AIoT water and fertilizer integrated regulation method and system based on crop growth dynamic model

Abstract

The application discloses an AIoT water and fertilizer integrated regulation method and system based on a crop growth dynamic model, relates to the technical field of water and fertilizer regulation, and first acquires multi-dimensional environmental parameters and soil characteristics by using an AIoT sensing network, and deeply analyzes the restriction mechanism of soil moisture on nutrient availability based on a crop growth dynamic model, and quantifies the nutrient availability coefficient. Then, the coefficient is input into a water and fertilizer coupling prediction model, the actual nutrient absorption and water evaporation of crops in a specific water and fertilizer interactive environment are dynamically deduced, and the digital representation of physiological needs is realized. Furthermore, combined with future weather information and cost constraints, a multi-objective collaborative optimization is carried out by using a rolling time domain optimization algorithm, an optimal regulation strategy considering growth needs and economic benefits is generated, and execution is driven. In this way, quantitative and accurate regulation of the synergistic effect of water and fertilizer can be realized, so that the resource utilization efficiency and the economic benefit of agricultural production in the whole growth cycle of crops can be greatly improved.

Description

Technical Field

[0001] This application relates to the field of water and fertilizer regulation technology, and more specifically, to an AIoT integrated water and fertilizer regulation method and system based on a crop growth dynamic model. Background Technology

[0002] With the rapid development of global smart agriculture technology and the widespread application of the Internet of Things (AIoT), precision agriculture has become a core driving force for improving crop yield and resource utilization. Water and fertilizer, as the two most critical environmental factors affecting crop growth, directly determine the economic benefits and ecological sustainability of agricultural production through their scientific management. Traditional extensive irrigation and fertilization methods in agriculture not only cause enormous waste of water resources and environmental problems such as soil compaction and salinization, but also fail to meet the demands of modern crops for refined control of their growth environment. To overcome this bottleneck, integrated water and fertilizer management technology, based on crop growth mechanisms and data-driven approaches, has emerged. This type of technology aims to dynamically match crop growth needs by sensing environmental parameters in real time and utilizing the efficient computing power of AIoT systems, thereby achieving the digital and intelligent transformation of agricultural production.

[0003] However, while some existing technologies have attempted to incorporate dynamic crop growth models to aid decision-making, they still face significant technical challenges. Most current mainstream control schemes employ a separate water and fertilizer management strategy, setting irrigation thresholds based solely on soil moisture and fertilization plans based solely on the crop growth cycle. This fragmented management model severely overlooks the strong coupling between water and fertilizer in soil transport and root absorption. In fact, water is not only a medium for crop physiological activities but also a carrier for nutrient dissolution and transport; the interaction of "water regulating fertilizer and fertilizer promoting water" exhibits highly nonlinear characteristics. Current technologies lack the ability to quantitatively model this synergistic effect, making it difficult to accurately quantify the specific impact of current water conditions on nutrient availability, resulting in an inability to accurately predict actual nutrient uptake and evapotranspiration under specific environmental and crop conditions. Furthermore, static control strategies relying solely on empirical values ​​cannot incorporate future weather changes and water and fertilizer costs for multi-objective optimization, causing the final irrigation and fertilization instructions to often deviate from the optimal trajectory of crop growth, making it difficult to achieve the economically optimal solution for resource input while ensuring crop yield.

[0004] Therefore, there is an urgent need for an optimized AIoT-based water and fertilizer integration regulation method and system based on crop growth dynamic models. Summary of the Invention

[0005] This application is made in order to solve the above-mentioned technical problems.

[0006] According to one aspect of this application, an AIoT-based water and fertilizer integration regulation method based on a crop growth dynamic model is provided, which includes: acquiring environmental data and soil hydraulic parameters, wherein the environmental data includes soil moisture content, soil electrical conductivity, air temperature, air humidity and photosynthetically active radiation, and the soil hydraulic parameters include soil field capacity, wilting point humidity and saturated water content.

[0007] Using a crop growth dynamics model, soil moisture content and soil hydraulic parameters are processed to obtain the nutrient availability coefficient at the current moment.

[0008] By using the nutrient availability coefficient at the current moment, the crop status at the previous moment, and environmental data, a coupled model is used to predict crop nutrient demand and water evapotranspiration to obtain the actual nutrient uptake and crop evapotranspiration at the current moment.

[0009] Based on preset future nutrient targets and soil moisture targets, a rolling time-domain optimized water and fertilizer synergistic decision is made based on the actual nutrient uptake at the current moment, the crop evapotranspiration at the current moment, the weather forecast for the next 24 hours, and the unit price of water and fertilizer to obtain the optimal irrigation amount and the optimal fertilizer amount at the current moment.

[0010] Transform the optimal irrigation amount and the optimal fertilization amount at the current moment into actuator instructions.

[0011] According to another aspect of this application, an AIoT integrated water and fertilizer regulation system based on a crop growth dynamic model is provided, which includes: an environmental sensing module for acquiring environmental data and soil hydraulic parameters. The environmental data includes soil moisture content, soil electrical conductivity, air temperature, air humidity and photosynthetically active radiation. The soil hydraulic parameters include soil field capacity, wilting point humidity and saturated water content.

[0012] The nutrient assessment module is used to process soil moisture content and soil hydraulic parameters using a crop growth dynamic model to obtain the nutrient availability coefficient at the current moment.

[0013] The demand forecasting module is used to predict crop nutrient demand and water evapotranspiration based on a coupled model, using the nutrient availability coefficient at the current moment, the crop status at the previous moment, and environmental data, in order to obtain the actual nutrient uptake and crop evapotranspiration at the current moment.

[0014] The water and fertilizer synergistic decision-making module is used to make water and fertilizer synergistic decisions based on preset future nutrient targets and soil moisture targets, and based on the actual nutrient absorption at the current moment, the crop evapotranspiration at the current moment, the weather forecast for the next 24 hours, and the unit price of water and fertilizer, in order to obtain the optimal irrigation amount and the optimal fertilizer amount at the current moment.

[0015] The instruction generation module is used to convert the optimal irrigation amount and the optimal fertilization amount at the current moment into actuator instructions.

