Real-time control method for precise pesticide application of plant protection unmanned aerial vehicle

Abstract

The invention provides a real-time control method for precise pesticide application of a plant protection unmanned aerial vehicle, and relates to the technical field of agricultural pesticide application, and the method specifically comprises the steps: dividing a farmland to be applied with pesticide into a plurality of sub-regions, collecting a plant spectrum image of each region, extracting reflectivity data of a key wave band, obtaining the average thickness and the average photosynthetic rate of leaves, and calculating the average thickness and the average photosynthetic rate of the leaves; and calculating a leaf structure density index, a plant growth health index and a water-light collaborative evaluation index in combination with reflected wave data so as to generate a regional pest evaluation index. And based on historical fertilization data and crop health states, establishing a pesticide application prediction model by using a long-short-term memory network, and predicting the pesticide application amount and mode by taking the regional disease and insect pest evaluation index as input. The environment correction factor is calculated in combination with the environment parameters, the pesticide application amount is corrected, it is ensured that the unmanned aerial vehicle accurately completes the pesticide application task, and the pesticide application accuracy and efficiency are effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of agricultural pesticide application, and in particular to a real-time control method for precise pesticide application by a crop protection unmanned aerial vehicle. Background Art

[0002] In modern agricultural production, the prevention and control of pests and diseases is not only related to the yield and quality of crops, but also directly affects the economic benefits of farmers. Traditional methods of applying pesticides usually rely on experience and often adopt fixed application rates and frequencies. However, this method has many shortcomings. First, uneven application of pesticides leads to excessive pesticides in some areas, while other areas are seriously affected by pests and diseases due to lack of adequate protection. Secondly, the waste of pesticides is widespread, resulting in waste of resources and environmental pollution. In addition, climate change and the heterogeneity of farmland make it difficult for fixed application patterns to adapt to different agricultural environments, thereby reducing the efficiency and effectiveness of pesticide use. The limitations of traditional methods have made farmers face greater challenges in pest and disease management, and a more precise and intelligent application solution is urgently needed.

[0003] In addition, existing technologies are relatively weak in real-time monitoring and data analysis capabilities. Although some technologies have been applied to plant growth monitoring and pest and disease identification, they are still insufficient in integrating multiple biophysical parameters and environmental factors. Existing technologies are often unable to obtain and process key data such as plant spectral information, leaf thickness, photosynthetic rate, etc. in real time, resulting in reduced accuracy in pesticide application decisions. At the same time, the lack of effective models to evaluate and predict the health status of different crops under different environmental conditions has led to a lack of scientific basis for pesticide application strategies. Existing systems are usually unable to achieve dynamic assessment of pest and disease risks, and are unable to adjust pesticide application strategies in a timely manner to cope with changing climate conditions and pest and disease development trends. This technical deficiency has seriously affected the sustainability and economic benefits of agricultural production, and there is an urgent need to introduce new technologies to improve the efficiency and accuracy of pest and disease management.

[0004] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not constitute the prior art that is already known to one of ordinary skill in the art. Summary of the invention

[0005] The purpose of the present invention is to provide a real-time control method for precise pesticide application by a plant protection UAV to solve the problems raised in the above background technology.

[0006] To achieve the above object, the present invention provides the following technical solutions: A real-time control method for precise pesticide application by a crop protection UAV, the specific steps include: S1, evenly divide the farmland to be sprayed into several sub-areas, use plant protection drones to collect plant spectral images in each sub-area, extract reflectance data of key bands from the spectral images, and obtain the average leaf thickness data and average photosynthetic rate of the sub-areas; S2, calculate the chlorophyll content, normalized vegetation index, plant water content and leaf area index of each sub-area based on the reflected wave data, and calculate the interaction term between the average leaf thickness data and the leaf area index to generate the leaf structure density index; generate the plant growth health index based on the chlorophyll content and the normalized vegetation index, and generate the water-light synergy evaluation index based on the plant water content and the average photosynthetic rate; S3, combining the plant growth health index, water-light synergy evaluation index and leaf structure density index to generate a regional pest and disease evaluation index; S4, collecting a number of historical farmland fertilization data and regional pest and disease assessment indexes, mapping the regional pest and disease assessment index in each sample with the corresponding farmland fertilization data one by one, generating a sample data set, wherein the farmland fertilization data includes the application amount and the spraying method, establishing a neural network prediction model based on a long short-term memory network LSTM model, taking the regional pest and disease assessment index in the sample data set as input, taking the corresponding farmland fertilization data in the sample data set as a label, training the neural network prediction model, and obtaining a pesticide application prediction model; S5, inputting the regional pest and disease assessment index of the sub-region into the trained pesticide application prediction model, the model outputs the predicted value of the farmland fertilization data of the sub-region to be pesticide applied, obtaining the environmental parameters of the farmland to be pesticide applied, calculating the environmental correction factor based on the obtained environmental parameters, correcting the pesticide application amount in the predicted value of the farmland fertilization data according to the environmental correction factor, and obtaining the precise value of the pesticide application amount, and operating the drone to complete the pesticide application of each sub-region in the farmland to be pesticide applied according to the precise value of the pesticide application amount and the fertilization method in the predicted value of the farmland fertilization data, wherein the environmental parameters include the average temperature and the average wind speed.

