A method and system for analyzing crop yields from bee pollination based on big data

Abstract

The invention discloses a method and system for analyzing crop yields based on bee pollination based on big data, and relates to the field of yield data analysis. The invention optimizes data of multiple yield influencing factors corresponding to a sample matrix of pollinated crops, maximizes the advantage of bee pollination, and can reflect the best effect that can be achieved when bee pollination is adopted; by comparing the actual optimal yield data set of pollinated crops with the optimal yield of a sample matrix of non-pollinated crops, the difference in the influence of bee pollination and non-bee pollination on the yields of multiple crops can be reflected; by adopting a matrix set of yield influencing factor data and a yield data matrix of pollinated crops to construct a final KNN classification model, the final KNN model can predict the corresponding crop yield data according to the yield influencing factor data adopted for crops, and simplifies the subsequent process of analyzing the yield of the sample set of pollinated crops to be analyzed.

Description

Technical Field

[0001] The present invention belongs to the field of yield data analysis, and in particular, relates to a method and system for analyzing crop yield through bee pollination based on big data. Background Art

[0002] Chinese patent CN117033810B discloses an agricultural data analysis management system and method based on big data. It uses historical data to understand the yield of the same crops in any area, and analyzes the growth status of the same type of crops in any area. It combines the yield of the same type of crops in different growth environments and different planting methods, and matches the optimal growth conditions of the corresponding type of crops according to the analysis results, thereby effectively improving the production quality and yield of crops.

[0003] Bee pollination can help crops to pollinate more efficiently, which is beneficial to increase crop yields; bee pollination can be especially applied to wolfberries. In the process of collecting wolfberry nectar, bees can promote cross-pollination of wolfberries and achieve extensive pollination, thereby increasing the fruit setting rate and the fullness of the fruit; however, plant diseases and insect pests may indirectly affect the health and behavior of pollinating insects, and increasing pesticide spraying can prevent the occurrence of diseases and pests; however, pesticide spraying also has a great impact on the health of bees; however, the current analysis of the relationship between bee pollination and crop yield has not taken into account the impact of pesticide spraying on bee pollination, making the analysis less accurate. Summary of the invention

[0004] In view of the problems in the related technology, the present invention proposes a method and system for analyzing crop yield based on bee pollination based on big data to overcome the above-mentioned technical problems existing in the existing related technology.

[0005] In order to solve the above technical problems, the present invention is achieved through the following technical solutions:

[0006] The present invention is a method for analyzing crop yields based on bee pollination based on big data, comprising the following steps:

[0007] S1. Collecting samples of multiple types of crops to obtain a pollinated crop sample matrix and a non-pollinated crop sample matrix;

[0008] S2, optimizing the pesticide application rate of each crop sample in the non-pollination crop sample matrix to obtain a final yield data set; constructing a plurality of yield influencing factors corresponding to the pollination crop sample matrix to obtain a yield influencing factor set;

[0009] S3, optimizing the various yield influencing factor data corresponding to the pollination crop sample matrix according to the yield influencing factor set, obtaining a yield influencing factor data matrix set, an actual optimal yield data set of pollination crops, and a pollination crop yield data matrix; comparing the actual optimal yield data set of pollination crops with the final yield data set to obtain an optimal yield growth degree data set; visually displaying the optimal yield growth degree data set and the corresponding yield influencing factor data matrix;

[0010] S4, using the yield influencing factors data matrix set and the pollinated crop yield data matrix to construct the final KNN classification model;

[0011] S5, setting a sample set of pollinated crops to be analyzed and a corresponding data set of yield influencing factors to be analyzed, using the final KNN classification model to classify the data set of yield influencing factors to be analyzed corresponding to the sample set of pollinated crops to be analyzed, and obtaining a classification result; and determining whether the data set of yield influencing factors to be analyzed needs to be adjusted according to the classification result;

[0012] The crop is preferably wolfberry;

[0013] By collecting samples of multiple types of crops, the samples are more abundant, which improves the versatility of the subsequent optimization scheme for the dosage of pesticides. The collected samples of multiple types of crops are equally divided into pollinated and non-pollinated types to analyze the yields of the crop samples in the two situations separately, which improves the pertinence and accuracy of the analysis. By optimizing the dosage of each crop sample in the non-pollinated crop sample matrix, each crop sample in the non-pollinated crop sample matrix can achieve the optimal yield when planted, so that the subsequent comparison between each crop sample and the pollinated crop sample matrix is ​​based on their respective optimal yields, making the comparison more scientific and the analysis more accurate. By constructing a variety of yield influencing factors corresponding to the pollinated crop sample matrix, the direction of subsequent optimization of the yield of each crop sample in the pollinated crop sample matrix is ​​determined. By comparing the data of a variety of yield influencing factors corresponding to the pollinated crop sample matrix, the optimization of the yield of each crop sample in the pollinated crop sample matrix is ​​carried out. Optimization can maximize the advantages of bee pollination, thereby reflecting the best effect that can be achieved when using bee pollination, and providing a data basis for subsequent comparison between the optimal yield of the non-pollinated crop sample matrix; by comparing the actual optimal yield data set of pollinated crops with the optimal yield of the non-pollinated crop sample matrix, it can reflect the difference in the impact of bee pollination and non-bee pollination on the yield of various crops; the comparison results are then visualized to facilitate visitors to understand this difference intuitively; the final KNN classification model is constructed by using the yield influencing factor data matrix set and the pollinated crop yield data matrix, so that the final KNN model can predict the corresponding crop yield data based on the yield influencing factor data of the crops, thereby eliminating the need for field measurement of the crop yield, simplifying the subsequent analysis of the yield of the pollinated crop sample set to be analyzed.

