Plant protection unmanned aerial vehicle spraying drift amount prediction method based on ann-gwo and related equipment

Abstract

The application discloses a plant protection unmanned aerial vehicle spraying drift amount prediction method based on an ANN-GWO and related equipment, test parameter data and drift amount data of a plurality of candidate characteristic parameters are collected in a test phase, and correlation analysis is performed by using a random forest algorithm to determine at least one target characteristic parameter for spraying drift amount prediction, the cooperation of the plurality of candidate characteristic parameters and the random forest algorithm is effectively used, the correlation and reliability of the selected target characteristic parameter for spraying drift amount prediction are ensured, and the data volume in model training and spraying drift amount prediction is reduced. By using an ANN-GWO artificial neural network model, the global search and local development capabilities of the GWO optimization algorithm are effectively utilized, the problem that convergence is difficult to be achieved by using part of the target characteristic parameters is solved, the uniformity of the weights of the plurality of target characteristic parameters is effectively considered, and the reliability of spraying drift amount prediction under different target downwind distances is improved.

Description

TECHNICAL FIELD

[0001] The present application generally relates to the technical field of plant protection unmanned aerial vehicle, and particularly relates to a plant protection unmanned aerial vehicle spray drift amount prediction method based on ANN-GWO and related equipment. BACKGROUND

[0002] In the process of crop planting, in order to effectively manage the diseases, pests and weeds in farmland, the growers usually spray chemical agents such as pesticides, fungicides and herbicides for disease, pest and weed control to ensure that the yield and quality of the crops are not affected. In recent years, plant protection unmanned aerial vehicles have developed rapidly in China due to their characteristics such as being not restricted by terrain and crop growth, low purchase and maintenance cost, simple operation and strong maneuverability, and are suitable for plant protection spraying operations in plots where manpower and ground machinery cannot enter, thereby expanding the application scenarios of agricultural and forestry aerial plant protection machinery and pesticide application technology. However, compared with ground plant protection machinery, the plant protection unmanned aerial vehicle has low pesticide application amount, fine atomized droplets, high flight height and fast operation speed, and is prone to pesticide drift under the influence of natural wind, resulting in pesticide damage to non-target crops and organisms and environmental pollution. In plant protection operations, the unmanned aerial vehicle is usually controlled to fly at a fixed distance from the edge of the planting area according to its effective spraying width. At the same time, the actual measurement of the spray drift amount often requires a large amount of time and manpower. Therefore, how to accurately predict the plant protection unmanned aerial vehicle spray drift amount at different distances in the wind direction under different characteristic parameters (flight parameters, pesticide application parameters and meteorological parameters) has important practical application value for pesticide application risk assessment, safe operation specification development, pesticide utilization rate improvement and pesticide reduction. SUMMARY

[0003] In view of the above defects or deficiencies in the prior art, it is desirable to provide a plant protection unmanned aerial vehicle spray drift amount prediction method and device based on ANN-GWO, which can use an ANN-GWO artificial neural network model to predict the plant protection unmanned aerial vehicle spray drift amount at a target downwind distance.

[0004] In a first aspect, an embodiment of the present application provides a plant protection unmanned aerial vehicle spray drift amount prediction method based on ANN-GWO. In a test phase, historical test data of plant protection unmanned aerial vehicle spray drift amount test are acquired, the historical test data at least including test parameter data and drift amount data of a plurality of candidate characteristic parameters under different test conditions, the candidate characteristic parameters at least including flight height, flight speed, droplet size, average environmental wind speed, average temperature, average humidity, nozzle rotation speed and downwind distance;

[0005] The historical test data are preprocessed to obtain preprocessed historical test data;

[0006] correlation analysis is performed on the pretreated historical test data by using a random forest algorithm to determine at least one target feature parameter for spray drift amount prediction, and pretreated test parameter data corresponding to the target feature parameter is obtained to construct a training set and a test set; the target feature parameter at least includes the downwind distance;

[0007] An initial ANN-GWO artificial neural network model is constructed, and the training set and the test set are used to train the initial ANN-GWO artificial neural network model to obtain a target prediction model for spray drift amount prediction;

[0008] The target prediction model is used to predict the spray drift amount of the plant protection unmanned aerial vehicle at a target downwind distance.

