Intelligent control method and system for broccoli root zone alternate irrigation device

Abstract

The invention discloses an intelligent control method of a broccoli root zone alternate irrigation device, and relates to the field of alternate irrigation, and the method comprises the steps: obtaining broccoli associated information; outputting an initial growth cycle through the initial cycle prediction model; acquiring broccoli remote sensing information through a remote sensing unmanned aerial vehicle; assimilating the initial period prediction model in combination with broccoli remote sensing information and broccoli associated information to obtain a standard period prediction model, and determining a standard growth period based on the standard period prediction model; and generating an alternate irrigation strategy based on the standard growth cycle, and performing intelligent regulation and control on the alternate irrigation device according to the alternate irrigation strategy. Waste of water resources can be effectively reduced.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of alternative irrigation, and particularly to an intelligent control method and system for an alternative irrigation device in the root zone of broccoli. Background Art

[0002] Alternative irrigation is a water-saving irrigation technology that optimizes plant water absorption by scientifically regulating the water distribution in the root zone of crops, thereby improving the water resource utilization efficiency. Its core lies in alternately wetting different root zones of crops and utilizing the physiological response mechanism of plants themselves to maintain the yield while reducing the irrigation water volume.

[0003] Existing alternative irrigation devices usually rely on past experience to design fixed irrigation plans. When unexpected situations such as sudden drought or precipitation occur, which are difficult to predict based on past experience, it may lead to insufficient or excessive irrigation of vegetables in the next growth cycle of vegetables due to the difficulty of timely adjusting the irrigation plan, affecting the growth of crops and causing water resource waste at the same time. Summary of the Invention

[0004] The embodiments of the present application provide an intelligent control method for an alternative irrigation device in the root zone of broccoli, which is used to solve the problem of easy water resource waste in the application of traditional alternative irrigation devices.

[0005] To achieve the above object, the embodiments of the present application adopt the following technical solutions:

[0006] In a first aspect, an intelligent control method for an alternative irrigation device in the root zone of broccoli is provided, and the method includes:

[0007] Obtain the broccoli-related information of the target vegetable field;

[0008] Input the broccoli-related information into a pre-constructed initial cycle prediction model, and output all the initial growth cycles of the target broccoli in the target vegetable field through the initial cycle prediction model;

[0009] When the initial growth cycle is reached for the first time, collect the broccoli remote sensing information of the target vegetable field through a remote sensing drone;

[0010] Assimilate the initial cycle prediction model by combining the broccoli remote sensing information and the broccoli-related information to obtain a standard cycle prediction model, and correct the initial growth cycle based on the standard cycle prediction model to obtain a standard growth cycle;

[0011] When any standard growth cycle is reached, generate an alternative irrigation strategy for the target vegetable field based on the standard growth cycle, and perform intelligent regulation on the alternative irrigation device preset in the target vegetable field according to the alternative irrigation strategy.

[0012] Optionally, the broccoli-related information includes the first meteorological information of the location of the target vegetable field, the broccoli information in the target vegetable field, and the soil information of the target vegetable field. The soil information includes the pre-acquired soil water absorption rate, soil water holding rate, and soil type, as well as the soil water content collected in real time by soil sensors preset in the target vegetable field.

[0013] Optionally, the initial cycle prediction model is assimilated with the broccoli remote sensing information and the broccoli-related information to obtain a standard cycle prediction model, and the initial growth cycle is corrected based on the standard cycle prediction model to obtain the standard growth cycle, including the following steps:

[0014] Preprocess the broccoli remote sensing information;

[0015] Use the inversion algorithm to invert and process the preprocessed broccoli remote sensing information, and extract the vegetable field remote sensing parameters in the target vegetable field according to the inversion results. The vegetable field remote sensing parameters include the broccoli index and the vegetable field water content;

[0016] Use the data fusion algorithm to complete the spatial fusion of the vegetable field water content and the soil water content to obtain the fusion water content of the target vegetable field;

[0017] Assimilate the initial cycle prediction model with the broccoli index and the fusion water content to obtain a standard cycle prediction model;

[0018] Input the first meteorological information, the broccoli information, and the soil information during the initial growth cycle into the standard cycle prediction model, and output the standard growth cycle of the target broccoli through the standard cycle prediction model.

[0019] [[ID=2l]]Optionally, using the data fusion algorithm to complete the spatial fusion of the vegetable field water content and the soil water content to obtain the fusion water content of the target vegetable field includes the following steps:

[0020] Take the soil water content collected during the initial growth cycle as the target water content;

[0021] Standardize the vegetable field water content and the target water content;

[0022] Use the correlation coefficient formula to calculate the data correlation between the standardized vegetable field water content and the target water content;

[0023] Use the preset variogram to analyze the spatial aggregation of the standardized vegetable field water content and the target water content;

[0024] Combine the spatial aggregation and the data correlation and use the Kriging interpolation method to interpolate and fuse the vegetable field water content and the target water content to obtain the fusion water content of the target vegetable field during the initial growth cycle.

[0025] Optionally, combining the broccoli index and the initial cycle prediction model for assimilating the water content, the steps for obtaining the standard cycle prediction model are as follows:

[0026] Align the spatio-temporal scales between the broccoli index and the assimilated water content;

[0027] Normalize the broccoli index and the assimilated water content after spatio-temporal scale alignment to obtain the standard index and the standard water content respectively;

[0028] Define the error covariance matrix by combining the standard index and the standard water content;

[0029] Generate the initial state set of the initial cycle prediction model according to the broccoli correlation information;

[0030] Run the initial cycle prediction model according to the initial state set and the ensemble Kalman filter algorithm to obtain the predicted state set;

[0031] Calculate the Kalman gain by combining the predicted state set and the error covariance matrix;

[0032] Adjust the model parameters of the initial cycle prediction model by combining the Kalman gain, the standard index and the standard water content to obtain the standard cycle prediction model.

[0033] Optionally, generating the alternative irrigation strategy for the target vegetable field based on the standard growth cycle includes the following steps:

[0034] Generate the alternative irrigation cycle for the target vegetable field according to the cycle type and cycle time in the standard growth cycle;

[0035] Re-obtain the second meteorological information of the location of the target vegetable field during the standard growth cycle, where the second meteorological information includes meteorological parameters and total precipitation;

[0036] Calculate the reference evapotranspiration of the target broccoli according to the meteorological parameters;

[0037] Determine the crop coefficient of the target broccoli according to the cycle type;

[0038] Calculate the theoretical water requirement of the target broccoli by combining the reference evapotranspiration and the crop coefficient;

[0039] Use the soil water content collected during the standard growth cycle as the standard water content;

[0040] Correct the theoretical water requirement of the broccoli by using the soil water absorption rate, soil water holding rate and the standard water content to obtain the actual water requirement of the broccoli;

[0041] Calculate the daily average irrigation amount of the target broccoli by combining the total precipitation and the actual water requirement of the broccoli;

[0042] Integrate the alternative irrigation cycle and the daily average irrigation amount to obtain the alternative irrigation strategy for the target vegetable field.

[0043] Optionally, the method further includes:

[0044] After completing the intelligent irrigation task within any standard growth cycle, calculate the irrigation qualification degree within the standard growth cycle using the irrigation qualification formula;

[0045] If the irrigation qualification degree is within the preset qualification degree threshold range, determine that the intelligent irrigation task is qualified;

[0046] If the irrigation qualification degree is not within the qualification degree threshold range, determine that the intelligent irrigation task is unqualified;

[0047] If the intelligent irrigation task is qualified, continue to complete the intelligent irrigation task within the next standard growth cycle;

[0048] If the intelligent irrigation task is unqualified, output a strategy correction factor according to the irrigation qualification degree, and use the strategy correction factor to correct the alternative irrigation strategy for the next standard growth cycle.

[0049] Optionally, the irrigation qualification formula includes:

[0050]

[0051] Where S x is the actual irrigation amount of the target broccoli, S y is the daily average irrigation amount of the target broccoli, γ is the irrigation correction factor, W l and W r are the daily average soil water contents on the left and right sides of the target broccoli respectively, is the preset water content difference threshold, N is the number of soil sensors, δ is the preset fixed integer parameter, K is the pre-acquired soil hydraulic conductivity, T is the daily average irrigation duration, and F is the preset temperature attenuation factor.

[0052] In a second aspect, the present application provides a machine-readable storage medium, characterized in that instructions are stored on the machine-readable storage medium, and the instructions are used to cause the machine to execute the method for intelligent control of the broccoli root zone alternative irrigation device according to any one of the first aspects.