[0014] Preferably, the S1 comprises the following steps:

[0015] S11. Set the crop type set represents the i-th crop type set, Indicates the total number of set crop types; Collect multiple crop samples to obtain a crop sample matrix; divide the crop sample matrix equally to obtain a pollination crop sample matrix and non-pollinating crop sample matrices ; They are as follows:

[0016] ; ;

[0017] in, represents the jth pollinated crop sample corresponding to the i-th crop type; represents the jth non-pollinating crop sample corresponding to the i-th crop type; represents the total number of pollinated crop samples and non-pollinated crop samples collected for each crop type;

[0018] S12, setting the first planting area set and the second planting area set ; , respectively represent the i-th planting area in the first planting area set and the i-th planting area in the second planting area set;

[0019] The pollinated crop sample matrix Each crop sample was planted in the first planting area according to its type. In each planting area, the non-pollinating crop sample matrix Each crop sample in the second planting area is planted separately according to its type. In each planting area; the pollination crop sample matrix and non-pollinating crop sample matrices The planting environment conditions are the same;

[0020] The crop types include cereals, beans, potatoes, fiber crops, oil crops, sugar crops, leafy vegetables, root vegetables, fruits, beans, etc.; by setting the first planting area set and the second planting area set, each crop sample in the pollination crop sample matrix and each crop sample in the non-pollination crop sample matrix are planted separately, so that subsequent bee pollination will not affect each other, thereby improving the accuracy of subsequent comparison results.

[0021] Preferably, the planting environment conditions include soil nutrients, soil moisture, air temperature, air humidity and light, etc.;

[0022] Preferably, S2 comprises the following steps:

[0023] S21. Obtaining a sample matrix of non-pollinating crops The optimal application amount and the corresponding optimal yield of each type of crop sample in the final yield data set are obtained. ; Indicates the sample matrix for non-pollinating crops The optimal yield value of the i-th type of crop sample;

[0024] S22, for the pollination crop sample matrix Set the set of factors affecting production , , , , , They represent the amount of pesticide applied before pollination, the amount of pesticide applied after pollination, the start time of pesticide removal, the end time of pesticide removal, and the number of bees;

[0025] By setting five yield influencing factors, including the amount of pesticide applied before pollination, the amount of pesticide applied after pollination, the start time of pesticide removal, the end time of pesticide removal, and the number of bees, these yield influencing factors have a great impact on the bee pollination process, providing optimization variables for the subsequent optimization of the yield of pollinated crops, thereby quantifying the subsequent optimization process.

[0026] Preferably, the S21 comprises the following steps:

[0027] S211. Establishing the first wild horse population , represents the i-th wild horse in the first wild horse population, c represents the size of the first wild horse population; the maximum number of iterations of the first wild horse population is set to , recorded as the first maximum number of iterations, the current number of iterations is , recorded as the first current iteration number, the search space dimension is The mating probability is , denoted as the first mating probability;

[0028] S212, setting the initial position of each wild horse in the first wild horse population and the initial position of the leader in the first wild horse population, and obtaining a first initial position matrix and the second initial position matrix ; They are as follows:

[0029] ; ;

[0030] in, represents the component of the initial position of the i-th common wild horse in the first wild horse population in the j-th dimension; represents the component of the initial position of the i-th leader in the first wild horse population in the j-th dimension; , They respectively represent the total number of common wild horses and leaders in the first wild horse population; the calculation formulas are as follows:

[0031] ; ;

[0032] In the formula, , Represents the non-pollinating crop sample matrix The minimum and maximum pesticide application rates of the i-th type of crop samples; , Respectively for , The generated random number, ;

[0033] S213: Setting the non-pollination crop sample matrix The actual yield of each type of crop sample in the , and the actual yield value set of non-pollination crops is obtained ; Represents the non-pollinating crop sample matrix The actual yield value of the i-th type of crop sample;

[0034] According to the actual yield value set of non-pollination crops Set the fitness function of the first wild horse population ;as follows:

[0035] ;

[0036] S214, start the iteration operation; before the iteration, the first current iteration number Set to 1; in each iteration, according to the fitness function of the first wild horse population Calculate the fitness value of each wild horse in the first wild horse population to obtain a first fitness value set; take the position of the wild horse individual corresponding to the maximum fitness value in the first fitness value set as the first global optimal position; and determine the first mating probability according to the first global optimal position and the first mating probability. The positions of the common wild horses and the leader in the first wild horse population are updated; after the position update is completed, the first current iteration number is Add 1 and enter the next iteration;

[0037] S215, when When , the iteration is stopped, and the first final global optimal position is obtained, and the first final global optimal position is used as the final pesticide dosage data set , Indicates the optimal pesticide application rate of the i-th type of crop sample; otherwise, continue to iterate until ;

[0038] The final pesticide dosage dataset Application to non-pollinating crop sample matrices Each type of crop sample in the final yield dataset ;

[0039] The wild horse optimization algorithm can guide the population to evolve towards the optimal solution and improve the optimization efficiency through the concept of leaders and the communication mechanism between leaders; therefore, in this scheme, the wild horse optimization algorithm is used to perform multiple iterative optimizations on the application rate of each type of crop sample in the non-pollination crop sample matrix, and the yield of each type of crop sample in the non-pollination crop sample matrix is ​​used as the fitness function. The fitness function can intuitively reflect the optimization effect, so that the optimal application rate of each type of crop sample in the non-pollination crop sample matrix can be efficiently searched.