[0009] In some embodiments, in the test phase, historical test data of the plant protection unmanned aerial vehicle spray drift amount test is obtained, including:

[0010] Nine ground mist droplet collection devices with a diameter of 15 cm are arranged at a downwind distance of 3 m, 5 m, 10 m, 15 m, 20 m, 30 m and 50 m from the edge of the plant protection unmanned aerial vehicle spray swath;

[0011] In the test phase, the plant protection unmanned aerial vehicle is sprayed with a simulated pesticide solution with an ABF fluorescent tracer and an OP-10 non-ionic surfactant concentration of 0.1%, and the ground mist droplets are collected by the ground mist droplet collection device;

[0012] The mist droplet collection device is eluted with deionized water to obtain the corresponding fluorescence value of each mist droplet collection device, and the drift amount data corresponding to the test is determined by the fluorescence value.

[0013] In some embodiments, the pretreatment of the historical test data to obtain pretreated historical test data includes:

[0014] The missing values in the historical test data are deleted; and

[0015] The historical test data after missing value processing is normalized and standardized;

[0016] Wherein, the normalization is performed by the following formula:

[0017]

[0018] Wherein, x i is the original historical test data to be normalized, is the normalized historical test data, and its value range is [0, 1], x maxMax (original historical test data)

[0019] The normalization processing is performed by using the following formula:

[0020]

[0021] wherein μ is the mean value of the normalized historical test data, σ is the standard deviation of the normalized historical test data, x standardized is the normalized historical test data, and the normalized historical test data conforms to the standard normal distribution with a mean value of 0 and a standard deviation of 1.

[0022] In some embodiments, the correlation analysis based on the preprocessed historical test data is performed by using the random forest algorithm to determine at least one target feature parameter for spray drift prediction, including:

[0023] The Gini index difference value of each candidate feature parameter before and after branching of the decision tree is obtained;

[0024] The feature contribution amount of each candidate feature parameter is determined by using the Gini index difference value of each candidate feature parameter;

[0025] The contribution value of the candidate feature parameter is determined by using the ratio of the feature contribution amount of each candidate feature parameter to the sum of the feature contribution amounts of all candidate feature parameters.

[0026] At least one candidate feature parameter satisfying the preset condition is taken as the target feature parameter.

[0027] In some embodiments, the Gini index difference value is calculated by using the following formula:

[0028] VIM jm = GI m - GI l - GI r

[0029] wherein VIM jm represents the Gini index change value of the candidate feature parameter j at the node m, GI m represents the Gini index before branching, GI l and GI r are the Gini indexes of the two new nodes generated after branching of the node m.

[0030] The Gini index is calculated by using the following formula:

[0031]

[0032] wherein p k represents the weight of the kth category.

[0033] In some embodiments, the initial ANN-GWO artificial neural network model is constructed, and the initial ANN-GWO artificial neural network model is trained by using the training set and the test set to obtain a target prediction model for predicting the spray drift.

[0034] An initial ANN artificial neural network model and GWO algorithm parameters are constructed.

[0035] The structure characteristics of the initial ANN artificial neural network model are taken as prey, and the position vector of the wolf pack is calculated.

[0036] The fitness corresponding to the current wolf pack position vector is calculated based on the training set.

[0037] The information of the three wolves with the optimal fitness is calculated, and the position vectors thereof and the position vector of the wolf pack are updated, and the iterative training is performed.

[0038] When the termination condition is met, the position vector of the wolf pack in the current iteration round is assigned to the ANN artificial neural network model to obtain the target prediction model.

[0039] In some embodiments, the method further comprises:

[0040] The fitness corresponding to the current wolf pack position vector is calculated by using the improved root mean square error.

[0041] In a second aspect, the embodiments of the present application provide an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the method described in the embodiments of the present application when executing the program.

[0042] In a third aspect, the embodiments of the present application provide a computer readable storage medium having a computer program stored thereon, and the program is executed by a processor to implement the method described in the embodiments of the present application.

[0043] In a fourth aspect, the embodiments of the present application provide a computer program product, including a computer program, and the computer program is executed by a processor to implement the method described in the embodiments of the present application.

