Fertilization control system and method for liquid fertilizer

Abstract

The invention relates to the technical field of fertilization, in particular to a fertilization control system and method for a liquid fertilizer, and the system comprises an interaction module which is used for inputting a crop type and loading a corresponding crop knowledge graph; an acquisition module; the instruction generation unit is used for constructing a liquid fertilizer loss dynamic model, performing dynamic optimization by using a depth deterministic strategy gradient reinforcement learning algorithm, predicting future meteorological parameters by using an LSTM time sequence model, generating a fertilization scheme based on the optimized liquid fertilizer loss dynamic model, a crop knowledge graph and the predicted future meteorological parameters, and sending the fertilization scheme to the control unit; generating a fertilization control instruction according to the fertilization scheme; and the fertilization control module is used for receiving the fertilization equipment control instruction, executing the fertilization equipment control instruction and feeding back in real time. Therefore, the problems of low precision, serious resource waste, single fertilization system data dimension, poor model adaptability, imperfect abnormal response mechanism and the like existing in agricultural fertilization in the prior art are solved.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of fertilization, in particular to a fertilization control system and method for liquid fertilizer. BACKGROUND

[0002] Traditional agricultural fertilization methods have long relied on manual experience, and have problems such as low precision and serious resource waste. On the one hand, fertilization schemes are mostly based on fixed periods or experience values, and cannot be adjusted dynamically according to the real-time growth state of crops, changes in soil fertility and meteorological conditions, resulting in excessive or insufficient fertilization, which affects crop yield and quality, and easily causes environmental problems such as soil compaction and water eutrophication. On the other hand, although existing simple fertilization systems introduce some sensors to monitor soil moisture or fertilizer liquid concentration, the data dimension is single, and the comprehensive consideration of meteorological factors such as light, wind speed and rainfall is lacking, especially the loss rules of liquid fertilizer in different environments (such as leaching caused by rainfall, and volatilization and drift caused by high temperature and strong wind) are ignored, so it is difficult to realize real precision fertilization.

[0003] With the development of agricultural intelligence, some fertilization control systems attempt to introduce data models to assist decision-making, but there are still obvious limitations. Most systems use static models or simple algorithms, which cannot adapt to complex and changeable field environments, and the model parameters need to be frequently calibrated manually, so the practicality is limited. At the same time, the device running state monitoring and abnormal response mechanism are not perfect, when the pump group fails, the pipeline pressure is abnormal or the environmental parameters mutate, it is difficult to give timely warning and processing, which not only affects the fertilization effect, but also may cause equipment damage or fertilizer waste. SUMMARY

[0004] The application provides a fertilization control system and method for liquid fertilizer to solve the problems of low precision, serious resource waste, single data dimension, poor model adaptability and imperfect abnormal response mechanism of existing agricultural fertilization.

[0005] In a first aspect, an embodiment of the present application provides a fertilization control system for liquid fertilizer, comprising: an interaction module, an acquisition module, an instruction generation unit, and a fertilization control module; wherein the interaction module is used to input a crop type and load a corresponding crop knowledge graph; the acquisition module is used to collect air temperature and humidity, wind speed, photosynthetic photon flux density, and rainfall, and deploy a sensor module to collect soil EC value, NPK content, fertilizer solution concentration, fertilizer solution flow rate, pipeline pressure, and crop growth status data; the instruction generation unit is used to construct a liquid fertilizer loss dynamics model, and dynamically optimize it using a deep deterministic policy gradient reinforcement learning algorithm, use an LSTM time series model to predict future meteorological parameters, and generate a fertilization plan based on the optimized liquid fertilizer loss dynamics model, the crop knowledge graph, and the predicted future meteorological parameters, and generate a fertilization control instruction according to the fertilization plan, wherein the liquid fertilizer loss dynamics model includes a rainfall correction function and a photosynthetic photon flux density-wind speed joint correction function, which are used to quantify the fertilizer liquid loss rate under different meteorological conditions; the fertilization control module is used to receive the fertilization equipment control instruction, execute the fertilization equipment control instruction, and provide real-time feedback.

[0006] Preferably, the interactive module includes: an interactive interface unit and a crop knowledge graph loading unit, wherein the interactive interface unit is used to support users to interactively modify the generated fertilization plan; the crop knowledge graph loading unit is used to load the corresponding crop knowledge graph according to the information, and the crop knowledge graph includes the nutrient requirements and fertilization response curves of crops in different growth stages.

[0007] Preferably, the acquisition module includes: a field weather station unit, a sensor unit, and a data processing unit, wherein the field weather station unit is used to collect air temperature and humidity, wind speed, photosynthetic photon flux density, and rainfall; the sensor unit is used to collect soil EC value, NPK content, fertilizer solution concentration, fertilizer solution flow, pipeline pressure and crop growth conditions; the data processing unit is used to reduce noise and standardize the data obtained by the field weather station unit and the sensor unit.