[0040] Preferably, S3 comprises the following steps:

[0041] S31, according to the output influencing factors set Sample matrix for pollinated crops Optimize the yield data and obtain the data matrix set of yield influencing factors And the corresponding optimal yield dataset of pollinated crops ; Represents the pollinated crop sample matrix The data matrix of factors affecting the yield of the i-th type of crop samples; Represents the pollinated crop sample matrix The actual optimal yield data of the i-th type of crop samples; as follows:

[0042] ;

[0043] in, , , , , Respectively represent the pollination crop sample matrix The pre-pollination pesticide application data, post-pollination pesticide application data, pesticide application and removal start time data, pesticide application and removal end time data, and bee quantity data in the j-th group of yield influencing factor data of the i-th type of crop sample; Represents the sample matrix of the pollinated crops The total number of groups of yield influencing factor data for each type of pollinated crop samples obtained after yield data optimization;

[0044] S32, the pollination crop actual optimal yield data set With the final output data set Compare and calculate to get the optimal yield growth data set , Represents the pollinated crop sample matrix The actual optimal yield data of the i-th crop sample is compared with the non-pollination crop sample matrix The growth rate data of the final yield data of the i-th type of crop samples; the calculation formula is as follows:

[0045] ;

[0046] S33, the optimal yield growth degree data set and the actual optimal yield dataset of pollinated crops The corresponding data matrix of production influencing factors is visualized;

[0047] By calculating the yield data growth rate of the actual optimal yield data set of pollinated crops relative to the final yield data set, and then visually displaying the yield data growth rate, the advantages of crops pollinated by bees over crops not pollinated by bees can be intuitively reflected.

[0048] Preferably, the S31 comprises the following steps:

[0049] S311. Building the second wild horse population , represents the i-th wild horse in the second wild horse population, represents the size of the second wild horse population; sets the maximum number of iterations of the second wild horse population to , recorded as the second maximum number of iterations, the current number of iterations is , recorded as the second current iteration number, the search space dimension is , the mating probability is , denoted as the second mating probability;

[0050] S312, according to the output influencing factor set Set the initial position of each wild horse in the second wild horse population and the initial position of the leader in the second wild horse population to obtain the first initial position matrix set and the second initial position matrix set ; represents the initial position matrix of the i-th common wild horse in the second wild horse population, represents the initial position matrix of the i-th leader in the second wild horse population, , represent the total number of common wild horses and leaders in the second wild horse population respectively; and They are as follows:

[0051] ; ;

[0052] in, , , , , They are The pollination crop sample matrix is ​​represented by The location component of the pre-pollination pesticide application data, post-pollination pesticide application data, pesticide removal start time data, pesticide removal end time data, and bee quantity data of the i-th type of crop samples; , , , , They are The pollination crop sample matrix is ​​represented by The location component of the pre-pollination pesticide application data, post-pollination pesticide application data, pesticide removal start time data, pesticide removal end time data, and bee quantity data of the i-th type of crop samples;

[0053] S313: Setting the pollination crop sample matrix The actual yield of each type of crop sample in the pollination is obtained to obtain the actual yield value set of pollinated crops. ; Represents the pollinated crop sample matrix The actual yield value of the i-th type of crop sample;

[0054] According to the actual yield value set of pollinated crops Set the fitness function of the second wild horse population ;as follows:

[0055] ;

[0056] S314, start the iteration operation; before the iteration, set the second current iteration number Set to 1; in each iteration, according to the fitness function of the second wild horse population Calculate the fitness value of each wild horse in the second wild horse population to obtain a second fitness value set; take the position of the wild horse individual corresponding to the maximum fitness value in the second fitness value set as the second global optimal position; and determine the second global optimal position and the second mating probability. The positions of the ordinary wild horses and the leader in the second wild horse population are updated; after the position update is completed, the second global best position and the corresponding fitness value are recorded and the second current iteration number is recorded. Add 1 and enter the next iteration;

[0057] S315, when When , the iteration is stopped, and the second final global optimal position is obtained, and the second final global optimal position is used as the optimal data matrix of the yield influencing factors Otherwise, continue iterating until ; Optimal data matrix of factors affecting production as follows:

[0058] ;

[0059] in, , , , , Respectively represent the pollination crop sample matrix The optimal pre-pollination pesticide application data, the optimal post-pollination pesticide application data, the optimal pesticide removal start time data, the optimal pesticide removal end time data and the optimal bee quantity data corresponding to the i-th type of crop samples;

[0060] The optimal data matrix of the yield influencing factors Application to pollinated crop sample matrices Each type of crop sample in the , to obtain the actual optimal yield data set of pollinated crops ; At the same time, we get the data matrix set of factors affecting production And the corresponding pollination crop yield data matrix ;as follows:

[0061] ;

[0062] in, Represents the sample matrix of the pollinated crops The yield data of pollinated crops corresponding to the j-th group of yield influencing factor data obtained by optimizing the yield influencing factor data of the i-th type of crop samples;

[0063] By using the wild horse optimization algorithm to perform multiple iterative optimizations on the pre-pollination pesticide application amount, post-pollination pesticide application amount, pesticide removal start time, pesticide removal end time and bee number of each type of crop sample in the pollination crop sample matrix, and at the same time using the yield of each type of crop sample in the pollination crop sample matrix as the fitness function, the optimal pre-pollination pesticide application amount, post-pollination pesticide application amount, pesticide removal start time, pesticide removal end time and bee number of each type of crop sample in the pollination crop sample matrix can be efficiently searched.