[0053] In a third aspect, the present application provides an intelligent control system for a broccoli root zone alternative irrigation device, characterized in that it includes:

[0054] A memory configured to store instructions; and

[0055] A processor configured to call instructions from the memory and be able to implement the method for intelligent control of the broccoli root zone alternative irrigation device according to any one of the first aspects when executing the instructions.

[0056] Through the above technical solution, the entire initial growth cycle of the target broccoli is predicted through the broccoli correlation information and the pre-constructed initial cycle prediction model. When the initial growth cycle is reached for the first time, the remote sensing information of the broccoli in the target vegetable field is collected, and the initial cycle prediction model is assimilated by combining the remote sensing information of the broccoli and the soil moisture content obtained in real time, so that the standard cycle prediction model obtained after assimilation has a more accurate prediction result, that is, the standard growth cycle more conforms to the actual growth condition of the target broccoli. Therefore, the alternative irrigation cycle of the irrigation device can be dynamically adjusted according to the different target initial growth cycles, thereby promoting the root development and physiological regulation of the target broccoli. In addition, according to the different target initial growth cycles and the different weather conditions in different growth cycles, the daily average irrigation amount of the target broccoli is determined, which can ensure that the target broccoli will not be over-irrigated or under-irrigated due to sudden drought or precipitation during the growth process. To sum up, compared with the existing alternative irrigation technology, the present application can effectively reduce water resource waste and improve irrigation accuracy while promoting the growth and development of the target broccoli and increasing its yield by dynamically regulating the alternative irrigation device, providing strong technical support for modern agricultural production.

[0057] Other features and advantages of the embodiments of the present application will be described in detail in the subsequent specific implementation part. Brief Description of the Drawings

[0058] Figure 1 It is a schematic flow chart of the intelligent control of an alternative irrigation device for the root zone of broccoli provided by an embodiment of the present application;

[0059] Figure 2 It is a schematic flow chart of generating an alternative irrigation strategy provided by an embodiment of the present application. Detailed Description of the Embodiments

[0060] To make the objectives, technical solutions, and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application. It should be understood that the specific implementation manners described herein are only used to illustrate and explain the embodiments of the present application, and are not used to limit the embodiments of the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present application without creative efforts shall fall within the protection scope of the present application.

[0061] It should be noted that if there are directional indications (such as up, down, left, right, front, back...) involved in the embodiments of the present application, the directional indications are only used to explain the relative position relationship and movement conditions between components in a specific posture (as shown in the drawings). If the specific posture changes, the directional indications will also change accordingly.

[0062] In addition, if the descriptions such as "first" and "second" are involved in the embodiments of the present application, these descriptions of "first", "second", etc. are for descriptive purposes only, and should not be construed as indicating or implying their relative importance or implicitly indicating the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include at least one such feature. In addition, the technical solutions between the various embodiments may be combined with each other, but it must be based on the ability of those of ordinary skill in the art to implement. When the combination of technical solutions results in contradictions or cannot be implemented, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection required by the present application.

[0063] Figure 1 Schematically shows a flow chart of a method for intelligent control of an alternating irrigation device in the root zone of broccoli according to an embodiment of the present application. As Figure 1 shown, the embodiments of the present application provide a method for intelligent control of an alternating irrigation device in the root zone of broccoli, and the method may include the following steps:

[0064] S101. Obtain the broccoli-related information of the target vegetable field;

[0065] In this embodiment, the broccoli-related information refers to the information that affects the initial growth cycle. The broccoli-related information includes the first meteorological information of the location of the target vegetable field, such as precipitation, temperature, wind speed, humidity, etc., the broccoli information in the target vegetable field, such as broccoli varieties (early-maturing varieties, late-maturing varieties, etc.), etc., and the soil information of the target vegetable field. The soil information includes soil water absorption rate, soil water holding rate, soil type, and soil water content, etc.

[0066] S102. Input the broccoli-related information into a pre-constructed initial cycle prediction model, and output all the initial growth cycles of the target broccoli in the target vegetable field through the initial cycle prediction model;

[0067] In this embodiment, the initial cycle prediction model can be constructed based on a neural network model, or can be constructed using existing WOFOST models or DSSAT models. Taking the neural network model as an example, historical broccoli planting information of the target vegetable field is obtained. The historical broccoli planting information includes the historical growth cycle and historical soil information (including historical average soil moisture content, historical soil type, etc.) of the target broccoli planted in the target vegetable field, as well as the historical meteorological information of the location of the target vegetable field during the historical time period. After data cleaning and information annotation of the historical growth cycle, historical meteorological information, and historical soil information, they are randomly divided into the model training set and model validation set of the initial cycle prediction model. The initial cycle prediction model includes an input layer for receiving information, a hidden layer for performing non-linear transformation and feature extraction on the input information, and an output layer for generating the final prediction result. The initial number of nodes is set for the input layer, hidden layer, and output layer according to the feature dimensions of the historical broccoli planting information. A suitable loss function and optimizer are selected. For example, the mean squared error (MSE) can be selected as the loss function, and the optimizer can be SGD (stochastic gradient descent). The loss function is used to measure the difference between the prediction result of the model and the true label, and to measure the performance of the model. The core function of the optimizer is to update the parameters of the model according to the gradient of the loss function, which can be used to select a suitable learning rate or dynamically adjust the learning rate during training, so as to accelerate the convergence speed and improve the model training speed. After multiple iterative trainings of broccoli cycle prediction using the model training set, the model parameters are continuously optimized, and parameters such as the model learning rate and batch size are adjusted until the maximum number of iterations is reached. Then, the model performance of the trained initial cycle prediction model is evaluated using the validation set. The parameters used for evaluation include accuracy, recall rate, etc. When each parameter reaches the preset parameter threshold, the training of the initial cycle prediction model is completed. The first meteorological information, broccoli information, soil type, and soil moisture content in the broccoli association information are input into the initial cycle prediction model, and all initial growth cycles of the target broccoli in the target vegetable field are output through the initial cycle prediction model. For example, the seedling stage of the target broccoli is 35 days, the growth stage is 71 days, and the maturity stage is 17 days.

[0068] In addition, existing WOFOST model or DSSAT model can be directly selected as the initial cycle prediction model. Both the WOFOST model and the DSSAT model are crop growth simulation models developed by relevant scientific researchers. Taking the WOFOST model as an example, it is jointly developed by Wageningen University of Agriculture in the Netherlands and the World Food Research Center and is used to simulate the growth process of annual crops under specific soil and climate conditions, including crop growth simulation under potential growth conditions, water-limited conditions, and nutrient-limited conditions. Since the WOFOST model adopts a modular structure, it can be parameterized and adapted to different crops and environmental conditions, and can simulate the growth cycles of broccoli, including various stages such as sowing, growth, and harvesting. Specifically, data such as temperature, light, soil water content, precipitation, soil water content, and soil type are converted into a fixed format and then input into the WOFOST model, and simulation scenarios are set, such as the alternate irrigation scenario, so as to use the WOFOST model to simulate and analyze the growth cycle of the target broccoli, that is, to obtain the predicted dates of the growth stages of the target broccoli such as the seedling stage, flowering stage, and maturity stage, that is, to obtain the initial growth cycle of the target broccoli. Therefore, if the above model construction steps are omitted and the WOFOST model or DSSAT model is directly used as the initial cycle prediction model, the time for model construction and prediction can be effectively saved.

[0069] S103. When the initial growth cycle is reached for the first time, collect the remote sensing information of the broccoli in the target vegetable field by using a remote sensing drone;

[0070] In this embodiment, the initial growth cycle refers to the cycle type and cycle time of each growth cycle of broccoli during the process from the emergence of broccoli seeds to the maturity of the seedlings. For example, the seedling stage of the target broccoli is 35 days, the growth stage is 71 days, and the maturity stage is 17 days. The seedling stage, flowering stage, and maturity stage are cycle types. Therefore, generally, the first initial growth cycle reached is the seedling stage. Since the area of the target vegetable field is large, a remote sensing drone equipped with remote sensing equipment can be used to collect the remote sensing information of the broccoli in the target vegetable field. Commonly used remote sensing equipment includes hyperspectral imagers, thermal infrared sensors, optical cameras, and infrared cameras, etc. At the predicted harvesting time node, control the remote sensing drone to take panoramic and sub-regional photos of the target vegetable field according to the pre-set monitoring method and monitoring path. After the shooting is completed, splice the taken remote sensing images according to the actual scene of the target vegetable field according to the real-time positioning information of the remote sensing drone to obtain the complete remote sensing information of the broccoli in the target vegetable field. The remote sensing information of broccoli refers to the optical remote sensing image, which is collected by devices such as a multispectral camera and a hyperspectral imager installed inside the remote sensing drone.

