the characteristic component information;

[0057] S204: introducing a multi-layer perceptron network, setting the environmental features as the input layer of the multi-layer perceptron network, and setting the historical release rates under various environmental features as the output layer of the multi-layer perceptron network;

[0058] S206: selecting an activation function of a multi-layer perceptron network, introducing a gradient descent optimization algorithm, and setting a gradient descent optimization algebra according to the gradient descent optimization algorithm;

[0059] S208: performing back propagation training on the input layer and the output layer of the multilayer perceptron network through the selected activation function, and obtaining the number of nodes in the hidden layer of the multilayer perceptron network after completing a training cycle;

[0060] S210: when the number of nodes in the hidden layer of the multilayer perceptron network is greater than the preset number of nodes, readjust the weights and biases of the input layer and the output layer of the multilayer perceptron network through gradient descent optimization algebra to minimize the loss function;

[0061] S212: until the number of nodes in the hidden layer of the multilayer perceptron network is no more than the preset number of nodes, the multilayer perceptron network model is output.

[0062] It should be noted that, first, by obtaining the characteristic component information of slow-release pesticides and using big data network retrieval technology, the historical release rate data of various slow-release pesticides under different environmental conditions are collected. Then, a multi-layer perceptron network model is introduced, in which environmental characteristics are used as the input layer and the historical release rate is used as the output layer. In order to optimize the performance of the model, a suitable activation function is selected, and the gradient descent optimization algorithm is applied to adjust the model parameters to minimize the loss function. The model is trained and learned by the back propagation algorithm, and the number of nodes in the hidden layer is recorded after each training cycle. When the number of nodes in the hidden layer exceeds the preset value, the weight and bias of the model are adjusted by the gradient descent optimization algorithm to further optimize the prediction results. The whole process continues until the number of nodes in the hidden layer does not exceed the preset value, and finally the optimized multi-layer perceptron network model is output. Through the deep learning model, the release rate of slow-release pesticides under specific environmental conditions can be predicted more accurately, providing the accuracy of the model prediction results. Accurate prediction results help optimize the application time and dosage of pesticides, reduce resource waste, and improve agricultural production efficiency.

[0063] Furthermore, according to the pre-application scheme of each sub-application area and combined with the multi-layer perceptron network model, the predicted release rate of the slow-release pesticide to be applied in each sub-application area at each sub-time node within the pre-application period is obtained, specifically:

[0064] Obtain the pre-pesticide installation plan for each sub-pesticide application area in the target pesticide application area, and obtain the pre-pesticide installation time period for each sub-pesticide application area in the target pesticide application area;

[0065] The pre-pesticide application time period of each sub-pesticide application area is divided into several sub-time nodes, and the predicted environmental characteristic data of each sub-pesticide application area at each sub-time node is obtained based on the meteorological software;

[0066] Acquire characteristic component information of the slow-release pesticide to be applied in each sub-application area according to the pre-application plan of each sub-application area;

[0067] The characteristic component information of the slow-release pesticide to be applied in each sub-application area and the predicted environmental characteristic data of each sub-application area at each sub-time node are imported into the multi-layer perceptron network model for prediction, so as to obtain the predicted release rate of the slow-release pesticide to be applied in each sub-application area in the target application area at each sub-time node within the pre-application time period.

[0068] It should be noted that the system first obtains the pre-drug application scheme and pre-drug application time period of each sub-drug application area in the target pesticide application area to prepare for the subsequent model prediction. Then, the system subdivides these pre-drug application time periods into multiple sub-time nodes, and uses meteorological software to predict the environmental characteristic data of each sub-drug application area at these sub-time nodes, such as temperature, light intensity, humidity, wind speed, etc., to reflect the climatic conditions during the actual application. Subsequently, the system further obtains the characteristic component information of the slow-release pesticides planned to be applied in each sub-drug application area, which is crucial for understanding the behavior of pesticides under different environmental conditions. Finally, the system inputs the above-mentioned characteristic component information together with the predicted environmental characteristic data into the pre-trained multi-layer perceptron network model. The multi-layer perceptron is an artificial neural network that can make predictions about unknown situations by learning patterns and laws in historical data. In this process, the model predicts the release rate of the slow-release pesticides in each sub-drug application area at each sub-time node during the pre-drug application time period through complex calculations and learning. By integrating detailed environmental data and pesticide characteristics, the model can provide highly accurate release rate predictions, which make the use of pesticides more efficient and reduce potential impacts on the environment. In addition, the model can adjust the prediction according to different environmental conditions, making the application strategy more flexible and adaptable.