[0064] Preferably, S4 comprises the following steps:

[0065] S41, constructing an initial KNN classification model and setting a training data ratio; and classifying a data matrix set of production influencing factors according to the training data ratio. and the pollinated crop yield data matrix Performing data division to obtain a yield influencing factor training data matrix set, a yield influencing factor test data matrix set, a pollination crop yield training data matrix, and a pollination crop yield test data matrix;

[0066] S42, setting a training error threshold; inputting the yield influencing factor training data matrix set and the pollination crop yield training data matrix into the initial KNN classification model for training; during the training process, when the training error is less than the training error threshold, stopping the training to obtain a trained KNN classification model; otherwise, continuing the training until the training error is less than the training error threshold;

[0067] S43, setting a test accuracy threshold; inputting the yield influencing factor test data matrix set and the pollination crop yield test data matrix into the trained KNN classification model for testing; after the test is completed, obtaining the test accuracy; when the test accuracy is greater than or equal to the test accuracy threshold, using the trained KNN classification model as the final KNN classification model; otherwise, returning to S42 to continue training until the test accuracy is greater than or equal to the test accuracy threshold, and obtaining the final KNN classification model;

[0068] By using the yield influencing factor data matrix set and the pollinated crop yield data matrix to train and test the initial KNN classification model, the final KNN classification model can directly predict the yield data based on the yield influencing factor data, providing a classification model for the subsequent classification of the pollinated crop sample set to be analyzed.

[0069] Preferably, the S5 comprises the following steps:

[0070] S51. Setting the pollination crop sample set to be analyzed and the corresponding yield influencing factor data set to be analyzed , , , , , Respectively represent the pre-pollination pesticide application amount data, post-pollination pesticide application amount data, pesticide application removal start time data, pesticide application removal end time data and bee quantity data set for the pollination crop sample set to be analyzed;

[0071] S52, setting an expected yield threshold; and Input into the final KNN classification model for classification, and obtain the yield data of the pollinated crop samples to be analyzed;

[0072] When the yield data of the pollinated crop sample to be analyzed is greater than or equal to the expected yield threshold, there is no need to analyze the yield influencing factor dataset. When the sample yield data of the pollinated crop to be analyzed is less than the expected yield threshold, the actual optimal yield data set of the pollinated crop is adjusted. Dataset of factors affecting yield to be analyzed Adjustment is performed to obtain an adjusted data set of production influencing factors to be analyzed;

[0073] By using the final KNN classification model to classify the dataset of yield influencing factors to be analyzed, the yield data of the pollinated crop sample set to be analyzed can be predicted. When the yield data does not meet the requirements, the dataset of yield influencing factors to be analyzed is adjusted, and an adjustment plan is given to facilitate growers to make timely adjustments and increase crop yields.

[0074] A crop yield analysis system based on bee pollination based on big data, including a crop sample collection module, a non-pollination crop pesticide application optimization module, a yield influencing factor construction module, a yield influencing factor data optimization module, a comparison module, a visualization display module, a classification model construction module, a classification module and a judgment module;

[0075] The crop sample collection module is used to collect samples of multiple types of crops to obtain a pollinated crop sample matrix and a non-pollinated crop sample matrix;

[0076] The non-pollination crop pesticide application rate optimization module is used to optimize the pesticide application rate of each crop sample in the non-pollination crop sample matrix to obtain a final yield data set;

[0077] The yield influencing factor construction module is used to construct a plurality of yield influencing factors corresponding to the pollination crop sample matrix to obtain a yield influencing factor set;

[0078] The yield influencing factor data optimization module is used to optimize the various yield influencing factor data corresponding to the pollination crop sample matrix according to the yield influencing factor set, to obtain a yield influencing factor data matrix set, an actual optimal yield data set of pollination crops, and a pollination crop yield data matrix;

[0079] The comparison module is used to compare the actual optimal yield data set of the pollinated crops with the final yield data set to obtain the optimal yield growth degree data set;

[0080] The visualization display module is used to visualize the optimal yield growth degree data set and the corresponding yield influencing factor data matrix;

[0081] The classification model building module is used to construct a final KNN classification model using a yield influencing factor data matrix set and a pollinated crop yield data matrix;

[0082] The classification module is used to classify the yield influencing factor data set to be analyzed corresponding to the pollinated crop sample set to be analyzed using the final KNN classification model to obtain a classification result;

[0083] The determination module is used to determine whether the dataset of yield influencing factors to be analyzed needs to be adjusted according to the classification result.

[0084] The present invention has the following beneficial effects:

[0085] 1. In the present invention, the advantages of bee pollination are maximized by optimizing the data of various yield influencing factors corresponding to the pollinated crop sample matrix, thereby reflecting the best effect that can be achieved when using bee pollination; by comparing the actual optimal yield data set of pollinated crops with the optimal yield of the non-pollinated crop sample matrix, the difference in the impact of bee pollination and non-bee pollination on the yield of various crops can be reflected; by using the yield influencing factor data matrix set and the pollinated crop yield data matrix to construct the final KNN classification model, the final KNN model can predict the corresponding crop yield data based on the yield influencing factor data of crops, thereby simplifying the subsequent process of analyzing the yield of the pollinated crop sample set to be analyzed.

[0086] 2. In the present invention, the wild horse optimization algorithm is used to perform multiple iterative optimizations on the yield influencing factor data of each type of crop sample in the pollination crop sample matrix. At the same time, the yield of each type of crop sample in the pollination crop sample matrix is ​​used as the fitness function, so that the optimal yield influencing factor data of each type of crop sample in the pollination crop sample matrix can be efficiently searched.

[0087] 3. In the present invention, the final KNN classification model is used to classify the data set of yield influencing factors to be analyzed, so that the yield data of the pollinated crop sample set to be analyzed can be predicted. When the yield data does not meet the requirements, the data set of yield influencing factors to be analyzed is adjusted, so that growers can make timely adjustments and improve crop yields.

[0088] Of course, any product implementing the present invention does not necessarily need to achieve all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0089] In order to more clearly illustrate the technical solutions of the embodiments of the invention, the drawings required for describing the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the invention. For ordinary technicians in this field, they can also obtain drawings based on these drawings without paying creative work.

