Agricultural fertilization and irrigation decision support method and system based on privacy intersection technology

Abstract

The invention belongs to the field of agricultural information decision, and provides an agricultural fertilization and irrigation decision support method and system based on a privacy intersection technology, and the method comprises the steps: receiving agricultural original data transmitted by a farmer terminal, an agricultural scientific research institution terminal and an agricultural material supplier terminal; performing encryption processing on the agricultural original data, and storing the agricultural encrypted data in a distributed database; after receiving an agricultural fertilization and irrigation decision request, determining agricultural key data matched with the agricultural fertilization and irrigation decision request based on a privacy intersection technology and according to agricultural encrypted data stored in a distributed database; and establishing an agricultural fertilization and irrigation decision model according to the agricultural key data, and sending an agricultural fertilization and irrigation decision result output by the agricultural fertilization and irrigation decision model to a farmer terminal. The security sharing of data of all parties is realized through encryption and privacy intersection technologies, and the data interaction security and decision reliability are improved in combination with an agricultural fertilization and irrigation decision model.

Description

Technical Field

[0001] The present invention relates to the technical field of agricultural information decision-making, and particularly to an agricultural fertilization and irrigation decision support method and system based on privacy intersection technology. Background Art

[0002] In modern agricultural production, precise fertilization and irrigation are crucial for increasing crop yields, reducing resource waste, and environmental pollution. However, achieving precise decision-making faces severe data privacy issues.

[0003] On the one hand, farmers have detailed information about their land, such as area, location, historical planting data, etc., which involve personal privacy and business secrets. On the other hand, agricultural research institutions possess a large amount of crop growth models, soil fertility research results, and data on the effects of different fertilizers and irrigation methods on crops obtained through experiments. At the same time, agricultural input suppliers know the composition, characteristics, and price information of fertilizers and water resources.

[0004] In the traditional process of agricultural data integration and analysis, if precise fertilization and irrigation decisions are to be achieved, all parties need to share data, but this often poses a risk of privacy leakage. For example, farmers are worried that their land information and planting habits will be obtained by competitors or unscrupulous merchants, and the research results of research institutions may be illegally used, and the product information of agricultural input suppliers may also lose their commercial advantages during sharing.

[0005] For the above reasons, traditional agricultural fertilization and irrigation decision-making schemes have problems of low data interaction security and low decision reliability. Summary of the Invention

[0006] The present invention provides an agricultural fertilization and irrigation decision support method and system based on privacy intersection technology to solve the defects of low data interaction security and low decision reliability in traditional agricultural fertilization and irrigation decision-making schemes.

[0007] On the one hand, the present invention provides an agricultural fertilization and irrigation decision support method based on privacy intersection technology. The method is executed by an agricultural data management server, and the agricultural data management server is respectively connected to a distributed database, a farmer terminal, an agricultural research institution terminal, and an agricultural input supplier terminal. The method includes: Receiving agricultural raw data sent by the farmer terminal, the agricultural research institution terminal, and the agricultural input supplier terminal; Performing encryption processing on the agricultural raw data, and storing the encrypted agricultural data obtained by the encryption processing in the distributed database; After receiving the agricultural fertilization and irrigation decision request initiated by the farmer terminal, based on the private set intersection technology and according to the agricultural encrypted data stored in the distributed database, determine the agricultural key data that matches the agricultural fertilization and irrigation decision request; Establish an agricultural fertilization and irrigation decision model based on the agricultural key data, and send the agricultural fertilization and irrigation decision result output by the agricultural fertilization and irrigation decision model to the farmer terminal.

[0008] According to the agricultural fertilization and irrigation decision support method based on the private set intersection technology provided by the present invention, the agricultural original data is encrypted, including: Assign corresponding data identifiers to each data element in the agricultural original data; Bind and encrypt each data element with the corresponding data identifier to obtain agricultural encrypted data.

[0009] According to the agricultural fertilization and irrigation decision support method based on the private set intersection technology provided by the present invention, assigning corresponding data identifiers to each data element in the agricultural original data includes: According to the data source, divide the agricultural original data into farmer data, scientific research institution data, and agricultural materials supplier data; Extract the land number and crop variety of each data element in the farmer data, and assign corresponding data identifiers to each data element in the farmer data according to the land number and crop variety; Extract the research project number and experimental conditions of each data element in the scientific research institution data, and assign corresponding data identifiers to the scientific research institution data according to the research project number and experimental conditions; Extract the product number and applicable scope of each data element in the agricultural materials supplier data, and assign corresponding data identifiers to the agricultural materials supplier data according to the product number and applicable scope.

[0010] According to the agricultural fertilization and irrigation decision support method based on the private set intersection technology provided by the present invention, based on the private set intersection technology and according to the agricultural encrypted data stored in the distributed database, determining the agricultural key data that matches the agricultural fertilization and irrigation decision request includes: Based on the agricultural fertilization and irrigation decision request, retrieve the target agricultural encrypted data in the distributed database; According to the data source, distribute the data elements in the target agricultural encrypted data to different computing nodes; Perform private set intersection calculation on the data elements on all the computing nodes to obtain the agricultural key data that matches the agricultural fertilization and irrigation decision request.

[0011] According to the agricultural fertilization and irrigation decision-making support method based on the private set intersection technology provided by the present invention, perform private set intersection calculations on the data elements on all the computing nodes to obtain agricultural key data that matches the agricultural fertilization and irrigation decision request, including: Based on the data identifiers of each data element, respectively obtain the intersection of the data elements between the computing nodes with the data source being farmers and other computing nodes, and obtain the intersection result data set; Comprehensively screen the data elements in the intersection result data set to obtain agricultural key data that matches the agricultural fertilization and irrigation decision request.

[0012] According to the agricultural fertilization and irrigation decision-making support method based on the private set intersection technology provided by the present invention, based on the data identifiers of each data element, respectively obtain the intersection of the data elements between the computing nodes corresponding to the farmer data and other computing nodes, and obtain the intersection result data set, including: Based on the data identifiers of each data element, respectively determine the correlation degrees between the data elements in the computing nodes corresponding to the farmer data and the data elements in other computing nodes; Extract the target data elements with the correlation degrees higher than the preset correlation degree threshold to obtain the intersection result data set.