[0069] Furthermore, the feasibility of the pre-pesticide treatment scheme for each sub-pesticide treatment area is analyzed according to the predicted release rate of the slow-release pesticide to be applied in each sub-pesticide treatment area at each sub-time node within the pre-pesticide treatment time period, and it is determined whether the pre-pesticide treatment can be performed on each sub-pesticide treatment area based on the corresponding pre-pesticide treatment scheme, specifically:

[0070] Constructing a predicted release rate curve of the slow-release pesticide to be applied in each sub-application area according to the predicted release rate of the slow-release pesticide to be applied in each sub-application area at each sub-time node within the pre-application time period;

[0071] Obtaining a preset release rate curve graph for each sub-application area, and calculating the degree of overlap between the preset release rate curve graph for each sub-application area and the corresponding predicted release rate curve graph based on the Euclidean distance algorithm;

[0072] If the overlap between the preset release rate curve of a sub-drug application area and the corresponding predicted release rate curve is greater than the preset overlap, it means that the sub-drug application area can be treated with pesticides based on the preset pesticide application plan;

[0073] If the degree of overlap between the preset release rate curve of a sub-application area and the corresponding predicted release rate curve is greater than the preset overlap, it means that the sub-application area cannot be sprayed based on the preset spraying plan.

[0074] It should be noted that by constructing a predicted release rate curve of the slow-release pesticide to be applied in each sub-application area, an intuitive visual prediction is provided for the application effect of each area. Then, by obtaining the preset release rate curve of each sub-application area, that is, the ideal or expected application effect curve, and comparing and analyzing it with the predicted release rate curve. This comparative analysis uses the Euclidean distance algorithm, which is a commonly used method for measuring the difference between two vectors and is suitable for distance calculation in multidimensional space. By calculating the overlap between the preset release rate curve and the predicted release rate curve, that is, the degree of similarity between them, the system can evaluate the accuracy of the prediction model and the consistency between the prediction results and the expected goals. If the overlap between the preset release rate curve and the predicted release rate curve of a sub-application area exceeds the preset overlap threshold, this indicates that the prediction result of the prediction model is highly consistent with the expected goal, and the system can therefore believe that the pre-application scheme is feasible for the area, thereby supporting the application of pesticides to the area. On the contrary, if the overlap does not reach the preset threshold, it means that there is a large difference between the predicted result and the expectation, and the system recommends not to apply pesticides to the area based on the current pre-installed pesticide plan. By providing a comparative analysis of the predicted and preset release rate curves, it provides a scientific basis for agricultural decision makers to help them make more accurate judgments on pesticide application strategies. Through the accuracy evaluation of the prediction model, the system can screen out areas where the pesticide application effect is expected to be good, optimize the allocation of pesticide application resources, and reduce the waste of resources and environmental impact caused by ineffective or excessive pesticide application. Accurate predictions and efficient decision-making processes shorten the cycle from prediction to execution and improve the overall efficiency of pesticide application. The system can flexibly adjust the pesticide application strategy according to the prediction results, adapt to the needs of different environments and crops, and improve the flexibility and adaptability of agricultural production.

[0075] Furthermore, if a sub-pesticide application area cannot be treated with pesticides based on the corresponding pre-pesticide application plan, the pre-pesticide application plan for the sub-pesticide application area is adjusted to obtain a recommended pesticide application plan, specifically:

[0076] If a sub-pesticide application area cannot be treated with pesticides based on the pre-application plan, then the pesticide application record of the sub-pesticide application area is obtained;

[0077] According to the pesticide application record, a historical release rate curve diagram of the sub-pesticide application area after various historical slow-release and controlled-release pesticide applications under various historical environmental characteristic data is obtained;

[0078] and obtaining the predicted environmental characteristic data of the sub-application area and a preset release rate curve, and calculating the Pearson correlation coefficient between the predicted environmental characteristic data of the sub-application area and various historical environmental characteristic data based on the Pearson correlation coefficient algorithm;

[0079] The historical release rate curve graphs corresponding to the historical environmental characteristic data whose Pearson correlation coefficient is not greater than the preset coefficient threshold are eliminated, and the historical release rate curve graphs corresponding to the historical environmental characteristic data whose Pearson correlation coefficient is greater than the preset coefficient threshold are retained, so as to obtain a class of historical release rate curve graphs;