[0071] For most vegetables, before the seedling stage, since their root systems are not formed, there is only one main root and no lateral root branches, and it is impossible to induce the growth of root-zone partitioning through alternative irrigation. At this time, vegetable seeds require a stable and moist environment, and alternative irrigation may cause local drought, which may lead to uneven emergence of vegetables. Therefore, before the first arrival at the initial growth cycle, alternative irrigation is not required, and there is no need to design an alternative irrigation plan. Just directly turn on all alternative irrigation equipment according to past experience to irrigate the target vegetable field as a whole. For all these reasons, it is only when the first arrival at the initial growth cycle that the initial cycle prediction model begins to be assimilated to generate an alternative irrigation strategy, because the alternative irrigation strategy generated at this time is meaningful and useful.

[0072] S104. Combine the broccoli remote sensing information and the broccoli correlation information to assimilate the initial cycle prediction model to obtain a standard cycle prediction model, and correct the initial growth cycle based on the standard cycle prediction model to obtain a standard growth cycle;

[0073] In this embodiment, first preprocess the broccoli remote sensing information. The preprocessing steps include geometric correction, image denoising, and radiometric calibration. Then use the response algorithm to invert the broccoli index and the water content of the vegetable field of the target broccoli. The broccoli index refers to the leaf area index of the target broccoli, which can be calculated according to the normalized difference vegetation index. The water content of the vegetable field is calculated by the echo intensity of the microwave signal in the broccoli remote sensing information. Then the Kriging interpolation method can be used to fuse the water content of the vegetable field and the target water content to obtain the fused water content of the target vegetable field during the initial growth cycle. Kriging interpolation is an optimal linear unbiased estimation method based on the variogram. It takes into account the spatial correlation of the data and can provide the optimal estimated value and estimation error for the unsampled points. Using the Kriging interpolation method can comprehensively fuse the water content of the vegetable field and the target water content and provide data support for the subsequent assimilation step.

[0074] Assimilation is a method of combining observational data with data considering the spatio-temporal distribution, observational field, and background field errors to optimize the model prediction results. The assimilation of the initial cycle prediction model can be realized by using the Ensemble Kalman Filter (EnKF). The Ensemble Kalman Filter (EnKF) is a data assimilation technique. It effectively fuses the model prediction and observational data by using a set of prediction samples (ensemble) to represent the state and uncertainty of the system. The core idea of the EnKF is to use these ensemble members to approximate the error covariance of the system and then update the estimate of the system state. The basic process of the EnKF includes initialization, observation step, analysis step, and prediction step. The standard cycle prediction model obtained after assimilation is more accurate in its prediction results compared to the initial cycle prediction model without assimilation.

[0075] Finally, convert the first meteorological information, broccoli information, and soil information during the initial growth period into a fixed format and input them into the standard cycle prediction model. Then, use the standard cycle prediction model to simulate the following process day by day with a step size of one day: determine the physiological parameters (such as photosynthetic rate and respiration efficiency) corresponding to different broccoli varieties according to the broccoli information, calculate the efficiency of converting daytime light energy into biomass based on the first meteorological information and physiological parameters, calculate the respiratory consumption of the target broccoli according to the physiological parameters, and simulate the effects of temperature and water stress on growth in combination with the first meteorological information. After completing the simulation process, output the standard growth cycle of the target broccoli through the standard cycle prediction model. At this time, the obtained standard growth cycle is more accurate than the initial growth cycle.

[0076] S105. When reaching any standard growth cycle, generate an alternative irrigation strategy for the target vegetable field based on the standard growth cycle, and intelligently control the alternative irrigation device preset in the target vegetable field according to the alternative irrigation strategy.

[0077] In this embodiment, when reaching any standard growth cycle, first generate an alternative irrigation cycle for the target vegetable field according to the cycle type and cycle time in the standard growth cycle. The cycle type includes growth stages of the target broccoli such as the seedling stage, flowering stage, and fruiting stage, and the cycle time is the duration of different growth stages of the target broccoli. For example, the flowering stage of the target broccoli is 14 days. First, obtain the approximate range of the alternative irrigation cycle of the target broccoli in different standard growth cycles according to historical experience, and then generate the final alternative irrigation cycle with the constraint of achieving irrigation balance on both sides of the target broccoli. At the same time, immediately obtain the second meteorological information of the location of the target vegetable field during this standard growth cycle through the meteorological website. The second meteorological information includes meteorological parameters and total precipitation. The meteorological parameters include parameters such as temperature, net radiation, soil heat flux, psychrometric constant, wind speed, saturation vapor pressure, and actual vapor pressure. Then, input the meteorological parameters into the Penman-Monteith formula (FAO Penman-Monteith formula), and use the Penman-Monteith formula to calculate the reference evapotranspiration of the target broccoli. Then, calculate the theoretical water requirement of the broccoli of the target broccoli in combination with the reference evapotranspiration and the crop coefficient of the target broccoli, and correct the theoretical water requirement of the broccoli according to the soil water content, soil water absorption rate, and soil water holding rate at this time to obtain the theoretical water requirement of the broccoli. Calculate the daily average irrigation amount of the target broccoli in combination with the theoretical water requirement of the broccoli and the total precipitation during this standard growth cycle. Through this method, the intelligent irrigation task can be effectively completed by using precipitation, thus saving water resources. Finally, integrate the alternative irrigation cycle and the daily average irrigation amount during the same standard growth cycle to obtain the alternative irrigation strategy for the target vegetable field, generate an irrigation instruction according to the alternative irrigation strategy, and intelligently control the alternative irrigation device preset in the target vegetable field according to the irrigation instruction. This can minimize unnecessary water waste while ensuring sufficient irrigation amount for the target broccoli.

[0078] In one of the embodiments, the broccoli-related information includes the first meteorological information of the location of the target vegetable field, the broccoli information in the target vegetable field, and the soil information of the target vegetable field. The soil information includes the pre-acquired soil water absorption rate, soil water holding rate, and soil type, as well as the soil water content collected in real time by soil sensors preset in the target vegetable field.

[0079] In this embodiment, the first meteorological information refers to the weather forecast obtained before growing broccoli in the target vegetable field, including but not limited to information such as precipitation, temperature, wind speed, and humidity. The broccoli information refers to information such as the broccoli varieties (early-maturing varieties, late-maturing varieties, etc.) planted in the target vegetable field. The soil information includes the soil water absorption rate, soil water holding rate, and soil type. Among them, the soil water absorption rate is measured by collecting soil samples from the target vegetable field and then using the capillary method or pressure plate method under different moisture conditions, which is used to reflect the ability of the soil to absorb water, that is, the speed of water infiltration into the soil per unit time, and can affect the daily average irrigation amount of the target broccoli. If the ability of the soil to absorb water is poor, then during the alternative irrigation process, due to the soil's inability to absorb in time, water loss may occur, resulting in insufficient irrigation amount for the target broccoli. Therefore, when calculating the daily average irrigation amount of the target broccoli, the soil water absorption rate of the target vegetable field needs to be considered. The soil water holding rate refers to the water content retained in the soil after gravity drainage, which is the upper limit of the available water for the target broccoli and is also one of the important factors affecting the daily average irrigation amount. If the daily average irrigation amount is too high and exceeds the upper limit of the available water for the target broccoli, it will cause waste of water resources. The soil type includes sandy soil, clay soil, etc. The soil type will affect the growth cycle of the target broccoli, so it needs to be pre-acquired. The soil water content can reflect the soil humidity and can be obtained by capacitance or time domain reflectometry sensors preset in the target vegetable field. The sensors cover the entire target vegetable field and are evenly distributed in a grid pattern, with one sensor installed at each grid point to collect and upload the soil water content of the target vegetable field in real time, providing key basic data for subsequent precise irrigation decision-making and also being able to be used to monitor whether there are abnormalities in the soil water content subsequently.