[0013] According to the agricultural fertilization and irrigation decision-making support method based on the private set intersection technology provided by the present invention, comprehensively screen the data elements in the intersection result data set to obtain agricultural key data that matches the agricultural fertilization and irrigation decision request, including: Calculate the data integrity of the intersection result data set, and determine whether the data integrity is higher than the preset integrity threshold to obtain the integrity judgment result; If the integrity judgment result is yes, extract the redundant elements and invalid elements in the intersection result data set; Remove the redundant elements and invalid elements in the intersection result data set to obtain agricultural key data that matches the agricultural fertilization and irrigation decision request.

[0014] According to the agricultural fertilization and irrigation decision-making support method based on the private set intersection technology provided by the present invention, establish an agricultural fertilization and irrigation decision model based on the agricultural key data, including: Extract the key feature data in the agricultural key data; Perform standardization and normalization processing on the key feature data to obtain intermediate feature data; Perform validity screening on the intermediate feature data to obtain sample data; Train a pre-constructed deep learning network model based on the sample data to obtain an agricultural fertilization and irrigation decision model.

[0015] On the other hand, the present invention also provides an agricultural fertilization and irrigation decision support system based on private set intersection technology, including: an agricultural data management server, a distributed database, a farmer terminal, an agricultural research institution terminal, and an agricultural supplies supplier terminal; The agricultural data management server is respectively connected to the distributed database, the farmer terminal, the agricultural research institution terminal, and the agricultural supplies supplier terminal; The agricultural data management server is used to receive the agricultural raw data sent by the farmer terminal, the agricultural research institution terminal, and the agricultural supplies supplier terminal, perform encryption processing on the agricultural raw data, and store the agricultural encrypted data obtained by the encryption processing in the distributed database; The agricultural data management server is further used to receive the agricultural fertilization and irrigation decision request initiated by the farmer terminal, determine the agricultural key data matching the agricultural fertilization and irrigation decision request based on the private set intersection technology and according to the agricultural encrypted data stored in the distributed database, establish an agricultural fertilization and irrigation decision model based on the agricultural key data, and send the agricultural fertilization and irrigation decision result output by the agricultural fertilization and irrigation decision model to the farmer terminal.

[0016] According to the agricultural fertilization and irrigation decision support system based on the private set intersection technology provided by the present invention, the agricultural data management server includes: A data encryption module, which is used to receive the agricultural raw data sent by the farmer terminal, the agricultural research institution terminal, and the agricultural supplies supplier terminal, assign a corresponding data identifier to each data element in the agricultural raw data, bind and encrypt each data element with the corresponding data identifier to obtain agricultural encrypted data, and store the agricultural encrypted data in the distributed database; A private set intersection calculation module, which is used to receive the agricultural fertilization and irrigation decision request initiated by the farmer terminal, retrieve the target agricultural encrypted data in the distributed database based on the agricultural fertilization and irrigation decision request, distribute the data elements in the target agricultural encrypted data to different calculation nodes, and perform private set intersection calculation on the data elements on all the calculation nodes to obtain the agricultural key data matching the agricultural fertilization and irrigation decision request; A model construction module, which is used to extract the key feature data from the agricultural key data, perform standardization and normalization processing on the key feature data to obtain intermediate feature data, perform validity screening on the intermediate feature data to obtain sample data, and train a pre-constructed deep learning network model based on the sample data to obtain an agricultural fertilization and irrigation decision model; A result output module, configured to send the agricultural fertilization and irrigation decision result output by the agricultural fertilization and irrigation decision model to the farmer terminal.

[0017] The agricultural fertilization and irrigation decision support method and system based on the private set intersection technology provided by the present invention encrypt the agricultural original data sent by the farmer terminal, the agricultural research institution terminal, and the agricultural materials supplier terminal, and store the encrypted agricultural data obtained by the encryption process in a distributed database. After receiving the agricultural fertilization and irrigation decision request initiated by the farmer terminal, based on the private set intersection technology and according to the agricultural encrypted data stored in the distributed database, determine the agricultural key data matching the agricultural fertilization and irrigation decision request, establish an agricultural fertilization and irrigation decision model based on the agricultural key data, and send the agricultural fertilization and irrigation decision result output by the agricultural fertilization and irrigation decision model to the farmer terminal. Since the decision-making process realizes the secure sharing of data among all parties through encryption and private set intersection technology, and combined with the agricultural fertilization and irrigation decision model, it can efficiently and accurately provide agricultural fertilization and irrigation decision support services for farmers, improving the data interaction security and decision reliability in the agricultural fertilization and irrigation decision-making process. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0019] Figure 1 is a schematic flowchart of the agricultural fertilization and irrigation decision support method based on the private set intersection technology provided by the embodiment of the present invention; Figure 2 is a schematic structural diagram of the agricultural fertilization and irrigation decision support system based on the private set intersection technology provided by the embodiment of the present invention; Figure 3 is a schematic structural diagram of the agricultural data management server in the embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0020] To make the objectives, technical solutions, and advantages of the present invention clearer, the following will clearly and completely describe the technical solutions in the present invention in conjunction with the drawings in the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the protection scope of the present invention.

[0021] The following will be combined with Figures 1 to 3Describe the detailed solution of the agricultural fertilization and irrigation decision-making support method and system based on the private set intersection technology provided by the embodiments of the present invention.

[0022] Figure 1 It is a schematic flowchart of the agricultural fertilization and irrigation decision-making support method based on the private set intersection technology provided by the embodiments of the present invention.

[0023] As Figure 1 shown, for the agricultural fertilization and irrigation decision-making support method based on the private set intersection technology provided by the embodiments of the present invention, the execution entity can be an agricultural data management server, and the agricultural data management server is respectively connected to a distributed database, a farmer terminal, an agricultural research institution terminal, and an agricultural supplies supplier terminal; the above method mainly includes the following steps: Step 110: Receive the agricultural original data sent by the farmer terminal, the agricultural research institution terminal, and the agricultural supplies supplier terminal.

[0024] In this embodiment, the farmer terminal, the agricultural research institution terminal, and the agricultural supplies supplier terminal can all be mobile terminals or fixed terminals.

[0025] It can be understood that the agricultural original data refers to the general term for the original data related to agricultural production uploaded by farmers, agricultural research institutions, and agricultural supplies suppliers.

[0026] Step 120: Encrypt the agricultural original data and store the encrypted agricultural data obtained by the encryption process in the distributed database.