[0080] Calculate the overlap between the preset release rate curve and each type of historical release rate curve, remove the type of historical release rate curve corresponding to the preset overlap, and retain the type of historical release rate curve corresponding to the preset overlap, to obtain the type II historical release rate curve;

[0081] Acquire characteristic component information of historical slow-release pesticides corresponding to the two types of historical release rate curves, and calculate component similarity of characteristic component information between the slow-release pesticide to be applied and the historical slow-release pesticides corresponding to the two types of historical release rate curves;

[0082] The historical controlled-release pesticides with the highest similarity in composition are taken as the recommended controlled-release pesticides for application in the sub-application area, and a recommended application plan for the sub-application area is generated.

[0083] It should be noted that, firstly, by obtaining the historical pesticide application records of the sub-application area, a historical release rate curve under various historical environmental characteristic data is constructed, which provides a basis for understanding the pattern of pesticide release under different environmental conditions. Then, the system uses the Pearson correlation coefficient algorithm to evaluate the correlation between the predicted environmental characteristic data and the historical environmental characteristic data, so as to screen out the historical environmental characteristic data that best matches the predicted environment, and then form a type of historical release rate curve. Furthermore, the system calculates the overlap between the preset release rate curve and various historical release rate curves, identifies the historical release rate curve with a high degree of match with the preset release rate curve, and forms a second type of historical release rate curve. Then, the system uses component similarity analysis to find the historical controlled-release pesticide that best matches the characteristic component information of the controlled-release pesticide to be applied, as the recommended controlled-release pesticide, and generates a recommended application plan based on this. By using a large amount of historical data and predictive analysis, the system can provide fact-based pesticide application decision support to reduce blindness and uncertainty. By identifying the historical release rate curve that best matches the predicted environment, the system can predict the optimal release rate of pesticides under specific conditions, thereby achieving precise application of pesticides, improving the utilization rate of pesticides and crop yields, and reducing negative impacts on the environment. Automated data analysis and recommendation systems reduce manual intervention in the decision-making process, speed up the speed of pesticide application decisions, and improve the efficiency of agricultural production management. In short, this intelligent pesticide application decision-making system based on historical data and predictive analysis, through a data-driven approach, realizes the scientific, precise and automated decision-making of pesticide application, significantly improving agricultural production efficiency and resource utilization efficiency.

[0084] The following steps are also included:

[0085] Obtain the predicted release rate of the slow-release pesticide to be applied in each sub-application area at each sub-time node within the pre-application time period;

[0086] The dissolution completion time node of the controlled-release pesticide to be applied in each sub-application area is calculated according to the predicted release rate of each sub-time node of the controlled-release pesticide to be applied in each sub-application area within the pre-application time period;

[0087] Determine whether the dissolution completion time node of the slow-release pesticide to be applied in each sub-application area is within the pre-application time period;

[0088] If the time node of completion of dissolution of the controlled-release pesticide to be applied in a certain sub-application area is within the pre-application time period, the time node of completion of dissolution of the controlled-release pesticide to be applied in the application area is generated, and early warning information is generated based on the time node of completion of dissolution of the controlled-release pesticide to be applied in the application area.

[0089] It should be noted that the predicted release rate of each sub-time node of the controlled-release pesticide to be applied in each sub-application area during the pre-application time period is obtained through the prediction model. This is the result of simulating the release behavior of the pesticide under different environmental conditions through the model. Then, the system calculates the dissolution completion time node of the controlled-release pesticide to be applied in each sub-application area based on these predicted release rates, that is, the time point when the pesticide is completely released or dissolved. After that, the system determines whether these dissolution completion time nodes fall within the preset application time period. If the dissolution completion time node of a sub-application area is within the preset application time period, it means that the pesticide has been completely released or dissolved during the application process, which may affect the application effect. In this case, the system will generate information on the dissolution completion time node of the controlled-release pesticide to be applied in the application area, and generate warning information based on the time node. The warning information usually includes the remaining effective time of the pesticide, whether the application time needs to be adjusted, or whether other measures need to be taken to ensure the effective application of the pesticide. In this way, the system can detect and deal with possible problems in time before applying pesticides, thereby improving the efficiency of pesticide use and reducing resource waste, protecting the environment and crop health. The system can accurately predict the dissolution and release time of pesticides, and ensure that pesticides work in the most suitable environment, avoiding the impact of pesticide effects due to poor environmental conditions, thereby ensuring the healthy growth of crops.