[0044] The ANN-GWO-based plant protection unmanned aerial vehicle spraying drift amount prediction method provided in the embodiment of the application collects test parameter data and drift amount data of a plurality of candidate feature parameters in a test phase, and correlation analysis is performed by using a random forest algorithm to determine at least one target feature parameter for spraying drift amount prediction.

[0045] Additional aspects and advantages of the application will be set forth in part in the description which follows, and in part will become apparent to those skilled in the art upon examination of the following and / or can be learned by practice of the application. BRIEF DESCRIPTION OF DRAWINGS

[0046] Other features, objects, and advantages of the application will become more apparent from the following detailed description when read in connection with the following drawings:

[0047] Figure 1 A flowchart of the ANN-GWO-based plant protection unmanned aerial vehicle spraying drift amount prediction method provided in an embodiment of the application is shown;

[0048] Figure 2 A flowchart of the ANN-GWO-based plant protection unmanned aerial vehicle spraying drift amount prediction method provided in another embodiment of the application is shown;

[0049] Figure 3 A structural schematic diagram of a computer system of an electronic device or a server suitable for being used to implement the embodiments of the application is shown. DETAILED DESCRIPTION

[0050] The application will be further described below in conjunction with the drawings and embodiments. It can be understood that the specific embodiments described herein are only used to explain the related application, and not to limit the application. In addition, it should be noted that only the parts related to the application are shown in the drawings for ease of description.

[0051] It should be noted that the embodiments in the application and the features in the embodiments can be combined with each other without conflict. The application will be described in detail below with reference to the drawings and in conjunction with the embodiments.

[0052] To further illustrate the technical solutions provided by the embodiments of the present application, the following will make a detailed description in combination with the drawings and specific embodiments. Although the embodiments of the present application provide the method operation instruction steps as shown in the following embodiments or drawings, more or less operation instruction steps can be included in the method based on conventional or non-creative labor. The execution order of the steps is not limited to the execution order provided by the embodiments of the present application in logically unnecessary causal relationship. The method can be executed in sequence or in parallel when the actual processing process or the device is executed according to the method sequence shown in the embodiments or drawings.

[0053] Please refer to Figure 1 , Figure 1 The flowchart of the ANN-GWO-based plant protection unmanned aerial vehicle spraying drift amount prediction method provided by an embodiment of the present application is shown. As shown in Figure 1 , the method comprises:

[0054] Step 101, in the test phase, the historical test data of the plant protection unmanned aerial vehicle spraying drift amount test is obtained.

[0055] The historical test data at least includes the test parameter data of the plurality of candidate characteristic parameters and the drift amount data under different test conditions, and the candidate characteristic parameters at least include the flight height, the flight speed, the droplet particle size, the average environmental wind speed, the average temperature, the average humidity, the nozzle rotating speed and the downwind distance.

[0056] That is, in the test phase, the plant protection unmanned aerial vehicle can be controlled to test multiple times according to different flight heights and different flight speeds, and the corresponding average environmental wind speed, average temperature and average humidity of each flight test process are obtained. The droplet particle size and the nozzle rotating speed of the plant protection unmanned aerial vehicle spraying can be the same or different in multiple times, and preferably, different nozzle rotating speeds and droplet particle sizes are selected in each test flight.

[0057] After each test flight, the drift amount data corresponding to the test flight is calculated. Specifically, before the test flight, 9 ground droplet collecting devices with a diameter of 15 cm are arranged at 3m, 5m, 10m, 15m, 20m, 30m and 50m from the edge of the plant protection unmanned aerial vehicle spraying range in the downwind direction. The opening of the ground droplet collecting device is not less than its diameter, and preferably, the ground droplet collecting device is a plastic culture dish.

[0058] In the test phase, the configured ABF fluorescent tracer and the simulated liquid medicine with a concentration of 10% of OP-10 nonionic surfactant are used for the plant protection unmanned aerial vehicle spraying operation, and the ground drifting droplets are collected by using the ground droplet collection device. After each test flight, the droplet collection device is eluted with deionized water to obtain the corresponding fluorescence value of each droplet collection device, and the drift amount data corresponding to the test is determined by the fluorescence value.