[0027] In this embodiment, a homomorphic encryption algorithm can be used to encrypt the agricultural original data. The homomorphic encryption algorithm allows specific computational operations to be performed on ciphertext without decrypting the data, thereby ensuring the privacy of the agricultural original data during storage and processing. For example, when calculating the correlation between soil fertility data and fertilizer component data, the encrypted data can be directly calculated to obtain the encrypted result, and the decryption operation is only performed when the result needs to be viewed finally. Even if the database is illegally accessed, the attacker cannot obtain the plaintext content of the data.

[0028] Step 130: After receiving the agricultural fertilization and irrigation decision-making request initiated by the farmer terminal, based on the private set intersection technology and according to the agricultural encrypted data stored in the distributed database, determine the agricultural key data that matches the agricultural fertilization and irrigation decision-making request.

[0029] It can be understood that the private set intersection technology specifically refers to a cryptographic protocol that allows two or more participating parties to calculate their intersection without revealing other information in their respective sets.

[0030] Step 140: Establish an agricultural fertilization and irrigation decision-making model based on key agricultural data, and send the agricultural fertilization and irrigation decision results output by the agricultural fertilization and irrigation decision-making model to the farmer terminal.

[0031] The solution provided in this embodiment can achieve data sharing among all parties without disclosing the privacy data of each party through the private intersection technology. At the same time, based on the agricultural fertilization and irrigation decision-making model, it can more efficiently and accurately provide decision support for agricultural fertilization and irrigation, improving the data interaction security and decision reliability in the decision-making process.

[0032] In one embodiment, the agricultural raw data is encrypted, which specifically includes: First, assign corresponding data identifiers to each data element in the agricultural raw data.

[0033] In a specific implementation, assigning corresponding data identifiers to each data element in the agricultural raw data specifically includes: In the first step, according to the data source, the agricultural raw data is divided into farmer data, scientific research institution data, and agricultural materials supplier data.

[0034] It can be understood that in this embodiment, the data sources of the agricultural raw data can be farmers, scientific research institutions, and agricultural materials suppliers. Therefore, the agricultural raw data can be divided into farmer data, scientific research institution data, and agricultural materials supplier data according to the data source.

[0035] In this embodiment, the farmer data can specifically include: land number, land location (such as the longitude and latitude information of the land), land area, soil type, crop variety, and planting history data.

[0036] Among them, the soil type can be obtained according to the soil type actively selected by the farmer terminal, or analyzed according to the soil test report uploaded by the farmer terminal. The planting history data specifically includes the yields of previously planted crops and the fertilization and irrigation conditions, etc.

[0037] In practical applications, the above-mentioned farmer data can be initially encrypted on the user terminal and then further encrypted after being uploaded to the agricultural data management server.

[0038] The agricultural data management server can also perform format verification and integrity check on the farmer data uploaded by the farmer terminal to ensure the accuracy and availability of the farmer data. For the farmer data that does not meet the requirements, the farmer terminal can be notified in time for correction or supplementation.

[0039] The data of scientific research institutions specifically includes: crop growth model data, soil fertility research data, and experimental data on the impact of different fertilizers and irrigation methods on crop growth, etc. Among them, the experimental data includes research project numbers and experimental conditions. Before uploading the data of scientific research institutions, the terminals of scientific research institutions can use an encryption algorithm compatible with the agricultural data management server to initially encrypt the data to ensure the security of the data of scientific research institutions.

[0040] After receiving the data of scientific research institutions, the agricultural data management server can classify and label the data of scientific research institutions for subsequent query and use. At the same time, the agricultural data management server can also evaluate the quality of the data of scientific research institutions and screen out high-quality data of scientific research institutions for labeling.

[0041] The data of agricultural input suppliers specifically includes: data such as product name, product number, product composition, composition content, scope of application, usage method, and product price. The data of agricultural input suppliers can also be initially encrypted before uploading. The data of agricultural input suppliers can also include product performance evaluation data and user feedback information to enrich the data resources.

[0042] The agricultural data management server can review and verify the data of agricultural input suppliers to ensure the authenticity and reliability of the data. For false or misleading data, the agricultural data management server has the right to reject it or require the terminal of the agricultural input supplier to make corrections.

[0043] In the second step, extract the land number and crop variety of each data element in the farmer data, and assign corresponding data identifiers to each data element in the farmer data according to the land number and crop variety.

[0044] In the third step, extract the research project number and experimental conditions of each data element in the data of scientific research institutions, and assign corresponding data identifiers to the data of scientific research institutions according to the research project number and experimental conditions.

[0045] In the fourth step, extract the product number and scope of application of each data element in the data of agricultural input suppliers, and assign corresponding data identifiers to the data of agricultural input suppliers according to the product number and scope of application.

[0046] It can be understood that in this embodiment, the data identifier is a fixed-length string of numbers. During the data identifier assignment process, the data on which it is based can be digitally transformed, and all or part of the numbers can be extracted as the basis for determining the data identifier.

[0047] Exemplarily, taking the data identifier of a data element in the farmer data as an example, the first three digits in the land number can be extracted, and then different crop varieties can be numbered by three digits, so as to obtain a data identifier composed of six digits.

[0048] Then, each data element is bound and encrypted with the corresponding data identifier to obtain agricultural encrypted data.

[0049] In one embodiment, based on the private set intersection technology and according to the agricultural encrypted data stored in the distributed database, the agricultural key data matching the agricultural fertilization and irrigation decision request is determined, which specifically includes: In the first step, based on the agricultural fertilization and irrigation decision request, the target agricultural encrypted data in the distributed database is retrieved.

[0050] It can be understood that the target agricultural encrypted data refers to the agricultural encrypted data related to the agricultural fertilization and irrigation decision request. In practical applications, the request content in the agricultural fertilization and irrigation decision request can be extracted, and the request content specifically includes key information such as the land location to be decided, crop variety, and soil type. Then, at least part of the key information in the request content can be used for data indexing in the distributed database, and thus the target agricultural encrypted data associated with the above key information can be queried.

[0051] In the second step, according to the data source, each data element in the target agricultural encrypted data is allocated to different computing nodes.

[0052] In this embodiment, at least three types of computing nodes can be set according to the data source, namely the computing nodes corresponding to farmers, the computing nodes corresponding to agricultural research institutions, and the computing nodes corresponding to agricultural input suppliers. Subsequently, each data element in the agricultural encrypted data can be allocated to the corresponding type of computing node according to its data identifier.