[0090] FIG1 is a schematic diagram of a process of analyzing crop yield using a crop yield analysis system based on bee pollination based on big data according to the present invention. DETAILED DESCRIPTION

[0091] The following will be combined with the drawings in the embodiments of the invention to clearly and completely describe the technical solutions in the embodiments of the invention. Obviously, the described embodiments are only part of the embodiments of the invention, not all of the embodiments. Based on the embodiments in the invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the invention.

[0092] In the description of the present invention, it is necessary to understand that the terms "opening", "upper", "lower", "top", "middle", "inside" and the like indicating orientation or positional relationship are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the components or elements referred to must have a specific orientation, be constructed and operate in a specific orientation, and therefore cannot be understood as limiting the invention.

[0093] Embodiment 1

[0094] This embodiment is a method for analyzing crop yields based on bee pollination based on big data, comprising the following steps:

[0095] S1. Collecting samples of multiple types of crops to obtain a pollinated crop sample matrix and a non-pollinated crop sample matrix;

[0096] The S1 comprises the following steps:

[0097] S11. Set the crop type set , represents the i-th crop type set, Indicates the total number of set crop types; Collect multiple crop samples to obtain a crop sample matrix; divide the crop sample matrix equally to obtain a pollination crop sample matrix and non-pollinating crop sample matrices ; They are as follows:

[0098] ; ;

[0099] in, represents the jth pollinated crop sample corresponding to the i-th crop type; represents the jth non-pollinating crop sample corresponding to the i-th crop type; represents the total number of pollinated crop samples and non-pollinated crop samples collected for each crop type;

[0100] S12, setting the first planting area set and the second planting area set ; , respectively represent the i-th planting area in the first planting area set and the i-th planting area in the second planting area set;

[0101] The pollinated crop sample matrix Each crop sample was planted in the first planting area according to its type. In each planting area, the non-pollinating crop sample matrix Each crop sample in the second planting area is planted separately according to its type. In each planting area; the pollination crop sample matrix and non-pollinating crop sample matrices The planting environment conditions are the same;

[0102] S2, optimizing the pesticide application rate of each crop sample in the non-pollination crop sample matrix to obtain a final yield data set; constructing a plurality of yield influencing factors corresponding to the pollination crop sample matrix to obtain a yield influencing factor set;

[0103] The S2 comprises the following steps:

[0104] S21. Obtaining a sample matrix of non-pollinating crops The optimal application amount and the corresponding optimal yield of each type of crop sample in the final yield data set are obtained. ; Indicates the sample matrix for non-pollinating crops The optimal yield value of the i-th type of crop sample;

[0105] The S21 comprises the following steps:

[0106] S211. Establishing the first wild horse population , represents the i-th wild horse in the first wild horse population, c represents the size of the first wild horse population; the maximum number of iterations of the first wild horse population is set to , recorded as the first maximum number of iterations, the current number of iterations is , recorded as the first current iteration number, the search space dimension is , the mating probability is , denoted as the first mating probability;

[0107] S212, setting the initial position of each wild horse in the first wild horse population and the initial position of the leader in the first wild horse population, and obtaining a first initial position matrix and the second initial position matrix ; They are as follows:

[0108] ; ;

[0109] in, represents the component of the initial position of the i-th common wild horse in the first wild horse population in the j-th dimension; represents the component of the initial position of the i-th leader in the first wild horse population in the j-th dimension; , They respectively represent the total number of common wild horses and leaders in the first wild horse population; the calculation formulas are as follows:

[0110] ; ;

[0111] In the formula, , Represents the non-pollinating crop sample matrix The minimum and maximum pesticide application rates of the i-th type of crop samples; , Respectively for , The generated random number, ;

[0112] S213: Setting the non-pollination crop sample matrix The actual yield of each type of crop sample in the , and the actual yield value set of non-pollination crops is obtained ; Represents the non-pollinating crop sample matrix The actual yield value of the i-th type of crop sample;

[0113] According to the actual yield value set of non-pollination crops Set the fitness function of the first wild horse population ;as follows:

[0114] ;

[0115] S214, start the iteration operation; before the iteration, the first current iteration number Set to 1; in each iteration, according to the fitness function of the first wild horse population Calculate the fitness value of each wild horse in the first wild horse population to obtain a first fitness value set; take the position of the wild horse individual corresponding to the maximum fitness value in the first fitness value set as the first global optimal position; and determine the first mating probability according to the first global optimal position and the first mating probability. The positions of the common wild horses and the leader in the first wild horse population are updated; after the position update is completed, the first current iteration number is Add 1 and enter the next iteration;

[0116] S215, when When , the iteration is stopped, and the first final global optimal position is obtained, and the first final global optimal position is used as the final pesticide dosage data set , Indicates the optimal pesticide application rate of the i-th type of crop sample; otherwise, continue to iterate until ;

[0117] The final pesticide dosage dataset Application to non-pollinating crop sample matrices Each type of crop sample in the final yield dataset ;

[0118] S22, for the pollination crop sample matrix Set the set of factors affecting production , , , , , They represent the amount of pesticide applied before pollination, the amount of pesticide applied after pollination, the start time of pesticide removal, the end time of pesticide removal, and the number of bees;

[0119] S3, optimizing the various yield influencing factor data corresponding to the pollination crop sample matrix according to the yield influencing factor set, obtaining a yield influencing factor data matrix set, an actual optimal yield data set of pollination crops, and a pollination crop yield data matrix; comparing the actual optimal yield data set of pollination crops with the final yield data set to obtain an optimal yield growth degree data set; visually displaying the optimal yield growth degree data set and the corresponding yield influencing factor data matrix;

[0120] The S3 comprises the following steps:

[0121] S31, according to the output influencing factors set Sample matrix for pollinated crops Optimize the yield data and obtain the data matrix set of yield influencing factors And the corresponding optimal yield dataset of pollinated crops ; Represents the pollinated crop sample matrix The data matrix of yield influencing factors of the i-th type of crop samples; Represents the pollination crop sample matrix The actual optimal yield data of the i-th type of crop samples; as follows:

[0122] ;

[0123] in, , , , , Respectively represent the pollination crop sample matrix The pre-pollination pesticide application data, post-pollination pesticide application data, pesticide application and removal start time data, pesticide application and removal end time data, and bee quantity data in the j-th group of yield influencing factor data of the i-th type of crop sample; Represents the sample matrix of the pollinated crops The total number of groups of yield influencing factor data for each type of pollinated crop samples obtained after yield data optimization;

[0124] The S31 comprises the following steps:

[0125] S311. Building the second wild horse population , represents the i-th wild horse in the second wild horse population, represents the size of the second wild horse population; sets the maximum number of iterations of the second wild horse population to , recorded as the second maximum number of iterations, the current number of iterations is , recorded as the second current iteration number, the search space dimension is , the mating probability is , denoted as the second mating probability;

[0126] S312, according to the output influencing factor set Set the initial position of each wild horse in the second wild horse population and the initial position of the leader in the second wild horse population to obtain the first initial position matrix set and the second initial position matrix set ; represents the initial position matrix of the i-th common wild horse in the second wild horse population, represents the initial position matrix of the i-th leader in the second wild horse population, , represent the total number of common wild horses and leaders in the second wild horse population respectively; and They are as follows:

[0127] ; ;

[0128] in, , , , , They are The pollination crop sample matrix is ​​represented by The location component of the pre-pollination pesticide application data, post-pollination pesticide application data, pesticide removal start time data, pesticide removal end time data, and bee quantity data of the i-th type of crop samples; , , , , They are The pollination crop sample matrix is ​​represented by The location component of the pre-pollination pesticide application data, post-pollination pesticide application data, pesticide removal start time data, pesticide removal end time data, and bee quantity data of the i-th type of crop samples;

[0129] S313: Setting the pollination crop sample matrix The actual yield of each type of crop sample in the pollination is obtained to obtain the actual yield value set of pollinated crops. ; Represents the pollination crop sample matrix The actual yield value of the i-th type of crop sample;

[0130] According to the actual yield value set of pollinated crops Set the fitness function of the second wild horse population ;as follows:

[0131] ;

[0132] S314, start the iteration operation; before the iteration, set the second current iteration number Set to 1; in each iteration, according to the fitness function of the second wild horse population Calculate the fitness value of each wild horse in the second wild horse population to obtain a second fitness value set; take the position of the wild horse individual corresponding to the maximum fitness value in the second fitness value set as the second global optimal position; and determine the second global optimal position and the second mating probability. The positions of the ordinary wild horses and the leader in the second wild horse population are updated; after the position update is completed, the second global best position and the corresponding fitness value are recorded and the second current iteration number is recorded. Add 1 and enter the next iteration;

[0133] S315, when When , the iteration is stopped, and the second final global optimal position is obtained, and the second final global optimal position is used as the optimal data matrix of the yield influencing factors Otherwise, continue iterating until ; Optimal data matrix of factors affecting production as follows:

[0134] ;

[0135] in, , , , , Respectively represent the pollination crop sample matrix The optimal pre-pollination pesticide application data, the optimal post-pollination pesticide application data, the optimal pesticide removal start time data, the optimal pesticide removal end time data and the optimal bee quantity data corresponding to the i-th type of crop samples;

[0136] The optimal data matrix of the yield influencing factors Application to pollination crop sample matrix Each type of crop sample in the , to obtain the actual optimal yield data set of pollinated crops ; At the same time, we get the data matrix set of production influencing factors And the corresponding pollination crop yield data matrix ;as follows:

[0137] ;

[0138] in, Represents the sample matrix of the pollinated crops The yield data of pollinated crops corresponding to the j-th group of yield influencing factor data obtained by optimizing the yield influencing factor data of the i-th type of crop samples;

[0139] S32, the pollination crop actual optimal yield data set With the final output data set Compare and calculate to get the optimal yield growth data set , Represents the pollinated crop sample matrix The actual optimal yield data of the i-th crop sample is compared with the non-pollination crop sample matrix The growth rate of the final yield data of the i-th type of crop samples; the calculation formula is as follows,

[0140] ;

[0141] S33, the optimal yield growth degree data set and the actual optimal yield dataset of pollinated crops The corresponding data matrix of production influencing factors is visualized;

[0142] S4, using the yield influencing factors data matrix set and the pollinated crop yield data matrix to construct the final KNN classification model;

[0143] The S4 comprises the following steps:

[0144] S41, constructing an initial KNN classification model and setting a training data ratio; and classifying a data matrix set of production influencing factors according to the training data ratio. and the pollinated crop yield data matrix Performing data division to obtain a yield influencing factor training data matrix set, a yield influencing factor test data matrix set, a pollination crop yield training data matrix, and a pollination crop yield test data matrix;

[0145] S42, setting a training error threshold; inputting the yield influencing factor training data matrix set and the pollination crop yield training data matrix into the initial KNN classification model for training; during the training process, when the training error is less than the training error threshold, stopping the training to obtain a trained KNN classification model; otherwise, continuing the training until the training error is less than the training error threshold;

[0146] S43, setting a test accuracy threshold; inputting the yield influencing factor test data matrix set and the pollination crop yield test data matrix into the trained KNN classification model for testing; after the test is completed, obtaining the test accuracy; when the test accuracy is greater than or equal to the test accuracy threshold, using the trained KNN classification model as the final KNN classification model; otherwise, returning to S42 to continue training until the test accuracy is greater than or equal to the test accuracy threshold, and obtaining the final KNN classification model;