[0008] Preferably, the instruction generation unit includes a loss model unit and an instruction generation engine unit, wherein the loss model unit is used to construct and continuously update the liquid fertilizer loss dynamics model based on the actual loss data after fertilization; the instruction generation engine unit is used to generate a fertilization plan, and use a deep deterministic policy gradient algorithm to dynamically optimize the parameters of the liquid fertilizer loss dynamics model with the optimization goal of minimizing the deviation between the predicted fertilizer loss and the actual loss. The instruction generation engine unit supports users to interactively adjust the optimized fertilization plan and generate fertilization control instructions based on the adjusted fertilization plan.

[0009] Preferably, the instruction generation engine unit further comprises: triggering an instruction regeneration process if the received real-time meteorological data or equipment feedback data deviates from the predicted value by more than a preset threshold during the fertilization execution process, re-optimizing the liquid fertilizer loss kinetic model based on real-time data and generating new fertilization control instructions.

[0010] Preferably, the formula of the liquid fertilizer loss kinetic model is:

[0011] wherein, is the loss amount of solute per unit volume, is time, is the rainfall correction function value, is the photosynthetic photon flux density and wind speed correction function value, is the fertilizer solution concentration, is the model error, is the photosynthetic photon flux density, is the wind speed; wherein, the formula of the rainfall correction function value is:

[0012] wherein, is the runoff critical rainfall, is the actual rainfall, is the non-linear influence intensity, is the linear tilt degree, is the basic correction value; wherein, the photosynthetic photon flux density and wind speed correction function expression is:

[0013] wherein, is the light compensation point density, is the light saturation point density, is the transpiration promotion critical wind speed, is the drift dominant critical wind speed, is the basic absorption efficiency coefficient, is the low wind speed transpiration promotion coefficient, is the medium wind speed drift loss coefficient, is the high wind speed absorption reference coefficient, is the high wind speed drift attenuation coefficient, is the high light inhibition coefficient, and e is the natural constant.

[0014] Preferably, the fertilization control module includes: a fertilization unit and an instruction execution unit, wherein the fertilization unit is used to arrange field fertilization equipment, and the field fertilization equipment includes a fertilizer storage tank, a water storage tank, a pipeline network, an adjustable nozzle, and a variable frequency pump; the instruction execution unit is used to fertilize according to the control instructions of the fertilization equipment.

[0015] Preferably, the instruction execution unit also includes: an abnormality judgment unit, wherein the abnormality judgment unit is used to collect the pump group operation status and valve opening and closing data in real time, receive the fertilizer liquid flow, pipeline pressure, air temperature and humidity, and soil EC value data from the data acquisition module, compare the collected data with the preset threshold, and when the preset conditions are met, it is judged as an abnormality and triggers an alarm.

[0016] Preferably, the preset conditions include: the vibration amplitude, temperature or current value of the pump group exceeds the normal operating range of the equipment; the valve fails to complete the opening and closing action within the specified time, or an abnormal opening and closing state occurs during the non-instruction period; the fertilizer liquid flow rate is continuously lower than the minimum design flow rate or continuously higher than the maximum design flow rate; the pipeline pressure continuously exceeds the safety pressure range or is lower than the minimum pressure to maintain normal operation of the system; the air temperature and humidity or soil EC value continuously deviates from the suitable range for crop growth or the system operation safety threshold.

[0017] Preferably, the fertilization plan includes the area number, fertilizer type, fertilizer amount and fertilization time; the fertilization control instruction includes the area number, fertilizer storage tank number, opening time, closing time, variable frequency pump frequency and pump group pressure valve control value.

[0018] The second embodiment of the present application provides a fertilization control method for liquid fertilizer, including: obtaining a crop knowledge map, crop growth status and growth cycle; obtaining the crop growth stage according to the crop growth status and growth cycle, obtaining the NPK amount required for the growth stage according to the crop knowledge map, and combining the soil NPK content and the concentration of the NPK fertilizer to calculate the total fertilizer amount and daily fertilizer amount required for the crop; constructing and optimizing a liquid fertilizer loss dynamics model, using an LSTM time series model to predict the air temperature and humidity, rainfall, photosynthetic photon flux density, and wind speed in the next 24 hours, and inputting the predicted rainfall, photosynthetic photon flux density, and wind speed into the optimal The optimized liquid fertilizer loss kinetic model is used to obtain the expected solute loss per unit volume, and combined with the predicted air temperature and humidity, the optimal fertilization time window is obtained, and the corresponding fertilization duration and the expected fertilizer loss are calculated based on the optimal fertilization time window; the actual fertilizer amount to be applied on the day is calculated based on the daily fertilizer amount and the expected fertilizer loss, the fertilizer liquid flow rate is calculated based on the actual fertilizer amount to be applied on the day and the fertilization duration, and mapped to the variable frequency pump frequency, and the optimal spraying pressure is determined based on the adjustable nozzle model, converted into the pump group pressure valve control value, and a fertilization plan is generated, and a fertilizer equipment control instruction is generated based on the fertilization plan, and fertilization is performed according to the fertilizer equipment control instruction.