[0053] In the third step, the private set intersection calculation is performed on the data elements on all the computing nodes to obtain the agricultural key data matching the agricultural fertilization and irrigation decision request.

[0054] In a specific implementation, performing the private set intersection calculation on the data elements on all the computing nodes to obtain the agricultural key data matching the agricultural fertilization and irrigation decision request specifically includes: First, based on the data identifier of each data element, the intersection of the data elements between the computing nodes with the data source of farmers and other computing nodes is obtained respectively to obtain the intersection result data set.

[0055] It can be understood that the intersection obtaining process in this embodiment can be a correlation matching process between data elements.

[0056] Exemplarily, based on the data identifier of each data element, the intersection of the data elements between the computing nodes corresponding to the farmer data and other computing nodes is obtained respectively to obtain the intersection result data set, which specifically includes: First, based on the data identifiers of each data element, determine the association degrees between the data elements in the computing node corresponding to the farmer data and the data elements in other computing nodes respectively.

[0057] In practical applications, an association comparison table can be established in advance between the data identifiers of farmer data, the data identifiers of scientific research institution data, and the data identifiers of agricultural input suppliers. For example, an association comparison table can be established between the crop varieties in the data identifiers of farmer data and the experimental conditions in the data identifiers of scientific research institution data, as well as the applicable ranges in the data identifiers of agricultural input suppliers. The association comparison table contains the corresponding relationships of each group of data identifiers and the values of the association degrees.

[0058] Subsequently, the data identifiers corresponding to the data elements in the computing node corresponding to the farmer data and the data elements in other computing nodes can be extracted respectively, and then the association degrees between the data elements in the computing node corresponding to the current farmer data and the data elements in other computing nodes can be directly queried according to the association comparison table.

[0059] Then, extract the target data elements with association degrees higher than the preset association degree threshold to obtain the intersection result data set.

[0060] In this embodiment, by aggregating the target data elements with association degrees higher than the preset association degree threshold together, the intersection result data set can be obtained.

[0061] Finally, comprehensively screen the data elements in the intersection result data set to obtain the agricultural key data that matches the agricultural fertilization and irrigation decision request.

[0062] In a specific implementation, comprehensively screening the data elements in the intersection result data set to obtain the agricultural key data that matches the agricultural fertilization and irrigation decision request specifically includes: In the first step, calculate the data integrity of the intersection result data set, and determine whether the data integrity is higher than the preset integrity threshold to obtain the integrity judgment result.

[0063] In this embodiment, the actual data volume of the intersection result data set can be obtained respectively, and the total data volume on all computing nodes can be obtained. Multiply the total data volume on all nodes by the preset allowable deviation coefficient to obtain the expected data volume, and divide the actual data volume by the expected data volume to obtain the data integrity of the intersection result data set.

[0064] It can be understood that the values of the preset allowable deviation coefficient and the data integrity are both within the data range of 0 to 1. The preset integrity threshold can be reasonably set according to the actual accuracy requirements.

[0065] In this embodiment, by first judging the data integrity of the intersection result dataset, it can be ensured that the amount of data for subsequent screening is sufficient, thus guaranteeing the effectiveness of the screening process.

[0066] In the second step, if the integrity judgment result is yes, redundant elements and invalid elements in the intersection result dataset are extracted.

[0067] In this embodiment, redundant elements refer to duplicate data elements, and invalid elements refer to data elements with obvious deviations.

[0068] In the third step, redundant elements and invalid elements in the intersection result dataset are removed to obtain the agricultural key data that matches the agricultural fertilization and irrigation decision request.

[0069] In this embodiment, by removing redundant elements and invalid elements in the intersection result dataset, the data quality of subsequent agricultural key data can be improved, thereby providing accurate and effective data basis for the establishment of the agricultural fertilization and irrigation decision model, and further improving the model accuracy of the agricultural fertilization and irrigation decision model.

[0070] In one embodiment, an agricultural fertilization and irrigation decision model is established based on the agricultural key data, which specifically includes: In the first step, key feature data in the agricultural key data is extracted.

[0071] In practical applications, in the key feature data extraction link, not only the agricultural key data can be relied on, but also other auxiliary data can be introduced, such as meteorological data and geographic information data, etc., so as to capture more comprehensive feature information.

[0072] In this embodiment, the key feature data specifically includes soil fertility indicators (such as nitrogen content, phosphorus content, potassium content, and organic matter content, etc.), crop growth stage characteristics (such as growth days, plant height, and leaf area index, etc.), climate characteristics (such as temperature, humidity, rainfall, sunshine duration, etc.), and agricultural input product characteristics (such as fertilizer composition, irrigation water volume, and irrigation frequency, etc.).

[0073] In the second step, the key feature data is standardized and normalized to obtain intermediate feature data.

[0074] It can be understood that by standardizing and normalizing the key feature data, comparability can be achieved between different types of feature data, facilitating subsequent model training and analysis operations.

[0075] In practical applications, for the standardization process of key feature data, Z-score standardization can be adopted. Z-score standardization calculates the standard deviation and mean of the original data and converts the original data into standardized Z-score values. Specifically, the Z-score value reflects the deviation degree between the original data and the mean, and is measured in units of the standard deviation.

[0076] In this embodiment, the formula for Z-score standardization is: (1) where is the observation value of any type of key feature data, is the mean of any type of key feature data, is the standard deviation of any type of key feature data.

[0077] Through the above formula, the associated feature data of different magnitudes can be converted into Z-score values with a unified measurement for comparison, improving the comparability of different types of key feature data.

[0078] The third step is to perform validity screening on the intermediate feature data to obtain sample data.

[0079] After obtaining the intermediate feature data, a screening value range can be set, and each intermediate feature data is compared with this screening value range. The data within this screening value range can be determined as valid, and the data outside this screening value range can be determined as invalid, thereby realizing the validity screening of the intermediate feature data.

[0080] The fourth step is to train the pre-constructed deep learning network model based on the sample data to obtain an agricultural fertilization and irrigation decision-making model.

[0081] In the model architecture selection step, a machine learning deep learning network model suitable for the characteristics of agricultural data, such as a deep learning network model with a neural network architecture, can be selected to construct an agricultural fertilization and irrigation decision-making model. Subsequently, the parameters and structure of the model can be reasonably adjusted according to the scale and complexity of the data to improve the performance and generalization ability of the model.