[0016] Compared with existing technologies, this application provides an AIoT-based integrated water and fertilizer regulation method and system based on a crop growth dynamic model. First, it utilizes an AIoT sensing network to acquire multi-dimensional environmental parameters and soil characteristics. Then, based on the crop growth dynamic model, it deeply analyzes the constraint mechanism of soil moisture on nutrient availability and quantifies the nutrient availability coefficient. Subsequently, this coefficient is input into a water and fertilizer coupling prediction model to dynamically predict the actual nutrient uptake and water evapotranspiration of crops under specific water and fertilizer interaction environments, achieving a digital representation of physiological needs. Furthermore, combining future meteorological information and cost constraints, a rolling time-domain optimization algorithm is used for multi-objective collaborative optimization to generate an optimal regulation strategy that balances growth needs and economic benefits, and drives its execution. This enables quantitative and precise regulation of the synergistic effect of water and fertilizer, thereby significantly improving resource utilization efficiency and agricultural production economic benefits throughout the entire crop growth cycle. Attached Figure Description

[0017] The above and other objects, features, and advantages of this application will become more apparent from the more detailed description of the embodiments of this application in conjunction with the accompanying drawings. The drawings are provided to further illustrate the embodiments of this application and form part of the specification. They are used together with the embodiments of this application to explain this application and do not constitute a limitation thereof. In the drawings, the same reference numerals generally represent the same components or steps.

[0018] Figure 1 This is a flowchart of an AIoT-based integrated water and fertilizer regulation method according to an embodiment of this application.

[0019] Figure 2 This is a data flow diagram of the AIoT-based integrated water and fertilizer regulation method according to an embodiment of this application.

[0020] Figure 3 This is a flowchart of sub-step S2 of the AIoT water and fertilizer integration regulation method based on a crop growth dynamic model according to an embodiment of this application.

[0021] Figure 4 This is a flowchart of sub-step S3 of the AIoT water and fertilizer integration regulation method based on a crop growth dynamic model according to an embodiment of this application.

[0022] Figure 5 This is a block diagram of an AIoT integrated water and fertilizer control system based on a crop growth dynamic model, according to an embodiment of this application. Detailed Implementation

[0023] Embodiments of this disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this disclosure. It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure.

[0024] To address the problems mentioned above, this application proposes an AIoT-based integrated water and fertilizer regulation method based on a crop growth dynamic model. Figure 1 This is a flowchart of an AIoT-based integrated water and fertilizer regulation method according to an embodiment of this application. Figure 2 This is a data flow diagram of an AIoT-based integrated water and fertilizer management method according to an embodiment of this application. Figure 1 and Figure 2 As shown, the AIoT-based integrated water and fertilizer regulation method based on a crop growth dynamic model includes the following steps: S1, acquiring environmental data and soil hydraulic parameters. The environmental data includes soil moisture content, soil electrical conductivity, air temperature, air humidity, and photosynthetically active radiation. The soil hydraulic parameters include field capacity, wilting point humidity, and saturated water content. S2, using a crop growth dynamic model, processing the soil moisture content and soil hydraulic parameters to obtain the nutrient availability coefficient at the current moment. S3, processing the nutrient availability coefficient at the current moment, the crop state at the previous moment, and the environmental data... S4. Based on the preset future nutrient target and soil moisture target, a rolling time-domain optimized water and fertilizer synergistic decision is made on the actual nutrient absorption and crop evapotranspiration at the current moment. S5. The optimal irrigation amount and optimal fertilizer application amount at the current moment are converted into actuator instructions.

[0025] In the aforementioned AIoT-based integrated water and fertilizer management method based on a crop growth dynamic model, step S1 involves acquiring environmental data and soil hydraulic parameters. Environmental data includes soil moisture content, soil electrical conductivity, air temperature, air humidity, and photosynthetically active radiation. Soil hydraulic parameters include field capacity, wilting point humidity, and saturation water content. It should be understood that since crop growth is directly regulated by environmental conditions and soil moisture status, the precise operation of the integrated water and fertilizer system must be based on the actual needs of the crop. Environmental data and soil hydraulic parameters are key indicators reflecting the crop growth microenvironment and soil water supply capacity. A lack of these data will lead to water and fertilizer management deviating from actual needs, causing problems of oversupply or undersupply. Therefore, this application acquires environmental data and soil hydraulic parameters by deploying sensing devices and combining them with specialized detection methods. This provides accurate input data for the crop growth dynamic model, achieving dynamic matching of water and fertilizer supply with crop physiological needs and soil water retention capacity. This ensures the scientific and timely nature of water and fertilizer regulation decisions, effectively avoids resource waste caused by blind irrigation and fertilization, maintains the stability of soil physical and chemical properties, promotes crop root development and nutrient absorption, improves crop yield and quality, and lays a data foundation for the intelligent operation of the AIoT integrated water and fertilizer system.

[0026] Specifically, in one possible embodiment, step S1 is implemented as follows: Regarding environmental data acquisition, according to the field crop planting layout, soil moisture sensors and soil conductivity sensors are deployed at the soil depth corresponding to the root distribution layer. Meteorological monitoring equipment is installed in open areas of the field. This equipment integrates air temperature and humidity detection modules and a photosynthetically active radiation sensor. All sensors are connected to the AIoT control platform via wireless communication modules, capturing various environmental data in real time according to a set acquisition cycle and transmitting them to the data processing unit. Regarding soil hydraulic parameter acquisition, soil samples are first collected from different areas of the field. The collected samples are sent to the laboratory, where soil moisture characteristic curves are used to determine field water holding capacity and wilting point humidity. Simultaneously, undisturbed soil samples are collected using the ring cutter method, and the saturated soil moisture content is determined using the saturated weighing method. The measurement results are then entered into the system to form a soil hydraulic parameter database, providing basic parameter support for subsequent control decisions.

[0027] In the aforementioned AIoT-based integrated water and fertilizer regulation method based on a crop growth dynamic model, step S2 involves using the crop growth dynamic model to process soil moisture content and soil hydraulic parameters to obtain the nutrient availability coefficient at the current moment. It should be understood that the actual availability of nutrients in the soil is not solely determined by nutrient content, but closely depends on soil moisture content and soil hydraulic characteristics, such as field capacity and wilting point. Excessive soil moisture prevents nutrients from dissolving and migrating to the roots, while excessive moisture causes root hypoxia and inhibits absorption. Static nutrient data alone cannot accurately determine the true level of available nutrients for the crop. Therefore, this application further introduces a crop growth dynamic model, using soil moisture content and soil hydraulic parameters as core inputs for synergistic processing to quantify the actual availability of nutrients to the crop under the current soil conditions, thus obtaining the nutrient availability coefficient. This provides key quantitative indicators that align with the crop's physiological absorption conditions for subsequent integrated water and fertilizer decisions, ensuring that regulatory commands accurately match the crop's current nutrient utilization capacity and avoiding over- or under-fertilization due to misjudgment of nutrient availability.