[0019] Therefore, this application has the following beneficial effects: This embodiment of the present application uses an acquisition module to collect multi-dimensional data such as air temperature and humidity, wind speed, soil EC values, and crop growth status. This data, combined with the nutrient requirements of different growth stages from the crop knowledge graph, avoids the problem of over- or under-fertilization caused by traditional fertilization, which relies on fixed cycles or empirical values. This significantly improves fertilizer utilization, reduces resource waste, and reduces negative environmental impacts. Furthermore, the system incorporates not only soil and crop data but also meteorological factors such as light, rainfall, and wind speed. A liquid fertilizer loss dynamics model characterizes the loss patterns of liquid fertilizer under different environments (such as rainfall leaching and strong wind drift). A deep deterministic policy gradient reinforcement learning algorithm is then used to dynamically optimize model parameters. This overcomes the limitations of existing systems, which suffer from a single data dimension and poor adaptability of static models, enabling fertilization plans to be precisely tailored to complex and changing field conditions. The fertilization control module, through a real-time feedback mechanism and combined with the abnormality detection function in the instruction execution unit, can promptly monitor abnormalities in equipment status, such as pumps, valves, and pipeline pressure, as well as environmental parameters. This addresses the imperfect abnormality response mechanisms of existing systems, prevents fertilization effectiveness from being affected or equipment from being damaged due to faults, and improves the stability and reliability of system operation. This solves the problems of low accuracy and serious waste of resources in agricultural fertilization in existing technologies, as well as the single data dimension of the fertilization system, poor model adaptability, and imperfect abnormal response mechanism.

[0020] Additional aspects and advantages of the application will be set forth in part in the description which follows, and in part will become apparent to those skilled in the art upon examination of the following and / or can be learned by practice of the application. BRIEF DESCRIPTION OF DRAWINGS

[0021] The above and / or additional aspects and advantages of the present application will become apparent and be more readily understood through reference to the following description, taken in conjunction with the accompanying drawings, wherein: Figure 1 A structural schematic diagram of a fertilization control system for liquid fertilizer according to an embodiment of the present application; Figure 2 A structural schematic diagram of a fertilization control system for liquid fertilizer according to an embodiment of the present application; Figure 3 A flow chart of a fertilization control method for liquid fertilizer according to an embodiment of the present application. DETAILED DESCRIPTION

[0022] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without any creative work fall within the protection scope of the present application.

[0023] The following describes a fertilization control system and method for liquid fertilizer according to an embodiment of the present application with reference to the accompanying drawings. To address the serious resource waste problem mentioned in the background art, the present application provides a fertilization control system for liquid fertilizer. In this system, an acquisition module collects multi-dimensional data such as air temperature and humidity, wind speed, soil EC value, and crop growth status. Combined with the nutrient requirements of different growth stages in the crop knowledge graph, this system avoids the problem of excessive or insufficient fertilizer caused by traditional fertilization relying on fixed cycles or empirical values, significantly improves fertilizer utilization, and reduces resource waste and negative environmental impacts. The system not only incorporates soil and crop data, but also focuses on integrating meteorological factors such as light, rainfall, and wind speed. A liquid fertilizer loss dynamics model is used to characterize the loss pattern of fertilizer liquid under different environments (such as rainfall leaching and strong wind drift). A deep deterministic policy gradient reinforcement learning algorithm is used to dynamically optimize model parameters. This system overcomes the limitations of existing systems with single data dimensions and poor adaptability of static models, enabling fertilization plans to accurately match the complex and changing actual conditions in the field. The fertilization control module uses a real-time feedback mechanism, combined with the abnormality detection function in the instruction execution unit, to promptly monitor abnormalities in equipment status and environmental parameters such as pumps, valves, and pipeline pressure. This addresses the existing system's imperfect abnormality response mechanism, preventing fertilization effects or equipment damage caused by faults, and improving the stability and reliability of system operation. This addresses existing agricultural fertilization issues such as low accuracy, severe resource waste, single-dimensional data, poor model adaptability, and imperfect abnormality response mechanisms.