[0082] The preprocessed sample data is divided into a training set and a test set. The deep learning network model is trained using the training set, and by continuously adjusting the parameters and weights of the model, the prediction results of the model are made as close as possible to the actual situation. Then, the trained agricultural fertilization and irrigation decision-making model is evaluated using the test set, and indicators such as the accuracy rate, recall rate, and F1 value of the model are calculated to determine whether the performance of the agricultural fertilization and irrigation decision-making model meets the requirements.

[0083] In the model architecture building phase, the appropriate number of neural network layers and the number of nodes in each layer can be selected according to the complexity of the data and the difficulty of the problem. For relatively simple agricultural decision-making problems, a neural network with 3 to 5 layers can be adopted, including an input layer, 1 to 3 hidden layers, and an output layer. The number of nodes in the input layer is equal to the number of features. For example, if there are 10 features after feature engineering, the number of nodes in the input layer is 10. The number of nodes in the hidden layer can be determined by an empirical formula or through experiments. In some embodiments, the empirical formula can be expressed as follows: (2) Wherein, is the number of nodes in the hidden layer, is the number of nodes in the input layer, is the number of nodes in the output layer.

[0084] In practical applications, a rough range can also be set first, such as 10 - 50 nodes, and by experimenting with different combinations of the number of nodes and observing the performance of the model on the validation set, the optimal number of nodes can be determined.

[0085] In the hidden layer, the ReLU (Rectified Linear Unit) activation function is usually selected. The ReLU function is simple to calculate, can effectively alleviate the vanishing gradient problem, and accelerate the training speed of the neural network.

[0086] For the output layer, the appropriate activation function can be selected according to the type of output data. If the output data is numerical data such as predicting continuous fertilization amount or irrigation amount, no activation function or a linear activation function can be used; if it is for classification decision-making (such as whether fertilization is required, which fertilizer to choose, etc.), the Sigmoid activation function or the Softmax activation function can be selected. For example, if it is a binary classification problem (such as whether irrigation is required), the Sigmoid activation function can be adopted; if it is a multi-classification problem (such as choosing different types of fertilizers), the Softmax activation function can be adopted, which can convert the output data into a probability distribution over each class.

[0087] In the model training phase, first, the weights and biases of the deep learning network model need to be initialized. The initialization methods can be random initialization and Xavier initialization, etc. Random initialization is to randomly generate the values of weights and biases within a certain range (such as uniform distribution or normal distribution). Xavier initialization determines the appropriate initialization range according to the number of input and output nodes, so that the weights have appropriate variances during initialization, which helps to accelerate the training speed and improve the stability of the model. For example, for the weight W, if Xavier initialization is adopted, its initialization formula is: (3) In the formula,n in and n out are the number of input nodes and output nodes of the current layer, respectively.

[0088] After that, an appropriate loss function can be selected according to the nature of the problem. If it is a regression problem (such as predicting the amount of fertilizer or irrigation water), the mean squared error (MSE) loss function is usually adopted. If it is a classification problem (such as determining whether fertilization is needed or selecting the type of fertilizer), the cross-entropy loss function can be used.

[0089] In the training loop phase, the preprocessed sample data can be input into the deep learning network model in batches. For example, set the batch size to 32, that is, take 32 sample data as a group to perform one forward propagation and backpropagation calculation.

[0090] During the training process of each batch, first perform forward propagation calculation, pass the input data through each layer of the neural network to obtain the output data, and then calculate the loss value according to the defined loss function.

[0091] Then perform backpropagation calculation. By calculating the gradients of the loss function with respect to the weights and biases, use optimization algorithms (such as Stochastic Gradient Descent SGD, Adagrad, Adadelta, Adam, etc.) to update the weights and biases.

[0092] During the process of updating the weights and biases, the selection of the learning rate is also crucial. If the learning rate is too large, the model may not converge or even diverge; if the learning rate is too small, the training speed will be very slow. Usually, a learning rate decay strategy can be adopted, gradually reducing the learning rate as the training progresses. For example, after a certain number of training steps or when the validation set loss no longer decreases, multiply the learning rate by a decay factor less than 1 (such as taking a value of 0.9).

[0093] Repeat the above process until the preset number of training epochs is reached or the loss value on the validation set no longer decreases, indicating that the model has converged or is close to the optimal solution.

[0094] After that, evaluate the trained agricultural fertilization and irrigation decision-making model on the validation set. During the training process, the validation set data can be regularly input into the trained agricultural fertilization and irrigation decision-making model to calculate the loss value and evaluation metrics on the validation set (such as accuracy, recall rate, F1 value, etc. If it is a regression problem, the root mean square error RMSE, etc. can be calculated). By observing the changes in the metrics on the validation set, determine whether the model has overfitting or underfitting phenomena. If the validation set loss continues to rise while the training set loss continues to fall, it indicates that the model may have overfitting and corresponding measures need to be taken for adjustment, such as increasing the data volume, adding regularization terms, or adjusting the model complexity, etc.

[0095] In the actual verification stage, the hyperparameters of the deep learning network model (such as learning rate, number of hidden layer nodes, number of layers, regularization parameters, etc.) can be adjusted and optimized. For example, grid search, random search, or more advanced hyperparameter optimization algorithms (such as genetic algorithms, simulated annealing algorithms, etc.) can be adopted.

[0096] For example, in grid search, for the number of hidden layer nodes, combined searches can be carried out among several values such as 10, 20, 30, etc. The model is trained under different hyperparameter combinations, and the optimal hyperparameter combination is selected according to the performance on the validation set.

[0097] In some embodiments, the agricultural fertilization and irrigation decision-making model can be optimized and updated according to the test results and the actual feedback of farmers. If the performance of the agricultural fertilization and irrigation decision-making model is not ideal, the reasons can be analyzed and corresponding measures can be taken, such as increasing the amount of training data, adjusting the feature selection strategy, or improving the model algorithm, etc.

[0098] Subsequently, the model can be updated regularly to adapt to the changes in the agricultural production environment and new data accumulation. For example, when new crop varieties, fertilizer products, or abnormal climate changes occur, the parameters and structure of the model can be adjusted in a timely manner to ensure the effectiveness and reliability of the agricultural fertilization and irrigation decision-making model.