[0028] In particular, in one specific embodiment, Figure 3 This is a flowchart of sub-step S2 of the AIoT-based integrated water and fertilizer regulation method based on a crop growth dynamic model according to an embodiment of this application. Figure 3 As shown, step S2 includes: S21, calculating the root zone water stress coefficient based on a piecewise function for soil moisture content and soil hydraulic parameters to obtain the water stress coefficient at the current moment; S22, calculating the nutrient availability coupling coefficient of salt stress combined with the water stress coefficient, soil electrical conductivity and fertilization parameters at the current moment to obtain the nutrient availability coefficient at the current moment.

[0029] Specifically, in step S21, the root zone water stress coefficient is calculated based on a piecewise function using soil moisture content and soil hydraulic parameters to obtain the water stress coefficient at the current moment. It should be understood that because the degree of water stress experienced by crop roots exhibits a piecewise characteristic with changes in soil moisture content—below the wilting point, the roots completely lose their absorption capacity (severe stress); between the wilting point and field capacity, the stress gradually eases with increasing moisture content; and between field capacity and saturated moisture content, the water is sufficient and suitable for crop absorption (no stress)—a single linear function cannot accurately characterize the stress differences in different moisture ranges, easily leading to significant deviations in subsequent nutrient availability calculations. Therefore, this application further uses a piecewise function, with soil moisture content and soil hydraulic parameters as inputs, to calculate the root zone water stress coefficient, thereby accurately quantifying the degree of inhibition of root physiological activity by current soil moisture conditions, and obtaining the water stress coefficient. In a specific example of this application, the piecewise function is expressed as:

[0030]

[0031] in, The water stress coefficient at the current moment. Soil moisture content, Humidity at the wilting point, This is the saturated water content. This refers to field holding capacity. In other words, it's determined by classifying soil moisture content. The 5 intervals ( , , , , This allows for precise quantification of the degree of water stress. The function in... The interval exhibits a monotonically increasing linear relationship (with a slope of 1 / ( This reflects that as soil moisture content increases from the wilting point to field capacity, water stress decreases linearly with increasing moisture content, consistent with the physiological pattern of gradual recovery of crop root water absorption activity; The interval exhibits a monotonically decreasing linear relationship (with a slope of -1 / ( When the water content exceeds the field capacity, soil pore oxygen deficiency leads to a linear decrease in root activity with increasing water content; or The value is set to 0 to achieve a hard constraint that completely inhibits root activity under extreme water stress, avoiding inaccurate regulation due to calculation deviations caused by extreme values. This avoids distortion in nutrient availability assessment caused by coarse quantification of water stress, providing accurate and reliable basic parameters for subsequent calculations of nutrient availability under salt stress, and ensuring the accuracy of the overall regulation model in characterizing the impact on soil moisture.

[0032] Specifically, step S22 involves calculating the nutrient availability coupling coefficient for salt stress by integrating the current water stress coefficient, soil electrical conductivity, and fertilization parameters to obtain the nutrient availability coefficient for the current moment. It should be understood that soil salinity (directly reflected by soil electrical conductivity) and water jointly constrain nutrient availability. High salinity environments increase soil solution osmotic pressure, leading to difficulties in root water and nutrient uptake even under suitable moisture conditions. Relying solely on the water stress coefficient cannot fully reflect the nutrient absorption limitations faced by crops, easily resulting in an overestimation of nutrient availability. Therefore, this application further integrates the current water stress coefficient, soil electrical conductivity, and fertilization parameters (optimal electrical conductivity, salt sensitivity coefficient) to construct a coupling model by introducing the salt stress effect, thereby obtaining a nutrient availability coupling coefficient that comprehensively reflects the dual stresses of water and salt. In a specific example of this application, step S22 includes: calculating the nutrient availability coupling coefficient for salt stress by integrating the current water stress coefficient, soil electrical conductivity, and fertilization parameters using the following formula:

[0033]

[0034] in, The water stress coefficient at the current moment. For soil electrical conductivity, The optimal soil electrical conductivity for the crop in the fertilization parameters. This refers to the salt sensitivity coefficient among fertilization parameters. It is the hyperbolic tangent function. This represents the nutrient availability coefficient at the current moment. In other words, it represents the water stress coefficient. Multiplying the salt stress penalty term achieves a mathematical fusion of the dual-stress effect. When When the salt stress penalty term is 1, the formula degenerates into: Mathematically, this represents "when salinity is within a suitable range, nutrient availability is determined solely by water stress"; when When, the hyperbolic tangent function is introduced. This function is an S-shaped monotonically increasing function (range [0,1)), which can measure the salinity deviation ( Smoothly transforming into a 0-1 penalty term: The closer The closer the penalty is to 0, the closer the nutrient availability is to the upper limit under water stress; Far exceeding At this point, the penalty term approaches 1, and nutrient availability approaches 0, mathematically accurately characterizing the "gradual inhibition of nutrient absorption by excessive salinity," avoiding the abrupt penalty changes caused by traditional step functions, and conforming to the gradual influence of soil salinity on crop root absorption. Simultaneously, The salt sensitivity coefficient, as a function parameter, needs to be determined during the model calibration phase before system deployment. For example, for salt-sensitive crops like lettuce... The value can be set between 0.15 and 0.25, resulting in a relatively steep function curve; however, for salt-tolerant crops such as cotton and barley, The value can be set between 0.05 and 0.1, resulting in a relatively flat function curve, thus achieving differentiated adaptation to the salt tolerance of different crops. This allows for a more realistic representation of the actual impact of the field soil environment on nutrient availability, avoiding deviations in regulatory instructions caused by ignoring salt stress, and ensuring that subsequent water and fertilizer decisions can simultaneously address the two key limiting factors of water and salt, thereby improving the scientific rigor and field applicability of the regulatory scheme.