[0024] Figure 1 A schematic structural diagram of a fertilization control system for liquid fertilizer provided in an embodiment of the present application.

[0025] An embodiment of the present application provides a fertilization control system for liquid fertilizer. The fertilization control system 10 for liquid fertilizer includes: an interaction module 100 , an acquisition module 200 , an instruction generation unit 300 and a fertilization control module 400 .

[0026] Among them, the interaction module 100 is used to input the crop type and load the corresponding crop knowledge graph; the acquisition module 200 is used to collect air temperature and humidity, wind speed, photosynthetic photon flux density, rainfall, and deploy sensor modules to collect soil EC value, NPK content, fertilizer liquid concentration, fertilizer liquid flow, pipeline pressure and crop growth status data; the instruction generation unit 300 is used to construct a liquid fertilizer loss dynamics model, and use a deep deterministic policy gradient reinforcement learning algorithm for dynamic optimization, use an LSTM time series model to predict future meteorological parameters, and generate a fertilization plan based on the optimized liquid fertilizer loss dynamics model, the crop knowledge graph and the predicted future meteorological parameters, and generate fertilization control instructions according to the fertilization plan; the fertilization control module 400 is used to receive fertilizer equipment control instructions, execute fertilizer equipment control instructions and provide real-time feedback.

[0027] The liquid fertilizer loss kinetics model can include a rainfall correction function and a photosynthetic photon flux density-wind speed joint correction function, which are used to quantify the fertilizer liquid loss rate under different meteorological conditions.

[0028] It can be understood that, in the embodiments of the present application, the interactive module provides the crop growth demand basis for the fertilization decision by loading the crop knowledge graph, the acquisition module constructs a comprehensive decision basis by collecting multi-dimensional meteorological, soil, crop and equipment operation data, the instruction generation unit realizes the precise adaptation of the fertilization scheme by means of the dynamically optimized liquid fertilizer loss kinetics model and the future weather forecast, and the fertilization control module ensures the efficient and reliable operation of the system through instruction execution and real-time feedback, thereby effectively improving the fertilization precision, solving the problems of resource waste, single data dimension, poor model adaptability and insufficient operation reliability in traditional fertilization, and realizing the intelligent and precise application of liquid fertilizer.

[0029] In the embodiments of the present application, the interactive module 100 includes an interactive interface unit and a crop knowledge graph loading unit.

[0030] The interactive interface unit is configured to support the user to interactively correct the generated fertilization scheme, and the crop knowledge graph loading unit is configured to load the corresponding crop knowledge graph according to the information, the crop knowledge graph containing the nutrient demand and the fertilization response curve of the crop in different growth stages.

[0031] It can be understood that, in the embodiments of the present application, the interactive module supports the user to interactively correct the generated fertilization scheme through the interactive interface unit, which not only retains the efficiency of intelligent decision-making, but also improves the flexibility and actual adaptability of the scheme through manual intervention; the crop knowledge graph loading unit loads the knowledge graph containing the nutrient demand and the fertilization response curve of the crop in different growth stages based on the crop type, thereby providing a scientific basis for the fertilization decision that is suitable for the biological characteristics of the crop, avoiding the blindness of the general scheme, and combining the two to make the fertilization scheme not only accurately match the growth law of the crop, but also flexibly cope with the complex actual situation in the field, thereby further improving the pertinence and reliability of the fertilization control.

[0032] Specifically, the user selects corn, A03 area, and jointing stage, the corn knowledge graph can be loaded, 15.9 kg of nitrogen, 4.1 kg of phosphorus and 13.8 kg of potassium are needed to produce 1 ton of corn kernels, 50% of nitrogen, 40% of phosphorus and 50% of potassium are needed in the jointing stage, 7.95 kg of nitrogen, 1.64 kg of phosphorus and 6.9 kg of potassium are calculated, the actual NPK amount is calculated according to the planting area and average yield, the fertilization response curve of corn in the jointing stage is in the shape of a parabola, before reaching the nutrient threshold, fertilization brings positive returns, which can promote the growth of corn, and after reaching the nutrient threshold, the returns drop sharply and soon become negative.

[0033] In the embodiment of the present application, the acquisition module 200 includes: a field weather station unit, a sensor unit, and a data processing unit.

[0034] Among them, the field weather station unit is used to collect air temperature and humidity, wind speed, photosynthetic photon flux density, and rainfall; the sensor unit is used to collect soil EC value, NPK content, fertilizer solution concentration, fertilizer solution flow, pipeline pressure and crop growth conditions; the data processing unit is used to reduce noise and standardize the data obtained by the field weather station unit and the sensor unit.