[0080] In one of the embodiments, combining the broccoli remote sensing information and the broccoli-related information to assimilate the initial cycle prediction model to obtain a standard cycle prediction model, and modifying the initial growth cycle based on the standard cycle prediction model to obtain the standard growth cycle includes the following steps:

[0081] Preprocess the broccoli remote sensing information;

[0082] Use the inversion algorithm to invert the preprocessed broccoli remote sensing information, and extract the vegetable field remote sensing parameters in the target vegetable field according to the inversion results. The vegetable field remote sensing parameters include the broccoli index and the vegetable field water content;

[0083] The spatial fusion of the water content in the vegetable field and the soil water content is completed using a data fusion algorithm to obtain the fused water content of the target vegetable field;

[0084] Combined with the broccoli index and the assimilation initial cycle prediction model of the fused water content, a standard cycle prediction model is obtained;

[0085] The first meteorological information, broccoli information, and soil information during the initial growth cycle are input into the standard cycle prediction model, and the standard growth cycle of the target broccoli is output through the standard cycle prediction model.

[0086] In this embodiment, the broccoli remote sensing information is preprocessed. The preprocessing steps include geometric correction, image denoising, and radiometric calibration. Geometric correction refers to the process of eliminating or correcting the geometric errors of the broccoli remote sensing information. The main purpose of image denoising is to remove the noise in the broccoli remote sensing information and improve the accuracy of the broccoli remote sensing information. Radiometric calibration is the process of converting the brightness and gray values of the broccoli remote sensing information into absolute radiance values. Through the above steps, the image quality of the broccoli remote sensing information is improved, and image distortion is reduced.

[0087] The inversion algorithm includes leaf area index inversion and soil water content inversion. Leaf area index inversion refers to using the broccoli remote sensing information to invert the broccoli index of the target broccoli. The broccoli index refers to the leaf area index of the target broccoli. Specifically, first, appropriate bands are selected from the broccoli remote sensing information for analysis. Usually, the red light band (NIR) and the near-infrared light band (RED) are used. The normalized difference vegetation index (NDVI) of the target broccoli is calculated based on the red light band and the near-infrared light band. The calculation formula is NDVI = (NIR – RED) / (NIR + RED). Then, the broccoli index of the target broccoli is calculated according to the pre-constructed linear regression equation between the normalized difference vegetation index and the leaf area index. Soil water content inversion is to analyze the echo intensity of the microwave signal through the broccoli remote sensing information, extract the dielectric constant of the soil, and thus calculate the water content of the vegetable field in the target vegetable field.

[0088] The water content of the vegetable field refers to the overall global water content of the target vegetable field, while the soil water content monitored by sensors refers to the local water content at multiple sampling points in the target vegetable field. Therefore, Kriging interpolation can be used to fuse the water content of the vegetable field and the target water content to obtain the fused water content of the target vegetable field during the initial growth period. Specifically, first retrieve the soil water content collected at the initial growth period and use it as the target water content for assimilating the prediction model of the initial period. Then, standardize the water content of the vegetable field and the target water content to eliminate the dimensional differences between different data sources and variables, making the data comparable. Next, use the preset variogram to analyze the spatial aggregation of the standardized water content of the vegetable field and the target water content. Spatial aggregation can also be referred to as spatial correlation. The variogram is an important tool for describing the correlation of spatial data, which reflects the influence of spatial distance on data similarity. Common variogram models include the spherical model, exponential model, and Gaussian model, etc. Finally, based on the data correlation and spatial aggregation analyzed previously, use Kriging interpolation to perform spatial interpolation on the standardized water content of the vegetable field and the target water content to obtain the fused water content of the target vegetable field during the initial growth period. Kriging interpolation is an optimal linear unbiased estimation method based on the variogram. It takes into account the spatial correlation of the data and can provide the optimal estimated value and estimation error for unsampled points. Using Kriging interpolation, the water content of the vegetable field and the target water content can be comprehensively fused to provide data support for subsequent assimilation steps.

[0089] Assimilation is a method that combines observational data with data considering spatio-temporal distribution, observational field, and background field errors to optimize the model prediction results. Ensemble Kalman Filter (EnKF) is a data assimilation technique that effectively fuses model predictions and observational data by using a set of prediction samples (ensemble) to represent the state and uncertainty of the system. The core idea of EnKF is to use these ensemble members to approximate the error covariance of the system and then update the estimate of the system state. The basic process of EnKF includes initialization, observation step, analysis step, and prediction step.

[0090] Specifically, first align the spatio-temporal scales between the broccoli index and the integrated water content, determine the sampling times of the broccoli index and the integrated water content, align their timestamps, and then ensure that the spatial resolutions of the broccoli index and the integrated water content are consistent. After that, normalize the broccoli index and the integrated water content that have completed spatio-temporal scale alignment to eliminate the dimensional differences between different data sources and variables. Then, construct the observation error covariance matrix of the broccoli index and the integrated water content according to the sampling accuracies of the sensors and the remote sensing drones. At the same time, construct the model error covariance matrix of the initial cycle prediction model according to the errors of the historical cycle prediction results, and integrate the model error covariance matrix and the observation error covariance matrix as the error covariance matrix. Then, randomly generate multiple state samples according to the broccoli correlation information and integrate them into the initial state set. Use the initial state set as the input of the initial cycle prediction model to drive the initial cycle prediction model to predict the predicted growth cycle of the current target broccoli, the predicted broccoli index, and the predicted water content of the target vegetable field. Finally, integrate all the prediction results to obtain the predicted state set. Statistically calculate the mean and covariance of all the predicted states in the predicted state set, and construct the predicted covariance matrix according to the mean and covariance. Correspond the predicted state set to the observation state set constructed by the standard index and the standard water content to obtain the observation operator. Then, combine the predicted covariance matrix, the observation operator, and the error covariance matrix and use the Kalman gain formula to calculate the Kalman gain. Use the Kalman gain, the standard index, and the standard water content to correct the model parameters of the initial cycle prediction model to obtain the standard cycle prediction model. The model parameters refer to the state variables such as the predicted leaf area index and the root zone soil water content in the initial cycle prediction model. Since the state variables can affect the prediction results of the initial cycle prediction model, by correcting the state variables through the observation state set, the prediction accuracy of the initial cycle prediction model for the target initial growth cycle can be indirectly improved, and a standard cycle prediction model with higher prediction accuracy can be obtained.

[0091] Convert the first meteorological information, broccoli information, and soil information during the initial growth cycle into a fixed format and input them into the standard cycle prediction model. Then, use the standard cycle prediction model to simulate the following process day by day with a step size of one day: Determine the physiological parameters (such as photosynthetic rate, respiration efficiency) corresponding to different broccoli varieties according to the broccoli information, calculate the efficiency of converting daytime light energy into biomass according to the first meteorological information and the physiological parameters, and calculate the respiratory consumption of the target broccoli according to the physiological parameters. Combine the first meteorological information to simulate the effects of temperature and water stress on growth. After completing the simulation process, output the standard growth cycle of the target broccoli through the standard cycle prediction model. At this time, the obtained standard growth cycle is more accurate than the initial growth cycle.

[0092] In one of the embodiments, the spatial fusion of the water content of the vegetable field and the soil water content is completed using a data fusion algorithm, and the steps for obtaining the fusion water content of the target vegetable field are as follows:

[0093] Take the soil water content collected during the initial growth period as the target water content;

[0094] Standardize the water content of the vegetable field and the target water content;

[0095] Use the correlation coefficient formula to calculate the data correlation between the standardized water content of the vegetable field and the target water content;

[0096] Use the preset variogram to analyze the spatial aggregation of the standardized water content of the vegetable field and the target water content;

[0097] Combine the spatial aggregation and data correlation and use the Kriging interpolation method to interpolate and fuse the water content of the vegetable field and the target water content to obtain the fusion water content of the target vegetable field during the initial growth period.

[0098] In this embodiment, first retrieve the soil water content collected at the initial growth period and use it as the target water content for assimilating the initial period prediction model. Then, standardize the water content of the vegetable field and the target water content to eliminate the dimensional differences between different data sources and variables, making the data comparable. The standardization process usually adopts the z-score standardization method, that is, subtracting the mean from the original data and then dividing by the standard deviation. The specific calculation formula is: Z = (X - μ) / σ, where X is the original data, μ is the data mean, and σ is the data standard deviation. The standardized data will show a distribution characteristic with a mean of 0 and a standard deviation of 1, which is conducive to subsequent data fusion and analysis. Then, use the correlation coefficient formula to calculate the data correlation between the water content of the vegetable field and the target water content. Commonly used correlation coefficient formulas include the Pearson correlation coefficient, the Spearman correlation coefficient, etc. Taking the Pearson correlation coefficient as an example, its mathematical expression is as follows:

[0099]

[0100] Where, X i and Y i are the water content of the i-th vegetable field and the target water content respectively; and are the sample means of the water content of the vegetable field and the target water content respectively, and n is the sample size.