[0090] In addition, the method may further comprise the following steps:

[0091] The normal characteristic data and failure characteristic data of the applied controlled-release pesticides in each sub-application area under various environmental characteristic conditions are obtained through the big data network;

[0092] Based on the fuzzy logic algorithm, the normal characteristic data and failure characteristic data of the applied slow-release pesticides in each sub-application area under various environmental characteristic conditions are fuzzily evaluated to obtain the fuzziness of the normal characteristic data and failure characteristic data of the applied slow-release pesticides under various environmental characteristic conditions;

[0093] A fuzzy matrix is ​​constructed based on the fuzziness of normal characteristic data and failure characteristic data of the applied slow-release pesticide under various environmental characteristic conditions;

[0094] A conditional random field is introduced, and based on the conditional random field, a probability random deduction is performed on the fuzzy matrix to obtain the state transition probability of the applied slow-release pesticide from a normal state to an ineffective state under various environmental characteristic conditions;

[0095] Construct a Markov model for the state transition probability of applied slow-release pesticides from normal state to failure state under various environmental characteristic conditions;

[0096] After applying the controlled-release pesticide to each sub-application area, obtaining the actual environmental characteristic data of each sub-application area at a preset time node, importing the actual environmental characteristic data of each sub-application area into the Markov model for prediction, and obtaining the state transition probability of the controlled-release pesticide in each sub-application area;

[0097] If the state transition probability of the controlled-release pesticide in a certain sub-application area is greater than the preset probability value, it is determined whether the state transition of the controlled-release pesticide in the sub-application area is to an invalid state; if it is to an invalid state, an early warning message is generated.

[0098] It should be noted that the normal characteristic data and failure characteristic data of the applied controlled-release pesticides in each sub-application area under different environmental characteristic conditions are obtained through the big data network. These data contain the performance of pesticides under ideal and non-ideal environmental conditions, providing a basis for subsequent analysis. Next, the system uses fuzzy logic algorithms to perform fuzzy evaluation on the acquired data and quantify the fuzziness of normal characteristic data and failure characteristic data. The degree of fuzziness reflects the uncertainty of the data and is a key indicator for further analysis of the system. On this basis, a fuzzy matrix is ​​constructed to organize the fuzziness information of normal characteristic data and failure characteristic data in the form of a matrix. Subsequently, the conditional random field (CRF) is introduced for probabilistic random deduction to predict the probability of pesticides transitioning from normal state to failure state under different environmental conditions. This step uses the characteristics of CRF and considers the influence of environmental characteristics on state transition, thereby obtaining an estimate of the state transition probability. A Markov model based on these state transition probabilities is constructed to describe the process of pesticide state change over time. The model can predict the dynamic changes of pesticides under different environmental conditions, especially the transition from normal state to failure state. After the pesticide is applied, the system will collect the actual environmental characteristic data of each sub-application area at the preset time node, and input it into the Markov model for prediction to obtain the pesticide state transition probability of each sub-application area. This step realizes the transformation from theoretical prediction to practical application, and improves the pertinence and practicality of the prediction. Finally, based on the prediction results, if the probability of pesticide state transition in a sub-application area exceeds the preset threshold, the system judges that its state is likely to change to an invalid state, and generates corresponding early warning information. This early warning mechanism helps to take measures in advance to prevent the adverse effects of pesticide failure. In general, this system realizes the dynamic prediction and monitoring of the state of slow-release pesticides by integrating big data analysis, fuzzy logic, probability deduction and Markov model, and improves the accuracy of state prediction through sophisticated data processing and model construction. Real-time monitoring of pesticide status after application helps to respond and adjust immediately, and through the early warning mechanism, potential problems can be discovered in time to reduce the risks of pesticide failure.

[0099] like Figure 3 As shown, the second aspect of the present invention discloses a controlled-release pesticide application system based on environmental characteristics, the controlled-release pesticide application system includes a memory 50 and a processor 100, the memory 50 stores a controlled-release pesticide application method program, when the controlled-release pesticide application method program is executed by the processor 100, the steps of any one of the controlled-release pesticide application methods are implemented.