[0099] In the actual application stage, according to the agricultural fertilization and irrigation decision results output by the agricultural fertilization and irrigation decision-making model, combined with the actual needs and production conditions of farmers, personalized fertilization and irrigation plans can be generated. The fertilization and irrigation plan specifically includes: fertilization time, fertilization amount, fertilizer type selection, as well as irrigation time, irrigation water volume, and irrigation frequency, etc.

[0100] Preferably, the fertilization and irrigation plan can be presented to farmers in a visual way (such as charts, reports, etc.), which is convenient for farmers to understand and operate. At the same time, detailed explanations and descriptions are provided to help farmers understand the basis and principle of the plan.

[0101] In a specific implementation, the effective feature data in the fertilization and irrigation plan can be extracted, such as concept names, numbers, etc. Then, visual charts and / or visual reports are generated according to the effective feature data.

[0102] Subsequently, farmers carry out agricultural production operations according to the fertilization and irrigation plan. During the implementation process, data such as soil fertility, crop growth status, and meteorological conditions can be monitored in real time through sensors and other devices. The above measured data is compared with the expected values in the fertilization and irrigation plan, and then the decision deviation information can be determined according to the comparison result, and the agricultural fertilization and irrigation decision-making model can be adjusted according to the decision deviation information.

[0103] If the decision deviation information exceeds the preset deviation threshold, deviation prompt information can be sent to the farmer's terminal in a timely manner to notify the farmer in time, and a new decision can be made based on the new agricultural fertilization and irrigation decision results output by the adjusted agricultural fertilization and irrigation decision model to ensure the smooth progress of agricultural production.

[0104] It should be noted that in the whole process of method implementation, this embodiment widely applies encryption technologies, specifically including data transmission encryption and data storage encryption. This embodiment adopts encryption protocols such as SSL / TLS to ensure the security of data of all parties during network transmission and prevent data from being stolen or tampered with. In terms of data storage, encryption algorithms can be used to encrypt the data to ensure that even if the data is illegally obtained, the specific content of the data cannot be obtained without decryption.

[0105] Exemplarily, in the data transmission encryption link, the SSL / TLS (Secure Sockets Layer / Transport Layer Security) protocol can be used to guarantee the security of data of all parties during network transmission. The SSL / TLS protocol prevents data from being stolen or tampered with by establishing a secure encryption channel between each user terminal and the agricultural data management server.

[0106] Before data transmission, a handshake process will be carried out between each user terminal and the agricultural data management server to negotiate encryption algorithms (such as AES, RSA, etc.) and key exchange methods. For example, the RSA algorithm is used for key exchange to generate a symmetric encryption key, and then the AES algorithm is used to encrypt the data for transmission with this symmetric key. This hybrid encryption method combines the key exchange security of asymmetric encryption (such as RSA) and the efficiency of symmetric encryption (such as AES) to ensure the confidentiality and integrity of data during transmission.

[0107] In the data storage encryption link, for the data stored in the distributed database, homomorphic encryption algorithms can be used for encryption. Homomorphic encryption allows specific calculation operations to be performed on ciphertext without decrypting the data, thus ensuring the privacy of data during storage and processing. For example, when calculating the correlation degree between soil fertility data and fertilizer component data, these encrypted data can be directly calculated to obtain the encrypted result, and decryption is only performed when the result needs to be viewed finally. In this way, even if the database is illegally accessed, the attacker cannot obtain the plaintext content of the data.

[0108] In some embodiments, a strict access control mechanism can also be established to authenticate and authorize the users of each user terminal. Different users (farmers, agricultural research institutions, agricultural supplies suppliers, etc.) have different access rights and operation rights. For example, farmers can only access and modify their own land and production data, agricultural research institutions can access relevant research data and experimental results, but cannot modify farmers' data, and agricultural supplies suppliers can only view and update their own product information.

[0109] In practical applications, the permissions of users can be reviewed regularly and adjusted according to the changes in users' roles and business requirements to ensure the effectiveness and security of permission management.

[0110] In some embodiments, a multi-factor authentication mechanism can be adopted to authenticate users by combining username and password, SMS verification code, biometric technologies (such as fingerprint recognition, face recognition, etc.). For example, when a farmer logs in to the system, the farmer first enters the username and password. After the system verification passes, an SMS verification code is sent to the mobile phone bound by the farmer. After the farmer enters the correct verification code, the farmer can also choose to perform fingerprint recognition or face recognition to further confirm the identity. Only users who pass all the authentication links can access the system resources.

[0111] In this embodiment, the role-based access control (RBAC) model is used to manage user permissions, specifically defining different roles such as farmers, agricultural research institution personnel, agricultural supplies suppliers, and administrators, etc. Each role is assigned a different set of permissions. For example, farmers can only access and modify their own land and production data, including viewing the soil test reports, historical planting records of their own land, and receiving fertilization and irrigation plans for their land; agricultural research institution personnel can access relevant research data and experimental results, but cannot modify farmers' data, and their access to research data is also restricted by projects and data levels; agricultural supplies suppliers can only view and update their own product information, such as the detailed parameters and prices of fertilizers and irrigation equipment, etc.; the administrator is responsible for the overall management and maintenance of the system, including the allocation and adjustment of user permissions, data backup and recovery, etc. The permission management system implements these permission restrictions through an access control list (ACL) or an attribute-based access control (ABAC) policy to ensure that users can only operate on data within their authorized scope.

[0112] In some embodiments, it is also possible to audit and record data operations in the entire data interaction and decision-making process, including operations such as data uploading, downloading, modification, deletion, etc. The audit log records information such as the time, user, and content of the operation, so as to be traceable and investigable in the event of data leakage or other security incidents.

[0113] In addition, the audit log can be analyzed regularly to discover potential security risks and abnormal behaviors, and timely measures can be taken for prevention and handling.

[0114] In some embodiments, the logging function of the distributed database and specialized audit software can be used to comprehensively audit and record data operations. The audit log includes information such as the timestamp of the operation, the identity of the user performing the operation, the specific content of the operation (such as detailed information on data uploading, downloading, modification, deletion, etc.), and the source IP address and destination IP address of the operation.

[0115] For example, when a farmer uploads new soil test data, the audit log will record information such as the upload time, the farmer's username, the name and size of the uploaded data file, and the IP address of the farmer's device. These audit logs are stored in an independent storage area with strict access control to prevent tampering or deletion.