[0059] Preferably, the drift amount data is the average value of the drift amount data corresponding to the 9 ground droplet collection devices at the distance position.

[0060] Then, a plurality of sets of historical test data are generated based on the test parameter data and the drift amount data corresponding to the candidate characteristic parameters of each test flight. Preferably, one set of test data includes the test parameter data of 7 candidate characteristic parameters and one test downwind distance and the corresponding drift amount data. In other words, one test flight can obtain 7 sets of historical test data corresponding to 7 downwind distances.

[0061] Step 102, preprocessing the historical test data to obtain preprocessed historical test data.

[0062] It should be noted that preprocessing the historical test data can reduce the interference of false data and improve the reliability and accuracy of the data.

[0063] In the embodiments of the present application, preprocessing the historical test data includes deleting the missing values in the historical test data. Specifically, a set of historical test data with missing values is deleted, and preferably, a set of historical test data with missing values in the test parameter data corresponding to the candidate characteristic parameters is deleted.

[0064] Further, the historical test data after processing the missing values is normalized and standardized in the embodiments of the present application.

[0065] It should be noted that normalizing the historical test data can scale the test parameter data corresponding to a plurality of candidate characteristic parameters to a unified measurement range, such as [0, 1], so as to avoid the abnormal weight influence caused by the large difference between the test parameter data in the subsequent analysis. In addition, standardizing the historical test data can make the processed data conform to the standard normal distribution, further improving the reliability and accuracy of the data in the subsequent use process.

[0066] Specifically, the following formula can be used for normalization:

[0067]

[0068] wherein, x ithe normalized original historical test data, the normalized historical test data, the value range of which is [0, 1], x max is a maximum value in the original historical test data;

[0069] The standardization processing is performed by using the following formula:

[0070]

[0071] wherein μ is a mean value of the normalized historical test data, σ is a standard deviation of the normalized historical test data, and x standardized is the standardized historical test data, the standardized historical test data conforms to a standard normal distribution, the mean value is 0, and the standard deviation is 1.

[0072] It should be further noted that when the historical test data is standardized, the test parameter data obtained after normalization of each candidate feature parameter can be standardized respectively to improve the feature expression ability of each candidate feature parameter itself.

[0073] In step 103, the random forest algorithm is used to perform correlation analysis based on the preprocessed historical test data, at least one target feature parameter for predicting the spray drift is determined, the preprocessed test parameter data corresponding to the target feature parameter is obtained, a training set and a test set are constructed.

[0074] The target feature parameter at least includes a downwind distance.

[0075] It should be noted that the random forest algorithm is an algorithm for judging the importance of features in a decision tree. Specifically, in the random forest algorithm, the Gini index is used to calculate the importance of each feature. The Gini index is an index for measuring the impurity or uncertainty of data.

[0076] Specifically, the Gini index is calculated by using the following formula:

[0077]

[0078] wherein p k represents a weight of the kth category.

[0079] Further, the random forest algorithm is used to perform correlation analysis based on the preprocessed historical test data to determine at least one target feature parameter for predicting the spray drift, including: obtaining a Gini index difference value of each candidate feature parameter before and after branching of a decision tree; determining a feature contribution of each candidate feature parameter using the Gini index difference value of each candidate feature parameter; determining a contribution value of the candidate feature parameter using a ratio of the feature contribution of each candidate feature parameter to a total sum of feature contributions of all candidate feature parameters; and taking at least one candidate feature parameter satisfying a preset condition as the target feature parameter.

[0080] For example, for the candidate feature parameter j, the Gini index change value at the node m can be obtained by calculating the difference between the Gini index of the candidate feature parameter j before branching of the node and the Gini index after branching. Specifically, the following formula can be used to express,

[0081] VIM jm = GI m -GI l -GI r

[0082] wherein VIM jm represents the Gini index change value of the candidate feature parameter j at the node m, GI m represents the Gini index before branching, GI l and GI r are the Gini indexes of the two new nodes generated after branching of the node m.

[0083] Therefore, the feature contribution of the candidate feature parameter j in the decision tree i is:

[0084] VIM ij =∑ m∈M VIM jm

[0085] If there are n decision trees in the random forest, the feature contribution of the candidate feature parameter j is:

[0086]

[0087] The contribution value of the candidate feature parameter j is the normalized value of the contribution of the candidate feature parameter j:

[0088]

[0089] wherein VIM' j represents the contribution of the feature j after normalization, represents the sum of the Gini index differences of all features.