[0100] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A method for applying a sustained-release pesticide based on environmental characteristics, characterized in that: The following steps are involved: Obtain the historical release rates of various slow-release pesticides under various environmental characteristics, and build a multi-layer perceptron network model of slow-release pesticides based on the historical release rates of various slow-release pesticides under various environmental characteristics; Obtain the target pesticide application area, divide the target pesticide application area into several sub-pesticide application areas, obtain the pre-pesticide application plan for each sub-pesticide application area, and predict the release rate of the slow-release pesticide to be applied in each sub-pesticide application area at each sub-time node within the pre-pesticide application time period according to the pre-pesticide application plan for each sub-pesticide application area and in combination with the multi-layer perceptron network model; Conduct feasibility analysis on the pre-pesticide treatment scheme for each sub-pesticide treatment area based on the predicted release rate of the slow-release pesticide to be applied in each sub-pesticide treatment area at each sub-time node within the pre-pesticide treatment time period, and determine whether the pre-pesticide treatment can be carried out on each sub-pesticide treatment area based on the corresponding pre-pesticide treatment scheme; If a sub-pesticide application area can be treated with pesticides based on the corresponding pre-pesticide installation plan, the pre-pesticide installation plan for the sub-pesticide application area will not be adjusted; if a sub-pesticide application area cannot be treated with pesticides based on the corresponding pre-pesticide installation plan, the pre-pesticide installation plan for the sub-pesticide application area will be adjusted to obtain a recommended pesticide application plan.

2. The method for applying a sustained-release pesticide based on environmental characteristics according to claim 1, characterized in that: The historical release rates of various slow-release pesticides under various environmental characteristics are obtained, and a multi-layer perceptron network model of slow-release pesticides is constructed according to the historical release rates of various slow-release pesticides under various environmental characteristics. Specifically: Acquire characteristic component information of various slow-release and controlled-release pesticides, and retrieve historical release rates of various slow-release and controlled-release pesticides under various environmental characteristics in a big data network according to the characteristic component information; A multi-layer perceptron network is introduced, and the environmental features are set as the input layer of the multi-layer perceptron network, and the historical release rates under various environmental features are set as the output layer of the multi-layer perceptron network; Selecting an activation function of a multi-layer perceptron network, introducing a gradient descent optimization algorithm, and setting a gradient descent optimization algebra according to the gradient descent optimization algorithm; The input layer and output layer of the multilayer perceptron network are trained and learned through back propagation through the selected activation function. After each training cycle is completed, the number of nodes in the hidden layer of the multilayer perceptron network is obtained. When the number of nodes in the hidden layer of the multilayer perceptron network is greater than the preset number of nodes, the weights and biases of the input layer and output layer of the multilayer perceptron network are readjusted through gradient descent optimization algebra to minimize the loss function; Until the number of nodes in the hidden layer of the multilayer perceptron network is no more than the preset number of nodes, the multilayer perceptron network model is output.

3. The method for applying a sustained-release pesticide based on environmental characteristics according to claim 1, characterized in that: According to the pre-application scheme of each sub-application area and combined with the multi-layer perceptron network model, the predicted release rate of the slow-release pesticide to be applied in each sub-application area at each sub-time node within the pre-application period is obtained, specifically: Obtain the pre-pesticide installation plan for each sub-pesticide application area in the target pesticide application area, and obtain the pre-pesticide installation time period for each sub-pesticide application area in the target pesticide application area; The pre-pesticide application time period of each sub-pesticide application area is divided into several sub-time nodes, and the predicted environmental characteristic data of each sub-pesticide application area at each sub-time node is obtained based on the meteorological software; Acquire characteristic component information of the slow-release pesticide to be applied in each sub-application area according to the pre-application plan of each sub-application area; The characteristic component information of the slow-release pesticide to be applied in each sub-application area and the predicted environmental characteristic data of each sub-application area at each sub-time node are imported into the multi-layer perceptron network model for prediction, so as to obtain the predicted release rate of the slow-release pesticide to be applied in each sub-application area in the target application area at each sub-time node within the pre-application time period.