[0101] Then, the preset variogram is used to analyze the spatial aggregation of the water content in the vegetable field after standardization and the target water content. Spatial aggregation can also be referred to as spatial correlation. The variogram is an important tool for describing the correlation of spatial data, which reflects the influence of spatial distance on data similarity. Commonly used variogram models include the spherical model, exponential model, Gaussian model, etc. Taking the spherical model as an example, its mathematical expression is: γ(h) = C0 + C * (1.5h / a - 0.5(h / a)^3) when h < a; γ(h) = C0 + C when h ≥ a. Where h is the spatial distance, C0 is the nugget value (representing the variation at a very small scale), C is the sill value (representing the overall variation), and a is the range (representing the effective distance of spatial correlation). In specific implementation, first calculate the experimental variogram values at different distances, and then fit the theoretical variogram model by the least squares method. For example, assuming that the experimental variogram values calculated at different distances are [(100, 0.2), (200, 0.4), (300, 0.6), (400, 0.7), (500, 0.75)], the spherical model parameters C0 = 0.1, C = 0.7, and a = 450 can be obtained through fitting.

[0102] Finally, based on the data correlation and spatial aggregation analyzed above, the Kriging interpolation method is used to perform spatial interpolation on the water content in the vegetable field after standardization and the target water content, so as to obtain the fused water content of the target vegetable field during the initial growth period. Kriging interpolation is an optimal linear unbiased estimation method based on the variogram. It takes into account the spatial correlation of the data and can provide the optimal estimated value and estimation error for unsampled points. The interpolation formula of Kriging is: Z*(x0) = Σ(λi * Z(xi)), where Z*(x0) is the predicted value of the point to be estimated, Z(xi) is the observed value of the known sampling point, and λi is the weight coefficient. The determination of the weight coefficient is the core of Kriging interpolation, and a system of linear equations needs to be solved: Σ(λj * γ(xi, xj)) + μ = γ(xi, x0), where γ(xi, xj) is the variogram value between two points, and μ is the Lagrange multiplier. In specific implementation, first construct the Kriging equations, then solve the weight coefficients, and finally perform interpolation calculations. For example, given that the observed values of three points are [7, 10, 5] respectively, and the distances between the point to be observed and these three points are [85, 100, 110] respectively, calculate the variogram values according to the previously obtained variogram model, and the solved weight coefficients are [0.5405, 0.4595, 0]. Then the predicted value of the point to be estimated is 0.5405×7 + 0.4595×10 + 0×5 ≈ 8.38. Using the Kriging interpolation method can comprehensively fuse the water content in the vegetable field and the target water content, providing data support for the subsequent assimilation step.

[0103] In one embodiment, combining the broccoli index and the initial cycle prediction model for assimilating the fusion water content, the steps for obtaining the standard cycle prediction model are as follows:

[0104] Align the spatio-temporal scales between the broccoli index and the fusion water content;

[0105] Normalize the broccoli index and the fusion water content with aligned spatio-temporal scales to obtain the standard index and the standard water content respectively;

[0106] Define the error covariance matrix by combining the standard index and the standard water content;

[0107] Generate the initial state set of the initial cycle prediction model according to the broccoli correlation information;

[0108] Run the initial cycle prediction model according to the initial state set and the ensemble Kalman filter algorithm to obtain the prediction state set;

[0109] Calculate the Kalman gain by combining the prediction state set and the error covariance matrix;

[0110] Adjust the model parameters of the initial cycle prediction model by combining the Kalman gain, the standard index and the standard water content to obtain the standard cycle prediction model.

[0111] In this embodiment, assimilation is a method of combining observed data with data considering spatio-temporal distribution, observation field and background field errors to optimize the model prediction results. The ensemble Kalman filter (EnKF) is a data assimilation technique that effectively fuses model predictions and observed data by using a set of prediction samples (ensemble) to represent the state and uncertainty of the system. The core idea of EnKF is to use these ensemble members to approximate the error covariance of the system and then update the estimate of the system state. The basic process of EnKF includes initialization, observation step, analysis step and prediction step. In the initialization stage, the state variables of the system, the initial error covariance matrix and the initial state prediction ensemble are set. In the observation step, the observation ensemble and the observation error covariance matrix are calculated. The analysis step involves calculating the Kalman gain matrix, the analysis ensemble, the analysis ensemble mean and the analysis error covariance matrix. Finally, in the prediction step, the prediction ensemble, the prediction ensemble mean and the prediction error covariance matrix are calculated.

[0112] Specifically, first align the spatio-temporal scales between the broccoli index and the fused water content, determine the sampling times of the broccoli index and the fused water content, perform timestamp alignment on the two, and then ensure that the spatial resolutions of the broccoli index and the fused water content are consistent. After that, normalize the broccoli index and the fused water content that have completed spatio-temporal scale alignment to eliminate the dimensional differences between different data sources and variables, making the data comparable. The normalization process usually adopts the z-score normalization method, that is, subtracting the mean from the original data and then dividing by the standard deviation. Since there may be certain errors in both the soil water content collected by sensors and the broccoli index and the water content of the vegetable field inverted from the remote sensing information of broccoli, it is necessary to construct an observation error covariance matrix, and the elements in the error covariance matrix are determined by the sampling accuracy of the sensors and the remote sensing drones. For example, the sensor error is ±0.02m 3 / m 3, the corresponding element is set to 0.0004. At the same time, the model error covariance matrix of the initial cycle prediction model is constructed according to the error of the historical cycle prediction result. If the historical initial growth cycle prediction error is ±3 days, the corresponding variance is set to 9, that is, the corresponding element in the model error covariance matrix is 9. The model error covariance matrix and the observation error covariance matrix are integrated into the error covariance matrix. Then, the broccoli correlation information is standardized. The first meteorological information, broccoli information, and soil information in the broccoli correlation information are standardized to 0-1, and the standardized broccoli correlation information is integrated into the state set. Then, multiple state samples are randomly generated according to the state set and integrated into the initial state set. The initial state set needs to cover the uncertainty range of each data in the state set to avoid over-concentration or divergence, that is, each data in the initial state set cannot be greater than the maximum value of each data in the initial state set and cannot be less than its minimum value. The maximum and minimum values are determined according to the historical broccoli planting information and historical meteorological information. The initial state set is used as the input of the initial cycle prediction model to drive the initial cycle prediction model to predict the predicted growth cycle, predicted broccoli index, and predicted water content of the target vegetable field of the current target broccoli. Taking the initial cycle prediction model constructed based on the WOFOST model as an example, by simulating physiological processes such as crop photosynthesis and dry matter accumulation and combining meteorological conditions such as temperature and light, the leaf area index of the vegetation canopy can be calculated day by day. The built-in crop growth stage division module in the model can accurately match the LAI change characteristics corresponding to each growth cycle of broccoli (such as emergence stage, flowering stage, maturity stage). Therefore, while predicting the growth cycle, the WOFOST model can predict the predicted broccoli index (LAI). In addition, the WOFOST model can calculate the soil water content in the root layer through the water balance module. This module integrates parameters such as precipitation, irrigation, evapotranspiration, and soil infiltration and can output daily soil water content data. Therefore, the corresponding predicted water content can also be predicted while predicting the growth cycle. Finally, all the prediction results are integrated to obtain the predicted state set. The mean and covariance of all the predicted states in the predicted state set are statistically calculated, and the prediction covariance matrix is constructed according to the mean and covariance. The predicted state set is put into one-to-one correspondence with the observation state set constructed by the standard index and standard water content. For example, the predicted broccoli index output by the model directly corresponds to the standard broccoli index to obtain the observation operator. The observation operator is the identity matrix and contains the parameters corresponding to the observation state set in the predicted state set. Then, combining the prediction covariance matrix P xy , the observation operator H, and the error covariance matrix R and using the Kalman gain formula to calculate the Kalman gain. The Kalman gain formula is: K = P xy ·(H·P xy + R) -1。The Kalman gain is a weighting coefficient used to weight between the predicted value and the measured value, ultimately attributed to the mean square error matrices of prediction and measurement. Its role is to adjust the weights of the predicted value and the measured value so that the final estimated value is as close as possible to the true value, thereby reducing the estimation error. The model parameters of the initial cycle prediction model are corrected using the Kalman gain, standard index, and standard water content to obtain the standard cycle prediction model. The correction formula is: F(x) = F(y) + K(Y - HF(y)), where F(y) is the model parameter before correction, K is the Kalman gain, and Y is the set of observed states obtained by integrating the standard index and standard water content. Model parameters refer to state variables such as the predicted leaf area index and root zone soil water content in the initial cycle prediction model. Since state variables can affect the prediction results of the initial cycle prediction model, correcting the state variables through the set of observed states can indirectly improve the prediction accuracy of the initial cycle prediction model for the target initial growth cycle.