[0090] It should be understood that the greater the contribution value of the candidate feature parameter, the greater the influence on the decision, and therefore, by calculating the Gini index of each candidate feature parameter, the importance of multiple candidate feature parameters in the historical test data can be analyzed, and then at least one candidate feature parameter with a higher importance is used as the target feature parameter.

[0091] Optionally, the N candidate feature parameters with the highest contribution value are used as the target feature parameter, or at least one candidate feature parameter with a contribution value of more than 50% is used as the target parameter.

[0092] Preferably, when the downwind distance of the candidate feature parameter meets the selection strategy of the target feature parameter, the downwind distance is directly used as the target feature parameter, and when the downwind distance of the candidate feature parameter does not meet the selection strategy of the target feature parameter, the downwind distance is used as the target feature parameter.

[0093] Further, the pre-processed test parameter data corresponding to the target feature parameter is obtained, and a training set and a test set are constructed. Preferably, the pre-processed test parameter data corresponding to the target feature parameter can be divided into a training set and a test set according to a ratio of 8:2.

[0094] Step 104, constructing an initial ANN-GWO artificial neural network model, and training the initial ANN-GWO artificial neural network model using the training set and the test set to obtain a target prediction model for predicting the drift amount.

[0095] It should be noted that GWO (Grey Wolf Optimizer) is an algorithm that uses the weights and biases of a neural network model as optimization variables, and performs global search and local development by simulating the hunting behavior of a grey wolf group, and then finds the optimal structure combination.

[0096] In the embodiments of the present application, different parameter combinations of ANN artificial neural network models are randomly initialized in the exploration space by the GWO algorithm to train the ANN artificial neural network models, and by iteratively updating these solutions, the optimal solution is gradually approached to determine the best ANN artificial neural network model structure for predicting the drift amount.

[0097] Specifically, as shown in Figure 2 the following steps are included:

[0098] Step 201, constructing an initial ANN artificial neural network model and GWO algorithm parameters.

[0099] It should be noted that the initial ANN artificial neural network model needs to be constructed according to the number of target feature parameters. The GWO algorithm parameters include the number of grey wolves and the maximum number of iterations, wherein the number of grey wolves is greater than or equal to 3.

[0100] Step 202, taking the structural features of the initial ANN artificial neural network model as prey, the position vector of the wolf pack is calculated.

[0101] Before optimization, all weights and biases of the initial ANN artificial neural network model need to be flattened into a one-dimensional vector and mapped to GWO:

[0102] D = (input_dim x hidden_dim) + hidden_dim + (hidden_dim x output_dim) + output_dim

[0103] Wherein, input_dim is the number of input target feature parameters, hidden_dim is the number of neurons in each layer, and output_dim is the number of output results.

[0104] Preferably, the initial position vector of the wolf pack is randomly generated. Wherein, the position vector of each gray wolf in the wolf pack corresponds to a set of ANN artificial neural network model parameters.

[0105] Step 203, calculate the fitness of the current wolf pack position vector based on the training set.

[0106] Specifically, the fitness of each gray wolf position vector in the wolf pack is calculated.

[0107] It should be noted that the fitness is used to evaluate the difference between the predicted value of the ANN artificial neural network model corresponding to the current wolf pack position vector based on the training set and the actual value corresponding to the training set. Wherein, the fitness can adopt the determination coefficient R 2 , root mean square error RMSE, mean absolute error MAE, etc.

[0108] In the embodiments of the present application, since the model prediction needs to be performed according to the target downwind distance during subsequent utilization, that is, the input of the downwind distance feature parameter is focused on, during model training, at least the downwind distance needs to be focused on, while other target feature parameters are also considered.

[0109] Preferably, in the embodiments of the present application, the improved root mean square error is used to calculate the fitness of the current wolf pack position vector.

[0110] Specifically, the present application adopts a multi-objective optimization strategy, that is, a balanced solution is selected through the Pareto frontier to realize simultaneous optimization of model performance and feature weight distribution. The fitness function is:

[0111]

[0112] Wherein, n is the number of training times, ω i is the feature weight, and Oi For the corresponding drift data in the training set, P i For the prediction results according to the test parameter data in the training set.