[0147] S5, setting a sample set of pollinated crops to be analyzed and a corresponding data set of yield influencing factors to be analyzed, using the final KNN classification model to classify the data set of yield influencing factors to be analyzed corresponding to the sample set of pollinated crops to be analyzed, and obtaining a classification result; and determining whether the data set of yield influencing factors to be analyzed needs to be adjusted according to the classification result;

[0148] The S5 comprises the following steps:

[0149] S51. Setting the pollination crop sample set to be analyzed and the corresponding yield influencing factor data set to be analyzed , , , , , Respectively represent the pre-pollination pesticide application amount data, post-pollination pesticide application amount data, pesticide application removal start time data, pesticide application removal end time data and bee quantity data set for the pollination crop sample set to be analyzed;

[0150] S52, setting an expected yield threshold; and Input into the final KNN classification model for classification, and obtain the yield data of the pollinated crop samples to be analyzed;

[0151] When the yield data of the pollinated crop sample to be analyzed is greater than or equal to the expected yield threshold, there is no need to analyze the yield influencing factor dataset. When the sample yield data of the pollinated crop to be analyzed is less than the expected yield threshold, the actual optimal yield data set of the pollinated crop is adjusted. Dataset of factors affecting yield to be analyzed Adjustments are made to obtain the adjusted data set of production influencing factors to be analyzed.

[0152] Embodiment 2

[0153] This embodiment discloses a system for analyzing crop yields based on bee pollination based on big data. The system can implement the method of the above embodiment, including a crop sample collection module, a non-pollination crop pesticide application optimization module, a yield influencing factor construction module, a yield influencing factor data optimization module, a comparison module, a visualization display module, a classification model construction module, a classification module and a determination module;

[0154] The crop sample collection module is used to collect samples of multiple types of crops to obtain a pollinated crop sample matrix and a non-pollinated crop sample matrix;

[0155] The non-pollination crop pesticide application rate optimization module is used to optimize the pesticide application rate of each crop sample in the non-pollination crop sample matrix to obtain a final yield data set;

[0156] The yield influencing factor construction module is used to construct a plurality of yield influencing factors corresponding to the pollination crop sample matrix to obtain a yield influencing factor set;

[0157] The yield influencing factor data optimization module is used to optimize the various yield influencing factor data corresponding to the pollination crop sample matrix according to the yield influencing factor set, to obtain a yield influencing factor data matrix set, an actual optimal yield data set of pollination crops, and a pollination crop yield data matrix;

[0158] The comparison module is used to compare the actual optimal yield data set of the pollinated crops with the final yield data set to obtain the optimal yield growth degree data set;

[0159] The visualization display module is used to visualize the optimal yield growth degree data set and the corresponding yield influencing factor data matrix;

[0160] The classification model building module is used to construct a final KNN classification model using a yield influencing factor data matrix set and a pollinated crop yield data matrix;

[0161] The classification module is used to classify the yield influencing factor data set to be analyzed corresponding to the pollinated crop sample set to be analyzed using the final KNN classification model to obtain a classification result;

[0162] The determination module is used to determine whether the dataset of yield influencing factors to be analyzed needs to be adjusted according to the classification result.

[0163] In the description of this specification, the description with reference to the terms "one embodiment", "example", "specific example", etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0164] The preferred embodiments of the invention disclosed above are only used to help explain the invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to the specific implementation methods described. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the invention, so that those skilled in the art can understand and use the invention well.

Claims

1. A method for analyzing crop yields based on bee pollination based on big data, characterized in that: The following steps are involved: S1. Collecting samples of multiple types of crops to obtain a pollinated crop sample matrix and a non-pollinated crop sample matrix; S2, optimizing the pesticide application rate of each crop sample in the non-pollination crop sample matrix to obtain a final yield data set; constructing a plurality of yield influencing factors corresponding to the pollination crop sample matrix to obtain a yield influencing factor set; S3, optimizing the various yield influencing factor data corresponding to the pollination crop sample matrix according to the yield influencing factor set, to obtain a yield influencing factor data matrix set, an actual optimal yield data set of pollination crops, and a pollination crop yield data matrix; Comparing the actual optimal yield data set of the pollinated crops with the final yield data set to obtain an optimal yield growth degree data set; visually displaying the optimal yield growth degree data set and the corresponding yield influencing factor data matrix; S4, using the yield influencing factors data matrix set and the pollinated crop yield data matrix to construct the final KNN classification model; S5. Set a sample set of pollinated crops to be analyzed and a corresponding data set of yield influencing factors to be analyzed, and use the final KNN classification model to classify the data set of yield influencing factors to be analyzed corresponding to the sample set of pollinated crops to be analyzed to obtain a classification result; and determine whether the data set of yield influencing factors to be analyzed needs to be adjusted according to the classification result.

2. The method for analyzing crop yield based on bee pollination based on big data according to claim 1, characterized in that: The S1 comprises the following steps: S11, setting a crop type set; collecting a plurality of crop samples according to the crop type set to obtain a crop sample matrix; equally dividing the crop sample matrix to obtain a pollinated crop sample matrix and a non-pollinated crop sample matrix; S12. Set a first planting area set and a second planting area set; plant each crop sample in the pollination crop sample matrix in each planting area of ​​the first planting area set according to its type, and then plant each crop sample in the non-pollination crop sample matrix in each planting area of ​​the second planting area set according to its type; the planting environment conditions of the pollination crop sample matrix and the non-pollination crop sample matrix are the same.

3. The method for analyzing crop yield based on bee pollination based on big data according to claim 2, characterized in that: The S2 comprises the following steps: S21, obtaining the optimal pesticide application amount and the corresponding optimal yield of each type of crop sample in the non-pollination crop sample matrix, and obtaining a final yield data set; S22. Setting a set of yield influencing factors for the pollination crop sample matrix.