[0007] Furthermore, the key band refers to a band in a spectral image that can reflect chlorophyll content, normalized vegetation index, plant water content and leaf area index; The chlorophyll content of plants in the sub-area is calculated based on the reflected wave data. The formula for calculating the chlorophyll content is as follows: ; in, is the chlorophyll content, is the reflectivity in the 660nm band, is the reflectivity in the 450nm band, is a constant, which is used to reflect the effect of chlorophyll content on reflectance ratio; The plant moisture content is calculated based on the reflected wave data. The formula for calculating the plant moisture content is as follows: ; in, is the plant water content, is the reflectivity in the 970nm band, is the reflectivity in the 660nm band, is a constant, which is used to reflect the effect of plant water content on reflectance ratio; The leaf area index is calculated based on the reflected wave according to the following formula: ; in, is the leaf area index, is the light attenuation index, obtained according to experimental data, is the reflectance of the leafless plant area in the 660nm band, is the reflectance of the leafy plant area at 660nm; The normalized vegetation index is calculated based on the reflected wave, and the formula is as follows: ; in, is the normalized difference vegetation index, is the near-infrared reflectivity, is the red light reflectance; The specific logic for collecting the average photosynthetic rate is as follows: collect three samples of plants of the same mass from the sub-area as samples, place them in a closed air chamber, apply the same lighting conditions, record the changes in carbon dioxide over a certain period of time, and calculate the photosynthetic rate. The average photosynthetic rate of the three samples is taken as the average photosynthetic rate, recorded as ; The formula for calculating the photosynthetic rate is: ; in, is the photosynthetic rate, The gas chamber in the measurement period The change in concentration, is the measurement time period, is the effective photosynthetic area of ​​the leaf; The specific logic for collecting the average leaf thickness data is as follows: collect the plant leaf thickness data in the sub-region multiple times, and take the average value as the average leaf thickness data of the sub-region, recorded as .

[0008] Furthermore, the interaction term of the dimensionless average leaf thickness data and leaf area index is calculated to generate the leaf structure density index, based on the following formula: ; In the formula, is the blade structure density index, is the average thickness data of the blade, is the leaf area index; The plant growth health index is generated based on the dimensionless chlorophyll content and the normalized vegetation index, and the formula is as follows: ; in, is the plant growth health index, is the chlorophyll content, is the normalized difference vegetation index, , are the weight coefficients of chlorophyll content and normalized difference vegetation index, , and satisfies ; The water-light synergy evaluation index is generated based on the dimensionless plant water content and average photosynthetic rate, and the formula is as follows: ; in, is the water-light synergy evaluation index, is the average photosynthetic rate, is the photosynthetic rate reference threshold, is the plant water content, is the reference threshold of moisture content.

[0009] Furthermore, the plant growth health index, water-light synergy evaluation index and leaf structure density index are combined to generate the regional pest and disease initial evaluation index based on the formula: ; in, is the regional pest assessment index, is a natural constant, is the blade structure density index, is the plant growth health index, is the water-light synergy evaluation index, , and are the weight coefficients of leaf structure density index, plant growth health index and water-light synergy evaluation index, respectively. , and satisfies .

[0010] Furthermore, a neural network prediction model is established based on the long short-term memory network LSTM model, and the activation function and optimization algorithm are selected. The Tanh function is selected as the activation function, and Adam is selected as the optimization algorithm of the LSTM model; the Tanh function formula is: ; In the formula, Represents the Tanh function, independent variable represents the weighted sum of the inputs of the neuron, that is, the result of the weighted summation of the inputs received by the neuron from the previous layer; at the same time, the hyperparameters of the LSTM model are set, and the hyperparameters of the LSTM model include: the number of network layers, the number of iterations, the learning rate, the batch size, the number of training times, the number of batches, and the number of neurons in the hidden layer; wherein the number of network layers is set to a 3-layer network structure, the number of iterations is set to 200, the learning rate is set to 0.001, the batch size is set to 32, the number of training times is set to 100, the batch size is set to 256, and the number of neurons in the hidden layer is 32; The regional pest and disease assessment index of the sub-region is input into the trained pesticide application prediction model, and the model outputs the predicted value of farmland fertilization data in the sub-region to be applied. The predicted value of farmland fertilization data includes the amount of pesticide applied. and fertilization method .