[0113] It should be noted that the optimization goal of the fitness function is to minimize the model error (root mean square error RMSE) and maximize the uniformity of the feature weight, so as to ensure that the ANN artificial neural network model can effectively consider the influence of other target feature parameters on the drift of the spray under the target downwind distance, so as to ensure that the drift of the spray under the target downwind distance can be predicted by relying on the uniform weight of other target feature parameters.

[0114] Step 204, calculate the fitness optimal three gray wolves information, and update the position vector and the position vector of the wolf pack, and iterate the training.

[0115] Step 205, when the termination condition is met, the position vector of the wolf pack in the current iteration round is assigned to the ANN artificial neural network model to obtain a target prediction model.

[0116] It should be noted that in the embodiment of the present application, the preset maximum iteration number of GWO is used as the training termination condition. Then, the position vector of the gray wolf with the optimal fitness is assigned to the ANN artificial neural network model to obtain a target prediction model. The target prediction model has multiple input ends, which correspond one by one to the number of target feature parameters.

[0117] Step 105, using the target prediction model to predict the drift of the plant protection unmanned aerial vehicle under the target downwind distance.

[0118] Specifically, after obtaining the target prediction model, based on the type of at least one target feature parameter determined by the random forest algorithm, at least one target parameter data to be predicted in flight is determined, such as one or more of flight height, flight speed, droplet size, average environmental wind speed, average temperature, average humidity, and nozzle speed, and then at least one target downwind distance is selected. The target prediction model is input into the target prediction model, and the predicted drift data of the target downwind distance under the target parameter data to be predicted can be obtained.

[0119] It should be understood that the target parameter data to be predicted can be determined according to the environment to be flown, for example, according to the predicted average environmental wind speed, average temperature, and average humidity of the actual to-be-flown period, the flight speed of the plant protection unmanned aerial vehicle within the highest wind speed of three levels of wind speed (wind speed 3.3 m / s) that the unmanned aerial vehicle can perform pesticide application operation, and the flight height that can be reliably flown, etc. The present application does not make specific limitation on this process.

[0120] Thus, the method for predicting the drift of the plant protection unmanned aerial vehicle spraying provided by the embodiment of the application based on the ANN-GWO, in the test stage, collects test parameter data and drift data of a plurality of candidate feature parameters, and uses a random forest algorithm to perform correlation analysis to determine at least one target feature parameter for predicting the drift of the spraying, effectively uses the plurality of candidate feature parameters in cooperation with the random forest algorithm, ensures the correlation and reliability of the selected target feature parameter for predicting the drift of the spraying, and reduces the amount of invalid data in the subsequent model training and the prediction of the drift of the spraying, improves the prediction accuracy of the target prediction model while improving the training calculation overhead and reducing the cost. Moreover, the ANN-GWO artificial neural network model is used, the global search and local development capabilities of the GWO optimization algorithm are effectively used, the problem of difficulty in convergence that is prone to occur when using part of the target feature parameters is solved, and the uniformity of the weights of the plurality of target feature parameters is effectively taken into account, and the reliability of the prediction of the drift of the spraying at different target downwind distances is improved.

[0121] It should be noted that although the operations of the method of the application are described in a specific order in the accompanying drawings, this does not require or imply that the operations must be performed in this specific order, or that all of the shown operations must be performed to achieve the desired result.

[0122] Reference will now be made to the following description Figure 3 , Figure 3 shows a structural schematic diagram of a computer system of an electronic device or a server suitable for being used to implement the embodiment of the application,

[0123] As shown in Figure 3 , the computer system includes a central processing unit (CPU) 301, which can perform various appropriate actions and processes according to programs stored in a read-only memory (ROM) 302 or programs loaded from a storage portion 308 into a random access memory (RAM) 303. In the RAM 303, various programs and data required for operation instructions of the system are also stored. The CPU 301, the ROM 302, and the RAM 303 are connected to each other through a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.