[0116] When a data leakage or other security incident occurs, through data analysis and mining of the audit log, the source and process of the incident can be traced. For example, by analyzing the operation time series and the access situation of relevant data, it can be determined whether there are abnormal large-scale data downloads or unauthorized modification operations. Using data mining algorithms (such as association rule mining, anomaly detection algorithms, etc.), potential security risks and malicious behavior patterns can be discovered. For example, association rule mining can find that certain users frequently access specific types of data within a specific time range, which may be an abnormal signal and requires further investigation. The anomaly detection algorithm can establish a normal behavior model based on the historical operation behavior of users. When an operation deviates from the normal model, an alarm is issued in a timely manner and a trace analysis is carried out to determine whether there is a security threat.

[0117] In summary, the core objective of the present invention is to use the private set intersection technology to break data islands and establish a secure and reliable agricultural data management server. Through this server, all parties can achieve effective data sharing and integration without exposing the original content of their sensitive data. Specifically, in the data interaction process, encryption algorithms and security protocols are used to process the data of all parties, and the intersection operation of the data is only carried out in the encrypted state to find the key data intersection related to specific land and crops, so as to provide accurate data support for subsequent decision-making analysis.

[0118] Based on these precise data, a personalized agricultural fertilization and irrigation decision-making model is further constructed. This model can generate customized agricultural fertilization and irrigation decision-making results for farmers according to the soil fertility differences of different lands, the changes in nutrient requirements of crops at different growth stages, and local climate conditions (such as rainfall, temperature, sunshine duration, etc.), combined with the characteristics of agricultural materials resources. For example, in areas with low soil fertility, the model will recommend increasing the application rate of specific fertilizers and reasonably arranging the irrigation time and water volume to ensure that crops can fully absorb nutrients and water and promote growth; while in seasons or regions with more rainfall, the model will correspondingly reduce the irrigation frequency and water volume to avoid water resource waste and disease problems caused by over-wet soil.

[0119] Finally, through the solution provided by the present invention, refined management of agricultural production can be achieved, the utilization efficiency of agricultural resources can be improved, waste of fertilizers and water resources can be reduced, the agricultural production cost can be lowered, and at the same time, the yield and quality of agricultural products can be improved, promoting the sustainable development of agriculture and making a positive contribution to ensuring food security and ecological environment protection.

[0120] Based on the same general inventive concept, the present invention also protects an agricultural fertilization and irrigation decision support system based on private set intersection technology. The agricultural fertilization and irrigation decision support system based on private set intersection technology provided by the present invention will be described below. The agricultural fertilization and irrigation decision support system based on private set intersection technology described below can be mutually corresponding and referred to the agricultural fertilization and irrigation decision support method based on private set intersection technology described above.

[0121] Figure 2 It is a schematic structural diagram of the agricultural fertilization and irrigation decision support system based on private set intersection technology provided by an embodiment of the present invention.

[0122] As Figure 2 shown, the agricultural fertilization and irrigation decision support system based on private set intersection technology provided by an embodiment of the present invention specifically includes: an agricultural data management server 210, a distributed database 220, a farmer terminal 230, an agricultural research institution terminal 240, and an agricultural materials supplier terminal 250.

[0123] The agricultural data management server 210 is respectively connected to the distributed database 220, the farmer terminal 230, the agricultural research institution terminal 240, and the agricultural materials supplier terminal 250.

[0124] The agricultural data management server 210 is used to receive the agricultural original data sent by the farmer terminal 230, the agricultural research institution terminal 240, and the agricultural materials supplier terminal 250, encrypt the agricultural original data, and store the encrypted agricultural data obtained by the encryption process into the distributed database 220.

[0125] The agricultural data management server 210 is further configured to receive an agricultural fertilization and irrigation decision request initiated by the farmer terminal 230, determine agricultural key data that matches the agricultural fertilization and irrigation decision request based on the private intersection technology and according to the agricultural encrypted data stored in the distributed database 220, establish an agricultural fertilization and irrigation decision model based on the agricultural key data, and send the agricultural fertilization and irrigation decision result output by the agricultural fertilization and irrigation decision model to the farmer terminal 230.

[0126] See Figure 3 , in one embodiment, the agricultural data management server specifically includes: A data encryption module 310, configured to receive agricultural raw data sent by the farmer terminal, the agricultural research institution terminal, and the agricultural materials supplier terminal, assign a corresponding data identifier to each data element in the agricultural raw data, bind and encrypt each data element with the corresponding data identifier to obtain agricultural encrypted data, and store the agricultural encrypted data in the distributed database.

[0127] A private intersection calculation module 320, configured to receive an agricultural fertilization and irrigation decision request initiated by the farmer terminal, retrieve target agricultural encrypted data from the distributed database based on the agricultural fertilization and irrigation decision request, distribute the data elements in the target agricultural encrypted data to different computing nodes, and perform private intersection calculation on the data elements on all the computing nodes to obtain agricultural key data that matches the agricultural fertilization and irrigation decision request.

[0128] A model construction module 330, configured to extract key feature data from the agricultural key data, perform standardization and normalization processing on the key feature data to obtain intermediate feature data, perform validity screening on the intermediate feature data to obtain sample data, and train a pre-constructed deep learning network model based on the sample data to obtain an agricultural fertilization and irrigation decision model; A result output module 340, configured to send the agricultural fertilization and irrigation decision result output by the agricultural fertilization and irrigation decision model to the farmer terminal.

[0129] Regarding the system in the above embodiments, the specific manners in which each module performs operations have been described in detail in the embodiments related to the method, and will not be elaborated herein.

[0130] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for agricultural fertilization and irrigation decision support based on privacy intersection technology, characterized in that: The method is executed by an agricultural data management service end, and the agricultural data management service end is respectively connected to a distributed database, a farmer terminal, an agricultural research institution terminal, and an agricultural material supplier terminal; the method includes: Receiving the original agricultural data sent by the farmer terminal, the agricultural research institution terminal and the agricultural material supplier terminal; Encrypting the agricultural raw data, and storing the encrypted agricultural data obtained by the encryption process in the distributed database; After receiving the agricultural fertilization and irrigation decision request initiated by the farmer terminal, based on the privacy intersection technology and according to the agricultural encrypted data stored in the distributed database, determine the agricultural key data matching the agricultural fertilization and irrigation decision request; An agricultural fertilization and irrigation decision model is established based on the agricultural key data, and the agricultural fertilization and irrigation decision results output by the agricultural fertilization and irrigation decision model are sent to the farmer terminal.