4. The method for applying a sustained-release pesticide based on environmental characteristics according to claim 1, characterized in that: According to the predicted release rate of the slow-release pesticide to be applied in each sub-application area at each sub-time node within the pre-application period, the feasibility analysis of the pre-application scheme for each sub-application area is carried out to determine whether the sub-application area can be sprayed based on the corresponding pre-application scheme. Specifically: Constructing a predicted release rate curve of the slow-release pesticide to be applied in each sub-application area according to the predicted release rate of the slow-release pesticide to be applied in each sub-application area at each sub-time node within the pre-application time period; Obtaining a preset release rate curve graph for each sub-application area, and calculating the degree of overlap between the preset release rate curve graph for each sub-application area and the corresponding predicted release rate curve graph based on the Euclidean distance algorithm; If the overlap between the preset release rate curve of a sub-drug application area and the corresponding predicted release rate curve is greater than the preset overlap, it means that the sub-drug application area can be treated with pesticides based on the preset pesticide application plan; If the degree of overlap between the preset release rate curve of a sub-application area and the corresponding predicted release rate curve is greater than the preset overlap, it means that the sub-application area cannot be sprayed based on the preset spraying plan.

5. The method for applying a sustained-release pesticide based on environmental characteristics according to claim 4, characterized in that: If a sub-application area cannot be treated with pesticides based on the corresponding pre-application plan, the pre-application plan for the sub-application area is adjusted to obtain a recommended pesticide plan, which is: If a sub-pesticide application area cannot be treated with pesticides based on the pre-application plan, then the pesticide application record of the sub-pesticide application area is obtained; According to the pesticide application record, a historical release rate curve diagram of the sub-pesticide application area after various historical slow-release and controlled-release pesticide applications under various historical environmental characteristic data is obtained; and obtaining the predicted environmental characteristic data of the sub-application area and a preset release rate curve, and calculating the Pearson correlation coefficient between the predicted environmental characteristic data of the sub-application area and various historical environmental characteristic data based on the Pearson correlation coefficient algorithm; The historical release rate curve graphs corresponding to the historical environmental characteristic data whose Pearson correlation coefficient is not greater than the preset coefficient threshold are eliminated, and the historical release rate curve graphs corresponding to the historical environmental characteristic data whose Pearson correlation coefficient is greater than the preset coefficient threshold are retained, so as to obtain a class of historical release rate curve graphs; Calculate the overlap between the preset release rate curve and each type of historical release rate curve, remove the type of historical release rate curve corresponding to the preset overlap, and retain the type of historical release rate curve corresponding to the preset overlap, to obtain the type II historical release rate curve; Acquire characteristic component information of historical slow-release pesticides corresponding to the two types of historical release rate curves, and calculate component similarity of characteristic component information between the slow-release pesticide to be applied and the historical slow-release pesticides corresponding to the two types of historical release rate curves; The historical controlled-release pesticides with the highest similarity in composition are taken as the recommended controlled-release pesticides for application in the sub-application area, and a recommended application plan for the sub-application area is generated.

6. The method for applying a sustained-release pesticide based on environmental characteristics according to claim 1, characterized in that: The following steps are also included: Obtain the predicted release rate of the slow-release pesticide to be applied in each sub-application area at each sub-time node within the pre-application time period; The dissolution completion time node of the controlled-release pesticide to be applied in each sub-application area is calculated according to the predicted release rate of each sub-time node of the controlled-release pesticide to be applied in each sub-application area within the pre-application time period; Determine whether the dissolution completion time node of the slow-release pesticide to be applied in each sub-application area is within the pre-application time period; If the time node of completion of dissolution of the controlled-release pesticide to be applied in a certain sub-application area is within the pre-application time period, the time node of completion of dissolution of the controlled-release pesticide to be applied in the application area is generated, and early warning information is generated based on the time node of completion of dissolution of the controlled-release pesticide to be applied in the application area.

7. A controlled-release pesticide application system based on environmental characteristics, characterized in that: The controlled-release pesticide application system comprises a memory and a processor. The memory stores a controlled-release pesticide application method program. When the controlled-release pesticide application method program is executed by the processor, the following steps are implemented: Obtain the historical release rates of various slow-release pesticides under various environmental characteristics, and build a multi-layer perceptron network model of slow-release pesticides based on the historical release rates of various slow-release pesticides under various environmental characteristics; Obtain the target pesticide application area, divide the target pesticide application area into several sub-pesticide application areas, obtain the pre-pesticide application plan for each sub-pesticide application area, and predict the release rate of the slow-release pesticide to be applied in each sub-pesticide application area at each sub-time node within the pre-pesticide application time period according to the pre-pesticide application plan for each sub-pesticide application area and in combination with the multi-layer perceptron network model; Conduct feasibility analysis on the pre-pesticide treatment scheme for each sub-pesticide treatment area based on the predicted release rate of the slow-release pesticide to be applied in each sub-pesticide treatment area at each sub-time node within the pre-pesticide treatment time period, and determine whether the pre-pesticide treatment can be carried out on each sub-pesticide treatment area based on the corresponding pre-pesticide treatment scheme; If a sub-pesticide application area can be treated with pesticides based on the corresponding pre-pesticide installation plan, the pre-pesticide installation plan for the sub-pesticide application area will not be adjusted; if a sub-pesticide application area cannot be treated with pesticides based on the corresponding pre-pesticide installation plan, the pre-pesticide installation plan for the sub-pesticide application area will be adjusted to obtain a recommended pesticide application plan.