[0011] Furthermore, the environmental parameters of the farmland to be treated are obtained based on the following specific logic: the environmental temperature and wind speed data of the farmland to be treated are collected multiple times, and the average environmental temperature is taken as the average temperature of the farmland, recorded as , the mean wind speed is taken as the average wind speed of the farmland, recorded as , the ambient temperature refers to the atmospheric temperature at a vertical height of 1.5-2m from the ground; After the obtained environmental parameters are dimensionless, the environmental correction factor is calculated based on the following formula: ; in, is the environmental correction factor, is the average temperature, is the average wind speed.

[0012] Furthermore, the amount of pesticide used in the predicted value of farmland fertilization data is adjusted according to the environmental correction factor. Correction is made to obtain the exact value of the application amount, based on the specific formula: ; in, For the exact value of the dosage, is the amount of pesticide used in the predicted value of farmland fertilization data, e is a natural constant, is the environmental correction factor, is the weight value set, and ; According to the exact value of the dosage and the fertilization method in the predicted value of farmland fertilization data , operate the drone to complete the spraying of each sub-area in the farmland to be sprayed.

[0013] Compared with the prior art, the present invention has the following beneficial effects: The present invention can ensure that the amount of pesticide applied in each sub-region matches the actual demand by performing detailed regional division and health assessment of the farmland to be applied, thereby reducing the waste of pesticides and the risk of environmental pollution. Secondly, the prediction model constructed based on the long short-term memory network can accurately predict the amount of pesticide applied and the spraying method, making the pesticide application process more scientific and systematic. This method not only improves the management efficiency of crop health, but also provides farmers with a basis for optimizing pesticide application decisions, ultimately achieving an increase in crop yield and quality. By correcting environmental parameters, the pesticide application process can better adapt to changing environmental conditions, further improving the accuracy and reliability of crop health management. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 It is a schematic diagram of the overall method flow of the present invention. DETAILED DESCRIPTION

[0015] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with specific embodiments.

[0016] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the present invention should be understood by people with ordinary skills in the field to which the present invention belongs. The words "first", "second" and similar words used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. "Include" or "comprise" and similar words mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0017] Example: See also Figure 1 , the present invention provides a technical solution: A real-time control method for precise pesticide application by a crop protection UAV, the specific steps include: S1, evenly divide the farmland to be sprayed into several sub-areas, use plant protection drones to collect plant spectral images in each sub-area, extract reflectance data of key bands from the spectral images, and obtain the average leaf thickness data and average photosynthetic rate of the sub-areas; In this embodiment, the key band refers to the band in the spectral image that can reflect the chlorophyll content, normalized vegetation index, plant water content and leaf area index; The chlorophyll content of plants in the sub-area is calculated based on the reflected wave data. The formula for calculating the chlorophyll content is as follows: ; in, is the chlorophyll content, is the reflectivity in the 660nm band, is the reflectivity in the 450nm band, is a constant used to reflect the effect of chlorophyll content on reflectance ratio, which is obtained by consulting relevant ecological and plant physiology research literature. reference value; the formula uses the reflectance ratio form, which can effectively eliminate the influence of external factors such as different lighting conditions and sensor differences on the reflectance, and the ratio can better reflect the relative chlorophyll content rather than the absolute value, thereby improving the stability and consistency of the data.

[0018] The plant moisture content is calculated based on the reflected wave data. The formula for calculating the plant moisture content is as follows: ; in, is the plant water content, is the reflectivity in the 970nm band, is the reflectivity in the 660nm band, is a constant used to reflect the effect of plant water content on reflectance ratio, which is obtained by consulting relevant ecological and plant physiology research literature. reference value; the formula uses the reflectance ratio form, which can effectively eliminate the impact of external factors such as different lighting conditions and sensor differences on the reflectance, and the ratio can better reflect the relative plant moisture content rather than the absolute value, thereby improving the stability and consistency of the data.

[0019] The leaf area index is calculated based on the reflected wave according to the following formula: ; in, is the leaf area index, is the light attenuation index, obtained according to experimental data, is the reflectance of the leafless plant area in the 660nm band, It is the reflectance of the leafy plant area in the 660nm band; this formula uses the natural logarithm to characterize the reflectance ratio because light attenuation is exponential. According to the Beer-Lambert law, the relationship between light intensity and distance can be described by an exponential function. Therefore, when calculating the leaf area index, taking the logarithm can linearize it, which is convenient for calculation and understanding; and It reflects the reflectivity changes with and without blades. The reflectivity ratio can eliminate the influence of some external factors, making the calculation results more stable and reliable.