[0035] In the aforementioned AIoT-based integrated water and fertilizer management method based on a crop growth dynamic model, step S3 involves using a coupled model to predict crop nutrient demand and water evapotranspiration based on the current nutrient availability coefficient, the previous crop state, and environmental data to obtain the current actual nutrient uptake and current crop evapotranspiration. In a specific example of this application, the previous crop state includes leaf area index, biomass, and developmental stage. It should be understood that the actual nutrient uptake of a crop is not solely determined by nutrient availability but is also closely related to the current growth stage of the crop (reflected by the previous state). For example, seedlings have low biomass and low nutrient demand, while vigorous growth stages have high leaf area index and a surge in demand. Simultaneously, water evapotranspiration is influenced by both environmental factors (temperature, humidity, radiation) and crop canopy characteristics (leaf area index), and prediction based on a single factor can lead to a disconnect between demand and reality. Therefore, this application further employs a coupled model to integrate the current nutrient availability coefficient, the previous crop state, and environmental data for prediction, thereby simultaneously obtaining the current actual nutrient uptake and crop evapotranspiration. This enables the coordinated quantification of crop physiological needs (nutrients) and environmental consumption (water), avoiding imbalances in water and fertilizer supply caused by fragmented nutrient and water predictions. It provides a unified demand benchmark for subsequent coordinated water and fertilizer decisions, ensuring that the control scheme matches both crop growth needs and environmental constraints.

[0036] In particular, in one specific embodiment, Figure 4 This is a flowchart of sub-step S3 of the AIoT-based integrated water and fertilizer regulation method according to an embodiment of this application. Figure 4 As shown, step S3 includes: S31, determining the actual nutrient uptake at the current moment based on the nutrient availability coefficient at the current moment, the crop status at the previous moment, photosynthetically active radiation and air temperature and crop parameters in the environmental data; S32, using the FAO Penman-Monteith model to process the leaf area index in the crop status at the previous moment, environmental data and site parameters to obtain the crop evapotranspiration at the current moment.

[0037] Specifically, step S31 determines the actual nutrient uptake at the current moment based on the nutrient availability coefficient at the current moment, the crop state at the previous moment, photosynthetically active radiation and air temperature from environmental data, and crop parameters. It should be understood that the actual nutrient uptake of crops is the result of the combined effects of "demand-driven and environmental constraints": the crop state at the previous moment (leaf area index, developmental stage) determines the basic scale of nutrient demand; photosynthetically active radiation and air temperature affect the rate of biomass growth through photosynthesis, thereby affecting the incremental nutrient demand; and the nutrient availability coefficient determines whether the demand can be actually met. The absence of any one of these factors will lead to inaccurate uptake predictions. Therefore, this application further integrates the above-mentioned multi-source parameters and determines the actual nutrient uptake through the synergistic calculation of physiological mechanisms and environmental constraints. In a specific example of this application, step S31 includes: determining the actual nutrient uptake at the current moment using the following formula:

[0038]

[0039]

[0040]

[0041] in, Photosynthetically active radiation, It is a natural constant. Extinction coefficient, Leaf area index, Light energy utilization efficiency is a crop parameter. Let be the temperature stress function for photosynthesis. This represents the potential daily biomass growth. The maximum nitrogen concentration associated with the developmental stage. For potential nutrient requirements, The nutrient availability coefficient, This represents the actual nutrient absorption at the current moment. That is, through multiplication and integration of photosynthetically active radiation (...). ), canopy light cutoff term exhibiting exponential saturation characteristics ( As the leaf area index increases, it approaches 1 from 0, which aligns with the canopy light interception pattern; light energy utilization efficiency (LUE) and a temperature stress term with a value of [0,1]. To avoid calculation distortion under extreme temperatures, environmental and crop conditions are converted into biomass potential values. Then, based on the biomass potential, the value is multiplied by the developmental stage (…). Dynamically adjusted maximum nitrogen concentration (e.g., high during the seedling stage, low during maturity), to realize the transformation of biomass growth into nutrient demand, providing a clear demand benchmark. Finally, through the nutrient availability coefficient with a value of [0,1] ( Environmental constraints are used to adjust potential demand. When the ratio is 1, the crop can fully absorb the potential demand. When the actual absorption rate is less than 1, the actual absorption rate decreases linearly, accurately quantifying the dual stress effects of water and salt. This allows for precise quantification of the amount of nutrients crops can truly absorb under current conditions, avoiding resource waste and soil pollution caused by over-fertilization, or crop growth stunts due to insufficient fertilization, thus providing a precise quantitative basis for fertilization decisions.

[0042] Preferably, currently based on potential nutrient requirements and nutrient availability coefficient When calculating actual nutrient absorption, it is assumed that water and fertilizer conditions permit (i.e.) This approach assumes that crops can absorb nutrients from the soil in proportion to their potential needs, without considering the limitation of the absolute concentration of nutrients in the soil on the absorption rate. This is clearly imprecise in agriculture and biology. Therefore, it is necessary to calculate the actual nutrient absorption. The model is a nonlinear process simultaneously constrained by crop demand, soil supply, and environmental regulation.

[0043] Specifically, two mechanisms need to be introduced: a dynamic model of the soil available nutrient pool, i.e., a simplified dynamic model of soil nutrients, to track changes in the concentration of available soil nutrients caused by irrigation, fertilization, absorption, and leaching; and the Michaelis-Menten equation to describe the actual rate at which crop roots absorb nutrients from the soil. This yields a classic enzyme-catalyzed reaction kinetic model that can simulate the physiological phenomenon where the absorption rate is proportional to the substrate (nutrient) concentration when it is low, and reaches saturation (maximum value) when the concentration is very high. Specifically, step S31 includes: updating the soil available nutrient concentration at the current moment based on the previous moment's soil available nutrient concentration, fertilization amount, actual nutrient absorption by the crop, and nutrient leaching loss; using the Michaelis-Menten equation and based on the current moment's soil available nutrient concentration, the crop's maximum nutrient absorption rate, and the Michaelis constant, calculating the nutrient absorption rate determined by soil supply capacity; and multiplying the nutrient absorption rate determined by soil supply capacity by the current moment's nutrient availability coefficient to obtain the actual nutrient absorption amount at the current moment.

[0044] That is, due to the nutrient availability coefficient While dimensionless discount coefficients can reflect the water-fertilizer coupling effect, they don't inherently include information on the absolute quantity of soil nutrients. This means that even if the actual nutrients in the soil are depleted, the model will still calculate a non-zero absorption rate if water conditions are favorable, which doesn't align with physical reality. In contrast, the Michaelis-Menten equation directly uses soil nutrient concentration as an input variable, linking the nutrient absorption rate directly to the amount of fertilizer in the soil. When soil nutrient concentration is low, even with high potential crop demand and favorable water and fertilizer conditions, actual absorption will be limited.

