Wine grape intelligent planting management method and system based on intelligent sensor network

Abstract

The invention relates to a wine grape intelligent planting management method and system based on an intelligent sensor network, and the method comprises the steps: carrying out the grid division of a target region for planting wine grapes, and obtaining the environment information in a plurality of grid units and the growth pictures of the wine grapes; inputting the environment information and the growth state of the wine grapes into a wine grape irrigation model, inputting growth pictures into a pest and disease damage detection model, obtaining irrigation water demand and pest and disease damage types, and marking the irrigation water demand and the pest and disease damage types in grid units; calculating an operation-production ratio numerical curve of the nitrogen fertilizer, the phosphate fertilizer and the potash fertilizer in the grid unit, and based on the minimum value of the operation-production ratio numerical curve, obtaining the fertilization amount and marking the fertilization amount in the grid unit; based on the irrigation water demand, the pest and disease damage types and the fertilization amount in the grid units, automatic planting management is conducted on the wine grapes. The system monitors the wine grapes, provides refined irrigation and fertilization decisions, and can accurately identify diseases and pests.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent planting management, and in particular to an intelligent planting management method and system for wine grapes based on an intelligent sensor network. Background Art

[0002] In the process of growing wine grapes, reasonable fertilization and irrigation as well as timely prevention and control of pests and diseases are the main factors affecting the benefits. The traditional wine grape planting model has many problems, such as waste of water resources and fertilizers, and the need to consume manpower in the process. Although agricultural intelligence has emerged to solve these problems, the current intelligent planting management of wine grapes still has the following problems: (1) It is impossible to provide refined irrigation and fertilization decisions for wine grapes; (2) It is impossible to accurately identify the pests and diseases that appear in the growth process of wine grapes.

[0003] Therefore, the present invention provides an intelligent planting management method and system for wine grapes based on an intelligent sensor network. Summary of the invention

[0004] The purpose of the present invention is to provide an intelligent planting management method and system for wine grapes based on an intelligent sensor network, to monitor wine grapes, to provide refined irrigation and fertilization decisions, and to accurately identify pests and diseases.

[0005] To achieve the above object, the present invention provides the following solutions:

[0006] The present invention provides a smart planting and management method for wine grapes based on an intelligent sensor network, comprising:

[0007] Grid-dividing the target area for growing wine grapes, and obtaining environmental information and wine grape growth pictures within several grid units;

[0008] Input the environmental information and the growth status of wine grapes into a wine grape irrigation model, and input the growth picture into a pest and disease detection model, obtain the irrigation water requirement and the type of pest and disease, and mark them in the grid unit, wherein the wine grape irrigation model is constructed based on a GA-BP neural network, and the pest and disease detection model is constructed by an AI classification algorithm;

[0009] Calculating the production ratio value curve of nitrogen fertilizer, phosphorus fertilizer and potash fertilizer in the grid unit, and obtaining the fertilizer amount based on the minimum value of the production ratio value curve and marking it in the grid unit;

[0010] Based on the irrigation water requirement, the pest and disease type and the fertilizer amount in the grid unit, the wine grapes are automatically planted and managed.

[0011] Optionally, the environmental information includes soil moisture data, light data, temperature data and daily precipitation;

[0012] The soil moisture data is obtained through a soil moisture sensor, the light data is obtained through a light sensor, and the temperature data is obtained through a temperature sensor.

[0013] Optionally, the wine grape irrigation model is obtained by training a training set, wherein the training set includes historical environmental information and corresponding historical irrigation water requirements;

[0014] Training the wine grape irrigation model by using the training set includes:

[0015] Initialize the population, randomly generate weights and biases of multiple neural networks, and use them as initial individuals, and use real number encoding for the initial individuals;

[0016] The error between the output obtained by training and the actual output is used to obtain the individual fitness value, and the selection, crossover and mutation operations of the genetic algorithm are repeated until the stopping condition is reached.

[0017] Optionally, the pest detection model includes a disease detection model and an insect pest detection model;

[0018] Acquiring the disease detection model includes: constructing a disease image data sample library, transmitting the disease image data sample library to AI, allowing AI to generate a disease classification algorithm, and constructing the disease detection model, wherein the disease image data sample library includes disease classification methods and results;

[0019] Acquiring the pest detection model includes: constructing a pest image data sample library, transmitting the pest image data sample library to AI, allowing AI to generate a pest classification algorithm, and constructing the pest detection model, wherein the pest image data sample library includes pest classification methods and results.