[0020] The normalized vegetation index is calculated based on the reflected wave, and the formula is as follows: ; in, is the normalized difference vegetation index, is the near-infrared reflectivity, is the red light reflectance; The specific logic for collecting the average photosynthetic rate is as follows: collect three samples of plants of the same mass from the sub-area as samples, place them in a closed air chamber, apply the same lighting conditions, record the changes in carbon dioxide over a certain period of time, and calculate the photosynthetic rate. The average photosynthetic rate of the three samples is taken as the average photosynthetic rate, recorded as ; The formula for calculating the photosynthetic rate is: ; in, is the photosynthetic rate, The gas chamber in the measurement period The change in concentration, is the measurement time period, is the effective photosynthetic area of ​​the leaf; The specific logic for collecting the average leaf thickness data is as follows: collect the plant leaf thickness data in the sub-region multiple times, and take the average value as the average leaf thickness data of the sub-region, recorded as .

[0021] The advantage of step 1 is that by evenly dividing the farmland to be sprayed into several sub-areas, it is possible to monitor and evaluate the growth conditions of crops in different areas in detail. Compared with the traditional overall spraying method, this regionalized management method can effectively identify the distribution of pests and diseases, avoid the waste of pesticides and unnecessary environmental impacts, and thus achieve a more precise spraying strategy.

[0022] Compared with the prior art, the beneficial effect of step 1 is that high-resolution plant spectral images are collected by plant protection drones, and reflectance data of key bands are extracted, combined with the acquisition of average leaf thickness and photosynthetic rate, to form rich biophysical parameter data. This data-driven approach makes subsequent pest and disease assessment and pesticide application decisions more scientific and precise, and can significantly improve the efficiency of crop health management. In the present invention, the use of step 1 not only provides basic data support for subsequent data analysis and model construction, but also lays a solid foundation for the implementation of the overall solution. Through targeted pesticide management, while ensuring crop health and increasing yields, the amount of pesticide used and environmental impact can be reduced, promoting the sustainable development of agriculture.

[0023] S2, calculate the chlorophyll content, normalized vegetation index, plant water content and leaf area index of each sub-area based on the reflected wave data, and calculate the interaction term between the average leaf thickness data and the leaf area index to generate the leaf structure density index; generate the plant growth health index based on the chlorophyll content and the normalized vegetation index, and generate the water-light synergy evaluation index based on the plant water content and the average photosynthetic rate; In this embodiment, the leaf average thickness data and the leaf area index after dimensionless processing are interactively calculated to generate the leaf structure density index, and the formula based on it is as follows: ; In the formula, It is the leaf structure density index, which is used to characterize the comprehensive effect of the relative thickness and number of leaves of plants per unit area. is the average thickness data of the blade, is the leaf area index. The formula uses a logarithmic function, which can effectively handle nonlinear relationships, and the logarithmic transformation can reduce the impact of outliers.

[0024] Thicker leaves generally mean that the plant is better able to store water and nutrients and is in a healthier state. Planting is also associated with good plant health and competitiveness because it can use light and other resources more efficiently, thus , and There is a positive correlation.

[0025] The plant growth health index is generated based on the dimensionless chlorophyll content and the normalized vegetation index, and the formula is as follows: ; in, is the plant growth health index, is the chlorophyll content, is the normalized difference vegetation index, , are the weight coefficients of chlorophyll content and normalized difference vegetation index, , and satisfies ; This is because the chlorophyll content is a direct indicator of plant photosynthesis and directly reflects the photosynthetic capacity of plants. The higher the chlorophyll content, the better the growth of the plant, so it is given a higher weight value. It is a vegetation index obtained through remote sensing technology, reflecting the vegetation coverage and growth status. It can provide useful information about the health of vegetation, but it is calculated based on several factors, so it is not as indicative as a direct measurement of chlorophyll content.

[0026] Using the cube root form of calculation, the plant growth health index is Smoother, avoiding excessive amplification of the results by large values. The higher the chlorophyll content, the better the growth of the plant; The value indicates that the plant growth is good; therefore, , and There is a positive correlation.

[0027] The water-light synergy evaluation index is generated based on the dimensionless plant water content and average photosynthetic rate, and the formula is as follows: ; in, is the water-light synergy evaluation index, is the average photosynthetic rate, is the photosynthetic rate reference threshold, is the plant water content, is the reference threshold for moisture content. This formula adds the two indicators of photosynthetic rate and moisture content, reflecting the synergistic effect of these two factors in plant growth. Photosynthetic rate and moisture content are two key factors in plant growth, and their interaction directly affects the health and growth of plants. Therefore, it is reasonable to combine them for evaluation. Squaring the result of the synthesis means that the impact on the synergistic evaluation index is nonlinear. Small deviations may lead to Larger changes in the values, and larger deviations lead to This nonlinear processing can more sensitively reflect changes in plant status, especially small changes when approaching the ideal state.