[0045] Additionally, if the demand for crops ( This is considered the driving force, while the soil's supply capacity is hidden. In the case of [the crop], the multiplicative relationship between the two is relatively weak. However, if the crop's potential absorption capacity (maximum absorption rate) is considered... ) and the soil's supply capacity (current nutrient concentration) As the core input of the Michaelis-Menten equation, crop demand no longer directly determines uptake; instead, a theoretical upper limit is set for uptake, while actual uptake is dynamically determined by soil supply within this limit. Therefore, the goal of fertilization decisions becomes maintaining soil nutrient concentration. Within the optimal range, the crop absorption rate can be ensured to be close to saturation without causing waste and environmental pollution due to excessive concentration, thus providing a more accurate and forward-looking control target for model predictive control.

[0046] First, based on the soil available nutrient concentration, fertilizer application rate, actual nutrient uptake by the crop, and nutrient leaching loss at the previous time step, the soil available nutrient concentration at the current time step is updated. That is, the actual nutrient uptake is calculated based on Michaelis-Menten dynamics and soil nutrient pool dynamics, obtaining the soil available nutrient concentration at time t-1 from the iterative calculation results of the previous time step. The amount of fertilizer applied at time t-1 is derived from the fertilization decision made by the control system at the previous time step. and the kinetic parameters of crop physiology. (Maximum nutrient uptake rate per unit root system or unit area of ​​land) and Michaelis constant Its absorption rate reaches The soil nutrient concentration at a given time reflects the root system's affinity for nutrients.

[0047] Before calculating uptake, update the current soil nutrient concentration based on the previous time step's fertilization, uptake, and leaching data:

[0048]

[0049] in, It is the concentration of available nutrients in the soil at the current time t (kg / m³) 3 ), It is the soil available nutrient concentration at the previous time t-1 (kg / m³) 3 ), This is the amount of fertilizer (kg) applied at the previous moment. It is the volume of soil in the crop root zone (m). 3 It can be dynamically adjusted according to the crop growth stage. It represents the amount of nutrients (kg) actually absorbed by the crop at the previous moment. It is a unit conversion factor that converts macroscopic absorption into concentration change, and This refers to the nutrient leaching loss (kg / m³) caused by excessive irrigation or other reasons at the previous moment. 3 This can be estimated by other sub-models or ignored in simplified models.

[0050] Then, using the Michaelis equation and based on the current soil available nutrient concentration, the crop's maximum nutrient uptake rate, and the Michaelis constant, the nutrient uptake rate determined by the soil's supply capacity is calculated:

[0051]

[0052] The nutrient uptake rate (kg / ha / day) is determined by the soil's supply capacity. This is the crop's maximum nutrient uptake rate (kg / ha / day), which can be correlated with the crop's potential growth requirements, for example... It can be The function or directly equal to This indicates that the crop's needs set an upper limit on its absorption capacity. It is the Michaelis constant, which is the absorption rate. Nutrient concentration at time (kg / m³) 3 ).

[0053] Thus, the final actual absorption is determined by the absorption rate, which is determined by the soil supply capacity. This nutrient absorption rate is then multiplied by the nutrient availability coefficient at the current moment (i.e., multiplied by...). The coefficient is used to obtain the actual nutrient absorption at the current moment. This process can be understood as the root system first determining whether there is sufficient nutrient in the soil (e.g., whether there is enough nutrient in the soil). ) determines the maximum amount it can absorb. ), and then the current moisture and salinity conditions ( This further reduces the efficiency of the absorption process.

[0054]

[0055] in, This is the final calculated actual nutrient absorption (kg / ha / day).

[0056] Thus, by incorporating state updates (soil nutrient pool), nonlinear dynamics (Michaelis equations), and multi-factor coupling, the model's predictive accuracy can be improved throughout the growing season, especially during periods when nutrients may become limiting factors (such as vigorous growth or improper fertilization). Simultaneously, because the model can more accurately identify the true timing and dosage of fertilization, avoiding over-fertilization when soil nutrients are abundant or under-fertilization when soil nutrients are deficient, water and fertilizer use efficiency can be optimized. Furthermore, by simulating the realistic response to fertilization events—that is, simulating the increase in soil nutrient concentration after fertilization, which in turn increases the crop absorption rate, followed by a gradual decrease in concentration due to absorption—this dynamic process makes the model more adaptable to external disturbances (such as incorrect fertilization), enhancing its robustness.

[0057] Specifically, step S32 uses a FAO Penman-Monteith model to process the leaf area index, environmental data, and site parameters from the previous crop state to obtain the crop evapotranspiration at the current time. It should be understood that crop evapotranspiration is a comprehensive result of "meteorological driving force - crop regulation - geographical influence": environmental data (temperature, humidity, radiation, wind speed) provides the energy and driving force for evapotranspiration; the leaf area index reflects the transpiration capacity of the crop canopy—the larger the leaf area, the stronger the transpiration; and site parameters (altitude, latitude) affect air pressure and solar radiation intensity, thus affecting the evapotranspiration rate. Traditional models that ignore any of these factors will lead to errors in evapotranspiration calculation. Therefore, this application further employs the FAO Penman-Monteith model to integrate the above parameters for processing to obtain crop evapotranspiration. In a specific example of this application, step S32 includes: processing the leaf area index, environmental data, and site parameters from the previous crop state using the following formula:

[0058]

[0059]