[0020] Optionally, constructing a disease image data sample library and the insect pest image data sample library includes:

[0021] Establish an initial disease image data sample library and an initial insect pest image data sample library, and use the SMOTE algorithm to balance the data;

[0022] The sample library after data balancing is subjected to data enhancement processing, wherein the data enhancement processing includes respectively performing scale transformation, cropping, flipping, rotating, translating, adding noise and changing light conditions on the sample data after data balancing.

[0023] Optionally, obtaining the growth status of the wine grapes includes:

[0024] Identify the height of the wine grapes, the trunk diameter of the wine grapes, and the leaf density of the wine grapes in the growth picture;

[0025] Substituting the height, the trunk diameter and the leaf density into a growth state calculation formula to calculate the growth state;

[0026] The growth state calculation formula is:

[0027] G=b×(a1×G1+a2×G2+a3×G3)

[0028] Wherein, G is the growth state value, a1, a2, a3 are the first coefficient, the second coefficient and the third coefficient respectively, G1, G2, G3 are the height, the trunk diameter and the leaf density respectively.

[0029] Optionally, based on the minimum value of the input-output ratio numerical curve, obtaining the fertilizer amount includes: taking the minimum value of the input-output ratio numerical curve as the initial fertilizer amount, and subtracting the original fertilizer amount in the grid unit from the initial fertilizer amount to obtain the fertilizer amount.

[0030] Optionally, the method for calculating the production ratio numerical curve of nitrogen fertilizer, phosphorus fertilizer and potash fertilizer in the grid unit is:

[0031]

[0032] Among them, Y p is the grape yield of the year, Y c is the grape yield of the previous year, P y is the current grape price, z is the residual fertilizer content in the soil after the grape harvest, z0 is the initial fertilizer content of the grid unit, P is the current fertilizer price, H is the amount of fertilizer used in this cultivation, and S is the amount of fertilizer loss.

[0033] The present invention also provides a system for implementing the intelligent planting management method of wine grapes based on an intelligent sensor network, comprising a data acquisition module, a data processing module, a calculation module and a planting management module;

[0034] The data acquisition module is used to grid the target area for growing wine grapes and obtain environmental information in a number of grid units and growth pictures of wine grapes;

[0035] The data processing module is used to input the environmental information and the growth status of wine grapes into a wine grape irrigation model, and the growth picture into a pest detection model, to obtain the irrigation water requirement and the type of pests and diseases, and mark them in the grid unit, wherein the wine grape irrigation model is constructed based on a GA-BP neural network, and the pest detection model is constructed by an A1 classification algorithm;

[0036] The calculation module is used to calculate the production ratio value curve of nitrogen fertilizer, phosphorus fertilizer and potash fertilizer in the grid unit, and obtain the fertilizer amount based on the minimum value of the production ratio value curve and mark it in the grid unit;

[0037] The planting management module is used to automatically manage the planting of the wine grapes based on the irrigation water requirement, the pest type and the fertilizer amount in the grid unit.

[0038] The beneficial effects of the present invention are as follows: the present invention accurately collects the growth environment of wine grapes through grid division and sensors, and constructs a wine grape irrigation model and a pest and disease detection model respectively, which can provide refined irrigation decisions and timely identify pests and diseases; by calculating the numerical curves of the production ratios of nitrogen fertilizers, phosphorus fertilizers and potassium fertilizers in grid units, it can provide refined fertilization decisions, avoid waste and environmental pollution caused by excessive fertilization, reduce the use of pesticides, and improve the quality and safety of agricultural products. At the same time, intelligent technology makes the agricultural production process more automated and improves work efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0040] Figure 1 This is a flow chart of a smart planting management method for wine grapes based on an intelligent sensor network according to an embodiment of the present invention;

[0041] Figure 2 Schematic diagram of an intelligent wine grape planting management system based on an intelligent sensor network according to an embodiment of the present invention. DETAILED DESCRIPTION

[0042] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0043] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0044] Embodiment 1:

[0045] like Figure 1 As shown, this embodiment provides a smart planting and management method for wine grapes based on an intelligent sensor network, including:

[0046] Grid-dividing the target area for growing wine grapes, and obtaining environmental information and wine grape growth pictures within several grid units;

[0047] Input environmental information and the growth status of wine grapes into the wine grape irrigation model, and input growth pictures into the pest and disease detection model to obtain irrigation water requirements and pest and disease types, and mark them in grid cells. The wine grape irrigation model is built based on the GA-BP neural network, and the pest and disease detection model is built using the AI ​​classification algorithm.