[0028] when When it increases, it means that the deviation between the average photosynthetic rate and the photosynthetic rate reference threshold increases. Too high or too low photosynthetic rate will have a negative impact on plant growth. Will follow decreases as the value of When increasing, will also decrease; that is, , and There is a negative correlation.

[0029] The advantage of step 2 is that through in-depth analysis of reflected wave data, plant growth indicators such as chlorophyll content, normalized vegetation index, plant moisture content and leaf area index can be accurately calculated for each sub-area. This detailed indicator calculation makes the assessment of crop health more comprehensive and scientific, and can better reflect the actual growth status of plants, thus providing a solid data foundation for subsequent pest and disease assessment and pesticide application decisions.

[0030] Compared with the prior art, the beneficial effect of step 2 is that it combines the calculation of multiple biophysical parameters to form a comprehensive evaluation system of plant growth health index and water-light synergy evaluation index. This systematic evaluation method can more effectively identify potential pest and disease risks. Compared with the single indicator relied on by traditional methods, it enhances the timeliness and accuracy of pesticide application and reduces the risk of crop losses caused by pests and diseases. In the present invention, the use of step 2 not only improves the accuracy and scientificity of the overall pesticide application decision, but also provides important support for the generation of subsequent regional pest and disease assessment indicators. Through a comprehensive assessment of plant health status, targeted pesticide management can be achieved to ensure the rational use of resources, thereby promoting the sustainable development of agriculture.

[0031] S3, combining the plant growth health index, water-light synergy evaluation index and leaf structure density index to generate a regional pest and disease evaluation index; In this embodiment, the plant growth health index, the water-light synergy evaluation index and the leaf structure density index are combined to generate the regional pest and disease initial evaluation index, and the formula based on it is: ; in, is the regional pest assessment index, is a natural constant, is the blade structure density index, is the plant growth health index, is the water-light synergy evaluation index, , and are the weight coefficients of leaf structure density index, plant growth health index and water-light synergy evaluation index, respectively. , and satisfies This is because Directly related to the overall health of the plant. Healthy plants are generally more resistant to pests and diseases, so when assessing the risk of disease and pests, should be given a higher weight to reflect its core role in pest and disease resistance; while the leaf structure density index Reflects the structural characteristics of plant leaves, affecting photosynthetic efficiency and water transpiration, etc. However, although the structural characteristics of leaves are important, they have a relatively small impact on overall plant health and pest and disease resistance, so they are given a lower weight value.

[0032] This formula makes full use of the advantages of the exponential function's rapid growth, positive value characteristics, and linear combination, which helps to comprehensively evaluate the relationship between plant health status and pest and disease risk. This form not only ensures the interpretability of the model, but also enhances its application effect in actual management. When plants are affected by pests and diseases, their leaves may become smaller or thinner, so a high AS value means that the plant is in good growth condition; similarly, when or When it increases, it means that the plant is currently in a relatively healthy growth state and the risk of diseases and insect pests is reduced. , , and There is a positive correlation.

[0033] The advantage of step 3 is that it can generate a regional pest and disease assessment index by combining the plant growth health index, water-light synergy assessment index and leaf structure density index. This comprehensive assessment method makes the judgment of crop health status more comprehensive and accurate. Compared with the traditional method that relies on a single indicator, it can better reflect the actual situation of pests and diseases, thereby providing a scientific basis for subsequent pesticide application decisions.

[0034] Compared with the prior art, the beneficial effect embodied in step 3 is that it reduces the misjudgment caused by insufficient or one-sided information through a comprehensive analysis of multiple dimensions, and improves the accuracy and reliability of pest and disease assessment. This method makes the application strategy more precise, can effectively reduce unnecessary use of pesticides, reduce the impact on the environment, and improve the yield and quality of crops. In the present invention, the use of step 3 not only enhances the scientificity and effectiveness of the overall application decision, but also provides important support for the precise management of subsequent application. By real-time evaluation of crop health status, the application plan can be adjusted in time to ensure efficient use of resources, thereby promoting the sustainable development of agricultural production.