[0124] The following components are connected to the I / O interface 305; an input part 306 including a keyboard, a mouse, etc.; an output part 307 including a display such as a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage part 308 including a hard disk, etc.; and a communication part 309 including a network interface card such as a LAN card, a modem, etc. The communication part 309 performs communication processing via a network such as the Internet. A drive 310 is also connected to the I / O interface 305 as necessary. A removable medium 311 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc. is attached to the drive 310 as necessary, so that a computer program read out therefrom is installed in the storage part 308 as necessary.

[0125] In particular, in accordance with embodiments of the present application, the above-referenced flowcharts Figure 2 The processes described can be implemented as computer software programs. For example, embodiments of the present application include a computer program product which includes a computer program carried on a computer-readable medium, the computer program containing program code for executing the methods illustrated by the flowcharts. In such embodiments, the computer program contains program code for executing the methods illustrated by the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network by the communication part 309, and / or installed from the removable medium 311. When the computer program is executed by the central processing unit (CPU) 301, the above-described functions defined in the system of the present application are executed.

[0126] It should be noted that the computer-readable medium shown in the application can be a computer-readable signal medium or a computer-readable storage medium or any combination of the two. The computer-readable storage medium may, for example, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or component, or any combination of the above. More specific examples of computer-readable storage media can include, but are not limited to, an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this application, the computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, device or component. In this application, the computer-readable signal medium can include a data signal carried in a baseband or as a carrier wave part of a carrier wave, which carries computer-readable program code. Such a propagated data signal can take many forms, including but not limited to an electromagnetic signal, an optical signal, or any suitable combination of the above. The computer-readable signal medium can also be any computer-readable medium that can send, propagate or transmit a program for use by or in conjunction with an instruction execution system, device or component. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to wireless, wire, optical cable, RF, etc., or any suitable combination of the above.

[0127] The flowcharts and block diagrams in the drawings illustrate the possible implementation architectures, functions and operation instructions of the systems, methods and computer program products according to various embodiments of the application. In this regard, each block in the flowchart or block diagram can represent a module, a program segment or a part of code, which contains one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions noted in the blocks can also occur in different order from that noted in the drawings. For example, two connected blocks can actually be executed substantially in parallel, and sometimes in reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, and the combination of blocks in the block diagram and / or flowchart, can be implemented by a dedicated hardware-based system that performs the specified functions or operation instructions, or can be implemented by a combination of dedicated hardware and computer instructions.

[0128] As another aspect, the present application also provides a computer readable storage medium, which can be contained in the electronic device described in the above embodiments, or can exist separately without being assembled into the electronic device. The computer readable storage medium stores one or more programs, which are used by one or more processors to execute the ANN-GWO-based plant protection unmanned aerial vehicle spray drift prediction method described in the present application.

[0129] The above description is merely the preferred embodiments of the present application and the explanation of the technical principles used. It should be understood by those skilled in the art that the disclosed scope of the present application is not limited to the technical solutions formed by the specific combinations of the above technical features, and should also cover other technical solutions formed by any combinations of the above technical features or their equivalent features without departing from the above disclosed concept. For example, the technical solutions formed by the mutual replacement of the above features and the technical features disclosed in the present application (but not limited to) having similar functions.

Claims

1. A method for predicting the spray drift of plant protection drones based on ANN-GWO, characterized by: include: During the testing phase, historical test data of spray drift tests of plant protection unmanned aircraft are obtained, wherein the historical test data includes at least test parameter data and drift data of multiple candidate characteristic parameters under different test conditions, wherein the candidate characteristic parameters include at least flight altitude, flight speed, droplet size, average ambient wind speed, average temperature, average humidity, nozzle rotation speed, and downwind distance; Preprocessing the historical test data to obtain preprocessed historical test data; performing a correlation analysis based on the preprocessed historical test data using a random forest algorithm to determine at least one target characteristic parameter for spray drift prediction, and obtaining preprocessed test parameter data corresponding to the target characteristic parameter to construct a training set and a test set; the target characteristic parameter at least includes the downwind distance; constructing an initial ANN-GWO artificial neural network model, and training the initial ANN-GWO artificial neural network model using the training set and the test set to obtain a target prediction model for spray drift prediction; The target prediction model is used to predict the spray drift of the plant protection unmanned aircraft at the target downwind distance.