[0060] in, For crop coefficients, This represents the minimum crop coefficient during the early stages of crop growth. This represents the maximum crop coefficient when the crop canopy is fully covered. To adjust LAI The coefficient of influence This represents the leaf area index of the crop at the previous time step. This represents the actual crop evapotranspiration. This represents the crop evapotranspiration at the current moment. In other words, it is expressed through an exponential term. Achieve crop coefficient With leaf area index The gradual growth, when =0 (before crop emergence) = , Increase Gradually approaching Mathematically, it precisely depicts the process of crop canopy growth from sparse to dense, and transpiration capacity from weak to strong until saturation, avoiding the unsaturated canopy conditions caused by traditional linear functions. Growth too fast or too slow, As an adjustment coefficient, it can be mathematically changed by adjusting its value. Approaching The rate is adjusted to suit the canopy growth characteristics of different crops, such as broadleaf crops. Large value Rapid growth. Then, through the crop coefficient... Reference crop evapotranspiration (Standardized grassland evapotranspiration) undergoes crop-specific modification to achieve the transformation from general evapotranspiration to target crop evapotranspiration. When the evapotranspiration of the target crop is greater than that of the grassland, it corresponds to a high-transpiration crop (such as maize). When evapotranspiration is less than 1, it is less than that of grassland, corresponding to low-transpiration crops (such as wheat). This linear relationship is mathematically simple and physically clear, ensuring that evapotranspiration calculations conform to FAO international standards and accurately match the characteristics of the target crop. Together, they construct a quantitative link between crop canopy characteristics, evapotranspiration correction coefficient, and actual evapotranspiration, providing a mathematical basis for irrigation decisions that aligns with the actual water consumption needs of the crop. In this way, relying on the scientific nature of international standard models, accurate evapotranspiration quantification can be achieved, avoiding crop water shortage stress due to underestimation of evapotranspiration or over-irrigation due to overestimation, which could lead to root hypoxia and nutrient leaching. This provides a benchmark for irrigation decisions that conforms to actual field water requirements, ensuring a dynamic balance between water supply and crop consumption.

[0061] In the aforementioned AIoT-based integrated water and fertilizer regulation method based on a crop growth dynamic model, step S4 involves making a rolling time-domain optimized water and fertilizer synergistic decision based on preset future nutrient targets and soil moisture targets, considering the current actual nutrient uptake, current crop evapotranspiration, the weather forecast for the next 24 hours, and the unit prices of water and fertilizer, to obtain the optimal irrigation and fertilization amounts for the current moment. It should be understood that because water and fertilizer regulation needs to simultaneously address the dynamics of crop growth (requiring continuous approximation of future nutrient and water targets), environmental uncertainties (weather changes in the next 24 hours directly alter water consumption and nutrient loss risks), and economic constraints (the unit prices of water and fertilizer determine input costs), static decision-making based solely on current data can easily lead to deviations from targets or cost overruns in subsequent periods. Therefore, this application further integrates current physiological needs, weather forecasts, and cost parameters, guided by future targets, and conducts rolling time-domain optimized water and fertilizer synergistic decision-making to balance the relationship between demand, environment, and cost within a dynamic time domain. In this way, losses caused by ignoring future weather (such as unforeseen rainfall leading to excessive irrigation and nutrient leaching) or costs (such as blindly applying expensive fertilizers and increasing input) can be avoided. This ensures that the water and fertilizer supply at each stage not only conforms to the plant's full-cycle growth trajectory but also meets the principle of economic efficiency. At the same time, deviations can be corrected in real time through rolling updates in the time domain, improving the dynamic adaptability of regulation.

[0062] Specifically, in one possible embodiment, step S4 is implemented as follows: First, the prediction time domain is set to the next 24 hours, and it is discretized into multiple consecutive time steps, such as 6 time steps, each lasting 4 hours. The control variables of the system are defined as the irrigation amount and fertilizer amount at each time step, and the state variables of the system are defined as the actual nutrient absorption and soil moisture content predicted based on the aforementioned steps. Next, a cost function for multi-objective collaborative optimization is established. Specifically, the calculation logic of this cost function includes three parts: the first part is to calculate the square of the deviation between the predicted actual nutrient absorption at each future time step and the preset nutrient target value, and multiply it by a set nutrient weight coefficient; the second part is to calculate the square of the deviation between the predicted soil moisture content and the soil moisture target value, and multiply it by a set moisture weight coefficient; the third part is to calculate the total economic cost of the irrigation amount and fertilizer amount at each future time step multiplied by the corresponding water and fertilizer unit prices. The optimization objective is to find a set of control sequences that minimizes the sum of the above three parts throughout the entire prediction time domain. The nutrient weight coefficient and water weight coefficient can be numerically set according to the crop's current growth stage and its priority in fertilizer or water requirements. For example, the nutrient weight can be set higher than the water weight, specifically 0.6 and 0.4 respectively. Simultaneously, to ensure execution safety, physical constraints are set for the system, explicitly limiting the irrigation and fertilization amounts at each time step to no less than zero and no more than the maximum single execution capacity of the field irrigation and fertilization systems. Finally, at each control time, a nonlinear programming solver, such as the IPOPT solver, is invoked. Based on the measured state at the current time, the coupled model is used as a predictor to perform online optimization of the aforementioned cost function, obtaining the optimal irrigation and fertilization control sequence covering the future prediction time domain. Ultimately, only the values ​​corresponding to the first time step in this sequence are extracted to obtain the optimal irrigation and fertilization amounts for the current time as the actual output.

[0063] In the aforementioned AIoT-based integrated water and fertilizer management method based on crop growth dynamics models, step S5 converts the optimal irrigation and fertilization amounts at the current moment into actuator commands. It should be understood that since the optimal irrigation and fertilization amounts at the current moment are decision results represented by physical units (such as cubic meters or kilograms), while field actuators (irrigation solenoid valves, fertilizer pumps) can only recognize control signals such as on / off duration and frequency of action, there is a disconnect between the units and the format. Without conversion, the decision results cannot directly drive the equipment's actions, leading to the problem of "effective decision, ineffective execution." Therefore, this application further converts the optimal water and fertilizer amounts into commands that the actuators can recognize, thereby establishing a connection channel between the decision-making and execution layers. This ensures the accurate implementation of the optimal solution, avoiding execution deviations caused by unit mismatches (such as directly using cubic meters as duration commands) or format errors, such as insufficient irrigation affecting nutrient absorption or excessive fertilization causing soil salinization. Simultaneously, the execution status can be monitored in real time through command feedback, providing execution-end data support for parameter calibration in subsequent control cycles and ensuring the integrity of the control loop.