[0035] It can be understood that the acquisition module in the embodiment of the present application comprehensively collects meteorological, soil, fertilizer liquid and crop growth status data through the field weather station unit and the sensor unit, constructs a basic data system for the entire scene, and breaks through the limitation of single data in traditional systems; the data processing unit performs noise reduction and standardization on the collected data, effectively improving the data quality, and providing accurate and consistent input for subsequent model calculations and decision generation, ensuring that the fertilization plan is based on real and reliable field information, and guaranteeing the scientificity and accuracy of the decision from the source.

[0036] Specifically, the collected air temperature data is (30℃, 32℃, 34℃, 20℃, 32℃). Through data denoising, we can get (30℃, 31℃, 33℃, 32℃, 32℃), eliminate the outlier 20℃, set the mean to 30℃, and after using Z-score standardization, we get (0, 0.5, 1.5, 1, 1).

[0037] In the embodiment of the present application, the instruction generation unit 300 includes a loss model unit and an instruction generation engine unit.

[0038] Among them, the loss model unit is used to construct and continuously update the liquid fertilizer loss dynamics model based on the actual loss data after fertilization; the instruction generation engine unit is used to generate a fertilization plan, and use a deep deterministic policy gradient algorithm to dynamically optimize the parameters of the liquid fertilizer loss dynamics model with the optimization goal of minimizing the deviation between the predicted fertilizer loss and the actual loss. The instruction generation engine unit supports users to interactively adjust the optimized fertilization plan and generate fertilization control instructions based on the adjusted fertilization plan.

[0039] It can be understood that the loss model unit in the embodiment of the present application takes the actual loss data after fertilization as the core, constructs and continuously updates the liquid fertilizer loss dynamics model, and provides a loss prediction basis that fits the actual scenario for subsequent scheme optimization, avoiding prediction deviations caused by model fixation; the instruction generation engine unit first generates a preliminary fertilization plan based on the model, and then uses a deep deterministic policy gradient algorithm to dynamically optimize the model parameters with the goal of minimizing the deviation between the predicted fertilizer loss and the actual loss, ensuring the scientificity and effectiveness of the plan in loss control, and at the same time supporting users to interactively adjust the optimized plan, taking into account the accuracy of the algorithm and the flexibility of actual operation, and finally generating fertilization control instructions based on the adjusted plan, which can not only reduce liquid fertilizer waste and reduce environmental burden, but also adapt to the personalized needs of different users and actual field conditions, and improve the intelligence and practicality of fertilization operations.

[0040] It should be noted that the deep deterministic policy gradient algorithm designs a continuous state space (soil moisture, temperature, real-time fertilizer loss rate, etc.) and action space (diffusion coefficient, adsorption rate and other parameter adjustment quantities), with the goal of minimizing the deviation between predicted and actual losses, and constructs a reward function (including deviation penalty and parameter change regularization terms) to achieve dynamic optimization of high-dimensional continuous control problems; the Actor network outputs parameter adjustment actions, the Critic network evaluates the Q value, and the target network soft updates (τ=0.001) to improve stability. With the continuous input of actual field loss data (such as leaching, volatilization) and environmental parameters (soil moisture, temperature, etc.), the parameters are iterated in real time to adapt to changes in different soil, crops, and climatic conditions, avoiding the limitations of static models.

[0041] In an embodiment of the present application, the instruction generation engine unit also includes: during the fertilization execution process, if the received real-time meteorological data or equipment feedback data deviates from the predicted value by more than a preset threshold, the instruction regeneration process is triggered, and the liquid fertilizer loss dynamics model is re-optimized based on the real-time data and a new fertilization control instruction is generated.

[0042] The preset threshold value may be determined according to actual conditions.

[0043] It is understood that in this embodiment of the application, when real-time meteorological data (such as sudden rainfall or high temperatures) or equipment feedback (such as abnormal flow or drip irrigation blockage) deviates from the predicted value by exceeding a preset threshold, a process is immediately initiated: the real-time abnormal data is incorporated into the liquid fertilizer loss dynamics model optimization, the model parameters are quickly corrected to match the actual operating conditions, and new fertilization control instructions are generated. This mechanism breaks the static execution loop and can resolve interference from uncontrollable factors in real time, avoiding fertilizer waste or nutrient deficiency. It enhances fertilization accuracy and equipment fault tolerance, ensuring crop needs while reducing environmental burdens, and provides support for the stable implementation of intelligent fertilization.

[0044] Specifically, a smart agriculture project implemented liquid nitrogen fertilizer drip irrigation during the jointing stage of summer corn. Based on the weather forecast data 12 hours before fertilization (forecast: "sunny, no precipitation, wind speed 2-3"), the initial liquid fertilizer loss dynamics model calculated that continuous fertilization at a drip irrigation flow rate of 8L / mu·hour was required for 4 hours to ensure that the nitrogen fertilizer retention rate in the soil root zone reached more than 85% to meet the nutrient needs of corn during the jointing stage.