2. The method for predicting spray drift of plant protection unmanned aircraft based on ANN-GWO according to claim 1, characterized in that: During the testing phase, historical test data of the spray drift test of the plant protection drone is obtained, including: Arrange nine ground droplet collection devices with a diameter of 15 cm in the downwind direction and at 3m, 5m, 10m, 15m, 20m, 30m and 50m from the edge of the spray pattern of the plant protection drone respectively; During the testing phase, a simulated pesticide solution containing 0.1% of both the ABF fluorescent tracer and the OP-10 non-ionic surfactant was used to spray the plant protection drone, and the ground droplet collection device was used to collect the droplets drifting on the ground. The droplet collecting devices are eluted with deionized water to obtain fluorescence values ​​corresponding to each of the droplet collecting devices, and the drift amount data corresponding to the test is determined by the fluorescence values.

3. The method for predicting spray drift of plant protection unmanned aircraft based on ANN-GWO according to claim 1, characterized in that: The preprocessing of the historical test data to obtain the preprocessed historical test data includes: Deleting missing values ​​in the historical test data; and Performing normalization and standardization on the historical test data after missing value processing; The following formula is used for normalization: Among them, x i is the normalized original historical test data, is the normalized historical test data, and its value range is [0,1], x max is the maximum value in the original historical test data; The following formula is used for standardization: Among them, μ is the mean of the normalized historical test data, σ is the standard deviation of the normalized historical test data, and x standardized The standardized historical test data conforms to the standard normal distribution with a mean of 0 and a standard deviation of 1.

4. The method for predicting spray drift of plant protection unmanned aircraft based on ANN-GWO according to claim 1, characterized in that: The method of using a random forest algorithm to perform correlation analysis based on the preprocessed historical test data to determine at least one target characteristic parameter for spray drift prediction includes: Obtaining the Gini index difference of each candidate feature parameter before and after the decision tree branch; Determining a feature contribution of each candidate feature parameter by using the Gini index difference of each candidate feature parameter; Determining a contribution value of the candidate feature parameter by using a ratio of the feature contribution amount of each candidate feature parameter to the sum of the feature contribution amounts of all candidate feature parameters; At least one of the candidate feature parameters whose contribution value meets a preset condition is used as the target feature parameter.

5. The method for predicting spray drift of plant protection unmanned aircraft based on ANN-GWO according to claim 4, characterized in that: The Gini index difference is calculated using the following formula: VIM jm =GI m -AND l -AND r Among them, VIM jm Indicates the Gini index change value of candidate feature parameter j on node m, GI m Represents the Gini index before branching, GI l and GI r Then it is the Gini index of the two new nodes generated after the node m branches; And, the Gini index is calculated using the following formula: Among them, p k represents the weight of the k-th category.

6. The method for predicting spray drift of plant protection unmanned aircraft based on ANN-GWO according to claim 1, characterized in that: The initial ANN-GWO artificial neural network model is constructed, and the initial ANN-GWO artificial neural network model is trained using the training set and the test set to obtain a target prediction model for spray drift amount prediction, including: Construct the initial ANN artificial neural network model and GWO algorithm parameters; The structural features of the initial ANN artificial neural network model are used as prey to calculate the position vector of the wolf pack; Calculating the fitness corresponding to the current wolf pack position vector based on the training set; Calculate the information of the three gray wolves with the best fitness and update their position vectors and the position vectors of the wolf pack to perform iterative training; When the termination condition is met, the position vector of the wolf pack in the current iteration round is assigned to the ANN artificial neural network model to obtain the target prediction model.

7. The method for predicting spray drift of plant protection unmanned aircraft based on ANN-GWO according to claim 6, characterized in that: Also includes: The improved root mean square error is used to calculate the fitness corresponding to the current wolf pack position vector.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, it implements the ANN-GWO-based plant protection unmanned aircraft spray drift prediction method as described in any one of claims 1 to 7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method for predicting the spray drift amount of a plant protection unmanned aircraft based on ANN-GWO as described in any one of claims 1 to 7 is implemented.

10. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the method for predicting the spray drift amount of a plant protection unmanned aircraft based on ANN-GWO as described in any one of claims 1 to 7 is implemented.

Images