2. The agricultural fertilization and irrigation decision support method based on privacy intersection technology according to claim 1 is characterized in that: The agricultural raw data is encrypted, including: Assigning a corresponding data identifier to each data element in the agricultural raw data; Each data element is bound and encrypted with the corresponding data identifier to obtain agricultural encrypted data.

3. The agricultural fertilization and irrigation decision support method based on privacy intersection technology according to claim 2 is characterized in that: Assigning a corresponding data identifier to each data element in the agricultural raw data includes: According to the data source, the agricultural raw data is divided into farmer data, scientific research institution data and agricultural input supplier data; Extracting the land number and crop variety of each data element in the farmer data, and assigning a corresponding data identifier to each data element in the farmer data according to the land number and crop variety; Extracting the research project number and experimental conditions of each data element in the scientific research institution data, and assigning a corresponding data identifier to the scientific research institution data according to the research project number and experimental conditions; The product number and applicable scope of each data element in the agricultural supplies supplier data are extracted, and a corresponding data identifier is assigned to the agricultural supplies supplier data based on the product number and applicable scope.

4. The agricultural fertilization and irrigation decision support method based on privacy intersection technology according to claim 2 is characterized in that: Based on the privacy intersection technology and according to the agricultural encrypted data stored in the distributed database, the agricultural key data matching the agricultural fertilization and irrigation decision request is determined, including: Based on the agricultural fertilization and irrigation decision request, retrieve the target agricultural encrypted data in the distributed database; Allocate each data element in the target agricultural encrypted data to different computing nodes according to the data source; A privacy intersection calculation is performed on the data elements on all the computing nodes to obtain key agricultural data that matches the agricultural fertilization and irrigation decision request.

5. The agricultural fertilization and irrigation decision support method based on privacy intersection technology according to claim 4 is characterized in that: Perform privacy intersection calculation on the data elements on all the computing nodes to obtain key agricultural data matching the agricultural fertilization and irrigation decision request, including: Based on the data identifier of each data element, the intersection of the data elements between the computing node whose data source is the farmer and other computing nodes is obtained to obtain the intersection result data set; Comprehensively screen the data elements in the intersection result data set to obtain key agricultural data that matches the agricultural fertilization and irrigation decision request.

6. The agricultural fertilization and irrigation decision support method based on privacy intersection technology according to claim 5 is characterized in that: Based on the data identifier of each data element, the intersection of the data elements between the computing node corresponding to the farmer data and other computing nodes is obtained to obtain the intersection result data set, including: Based on the data identifier of each data element, determine the correlation between each data element in the computing node corresponding to the farmer data and the data elements in other computing nodes; The target data elements whose correlation degree is higher than a preset correlation degree threshold are extracted to obtain an intersection result data set.

7. The agricultural fertilization and irrigation decision support method based on privacy intersection technology according to claim 5 is characterized in that: Comprehensively screen the data elements in the intersection result data set to obtain key agricultural data that matches the agricultural fertilization and irrigation decision request, including: Calculating the data integrity of the intersection result data set, and determining whether the data integrity is higher than a preset integrity threshold, to obtain a integrity determination result; If the completeness judgment result is yes, then extracting redundant elements and invalid elements in the intersection result data set; Redundant elements and invalid elements in the intersection result data set are removed to obtain key agricultural data that matches the agricultural fertilization and irrigation decision request.

8. The agricultural fertilization and irrigation decision support method based on privacy intersection technology according to claim 1 is characterized in that: An agricultural fertilization and irrigation decision model is established based on the key agricultural data, including: Extracting key feature data from the agricultural key data; Standardizing and normalizing the key feature data to obtain intermediate feature data; Performing validity screening on the intermediate feature data to obtain sample data; The pre-constructed deep learning network model is trained based on the sample data to obtain an agricultural fertilization and irrigation decision-making model.

9. An agricultural fertilization and irrigation decision support system based on privacy intersection technology, characterized in that: include: Agricultural data management server, distributed database, farmer terminal, agricultural research institution terminal and agricultural input supplier terminal; The agricultural data management service end is respectively connected to the distributed database, the farmer terminal, the agricultural research institution terminal and the agricultural materials supplier terminal; The agricultural data management service end is used to receive the original agricultural data sent by the farmer terminal, the agricultural research institution terminal and the agricultural material supplier terminal, encrypt the original agricultural data, and store the encrypted agricultural data obtained by the encryption process in the distributed database; The agricultural data management service end is also used to receive the agricultural fertilization and irrigation decision request initiated by the farmer terminal, determine the agricultural key data matching the agricultural fertilization and irrigation decision request based on the privacy intersection technology and according to the agricultural encrypted data stored in the distributed database, establish an agricultural fertilization and irrigation decision model based on the agricultural key data, and send the agricultural fertilization and irrigation decision results output by the agricultural fertilization and irrigation decision model to the farmer terminal.

10. The agricultural fertilization and irrigation decision support system based on privacy intersection technology according to claim 1 is characterized in that: The agricultural data management server includes: A data encryption module is used to receive the agricultural raw data sent by the farmer terminal, the agricultural research institution terminal and the agricultural material supplier terminal, assign a corresponding data identifier to each data element in the agricultural raw data, bind and encrypt each data element with the corresponding data identifier to obtain agricultural encrypted data, and store the agricultural encrypted data in the distributed database; A privacy intersection calculation module is used to receive the agricultural fertilization and irrigation decision request initiated by the farmer terminal, retrieve the target agricultural encrypted data in the distributed database based on the agricultural fertilization and irrigation decision request, distribute each data element in the target agricultural encrypted data to different computing nodes, and perform privacy intersection calculation on the data elements on all the computing nodes to obtain agricultural key data matching the agricultural fertilization and irrigation decision request; A model building module, used to extract key feature data from the agricultural key data, standardize and normalize the key feature data to obtain intermediate feature data, screen the intermediate feature data for effectiveness to obtain sample data, and train a pre-built deep learning network model based on the sample data to obtain an agricultural fertilization and irrigation decision model; The result output module is used to send the agricultural fertilization and irrigation decision results output by the agricultural fertilization and irrigation decision model to the farmer terminal.

Images