[0045] Two hours into the fertilization process, the meteorological monitoring equipment reported real-time data: a sudden, short-term, heavy rainfall (25 mm / hour) and a sudden increase in wind speed to level 6-7. At this point, it was necessary to determine whether the deviation exceeded the preset threshold. The project preset "rainfall deviation ≥ 15 mm / hour, wind speed deviation ≥ level 3" triggers the regeneration process. The current real-time data clearly exceeded the threshold, so the regeneration command was immediately initiated: Model optimization: Real-time data such as "short-term heavy rainfall (which will accelerate the loss of liquid fertilizer with surface runoff) and high wind speed (which will increase the evaporation loss of liquid fertilizer during drip irrigation)" are input into the model, and core parameters such as "rainfall loss coefficient" and "evaporation loss coefficient" are recalibrated. The revised model predicts that if fertilizer is continued at the original flow rate, the nitrogen fertilizer root zone retention rate will drop below 50%, and fertilizer damage may occur due to soil waterlogging.

[0046] New instruction generation: Based on the optimized model, new fertilization control instructions are generated: ① Immediately suspend drip irrigation to prevent fertilizer loss with rainwater; ② After the rainfall stops (real-time meteorological monitoring shows that the rain will stop in 1 hour), adjust the drip irrigation flow rate to 5L / mu·hour and shorten the fertilization time to 1.5 hours (to replenish some lost nutrients and prevent the soil from becoming overly wet); ③ Synchronously turn on the soil moisture sensor for real-time monitoring. If the soil moisture content exceeds 80% of the field water holding capacity, further delay fertilization.

[0047] In the embodiment of the present application, the formula of the liquid fertilizer loss kinetic model is:

[0048] in, is the solute loss per unit volume, For time, is the rainfall correction function value, is the correction function value of photosynthetic photon flux density and wind speed, is the concentration of fertilizer solution, is the model error, is the photosynthetic photon flux density, is the wind speed; The formula for the rainfall correction function value is:

[0049] in, is the critical rainfall for runoff generation, is the actual rainfall, is the nonlinear influence intensity, is the linear slope, is the basic correction value; Among them, the photosynthetic photon flux density and wind speed correction function expressions are:

[0050] in, is the light compensation point density, is the light saturation point density, The critical wind speed for transpiration promotion is is the drift-dominated critical wind speed, is the basic absorption efficiency coefficient, is the low wind speed transpiration promotion coefficient, is the drift loss coefficient at medium wind speed, is the high wind speed absorption reference coefficient, is the high wind speed drift attenuation coefficient, is the high light suppression coefficient, and e is a natural constant.

[0051] It should be noted that It is the lighting in the low light area. is the beneficial ratio of low wind speed to fertilizer absorption rate, is the beneficial ratio of wind speed to fertilizer absorption rate, is the offset rate of fertilizer absorption rate due to high wind speed, It is the lighting in the highlight area.

[0052] In addition, the wind speed is negligible during strong light conditions because plants do not carry out photosynthesis at this time and the absorption rate of fertilizer is negligible.

[0053] Specifically, for example, in greenhouse tomato cultivation, the fertilizer concentration is 200g / L, the greenhouse rainfall is 0, the photosynthetic photon flux density is 600μmol / m²・s, and the wind speed is 1m / s. It can be calculated that L(t)≈53.6g / L, and the loss rate is about 26.8%. The initialization values ​​of the coefficients in the photosynthetic photon flux density and wind speed correction function can be obtained based on laboratory experiments. Taking greenhouse tomatoes as an example, a=0.002, b=0.01, c=0.05, =100μmol / m²・s, =800μmol / m²・s, =1.5m / s, =3.0 m / s, d=0.85, g=0.3, h=1.2, j=0.6, k=1.5, m=0.002. Assuming that 12 kg of nitrogen is required after deducting the nitrogen provided by the soil, and the fertilizer solution concentration is 200 g / L, a total of 600 L of nitrogen fertilizer is required. Assuming this growth phase lasts 30 days, 20 L of fertilizer is required per day. Based on weather forecasts and a liquid fertilizer loss kinetics model, the optimal fertilization window is between 8:00 AM and 10:00 AM, requiring approximately 27.32 L of fertilizer at a fertilizer flow rate of approximately 0.227 L / min. A fertilization plan can be generated (Zone 1, 27.32 L, 8:00 AM - 10:00 AM). The user modified this to (Zone 1, 27.32 L, 8:00 AM - 9:00 AM), generating the following control instructions for the fertilization equipment: (Zone 1, Fertilizer Tank No. 1 (N fertilizer), 7.58-9.02 (open 2 minutes early, close 2 minutes late), 35 Hz, 0.3 MPa). The device was turned on in 7.58 and turned off in 9.02, with no abnormalities.