[0113] Using the assimilated standard cycle prediction model, it is possible to more accurately predict each growth stage of the target broccoli and the corresponding time for each growth stage. For example, the seedling stage of tomatoes is 15 days, the flowering stage is 14 days, and the fruiting stage is 45 days. The alternative irrigation strategy generated based on the accurate cycle prediction results will also be more suitable for the irrigation requirements of the target broccoli. The specific reasons are as follows: The target broccoli is in different standard growth cycles, and the suitable alternative cycle during the alternative irrigation process is also different. For example, during the seedling stage of tomatoes, the irrigation direction is adjusted every 2 - 3 days, and during the flowering stage of tomatoes, it is adjusted every 4 - 5 days. When reaching different growth stages, it is necessary to promptly change the alternative irrigation cycle to ensure the normal growth of the target broccoli. In addition, when designing the alternative irrigation cycle, it is also necessary to try to maintain the irrigation balance on both sides of the target broccoli to prevent the total irrigation amount on one side from being less than the other side during a certain cycle. For example, the seedling stage of tomatoes is 15 days, and generally, alternative irrigation is required every 2 - 3 days. To ensure irrigation balance on both sides during the seedling stage, it can be set to alternative irrigation every 2.5 days. In addition, the target broccoli is in different standard growth cycles, and the required irrigation amount is also different. At the same time, the precipitation amounts in different standard growth cycles are also different. Therefore, the required daily average irrigation amount also needs to be dynamically adjusted according to the different growth stages of the target broccoli to ensure that the target broccoli does not experience under-irrigation or over-irrigation. Based on this, it is necessary to accurately predict each growth stage of the target broccoli and the corresponding time for each growth stage in order to dynamically adjust the alternative cycle and daily average irrigation amount according to the different stages of the target broccoli, and to achieve precise regulation of water resources while ensuring the normal growth of the target broccoli, thereby achieving the goal of saving water resources.

[0114] In one embodiment, referring to Figure 2, the alternating irrigation strategy for the target vegetable field generated based on the standard growth cycle includes the following steps:

[0115] Generate the alternating irrigation cycle of the target vegetable field according to the cycle type and cycle time in the standard growth cycle;

[0116] Re-obtain the second meteorological information of the location of the target vegetable field within the standard growth cycle, where the second meteorological information includes meteorological parameters and total precipitation;

[0117] Calculate the reference evapotranspiration of the target broccoli according to the meteorological parameters;

[0118] Determine the crop coefficient of the target broccoli according to the cycle type;

[0119] Calculate the theoretical water requirement of the target broccoli by combining the reference evapotranspiration and the crop coefficient;

[0120] Take the soil water content collected within the standard growth cycle as the standard water content;

[0121] Modify the theoretical water requirement of broccoli by using the soil water absorption rate, soil water holding rate and standard water content to obtain the actual water requirement of broccoli;

[0122] Calculate the daily average irrigation amount of the target broccoli by combining the total precipitation and the actual water requirement of broccoli;

[0123] Integrate the alternating irrigation cycle and the daily average irrigation amount to obtain the alternating irrigation strategy of the target vegetable field.

[0124] In this embodiment, the alternating irrigation cycle of the target vegetable field is generated according to the cycle type and cycle time in the standard growth cycle. The cycle type includes the growth stages of the target broccoli such as the seedling stage, flowering stage and fruiting stage, and the cycle time is the duration of different growth stages of the target broccoli. For example, the flowering stage of the target broccoli is 14 days. First, obtain the approximate range of the alternating irrigation cycle of the target broccoli in different standard growth cycles according to historical experience, and then generate the final alternating irrigation cycle with the constraint of achieving irrigation balance on both sides of the target broccoli. For example, the seedling stage of tomatoes is 15 days, the flowering stage is 14 days, and the fruiting stage is 45 days. According to historical experience, it can be obtained that the irrigation direction of tomatoes is adjusted every 2 - 3 days during the seedling stage, every 4 - 5 days during the flowering stage, and every 5 - 7 days during the fruiting stage. The generated alternating irrigation cycle is: the irrigation direction is adjusted every 2.5 days within the 0 - 15th day, every 4 days within the 16 - 29th day, and every 5.5 days within the 30 - 74th day.

[0125] Upon reaching any standard growth period, immediately obtain the second meteorological information of the location of the target vegetable field during this standard growth period through a meteorological website. The second meteorological information includes meteorological parameters and total precipitation. The meteorological parameters include parameters such as temperature, net radiation, soil heat flux, psychrometric constant, wind speed, saturation vapor pressure, and actual vapor pressure. Then, input the meteorological parameters into the Penman-Monteith formula (FAO Penman-Monteith formula), and use the Penman-Monteith formula to calculate the reference evapotranspiration of the target broccoli. The Penman-Monteith formula is as follows:

[0126]

[0127] where, Δ is the slope of saturation vapor pressure, R n is net radiation, G is soil heat flux (usually negligible and set to 0 for short-term calculations), γ is the psychrometric constant, T is the daily average temperature, u2 is the wind speed at 2 meters height, e s is saturation vapor pressure, e a is actual vapor pressure.

[0128] The slope of saturation vapor pressure refers to the change in saturation vapor pressure when the temperature changes by 1 degree, and it can be calculated based on temperature and saturation vapor pressure. The daily average temperature can be calculated based on temperature and the cycle time of the standard growth period.

[0129] Next, determine the crop coefficient of the target broccoli according to the cycle type. The crop coefficient can reflect the difference in water requirements of broccoli at different growth stages. You can directly access the FAO (Food and Agriculture Organization of the United Nations) database or the official website of the relevant agricultural department to query the corresponding crop coefficient according to the cycle type of the target broccoli. For example, the crop coefficient during the emergence period of tomatoes is 0.4. Multiply the reference evapotranspiration by the crop coefficient to obtain the theoretical water requirement of the target broccoli. Next, extract the soil water content collected by the sensor during the standard growth cycle as the standard water content. Correct the theoretical water requirement of broccoli according to the standard water content, soil water absorption rate, and soil water holding rate. If any one of the standard water content, soil water absorption rate, and soil water holding rate is less than the corresponding preset threshold (for example, the soil water absorption rate is less than the preset water absorption threshold), the theoretical water requirement of broccoli can be appropriately increased to obtain the increased actual water requirement of broccoli. The increase range cannot exceed a fixed percentage of the theoretical water requirement of broccoli (for example, it cannot exceed 10% of the theoretical water requirement of broccoli) to prevent over-irrigation. If the standard water content, soil water absorption rate, and soil water holding rate are all greater than or equal to the corresponding preset thresholds, there is no need to adjust the theoretical water requirement of broccoli, and the theoretical water requirement of broccoli can be directly used as the actual water requirement of broccoli. Then, calculate the daily average irrigation amount of the target broccoli by combining the total precipitation and the actual water requirement of broccoli. Subtract the total precipitation from the actual water requirement of broccoli and divide by the cycle time to obtain the daily average irrigation amount of the target broccoli. Through this method, precipitation can be effectively utilized to complete the intelligent irrigation task, thereby saving water resources. Finally, integrate the alternate irrigation cycle and the daily average irrigation amount within the same standard growth cycle to obtain the alternate irrigation strategy for the target vegetable field, and generate irrigation instructions according to the alternate irrigation strategy. The alternate irrigation device executes the intelligent irrigation task according to the irrigation instructions.

[0130] In one embodiment, the method further includes:

[0131] After completing the intelligent irrigation task within any standard growth cycle, calculate the irrigation qualification degree within the standard growth cycle using the irrigation qualification formula;

[0132] If the irrigation qualification degree is within the preset qualification threshold range, it is determined that the intelligent irrigation task is qualified;

[0133] If the irrigation qualification degree is not within the qualification threshold range, it is determined that the intelligent irrigation task is unqualified;

[0134] If the intelligent irrigation task is qualified, continue to complete the intelligent irrigation task within the next standard growth cycle;

[0135] If the intelligent irrigation task is unqualified, output a strategy correction factor according to the irrigation qualification degree, and use the strategy correction factor to correct the alternate irrigation strategy for the next standard growth cycle. The irrigation qualification formula includes:

[0136]

[0137] Among them, S x is the actual irrigation amount of the target broccoli, S y is the average daily irrigation amount of the target broccoli, γ is the irrigation correction factor, W l and W r are respectively the average daily soil water contents on the left and right sides of the target broccoli, is the preset water content difference threshold, N is the number of soil sensors, δ is the preset fixed integer parameter, K is the pre-acquired soil hydraulic conductivity, T is the average daily irrigation duration, and F is the preset air temperature attenuation factor.