[0048] Calculate the input-output ratio curve of nitrogen fertilizer, phosphorus fertilizer and potassium fertilizer in the grid unit, obtain the fertilizer amount based on the minimum value of the input-output ratio curve and mark it in the grid unit;

[0049] Automatically manage the cultivation of wine grapes based on irrigation water requirements, pest and disease types, and fertilizer application rates within the grid cells.

[0050] Furthermore, the environmental information includes soil moisture data, light data, temperature data and daily precipitation; wherein the soil moisture data is obtained through a soil moisture sensor, the light data is obtained through a light sensor, and the temperature data is obtained through a temperature sensor.

[0051] Specifically, in this embodiment, the sensors are all equipped with GPS positioning inside and are provided with a unique device identifier, which is uniformly generated by the server and has a unique QR code corresponding to the device identifier. Users scan the QR code through the client application device (APP and Web), and after successful binding, they can determine the location of each sensor.

[0052] Furthermore, the wine grape irrigation model is obtained by training with a training set that includes historical environmental information and the corresponding historical irrigation water requirements;

[0053] Training the wine grape irrigation model with the training set includes:

[0054] Initialize the population, randomly generate weights and biases of multiple neural networks, and use real number encoding as the initial individuals;

[0055] The error between the output obtained by training and the actual output is used to obtain the individual fitness value, and the selection, crossover and mutation operations of the genetic algorithm are repeated until the stopping condition is reached.

[0056] Specifically, for each individual, the input data is input into the network, the output result is calculated and the fitness is calculated, the selection and crossover operations in the genetic algorithm are used to select individuals with high fitness and perform crossover operations to generate new individuals, and the mutation operation in the genetic algorithm is used to mutate the new individuals to increase the diversity of the population. Finally, for each new individual, the back propagation algorithm is used to update the weights and biases, and the network with the best output according to the fitness evaluation can be used to predict new data. The GA-BP neural network model combines the advantages of the genetic algorithm and the BP neural network. The genetic algorithm can search for the optimal solution in multiple parameter combinations, while the BP neural network can learn and fit the complex nonlinear relationship in the data, thereby improving the prediction accuracy and stability of the model.

[0057] Further, the pest detection model includes a disease detection model and an insect pest detection model;

[0058] Acquiring a disease detection model includes: building a disease image data sample library, transmitting the disease image data sample library to AI, allowing AI to generate a disease classification algorithm, and building a disease detection model, wherein the disease image data sample library includes disease classification methods and results;

[0059] Acquiring a pest detection model includes: building a pest image data sample library, transmitting the pest image data sample library to AI, allowing AI to generate a pest classification algorithm, and building a pest detection model, wherein the pest image data sample library includes pest classification methods and results.

[0060] Furthermore, building a disease image data sample library and an insect pest image data sample library includes:

[0061] Establish an initial disease image data sample library and an initial insect pest image data sample library, and use the SMOTE algorithm to balance the data;

[0062] The sample library after data balancing is subjected to data enhancement processing, which includes scaling, cropping, flipping, rotating, translating, adding noise and changing lighting conditions of the sample data after data balancing.

[0063] Further, obtaining the growth status of the wine grapes includes: identifying the height of the wine grapes in the growth picture, the trunk diameter of the wine grapes, and the leaf density of the wine grapes; substituting the height, the trunk diameter, and the leaf density into a growth status calculation formula to calculate the growth status;

[0064] The growth state calculation formula is:

[0065] G=b×(a1×G1+a2×G2+a3×G3)

[0066] Wherein, G is the growth state value, a1, a2, a3 are the first coefficient, the second coefficient and the third coefficient respectively, G1, G2, G3 are the height, the trunk diameter and the leaf density respectively.