8. The controlled-release pesticide application system based on environmental characteristics according to claim 7, characterized in that: The historical release rates of various slow-release pesticides under various environmental characteristics are obtained, and a multi-layer perceptron network model of slow-release pesticides is constructed according to the historical release rates of various slow-release pesticides under various environmental characteristics. Specifically: Acquire characteristic component information of various slow-release and controlled-release pesticides, and retrieve historical release rates of various slow-release and controlled-release pesticides under various environmental characteristics in a big data network according to the characteristic component information; A multi-layer perceptron network is introduced, and the environmental features are set as the input layer of the multi-layer perceptron network, and the historical release rates under various environmental features are set as the output layer of the multi-layer perceptron network; Selecting an activation function of a multi-layer perceptron network, introducing a gradient descent optimization algorithm, and setting a gradient descent optimization algebra according to the gradient descent optimization algorithm; The input layer and output layer of the multilayer perceptron network are trained and learned through back propagation through the selected activation function. After each training cycle is completed, the number of nodes in the hidden layer of the multilayer perceptron network is obtained. When the number of nodes in the hidden layer of the multilayer perceptron network is greater than the preset number of nodes, the weights and biases of the input layer and output layer of the multilayer perceptron network are readjusted through gradient descent optimization algebra to minimize the loss function; Until the number of nodes in the hidden layer of the multilayer perceptron network is no more than the preset number of nodes, the multilayer perceptron network model is output.

9. The environmental characteristics-based sustained-release pesticide application system according to claim 7, characterized in that: According to the pre-application scheme of each sub-application area and combined with the multi-layer perceptron network model, the predicted release rate of the slow-release pesticide to be applied in each sub-application area at each sub-time node within the pre-application period is obtained, specifically: Obtain the pre-pesticide installation plan for each sub-pesticide application area in the target pesticide application area, and obtain the pre-pesticide installation time period for each sub-pesticide application area in the target pesticide application area; The pre-pesticide application time period of each sub-pesticide application area is divided into several sub-time nodes, and the predicted environmental characteristic data of each sub-pesticide application area at each sub-time node is obtained based on the meteorological software; Acquire characteristic component information of the slow-release pesticide to be applied in each sub-application area according to the pre-application plan of each sub-application area; The characteristic component information of the slow-release pesticide to be applied in each sub-application area and the predicted environmental characteristic data of each sub-application area at each sub-time node are imported into the multi-layer perceptron network model for prediction, so as to obtain the predicted release rate of the slow-release pesticide to be applied in each sub-application area in the target application area at each sub-time node within the pre-application time period.

10. The controlled-release pesticide application system based on environmental characteristics according to claim 7, characterized in that: According to the predicted release rate of the slow-release pesticide to be applied in each sub-application area at each sub-time node within the pre-application period, the feasibility analysis of the pre-application scheme for each sub-application area is carried out to determine whether the sub-application area can be sprayed based on the corresponding pre-application scheme. Specifically: Constructing a predicted release rate curve of the slow-release pesticide to be applied in each sub-application area according to the predicted release rate of the slow-release pesticide to be applied in each sub-application area at each sub-time node within the pre-application time period; Obtaining a preset release rate curve graph for each sub-application area, and calculating the degree of overlap between the preset release rate curve graph for each sub-application area and the corresponding predicted release rate curve graph based on the Euclidean distance algorithm; If the overlap between the preset release rate curve of a sub-drug application area and the corresponding predicted release rate curve is greater than the preset overlap, it means that the sub-drug application area can be treated with pesticides based on the preset pesticide application plan; If the degree of overlap between the preset release rate curve of a sub-application area and the corresponding predicted release rate curve is greater than the preset overlap, it means that the sub-application area cannot be sprayed based on the preset spraying plan.

Images