[0035] S4, collecting a number of historical farmland fertilization data and regional pest and disease assessment indexes, mapping the regional pest and disease assessment index in each sample with the corresponding farmland fertilization data one by one, generating a sample data set, wherein the farmland fertilization data includes the application amount and the spraying method, establishing a neural network prediction model based on a long short-term memory network LSTM model, taking the regional pest and disease assessment index in the sample data set as input, taking the corresponding farmland fertilization data in the sample data set as a label, training the neural network prediction model, and obtaining a pesticide application prediction model; In this embodiment, a neural network prediction model is established based on a long short-term memory network LSTM model, and an activation function and an optimization algorithm are selected, wherein the Tanh function is selected as the activation function, and Adam is selected as the optimization algorithm of the LSTM model; the Tanh function formula is: ; In the formula, Represents the Tanh function, independent variable represents the weighted sum of the inputs of the neuron, that is, the result of the weighted summation of the inputs received by the neuron from the previous layer; at the same time, the hyperparameters of the LSTM model are set, and the hyperparameters of the LSTM model include: the number of network layers, the number of iterations, the learning rate, the batch size, the number of training times, the number of batches, and the number of neurons in the hidden layer; wherein the number of network layers is set to a 3-layer network structure, the number of iterations is set to 200, the learning rate is set to 0.001, the batch size is set to 32, the number of training times is set to 100, the batch size is set to 256, and the number of neurons in the hidden layer is 32; The regional pest and disease assessment index of the sub-region is input into the trained pesticide application prediction model, and the model outputs the predicted value of farmland fertilization data in the sub-region to be applied. The predicted value of farmland fertilization data includes the amount of pesticide applied. and fertilization method .

[0036] The advantage of step 4 is that by building a long-term and short-term memory network model, it can achieve accurate pesticide application prediction based on historical farmland fertilization data and crop health status. This process associates historical data with the current regional pest and disease assessment index to generate highly accurate predictions of pesticide application amount and spraying method, providing strong data support for pesticide application decisions.

[0037] Compared with the prior art, the beneficial effect of step 4 is that the intelligent and precise application of plant protection pesticides is achieved by using a deep learning model. In traditional methods, pesticide application decisions often rely on experience or simple rules, while the use of an LSTM model can effectively process time series data, capture potential trends and changes, significantly improve the scientificity and adaptability of pesticide application, and reduce errors caused by human factors. In the present invention, the use of step 4 not only improves the accuracy of pesticide application prediction, but also lays the foundation for the automation and intelligence of the entire pesticide application process. By obtaining the regional pest and disease assessment index in real time and combining it with historical data for dynamic adjustment, we can better cope with the ever-changing agricultural environment, ensure the rational use of resources and the healthy growth of crops, and thus promote the sustainable development of agricultural production.

[0038] S5, inputting the regional pest and disease assessment index of the sub-region into the trained pesticide application prediction model, the model outputs the predicted value of the farmland fertilization data of the sub-region to be pesticide applied, obtaining the environmental parameters of the farmland to be pesticide applied, calculating the environmental correction factor based on the obtained environmental parameters, correcting the pesticide application amount in the predicted value of the farmland fertilization data according to the environmental correction factor, and obtaining the precise value of the pesticide application amount, and operating the drone to complete the pesticide application of each sub-region in the farmland to be pesticide applied according to the precise value of the pesticide application amount and the fertilization method in the predicted value of the farmland fertilization data, wherein the environmental parameters include the average temperature and the average wind speed.

[0039] In this embodiment, the environmental parameters of the farmland to be treated are obtained based on the following specific logic: the environmental temperature and wind speed data of the farmland to be treated are collected multiple times, and the average environmental temperature is taken as the average temperature of the farmland, which is recorded as , the mean wind speed is taken as the average wind speed of the farmland, recorded as , the ambient temperature refers to the atmospheric temperature at a vertical height of 1.5-2m from the ground; After the obtained environmental parameters are dimensionless, the environmental correction factor is calculated based on the following formula: ; in, is the environmental correction factor, is the average temperature, is the average wind speed. Squaring the values ​​emphasizes the nonlinear nature of the combined effects of temperature and wind speed on the environment. Squaring is often used to characterize a certain degree of amplification, which is reasonable in ecology because environmental factors can have nonlinear effects on plant growth.

[0040] The amount of pesticide used in the predicted value of farmland fertilization data according to the environmental correction factor Correction is made to obtain the exact value of the application amount, based on the specific formula: ; in, For the exact value of the dosage, is the amount of pesticide used in the predicted value of farmland fertilization data, e is a natural constant, is the environmental correction factor, is the weight value set, and By setting up multiple experimental areas, the dosage of pesticide is calculated and applied based on different weight values ​​in different experimental areas, and the weight value corresponding to the area with the best insecticide removal effect is selected as The formula uses the form of an exponential function to effectively capture the nonlinear effects of environmental impacts. The properties of the exponential function make it possible to When increasing, the correction factor It will increase in a rapid growth manner, thus more clearly reflecting the impact of environmental factors on the amount of fertilizer applied.

[0041] When the average wind speed or average temperature increases, the kinetic energy of the drug molecules will increase, the volatility of the drug will increase, and thus affect the effect of the drug. Therefore, when other conditions remain unchanged, When it increases, the dosage needs to be increased to maintain the drug effect, i.e. and There is a positive correlation.

[0042] According to the exact value of the dosage and the fertilization method in the predicted value of farmland fertilization data , operate the drone to complete the spraying of each sub-area in the farmland to be sprayed.