[0054] It can be understood that the present application introduces a liquid fertilizer loss dynamics model in the liquid fertilizer fertilization process in real time, wherein the rainfall correction function fully considers the impact of rainfall on fertilizer loss before and after reaching the critical value. Before reaching the critical value, the soil moisture content is not saturated and no effective runoff is formed, so the contribution of rainfall to fertilizer loss can be ignored. When the rainfall reaches the critical value, surface runoff is generated, and the rainwater washes away the residual fertilizer liquid on the surface and even entrains the surface soil particles. As the rainfall increases, the fertilizer loss increases rapidly with the rainfall intensity; the photosynthetic photon flux density and wind speed correction function, taking into account that light and photon flux density I and wind speed V will affect the plant's absorption rate of fertilizer, so the function is divided into four stages, photophilic low wind zone, , I exceeds the compensation point but is less than the saturation point, and the wind speed V is lower than the critical wind speed for transpiration promotion. At this time, the plant's absorption rate of fertilizer is basically derived from basic photosynthesis, with no obvious drift loss, in the light and breeze zone. , I exceeds the compensation point but is less than the saturation point, the wind speed exceeds the critical wind speed for transpiration promotion but is lower than the critical wind speed for drift dominance, at this time the plant has photosynthesis, by enhancing leaf transpiration, the foliar fertilizer droplets are more evenly attached to the stomata, and at the same time promote the flow of water from the root zone to the root system (nutrients migrate with water), indirectly improving the absorption efficiency, suitable for light and strong wind areas, , I exceeds the compensation point but is less than the saturation point, the wind speed exceeds the drift-dominated critical wind speed, high wind speed causes a large amount of foliar fertilizer droplets to drift (deviate from the target leaves), the residual droplets evaporate and concentrate rapidly due to strong transpiration, and the absorption efficiency drops sharply, high light area, When I exceeds the saturation point, the chloroplast thylakoid membranes are damaged, the PSⅡ reaction center is inactivated, and the photosynthetic electron transport chain is blocked, which leads to reduced ATP production, exponentially decreasing active absorption efficiency. The effect of wind speed on absorption efficiency is essentially negated. Dynamic model optimization using a deep deterministic policy gradient reinforcement learning algorithm ensures the model's adaptability to diverse environments in different regions.

[0055] In the embodiment of the present application, the fertilization control module 400 includes: a fertilization unit and an instruction execution unit.

[0056] Among them, the fertilization unit is used to arrange field fertilization equipment, which includes fertilizer storage tanks, water storage tanks, pipeline networks, adjustable nozzles, and variable frequency pumps; the instruction execution unit is used to fertilize according to the control instructions of the fertilization equipment.

[0057] It can be understood that the fertilization unit in the embodiment of the present application constructs a complete field fertilization execution system by deploying fertilizer storage tanks, variable frequency pumps, adjustable nozzles and other equipment, providing hardware support for precise fertilization; the instruction execution unit drives the equipment to operate according to the control instructions, realizes the precise control of parameters such as fertilizer amount, spraying pressure, and range, and converts the decision-making plan of the instruction generation unit into actual operation, which not only ensures the automation and standardization of the fertilization process, but also adapts to the needs of different crops and different growth stages through adjustable equipment, thereby improving the practicality of the system and the stability of the fertilization effect.

[0058] In the embodiment of the present application, the instruction execution unit 400 further includes: an exception determination unit.

[0059] Among them, the abnormality judgment unit is used to collect the operating status of the pump group and valve opening and closing data in real time, receive the fertilizer liquid flow, pipeline pressure, air temperature and humidity, and soil EC value data from the data acquisition module, and compare the collected data with the preset threshold. When the preset conditions are met, it is judged as an abnormality and an alarm is triggered.

[0060] It can be understood that the abnormality judgment unit in the instruction execution unit in the embodiment of the present application realizes abnormality monitoring and alarm by collecting the operating data of equipment such as pump groups and valves in real time, as well as parameters such as fertilizer liquid flow, pipeline pressure, ambient temperature and humidity, and comparing them with preset thresholds. It can not only timely detect equipment failures (such as pump group abnormalities, valve failures) and environmental parameter abnormalities (such as pressure exceeding the limit, temperature and humidity inappropriateness), avoid equipment damage or fertilization interruption due to the expansion of faults, but also ensure the safety and continuity of the fertilization process through early warning, making up for the shortcomings of the traditional system's delayed abnormal response.