[0067] Specifically, in this embodiment, a plurality of cameras are installed in each grid unit, and the cameras are controlled to shoot the wine grapes in the grid unit according to the growth cycle of the wine grapes, and the growth pictures are obtained. The plurality of pictures are analyzed respectively to obtain the average value of the growth status. When identifying the height, a horizontal line is drawn at the highest point and the lowest point of the wine grape outline respectively, and the distance between the two horizontal lines is set as the height. When identifying the diameter, the trunk diameter at the preset position is identified.

[0068] Furthermore, based on the minimum value of the input-output ratio numerical curve, obtaining the fertilizer amount includes: taking the minimum value of the input-output ratio numerical curve as the initial fertilizer amount, and subtracting the original fertilizer amount in the grid unit from the initial fertilizer amount to obtain the fertilizer amount.

[0069] Furthermore, the method for calculating the numerical curve of the input-output ratio of nitrogen fertilizer, phosphorus fertilizer and potassium fertilizer in the grid unit is:

[0070]

[0071] Among them, Y p is the grape yield of the year, Y c is the grape yield of the previous year, P y is the current grape price, Z is the residual fertilizer content in the soil after the grape harvest, Z0 is the initial fertilizer content of the grid unit, P is the current fertilizer price, H is the amount of fertilizer used in this cultivation, and S is the amount of fertilizer loss.

[0072] Specifically, calculate the amount of fertilizer loss, fertilizer loss S = (1-I%) × total amount of fertilizer absorbed by crops Xp / regional area; fertilizer utilization rate I% = (total amount of fertilizer absorbed by crops Xp-remaining fertilizer content in the soil after harvest Zi) / amount of fertilizer used during farming H×100%.

[0073] Embodiment 2:

[0074] like Figure 2 As shown, the intelligent planting management system for wine grapes based on the intelligent sensor network includes a data acquisition module, a data processing module, a calculation module and a planting management module;

[0075] The data acquisition module is used to grid the target area for growing wine grapes and obtain the environmental information in several grid units and the growth pictures of wine grapes;

[0076] The data processing module is used to input environmental information and the growth status of wine grapes into the wine grape irrigation model, and the growth pictures into the pest and disease detection model, obtain the irrigation water requirement and the type of pests and diseases, and mark them in the grid unit. The wine grape irrigation model is built based on the GA-BP neural network, and the pest and disease detection model is built through the AI ​​classification algorithm;

[0077] A calculation module is used to calculate the input-output ratio value curve of nitrogen fertilizer, phosphorus fertilizer and potassium fertilizer in the grid unit, and obtain the fertilizer amount based on the minimum value of the input-output ratio value curve and mark it in the grid unit;

[0078] The planting management module is used to automatically manage the planting of wine grapes based on the irrigation water demand, pest and disease types, and fertilizer application amount within the grid unit.

[0079] The embodiments described above are only descriptions of the preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Without departing from the design spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary technicians in this field should all fall within the protection scope determined by the claims of the present invention.

Claims

1. An intelligent planting and management method for wine grapes based on an intelligent sensor network, characterized in that: include: Grid-dividing the target area for growing wine grapes, and obtaining environmental information and wine grape growth pictures within several grid units; Input the environmental information and the growth status of wine grapes into a wine grape irrigation model, and input the growth picture into a pest and disease detection model, obtain the irrigation water requirement and the type of pest and disease, and mark them in the grid unit, wherein the wine grape irrigation model is constructed based on a GA-BP neural network, and the pest and disease detection model is constructed by an AI classification algorithm; Calculating the production ratio value curve of nitrogen fertilizer, phosphorus fertilizer and potash fertilizer in the grid unit, and obtaining the fertilizer amount based on the minimum value of the production ratio value curve and marking it in the grid unit; Based on the irrigation water requirement, the pest and disease type and the fertilizer amount in the grid unit, the wine grapes are automatically planted and managed.

2. The intelligent planting and management method of wine grapes based on intelligent sensor network according to claim 1 is characterized in that: The environmental information includes soil moisture data, light data, temperature data and daily precipitation; The soil moisture data is obtained through a soil moisture sensor, the light data is obtained through a light sensor, and the temperature data is obtained through a temperature sensor.