[0043] The advantage of step S5 is that by inputting the regional pest assessment index into the trained pesticide application prediction model, the farmland fertilization data of the sub-area to be applied can be accurately predicted, including the amount of application and the method of fertilization. This method not only improves the accuracy of pesticide application, but also effectively reduces the cost of pesticide application and reduces resource waste, while ensuring that crops receive the optimal nutrient supply under specific environmental conditions. Through data-driven prediction, farmland managers can make more scientific decisions based on real-time environmental and crop conditions.

[0044] Compared with the existing technology, step S5 can more comprehensively consider the dynamic changes of farmland environment and the occurrence patterns of pests and diseases by introducing a deep learning model to process complex time series data. Compared with the traditional experience-based pesticide application method, the model can update and adjust the pesticide application strategy in real time, significantly improving the adaptability and effect of pesticide application. At the same time, the correction mechanism based on the environmental correction factor makes the application amount more in line with actual needs, thereby improving crop yield and quality. In this scheme, step S5 effectively integrates the regional pest and disease assessment index and environmental parameters to form a scientific pesticide application decision support system, which promotes the development of precision agriculture. This step not only optimizes the pesticide application process, but also improves the scientificity and adaptability of fertilization, making farmland management more intelligent and data-based. The introduction of this method can significantly improve the growth, health and yield of crops, and promote agricultural production to develop in an efficient and sustainable direction.

[0045] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0046] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination thereof. When implemented by software, the above embodiments may be implemented in whole or in part in the form of a computer program product. Those skilled in the art may appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein may be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed by hardware or software methods depends on the specific application and design constraints of the technical solution.

[0047] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, and may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0048] The above description is only a specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application.

Claims

1. A real-time control method for precise pesticide application by a crop protection drone, characterized in that: The specific steps include: S1, evenly divide the farmland to be sprayed into several sub-areas, use plant protection drones to collect plant spectral images in each sub-area, extract reflectance data of key bands from the spectral images, and obtain the average leaf thickness data and average photosynthetic rate of the sub-areas; S2, calculate the chlorophyll content, normalized vegetation index, plant water content and leaf area index of each sub-area based on the reflected wave data, and calculate the interaction term between the average leaf thickness data and the leaf area index to generate the leaf structure density index; generate the plant growth health index based on the chlorophyll content and the normalized vegetation index, and generate the water-light synergy evaluation index based on the plant water content and the average photosynthetic rate; S3, combining the plant growth health index, water-light synergy evaluation index and leaf structure density index to generate a regional pest and disease evaluation index; S4, collecting a number of historical farmland fertilization data and regional pest and disease assessment indexes, mapping the regional pest and disease assessment index in each sample with the corresponding farmland fertilization data one by one, generating a sample data set, wherein the farmland fertilization data includes the application amount and the spraying method, establishing a neural network prediction model based on a long short-term memory network LSTM model, taking the regional pest and disease assessment index in the sample data set as input, taking the corresponding farmland fertilization data in the sample data set as a label, training the neural network prediction model, and obtaining a pesticide application prediction model; S5, inputting the regional pest and disease assessment index of the sub-region into the trained pesticide application prediction model, the model outputs the predicted value of the farmland fertilization data of the sub-region to be pesticide applied, obtaining the environmental parameters of the farmland to be pesticide applied, calculating the environmental correction factor based on the obtained environmental parameters, correcting the pesticide application amount in the predicted value of the farmland fertilization data according to the environmental correction factor, and obtaining the precise value of the pesticide application amount, and operating the drone to complete the pesticide application of each sub-region in the farmland to be pesticide applied according to the precise value of the pesticide application amount and the fertilization method in the predicted value of the farmland fertilization data, wherein the environmental parameters include the average temperature and the average wind speed.

2. The method for real-time control of precise pesticide application by a crop protection drone according to claim 1, characterized in that: The key band refers to the band in the spectral image that can reflect the chlorophyll content, normalized vegetation index, plant water content and leaf area index; The chlorophyll content of plants in the sub-area is calculated based on the reflected wave data. The formula for calculating the chlorophyll content is as follows: ； in, is the chlorophyll content, is the reflectivity in the 660nm band, is the reflectivity in the 450nm band, is a constant, which is used to reflect the effect of chlorophyll content on reflectance ratio; The plant moisture content is calculated based on the reflected wave data. The formula for calculating the plant moisture content is as follows: ； in, is the plant water content, is the reflectivity in the 970nm band, is the reflectivity in the 660nm band, is a constant, which is used to reflect the effect of plant water content on reflectance ratio; The leaf area index is calculated based on the reflected wave according to the following formula: ； in, is the leaf area index, is the light attenuation index, obtained according to experimental data, is the reflectance of the leafless plant area in the 660nm band, is the reflectance of the leafy plant area at 660nm; The normalized vegetation index is calculated based on the reflected wave, and the formula is as follows: ； in, is the normalized difference vegetation index, is the near-infrared reflectivity, is the red light reflectance; The specific logic for collecting the average photosynthetic rate is as follows: collect three samples of plants of the same mass from the sub-area as samples, place them in a closed air chamber, apply the same lighting conditions, record the changes in carbon dioxide over a certain period of time, and calculate the photosynthetic rate. The average photosynthetic rate of the three samples is taken as the average photosynthetic rate, recorded as ; The formula for calculating the photosynthetic rate is: ； in, is the photosynthetic rate, The gas chamber in the measurement period The change in concentration, is the measurement time period, is the effective photosynthetic area of ​​the leaf; The specific logic for collecting the average leaf thickness data is as follows: collect the plant leaf thickness data in the sub-region multiple times, and take the average value as the average leaf thickness data of the sub-region, recorded as .