[0138] In this embodiment, when the intelligent irrigation task in each standard growth cycle is completed, the irrigation qualification degree in the standard growth cycle is immediately calculated using the irrigation qualification degree formula. The actual irrigation amount in the irrigation qualification degree formula is the average daily actual irrigation amount calculated based on the daily irrigation amount and the cycle time. The daily irrigation amount can be detected by sensors (such as a flow meter and a pressure sensor) preset in the alternative irrigation device. The irrigation correction factor is set based on the total actual precipitation. The total actual precipitation is the precipitation obtained at the end of the standard growth cycle. Compared with the previously obtained total precipitation, the total actual precipitation obtained at this time is more accurate because unexpected sudden precipitation (such as artificial rainfall) may have occurred during the standard growth cycle. Since unexpected sudden precipitation or sudden drought is not considered during actual irrigation, an irrigation correction factor needs to be introduced when calculating the irrigation qualification degree. The magnitude of the irrigation correction factor depends on the difference between the total actual precipitation and the total precipitation. For example, if the calculated difference is within the difference range (such as [0.3, -0.3]), the irrigation correction factor can be 1. If the calculated difference is not within the difference range and is greater than the preset first difference threshold (such as 0.3), the irrigation correction factor is 1.2. If the difference is not within the difference range and is less than the preset second difference threshold (such as -0.3), the irrigation correction factor is 0.8. The first difference threshold is greater than the second difference threshold. The average daily soil water content is collected in real time by soil sensors preset in the target vegetable field. The number of soil sensors and the fixed integer parameter are used to introduce a correction factor for sensor density. The closer the number of soil sensors is to the fixed integer parameter, the more reliable the monitored data, and the closer the correction factor for sensor density is to 1, and the smaller the impact on the irrigation qualification degree. The water content difference threshold is generally set according to the soil type. For example, for sandy soil, it is 10% of the actual irrigation amount, and for clay soil, it is 6% of the actual irrigation amount. The soil hydraulic conductivity depends on the soil's infiltration ability and aquifer thickness and can be queried on the website of the relevant agricultural department according to the soil type. For example, for sandy soil, it is 20 mm / h, and for clay soil, it is 8 mm / h. The air temperature attenuation factor is set according to the average daily air temperature in the standard growth cycle. For example, when the average daily air temperature is less than or equal to 30 degrees Celsius, the air temperature attenuation factor is 1. When the average daily air temperature u is greater than 30 degrees Celsius, the air temperature attenuation factor is 1 - 0.05(u - 30). When u = 35 degrees Celsius, the air temperature attenuation factor is 0.925.

[0139] Calculate the irrigation compliance within the standard growth period. If the irrigation compliance is within the preset compliance threshold range, such as [1.2, 0.8], it indicates that the intelligent irrigation task is qualified, and the irrigation task for the next standard growth period can be directly executed. If the irrigation compliance is not within the compliance threshold range, it is determined that the intelligent irrigation task is unqualified. If the irrigation compliance is greater than the preset first compliance threshold (such as 1.2), it indicates that there is over-irrigation within the standard growth period. If the irrigation compliance is less than the preset second compliance threshold (such as 0.8), it indicates that there is insufficient irrigation within the standard growth period. When over-irrigation or insufficient irrigation occurs in the previous standard growth period, the actual water requirement of the broccoli needs to be adjusted in its next standard growth period. When over-irrigation occurs, the actual water requirement of the broccoli needs to be reduced. When insufficient irrigation occurs, the actual water requirement of the broccoli needs to be increased to ensure the normal growth of the target broccoli while reducing water resource waste.

[0140] The above method further includes the following steps:

[0141] When the number of unqualified intelligent irrigation tasks is greater than the preset number threshold, collect the irrigation water flow through the irrigation sensor preset in the alternate irrigation device;

[0142] Mark the irrigation sensor with an irrigation water flow greater than the preset first flow threshold or less than the preset second flow threshold as an abnormal irrigation sensor;

[0143] Obtain the irrigation position information of the abnormal irrigation sensor;

[0144] Screen out all adjacent soil sensors of the abnormal irrigation sensor according to the irrigation position information;

[0145] For any adjacent soil sensor, retrieve the soil moisture content collected by the adjacent soil sensor as the adjacent moisture content;

[0146] Mark the adjacent soil sensor with an adjacent moisture content higher than the preset first moisture content threshold or lower than the preset second moisture content threshold as an abnormal soil sensor;

[0147] Obtain the soil position information of the abnormal soil sensor;

[0148] Locate the abnormal position information of the alternate irrigation device by combining the irrigation position information and the soil position information.

[0149] There is also a situation where when consecutive multiple (which can be two) standard growth cycles all show unqualified intelligent irrigation tasks, and they are all over-irrigation or all under-irrigation, the irrigation water flow is collected through the irrigation sensor preset in the alternative irrigation device. The irrigation sensor can be a flow meter. When the irrigation water flow is greater than the preset first flow threshold or less than the preset second flow threshold, and the first flow threshold is greater than the second flow threshold, the corresponding irrigation sensor is immediately marked as an abnormal irrigation sensor, and the sensor number of the abnormal irrigation sensor is immediately uploaded. According to the sensor number, the irrigation position information of the abnormal irrigation sensor with the corresponding number can be directly queried from the pre-constructed database. According to the irrigation position information, the soil water content collected by all adjacent soil sensors within a preset range around the irrigation sensor is retrieved, and an abnormal analysis is performed on all the soil water contents. According to the abnormal analysis result, the soil sensors with abnormalities are marked as abnormal soil sensors, and the position information of the abnormal soil sensors, that is, the soil position information, is obtained. Combining the irrigation position information and the soil position information can locate the abnormal position information of the alternative irrigation device, that is, the position where the alternative irrigation device has an abnormality is located between the abnormal soil sensor and the abnormal irrigation sensor. Finally, the abnormal position information is uploaded to the alternative irrigation management system to remind the relevant staff to check for faults in a timely manner. Additionally, the above-mentioned abnormal analysis refers to judging whether there is data with a water content higher than the preset first water content threshold or lower than the preset second water content threshold among all the soil water contents collected by the soil sensors within any preset range, that is, the adjacent water contents, and the first water content threshold is greater than the second water content threshold. If there is, the adjacent soil sensor is marked as an abnormal soil sensor. This is because when the irrigation device is damaged, its actual irrigation amount is higher than the daily average irrigation amount, so the soil water content will be higher than the first water content threshold. When the irrigation device is blocked, its actual irrigation amount is lower than the daily average irrigation amount, so the soil water content will be lower than the second water content threshold. Through this method, it is possible to quickly and conveniently locate the corresponding abnormal position information when there is a blockage or damage in the alternative irrigation device, and remind the relevant staff to conduct abnormal inspections in a timely manner, preventing water resource waste while also preventing the target broccoli from having a reduced yield due to under-irrigation.

[0150] The present application also discloses a machine-readable storage medium, characterized in that instructions are stored on the machine-readable storage medium, and the instructions are used to cause the machine to execute the method for intelligent control of the broccoli root zone alternative irrigation device according to any one of the above.

[0151] The present application also discloses an intelligent control system for a broccoli root zone alternative irrigation device, characterized by including:

[0152] A memory configured to store instructions; and

[0153] A processor configured to call instructions from a memory and capable of implementing a method for intelligent control of the broccoli root zone alternate irrigation device according to any one of the above.

[0154] Among them, the processor may adopt a central processing unit (CPU). Of course, according to the actual usage situation, other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), off-the-shelf programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. may also be adopted. The general-purpose processor may adopt a microprocessor or any conventional processor, etc. The present application does not make any restrictions in this regard.

[0155] Among them, the memory may be an internal storage unit of a computer device, for example, the hard disk or memory of a computer device, or may also be an external storage device of a computer device, for example, a plug-in hard disk, a smart media card (SMC), a secure digital card (SD), or a flash memory card (FC) equipped on the computer device, etc. Moreover, the memory may also be a combination of the internal storage unit and the external storage device of the computer device. The memory is used to store computer programs and other programs and data required by the computer device. The memory may also be used to temporarily store data that has been output or is to be output. The present application does not make any restrictions in this regard.