3. The intelligent planting and management method of wine grapes based on intelligent sensor network according to claim 1 is characterized in that: The wine grape irrigation model is obtained by training a training set, wherein the training set includes historical environmental information and corresponding historical irrigation water requirements; Training the wine grape irrigation model by using the training set includes: Initialize the population, randomly generate weights and biases of multiple neural networks, and use them as initial individuals, and use real number encoding for the initial individuals; The error between the output obtained by training and the actual output is used to obtain the individual fitness value, and the selection, crossover and mutation operations of the genetic algorithm are repeated until the stopping condition is reached.

4. The intelligent planting and management method of wine grapes based on intelligent sensor network according to claim 1 is characterized in that: The pest detection model includes a disease detection model and an insect pest detection model; Acquiring the disease detection model includes: constructing a disease image data sample library, transmitting the disease image data sample library to AI, allowing AI to generate a disease classification algorithm, and constructing the disease detection model, wherein the disease image data sample library includes disease classification methods and results; Acquiring the pest detection model includes: constructing a pest image data sample library, transmitting the pest image data sample library to AI, allowing AI to generate a pest classification algorithm, and constructing the pest detection model, wherein the pest image data sample library includes pest classification methods and results.

5. The intelligent planting and management method of wine grapes based on intelligent sensor network according to claim 4 is characterized in that: Constructing a disease image data sample library and the insect pest image data sample library includes: Establish an initial disease image data sample library and an initial insect pest image data sample library, and use the SMOTE algorithm to balance the data; The sample library after data balancing is subjected to data enhancement processing, wherein the data enhancement processing includes respectively performing scale transformation, cropping, flipping, rotating, translating, adding noise and changing light conditions on the sample data after data balancing.

6. The intelligent planting and management method of wine grapes based on intelligent sensor network according to claim 1 is characterized in that: Obtaining the growth status of the wine grapes includes: Identify the height of the wine grapes, the trunk diameter of the wine grapes, and the leaf density of the wine grapes in the growth picture; Substituting the height, the trunk diameter and the leaf density into a growth state calculation formula to calculate the growth state; The growth state calculation formula is: G=b×(a1×G1+a2×G2+a3×G3) Wherein, G is the growth state value, a1, a2, a3 are the first coefficient, the second coefficient and the third coefficient respectively, G1, G2, G3 are the height, the trunk diameter and the leaf density respectively.

7. The intelligent planting and management method of wine grapes based on intelligent sensor network according to claim 1 is characterized in that: Based on the minimum value of the input-output ratio numerical curve, obtaining the fertilizer amount includes: taking the minimum value of the input-output ratio numerical curve as the initial fertilizer amount, and subtracting the original fertilizer amount in the grid unit from the initial fertilizer amount to obtain the fertilizer amount.

8. The intelligent planting and management method for wine grapes based on intelligent sensor network according to claim 1 is characterized in that: The method for calculating the production ratio curve of nitrogen fertilizer, phosphorus fertilizer and potash fertilizer in the grid unit is: Among them, Y p is the grape yield of the year, Y c is the grape yield of the previous year, P y is the current grape price, Z is the residual fertilizer content in the soil after the grape harvest, Z0 is the initial fertilizer content of the grid unit, P is the current fertilizer price, H is the amount of fertilizer used in this cultivation, and S is the amount of fertilizer loss.

9. A system for implementing the intelligent planting and management method of wine grapes based on an intelligent sensor network according to any one of claims 1 to 8, characterized in that: It includes data acquisition module, data processing module, calculation module and planting management module; The data acquisition module is used to grid the target area for growing wine grapes and obtain environmental information in a number of grid units and growth pictures of wine grapes; The data processing module is used to input the environmental information and the growth status of wine grapes into a wine grape irrigation model, and the growth picture into a pest detection model, to obtain the irrigation water requirement and the type of pests and diseases, and mark them in the grid unit, wherein the wine grape irrigation model is constructed based on a GA-BP neural network, and the pest detection model is constructed by an AI classification algorithm; The calculation module is used to calculate the input-output ratio value curve of nitrogen fertilizer, phosphorus fertilizer and potassium fertilizer in the grid unit, and based on the minimum value of the input-output ratio value curve, obtain the fertilizer amount and mark it in the grid unit; The planting management module is used to automatically manage the planting of the wine grapes based on the irrigation water requirement, the pest type and the fertilizer amount in the grid unit.

Images