3. The method for real-time control of precise pesticide application by a crop protection drone according to claim 1, characterized in that: The interaction term of the dimensionless average leaf thickness data and leaf area index is calculated to generate the leaf structure density index. The formula is as follows: ； In the formula, is the blade structure density index, is the average thickness data of the blade, is the leaf area index; The plant growth health index is generated based on the dimensionless chlorophyll content and the normalized vegetation index, and the formula is as follows: ； in, is the plant growth health index, is the chlorophyll content, is the normalized difference vegetation index, , are the weight coefficients of chlorophyll content and normalized difference vegetation index, , and satisfies ; The water-light synergy evaluation index is generated based on the dimensionless plant water content and average photosynthetic rate, and the formula is as follows: ； in, is the water-light synergy evaluation index, is the average photosynthetic rate, is the photosynthetic rate reference threshold, is the plant water content, is the reference threshold value of moisture content.

4. The method for real-time control of precise pesticide application by a crop protection drone according to claim 3, characterized in that: The plant growth health index, water-light synergy assessment index and leaf structure density index are combined to generate the regional pest and disease initial assessment index based on the formula: ； in, is the regional pest assessment index, is a natural constant, is the blade structure density index, is the plant growth health index, is the water-light synergy evaluation index, , and are the weight coefficients of leaf structure density index, plant growth health index and water-light synergy evaluation index, respectively. , and satisfies .

5. The real-time control method for precise pesticide application by a crop protection drone according to claim 1, characterized in that: A neural network prediction model is established based on the long short-term memory network LSTM model, and the activation function and optimization algorithm are selected. The Tanh function is selected as the activation function, and Adam is selected as the optimization algorithm of the LSTM model; the Tanh function formula is: ； In the formula, Represents the Tanh function, independent variable represents the weighted sum of the inputs of the neuron, that is, the result of the weighted summation of the inputs received by the neuron from the previous layer; at the same time, the hyperparameters of the LSTM model are set, and the hyperparameters of the LSTM model include: the number of network layers, the number of iterations, the learning rate, the batch size, the number of training times, the number of batches, and the number of neurons in the hidden layer; wherein the number of network layers is set to a 3-layer network structure, the number of iterations is set to 200, the learning rate is set to 0.001, the batch size is set to 32, the number of training times is set to 100, the batch size is set to 256, and the number of neurons in the hidden layer is 32; The regional pest and disease assessment index of the sub-region is input into the trained pesticide application prediction model, and the model outputs the predicted value of farmland fertilization data in the sub-region to be applied. The predicted value of farmland fertilization data includes the amount of pesticide applied. and fertilization method .

6. The real-time control method for precise pesticide application by a crop protection drone according to claim 1, characterized in that: The specific logic for obtaining the environmental parameters of the farmland to be treated is as follows: collect the ambient temperature and wind speed data of the farmland to be treated multiple times, and take the average ambient temperature as the average temperature of the farmland, recorded as , the mean wind speed is taken as the average wind speed of the farmland, recorded as , the ambient temperature refers to the atmospheric temperature at a vertical height of 1.5-2m from the ground; After the obtained environmental parameters are dimensionless, the environmental correction factor is calculated based on the following formula: ； in, is the environmental correction factor, is the average temperature, is the average wind speed.

7. The real-time control method for precise pesticide application by a crop protection drone according to claim 6, characterized in that: The amount of pesticide used in the predicted value of farmland fertilization data according to the environmental correction factor Correction is made to obtain the exact value of the application amount, based on the specific formula: ； in, For the exact value of the dosage, is the amount of pesticide used in the predicted value of farmland fertilization data, e is a natural constant, is the environmental correction factor, is the weight value set, and ; According to the exact value of the dosage and the fertilization method in the predicted value of farmland fertilization data , operate the drone to complete the spraying of each sub-area in the farmland to be sprayed.

Images