4. The method for analyzing crop yield based on bee pollination based on big data according to claim 3, characterized in that: The S21 comprises the following steps: S211, constructing a first wild horse population; setting the maximum number of iterations of the first wild horse population to , recorded as the first maximum number of iterations, the current number of iterations is , recorded as the first current iteration number; the mating probability of the first wild horse population is recorded as the first mating probability; S212, setting the initial position of each wild horse in the first wild horse population and the initial position of the leader in the first wild horse population to obtain a first initial position matrix and a second initial position matrix; S213, setting the actual yield of each type of crop sample in the non-pollination crop sample matrix to obtain a non-pollination crop actual yield value set; setting a fitness function of the first wild horse population according to the non-pollination crop actual yield value set; S214, start an iterative operation; in each round of iteration, calculate the fitness value of each wild horse in the first wild horse population according to the fitness function of the first wild horse population to obtain a first fitness value set; take the position of the wild horse individual corresponding to the maximum fitness value in the first fitness value set as the first global optimal position; update the position of the common wild horses and the leader position in the first wild horse population according to the first global optimal position and the first mating probability; after the position update is completed, enter the next round of iteration; S215, when When , the iteration is stopped, and the first final global optimal position is obtained, and the first final global optimal position is used as the final pesticide dosage data set; otherwise, the iteration is continued until ; Apply the final pesticide dosage data set to each type of crop sample in the non-pollination crop sample matrix to obtain a final yield data set.

5. The method for analyzing crop yield based on bee pollination based on big data according to claim 4, characterized in that: The S3 comprises the following steps: S31, optimizing the yield data of the pollinated crop sample matrix according to the yield influencing factor set, to obtain a yield influencing factor data matrix set and a corresponding pollinated crop actual optimal yield data set; S32, comparing and calculating the actual optimal yield data set of the pollinated crops with the final yield data set to obtain an optimal yield growth degree data set; S33, visually displaying the yield influencing factor data matrix corresponding to the optimal yield growth degree data set and the actual optimal yield data set of pollinated crops.

6. The method for analyzing crop yield based on bee pollination based on big data according to claim 5, characterized in that: The S31 comprises the following steps: S311, construct a second wild horse population; set the maximum number of iterations of the second wild horse population to , recorded as the second maximum number of iterations, the current number of iterations is , recorded as the second current iteration number; the mating probability of the second wild horse population is recorded as the second mating probability; S312, setting the initial position of each wild horse in the second wild horse population and the initial position of the leader in the second wild horse population according to the set of production influencing factors, to obtain a first initial position matrix set and a second initial position matrix set; S313, setting the actual yield of each type of crop sample in the pollination crop sample matrix to obtain a pollination crop actual yield value set; setting the fitness function of the second wild horse population according to the pollination crop actual yield value set; S314, start iterative operation; in each round of iteration, calculate the fitness value of each wild horse in the second wild horse population according to the fitness function of the second wild horse population, and obtain a second fitness value set; take the position of the wild horse individual corresponding to the maximum fitness value in the second fitness value set as the second global optimal position; update the position of the common wild horses and the leader position in the second wild horse population according to the second global optimal position and the second mating probability; after the position update is completed, record the second global optimal position and the corresponding fitness value, and enter the next round of iteration; S315, when When , the iteration is stopped, and the second final global optimal position is obtained, and the second final global optimal position is used as the optimal data matrix of the yield influencing factors; otherwise, the iteration is continued until ; The optimal data matrix of yield influencing factors is applied to each type of crop sample in the pollinated crop sample matrix to obtain the actual optimal yield data set of pollinated crops; at the same time, a yield influencing factor data matrix set and a corresponding pollinated crop yield data matrix are obtained.

7. The method for analyzing crop yield based on bee pollination based on big data according to claim 6, characterized in that: The S4 comprises the following steps: S41, constructing an initial KNN classification model and setting a training data ratio; dividing the yield influencing factor data matrix set and the pollination crop yield data matrix according to the training data ratio to obtain a yield influencing factor training data matrix set, a yield influencing factor test data matrix set, a pollination crop yield training data matrix, and a pollination crop yield test data matrix; S42, setting a training error threshold; inputting the yield influencing factor training data matrix set and the pollination crop yield training data matrix into the initial KNN classification model for training; during the training process, when the training error is less than the training error threshold, stopping the training to obtain a trained KNN classification model; otherwise, continuing the training until the training error is less than the training error threshold; S43, setting a test accuracy threshold; inputting the yield influencing factor test data matrix set and the pollination crop yield test data matrix into the trained KNN classification model for testing; after the test is completed, obtaining the test accuracy; when the test accuracy is greater than or equal to the test accuracy threshold, using the trained KNN classification model as the final KNN classification model; otherwise, returning to S42 to continue training until the test accuracy is greater than or equal to the test accuracy threshold, and obtaining the final KNN classification model.

8. The method for analyzing crop yield based on bee pollination based on big data according to claim 7, characterized in that: The S5 comprises the following steps: S51, setting a sample set of pollinated crops to be analyzed and a corresponding data set of yield influencing factors to be analyzed; S52, setting an expected yield threshold; inputting the yield influencing factor data set to be analyzed into the final KNN classification model for classification, and obtaining the yield data of the pollinated crop samples to be analyzed; When the yield data of the pollinated crop samples to be analyzed is greater than or equal to the expected yield threshold, there is no need to adjust the yield influencing factor dataset to be analyzed; when the yield data of the pollinated crop samples to be analyzed is less than the expected yield threshold, the yield influencing factor dataset to be analyzed is adjusted according to the actual optimal yield dataset of the pollinated crops to obtain an adjusted yield influencing factor dataset to be analyzed.

9. A system for implementing the method for analyzing crop yield through bee pollination based on big data as described in any one of claims 1 to 8.

Images