[0064] Specifically, in one possible embodiment, step S5 is implemented as follows: First, the basic parameters of the equipment are retrieved. The inner diameter of the irrigation system's pipes, rated working pressure (to calculate the actual flow rate), rated flow rate of the fertilizer pump, and fertilizer concentrate concentration (to calculate the amount of fertilizer applied per unit time) are obtained from the system equipment library, along with the actuator's communication protocol, such as LoRaWAN. Next, the execution duration is calculated. Based on the ratio of the optimal irrigation volume to the actual flow rate of the pipe, the single opening duration of the irrigation solenoid valve is obtained. For example, if the optimal irrigation volume is 0.5 cubic meters and the flow rate is 1 cubic meter per hour, the opening duration is 30 minutes. Based on the ratio of the optimal fertilizer application volume to the fertilizer pump's fertilizer application per unit time, the opening duration of the fertilizer pump is obtained. If the duration exceeds the equipment's continuous operation threshold, such as a maximum of 60 minutes, it is split into two 30-minute actions with a 5-minute interval. Then, instruction encoding is performed, converting the opening duration, device address, number of actions, and other information into a binary instruction frame conforming to the LoRaWAN protocol, including a frame header, data segment, and checksum. Finally, the command frame is sent to the field actuator controller through the gateway. The controller drives the solenoid valve and fertilizer pump to act according to the command. At the same time, the current and voltage signals of the actuator are collected in real time (to determine whether it is operating normally). The operating status is encoded and sent back to the control platform. If the status is abnormal, such as the current exceeding the limit, a stop command is immediately issued and an alarm is triggered.

[0065] In summary, the AIoT-based water and fertilizer integration regulation method based on a crop growth dynamic model, as described in this application, is elucidated. First, it utilizes an AIoT sensing network to acquire multi-dimensional environmental parameters and soil characteristics. Then, based on the crop growth dynamic model, it deeply analyzes the constraint mechanism of soil moisture on nutrient availability, quantifying the nutrient availability coefficient. Subsequently, this coefficient is input into a water and fertilizer coupling prediction model to dynamically predict the actual nutrient uptake and water evapotranspiration of crops under specific water and fertilizer interaction environments, achieving a digital representation of physiological needs. Furthermore, combining future meteorological information and cost constraints, a rolling time-domain optimization algorithm is used for multi-objective collaborative optimization, generating an optimal regulation strategy that balances growth needs and economic benefits, and driving its execution. This enables quantitative and precise regulation of the synergistic effect of water and fertilizer, thereby significantly improving resource utilization efficiency and agricultural production economic benefits throughout the entire crop growth cycle.

[0066] Figure 5 This is a block diagram of an AIoT-based integrated water and fertilizer management system according to an embodiment of this application. Figure 5 As shown, the AIoT integrated water and fertilizer control system 100 based on a crop growth dynamic model according to an embodiment of this application includes: an environmental sensing module 110, used to acquire environmental data and soil hydraulic parameters, the environmental data including soil moisture content, soil electrical conductivity, air temperature, air humidity, and photosynthetically active radiation, and the soil hydraulic parameters including field capacity, wilting point humidity, and saturated water content; a nutrient assessment module 120, used to process the soil moisture content and soil hydraulic parameters using a crop growth dynamic model to obtain the nutrient availability coefficient at the current moment; and a demand prediction module 130, used to analyze the nutrient availability coefficient at the current moment and the nutrient availability coefficient at the previous moment. The crop status and environmental data are used to predict crop nutrient requirements and water evapotranspiration based on a coupled model to obtain the actual nutrient uptake and crop evapotranspiration at the current moment; the water and fertilizer synergistic decision module 140 is used to make water and fertilizer synergistic decisions based on a rolling time domain optimization, based on preset future nutrient targets and soil moisture targets, the actual nutrient uptake at the current moment, the crop evapotranspiration at the current moment, the weather forecast for the next 24 hours, and the unit price of water and fertilizer, to obtain the optimal irrigation amount and the optimal fertilization amount at the current moment; the instruction generation module 150 is used to convert the optimal irrigation amount and the optimal fertilization amount at the current moment into executor instructions.

[0067] As described above, the AIoT water and fertilizer integrated control system 100 based on a crop growth dynamic model according to the embodiments of this application can be implemented in various wireless terminals, such as servers with an AIoT water and fertilizer integrated control algorithm based on a crop growth dynamic model. In one possible implementation, the AIoT water and fertilizer integrated control system 100 based on a crop growth dynamic model according to the embodiments of this application can be integrated into the wireless terminal as a software module and / or a hardware module. For example, the AIoT water and fertilizer integrated control system 100 based on a crop growth dynamic model can be a software module in the operating system of the wireless terminal, or it can be an application developed for the wireless terminal; of course, the AIoT water and fertilizer integrated control system 100 based on a crop growth dynamic model can also be one of many hardware modules of the wireless terminal.

[0068] Alternatively, in another example, the AIoT water and fertilizer integrated control system 100 based on the crop growth dynamic model and the wireless terminal can also be separate devices, and the AIoT water and fertilizer integrated control system 100 based on the crop growth dynamic model can be connected to the wireless terminal via wired and / or wireless networks, and transmit interactive information in accordance with the agreed data format.

[0069] Here, those skilled in the art will understand that the specific operations of each step in the above-mentioned AIoT water and fertilizer integration control system based on crop growth dynamics models have been referenced above. Figures 1 to 4 The AIoT-based water and fertilizer integration regulation method based on crop growth dynamic models has been described in detail, and therefore, its repeated description will be omitted.

Claims

1. An AIoT-based integrated water and fertilizer management method based on a crop growth dynamic model, characterized in that, include: Acquire environmental data and soil hydraulic parameters. Environmental data includes soil moisture content, soil electrical conductivity, air temperature, air humidity, and photosynthetically active radiation. Soil hydraulic parameters include soil field capacity, wilting point humidity, and saturated water content. Using a crop growth dynamic model, soil moisture content and soil hydraulic parameters are processed to obtain the nutrient availability coefficient at the current moment. This includes: calculating the root zone water stress coefficient based on a piecewise function of soil moisture content and soil hydraulic parameters to obtain the water stress coefficient at the current moment, wherein the piecewise function is expressed as: ； in, The water stress coefficient at the current moment. Soil moisture content, Humidity at the wilting point, This is the saturated water content. The field capacity is used as the reference value. The nutrient availability coefficient at the current moment is calculated by integrating the water stress coefficient, soil electrical conductivity, and fertilization parameters with the nutrient availability coefficient under salt stress. This calculation includes: using the following formula to calculate the nutrient availability coefficient under salt stress based on the current water stress coefficient, soil electrical conductivity, and fertilization parameters, where the formula is: ； in, The water stress coefficient at the current moment. For soil electrical conductivity, The optimal soil electrical conductivity for the crop in the fertilization parameters. This refers to the salt sensitivity coefficient among fertilization parameters. It is the hyperbolic tangent function. This represents the nutrient availability coefficient at the current moment. Based on the nutrient availability coefficient at the current moment, the crop status at the previous moment, and environmental data, a coupled model is used to predict crop nutrient demand and water evapotranspiration to obtain the actual nutrient uptake and crop evapotranspiration at the current moment. Based on the preset future nutrient targets and soil moisture targets, the system makes a rolling time-domain optimized water and fertilizer synergistic decision based on the actual nutrient uptake at the current moment, the crop evapotranspiration at the current moment, the weather forecast for the next 24 hours, and the unit price of water and fertilizer to obtain the optimal irrigation amount and the optimal fertilizer amount at the current moment. Transform the optimal irrigation amount and the optimal fertilization amount at the current moment into actuator instructions.