[0156] The embodiment of the present application also provides a machine-readable storage medium, on which instructions are stored, and the instructions are used to cause the machine to execute the method for intelligent control of the above-mentioned broccoli root zone alternate irrigation device.

[0157] Those skilled in the art should understand that the embodiments of the present application may be provided as a method, a system, or a computer program product. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0158] The present application is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams can be implemented by computer program instructions, and the combination of the flows and / or blocks in the flowcharts and / or block diagrams can also be implemented. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate for implementation in the process Figure 1 one process or multiple processes and / or blocksFigure 1 means for the functions specified in one or more boxes.

[0159] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory produce a manufacture including an instruction means that implements in the process Figure 1 one process or more processes and / or boxes Figure 1 the functions specified in one or more boxes.

[0160] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operation steps are executed on the computer or other programmable device to produce a computer-implemented process, so that the instructions executed on the computer or other programmable device provide for implementing in the process Figure 1 one process or more processes and / or boxes Figure 1 the steps of the functions specified in one or more boxes.

[0161] In a typical configuration, a computing device includes one or more processors (CPUs), an input / output interface, a network interface, and memory.

[0162] The memory may include non-permanent memory in the form of computer-readable media, random access memory (RAM) and / or non-volatile memory such as read-only memory (ROM) or flash memory (flash RAM). The memory is an example of computer-readable media.

[0163] Computer-readable media includes permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology for information storage. The information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassette tapes, magnetic disk storage or other magnetic storage devices, or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable media does not include transitory computer-readable media such as modulated data signals and carrier waves.

[0164] It should also be noted that the term "including", "comprising" or any other variant thereof is intended to cover non-exclusive inclusion, such that a process, method, commodity or device comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, commodity or device. Without further limitation, an element defined by the statement "comprising an..." does not exclude the presence of additional identical elements in the process, method, commodity or device comprising the element.

[0165] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, various changes and modifications can be made to the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present application shall be included within the scope of the claims of the present application.

Claims

1. An intelligent control method for an alternative irrigation device in the root zone of broccoli, characterized in that, The method includes the following steps: Obtain the broccoli-related information of the target vegetable field; Input the broccoli-related information into a pre-constructed initial cycle prediction model, and output all the initial growth cycles of the target broccoli in the target vegetable field through the initial cycle prediction model; When the initial growth cycle is reached for the first time, collect the broccoli remote sensing information of the target vegetable field through a remote sensing drone; Integrate the broccoli remote sensing information and the broccoli-related information to assimilate the initial cycle prediction model, obtain a standard cycle prediction model, and correct the initial growth cycle based on the standard cycle prediction model to obtain a standard growth cycle; When any standard growth cycle is reached, generate an alternative irrigation strategy for the target vegetable field based on the standard growth cycle, and perform intelligent control on the alternative irrigation device preset in the target vegetable field according to the alternative irrigation strategy.

2. The method according to claim 1, wherein The broccoli-related information includes the first meteorological information of the location of the target vegetable field, the broccoli information in the target vegetable field, and the soil information of the target vegetable field. The soil information includes the pre-obtained soil water absorption rate, soil water holding rate, and soil type, as well as the soil water content collected in real time through the soil sensors preset in the target vegetable field.

3. The method according to claim 2, characterized in that The step of integrating the broccoli remote sensing information and the broccoli-related information to assimilate the initial cycle prediction model, obtain a standard cycle prediction model, and correct the initial growth cycle based on the standard cycle prediction model to obtain a standard growth cycle includes the following steps: Preprocess the broccoli remote sensing information; Invert the preprocessed broccoli remote sensing information using an inversion algorithm, and extract the vegetable field remote sensing parameters in the target vegetable field according to the inversion results. The vegetable field remote sensing parameters include the broccoli index and the vegetable field water content; Use a data fusion algorithm to complete the spatial fusion of the vegetable field water content and the soil water content to obtain the fusion water content of the target vegetable field; Integrate the broccoli index and the fusion water content to assimilate the initial cycle prediction model to obtain a standard cycle prediction model; Input the first meteorological information, the broccoli information, and the soil information during the initial growth cycle into the standard cycle prediction model, and output the standard growth cycle of the target broccoli through the standard cycle prediction model.

4. The method according to claim 3, characterized in that, The step of using a data fusion algorithm to complete the spatial fusion of the vegetable field water content and the soil water content to obtain the fusion water content of the target vegetable field includes the following steps: Take the soil water content collected during the initial growth cycle as the target water content; Standardize the vegetable field water content and the target water content; Use the correlation coefficient formula to calculate the data correlation between the standardized vegetable field water content and the target water content; Use a preset variogram to analyze the spatial aggregation of the standardized vegetable field water content and the target water content; Integrate the spatial aggregation and the data correlation and use the Kriging interpolation method to interpolate and fuse the vegetable field water content and the target water content to obtain the fusion water content of the target vegetable field during the initial growth cycle.

5. The method according to claim 3, wherein The step of integrating the broccoli index and the fusion water content to assimilate the initial cycle prediction model to obtain a standard cycle prediction model includes the following steps: Align the spatio-temporal scales between the broccoli index and the fusion water content; Normalize the broccoli index and the fusion water content after spatio-temporal scale alignment to obtain a standard index and a standard water content respectively; Define an error covariance matrix by combining the standard index and the standard water content; Generate an initial state set of the initial cycle prediction model based on the broccoli-related information; Run the initial cycle prediction model according to the initial state set and the ensemble Kalman filter algorithm to obtain a predicted state set; Calculate the Kalman gain by combining the predicted state set and the error covariance matrix; Adjust the model parameters of the initial cycle prediction model by combining the Kalman gain, the standard exponent, and the standard water content to obtain the standard cycle prediction model.

6. The method according to claim 2, characterized in that, The alternate irrigation strategy for the target vegetable field based on the standard growth cycle includes the following steps: Generate an alternate irrigation cycle for the target vegetable field according to the cycle type and cycle time in the standard growth cycle; Re-obtain the second meteorological information of the location of the target vegetable field during the standard growth cycle, where the second meteorological information includes meteorological parameters and total precipitation; Calculate the reference evapotranspiration of the target broccoli according to the meteorological parameters; Determine the crop coefficient of the target broccoli according to the cycle type; Calculate the theoretical water requirement of the target broccoli by combining the reference evapotranspiration and the crop coefficient; Use the soil water content collected during the standard growth cycle as the standard water content; Correct the theoretical water requirement of broccoli using the soil water absorption rate, soil water holding rate, and standard water content to obtain the actual water requirement of broccoli; Calculate the daily average irrigation amount of the target broccoli by combining the total precipitation and the actual water requirement of broccoli; Integrate the alternate irrigation cycle and the daily average irrigation amount to obtain the alternate irrigation strategy for the target vegetable field.

7. The method according to claim 1, wherein The method further includes: After completing the intelligent irrigation task within any standard growth cycle, calculate the irrigation qualification degree within the standard growth cycle using the irrigation qualification formula; If the irrigation qualification degree is within the preset qualification degree threshold interval, determine that the intelligent irrigation task is qualified; If the irrigation qualification degree is not within the qualification degree threshold interval, determine that the intelligent irrigation task is unqualified; If the intelligent irrigation task is qualified, continue to complete the intelligent irrigation task within the next standard growth cycle; If the intelligent irrigation task is unqualified, output a strategy correction factor according to the irrigation qualification degree, and use the strategy correction factor to correct the alternate irrigation strategy for the next standard growth cycle.

8. The method according to claim 7, wherein The irrigation qualification formula includes: Among them, S x is the actual irrigation amount of the target broccoli, S y is the daily average irrigation amount of the target broccoli, γ is the irrigation correction factor, W1 and W r are respectively the daily average soil water contents on the left and right sides of the target broccoli, is the preset water content difference threshold, N is the number of soil sensors, δ is the preset fixed integer parameter, K is the pre-obtained soil hydraulic conductivity, T is the daily average irrigation duration, and F is the preset air temperature attenuation factor.

9. A machine-readable storage medium, characterized in that, Instructions are stored on the machine-readable storage medium, and the instructions are used to cause the machine to execute the method for intelligent control of the alternate irrigation device for broccoli roots according to any one of claims 1 to 8.

10. An intelligent control system for an alternative irrigation device in the root zone of broccoli, characterized in that, Including: A memory configured to store instructions; And A processor configured to call instructions from the memory and capable of implementing the method for intelligent control of the alternate irrigation device for broccoli roots according to any one of claims 1 to 8 when executing the instructions.

Images