2. The AIoT-based integrated water and fertilizer management method based on a crop growth dynamic model according to claim 1, characterized in that, The crop status at the previous moment includes leaf area index, biomass, and developmental stage.

3. The AIoT-based integrated water and fertilizer management method based on a crop growth dynamic model according to claim 2, characterized in that, Using a coupled model, crop nutrient demand and water evapotranspiration are predicted based on the current nutrient availability coefficient, the crop status at the previous time step, and environmental data to obtain the actual nutrient uptake and crop evapotranspiration at the current time step, including: Based on the nutrient availability coefficient at the current moment, the crop status at the previous moment, the photosynthetically active radiation and air temperature and crop parameters in the environmental data, determine the actual nutrient uptake at the current moment; Using the FAO Penman-Monteith model, the leaf area index, environmental data, and site parameters of the crop at the previous time step are processed to obtain the crop evapotranspiration at the current time step.

4. The AIoT-based integrated water and fertilizer management method based on a crop growth dynamic model according to claim 3, characterized in that, Based on the current nutrient availability coefficient, the crop status at the previous time, photosynthetically active radiation and air temperature from environmental data, and crop parameters, the actual nutrient uptake at the current time is determined, including: determining the actual nutrient uptake at the current time using the following formula, where the formula is: ； ； ； in, Photosynthetically active radiation, It is a natural constant. Extinction coefficient, Leaf area index, Light energy utilization efficiency is a crop parameter. Let be the temperature stress function for photosynthesis. This represents the potential daily biomass growth. The maximum nitrogen concentration associated with the developmental stage. For potential nutrient requirements, The nutrient availability coefficient, This represents the actual nutrient absorption at the current moment.

5. The AIoT-based integrated water and fertilizer management method based on a crop growth dynamic model according to claim 3, characterized in that, Using the FAO Penman-Monteith model, the leaf area index, environmental data, and site parameters from the previous time step are processed to obtain the crop evapotranspiration at the current time step. This includes processing the leaf area index, environmental data, and site parameters from the previous time step using the following formula: ； ； in, For crop coefficients, This represents the minimum crop coefficient during the early stages of crop growth. This represents the maximum crop coefficient when the crop canopy is fully covered. To adjust LAI The coefficient of influence This represents the leaf area index of the crop at the previous time step. This represents the actual crop evapotranspiration. This represents the crop evapotranspiration at the current moment.

6. The AIoT-based integrated water and fertilizer management method based on a crop growth dynamic model according to claim 3, characterized in that, Based on the current nutrient availability coefficient, the crop status at the previous time, photosynthetically active radiation and air temperature from environmental data, and crop parameters, determine the actual nutrient uptake at the current time, including: Based on the soil available nutrient concentration, fertilizer application rate, actual nutrient absorption by the crop, and nutrient leaching loss at the previous moment, the soil available nutrient concentration at the current moment is updated. The nutrient absorption rate determined by the soil supply capacity is calculated using the Michaelis equation and based on the current soil available nutrient concentration, the crop's maximum nutrient absorption rate, and the Michaelis constant. The actual nutrient uptake at the current moment is obtained by multiplying the nutrient uptake rate, which is determined by the soil's supply capacity, by the nutrient availability coefficient at the current moment.

7. An AIoT-based integrated water and fertilizer management system based on a crop growth dynamic model, characterized in that, include: The environmental sensing module is used to acquire environmental data and soil hydraulic parameters. The environmental data includes soil moisture content, soil electrical conductivity, air temperature, air humidity and photosynthetically active radiation. The soil hydraulic parameters include soil field capacity, wilting point humidity and saturation water content. The nutrient assessment module is used to process soil moisture content and soil hydraulic parameters using a crop growth dynamic model to obtain the nutrient availability coefficient at the current moment. This includes: calculating the root zone water stress coefficient based on a piecewise function of soil moisture content and soil hydraulic parameters to obtain the water stress coefficient at the current moment, wherein the piecewise function is expressed as: ； in, The water stress coefficient at the current moment. Soil moisture content, Humidity at the wilting point, This is the saturated water content. The field capacity is used as the reference value. The nutrient availability coefficient at the current moment is calculated by integrating the water stress coefficient, soil electrical conductivity, and fertilization parameters with the nutrient availability coefficient under salt stress. This calculation includes: using the following formula to calculate the nutrient availability coefficient under salt stress based on the current water stress coefficient, soil electrical conductivity, and fertilization parameters, where the formula is: ； in, The water stress coefficient at the current moment. For soil electrical conductivity, The optimal soil electrical conductivity for the crop in the fertilization parameters. This refers to the salt sensitivity coefficient among fertilization parameters. It is the hyperbolic tangent function. This represents the nutrient availability coefficient at the current moment. The demand forecasting module is used to predict crop nutrient demand and water evapotranspiration based on the nutrient availability coefficient at the current moment, the crop status at the previous moment, and environmental data, in order to obtain the actual nutrient uptake and crop evapotranspiration at the current moment. The water and fertilizer synergistic decision-making module is used to make water and fertilizer synergistic decisions based on the preset future nutrient targets and soil moisture targets, the actual nutrient absorption at the current moment, the crop evapotranspiration at the current moment, the weather forecast for the next 24 hours, and the unit price of water and fertilizer, based on rolling time domain optimization, so as to obtain the optimal irrigation amount and the optimal fertilizer amount at the current moment. The instruction generation module is used to convert the optimal irrigation amount and the optimal fertilization amount at the current moment into actuator instructions.

Images