[0061] In the embodiments of the present application, the preset conditions include: the vibration amplitude, temperature or current value of the pump group exceeds the normal operation range of the equipment; the valve does not complete the opening and closing action within the specified time, or appears abnormal opening and closing state at non-command time; the fertilizer solution flow rate continuously below the minimum design flow rate or continuously above the maximum design flow rate; the pipeline pressure continuously exceeds the safe pressure range or is below the minimum pressure required to maintain normal operation of the system; the air temperature and humidity or the soil EC value continuously deviates from the suitable range for crop growth or the safe threshold of system operation.

[0062] Specifically, the pump group operation state monitoring: real-time acquisition of vibration sensor data, temperature sensor data and current transformer data of the pump group, when the vibration amplitude exceeds the stable running threshold set by the equipment manufacturer (such as abnormal vibration caused by impeller wear or bearing failure), the shell temperature exceeds the upper limit of the motor allowable working temperature (such as the risk of overload caused by continuous high temperature), or the working current deviates from the rated current ± 20% (such as sudden rise of current caused by pipeline blockage or sudden drop of current caused by impeller idling), it is immediately determined that the pump group is abnormal.

[0063] Valve state monitoring: obtain the opening and closing state of the valve through the stroke sensor or position feedback module, when receiving the control command, the valve does not complete the action within its designed full stroke time (such as 1-3 seconds for electromagnetic valve, 5-15 seconds for electric valve according to the diameter), or appears unexpected opening and closing without control command (such as misoperation caused by valve rod jam), and the state lasts more than 5 sampling periods (about 10 seconds), it is determined that the valve is abnormal.

[0064] Fertilizer solution flow rate and pipeline pressure monitoring: based on ultrasonic flowmeter, real-time acquisition of flow data, when the flow rate continuously below the minimum irrigation flow rate designed by the system (such as the lower limit value calculated according to the water requirement of single plant) or above the maximum safe flow rate (such as the upper limit value to avoid pipeline overload) for 3 minutes, it is determined that the flow rate is abnormal; at the same time, the pipeline pressure is monitored by pressure transmitter, if the pressure continuously exceeds 1.1 times of the rated pressure of the pipeline (such as 0.6 MPa for PVC pipeline, exceeding the limit may cause pipe burst) or is below 0.1 MPa (such as pressure drop caused by pipeline leakage) for 1 minute, it is determined that the pressure is abnormal.

[0065] Environmental and soil parameter monitoring: for air temperature and humidity, when the temperature continuously above the critical high temperature for crop growth (such as 35℃ for vegetables) or below 5℃ (may cause fertilizer solution crystallization) for 1 hour, the relative humidity continuously above 90% (easy to cause equipment condensation short circuit) or below 15% (extreme drought affects fertilizer absorption) for 2 hours, it is determined that the temperature and humidity are abnormal; for soil EC value, when continuously above the upper limit of crop salt tolerance (such as 2.5 mS / cm for strawberries) or below the soil basic fertility threshold (such as 0.5 mS / cm below, reflecting soil barrenness) for 1 hour, it is determined that the EC value is abnormal.

[0066] It can be understood that, by means of the explicit parameter correlation and the time threshold setting, the embodiments of the present application can not only avoid false positives caused by instantaneous fluctuations, but also respond in time at the initial stage of a fault or abnormal trend, so as to ensure that the system forms effective protection between device safety, fertilization effect and crop growth environment, and further strengthens the reliability of the entire fertilization control system.

[0067] In the embodiments of the present application, the fertilization scheme includes region number, fertilizer type, fertilization amount and fertilization time; and the fertilization control instruction includes region number, fertilizer tank number, opening time, closing time, frequency of the variable frequency pump and control value of the pump group pressure valve.

[0068] Specifically, the fertilizer storage tank includes a nitrogen fertilizer storage tank, a phosphorus fertilizer storage tank, a potassium fertilizer storage tank and an organic fertilizer storage tank. Each time of fertilization should ensure that the pipeline works independently, and the pipeline is flushed with clean water in the water storage tank before and after work to avoid cross contamination. The fertilizer storage tanks numbered 1, 2 and 3 are opened when the corresponding fertilizer is applied.

[0069] The fertilization control system for liquid fertilizer provided in the embodiments of the present application collects multidimensional data such as air temperature and humidity, wind speed, soil EC value and crop growth conditions through the acquisition module, and combines the nutrient requirements of different growth stages in the crop knowledge graph, avoiding the problem of excessive or insufficient fertilizer caused by the traditional fertilization relying on fixed cycle or experience value, significantly improving the fertilizer utilization rate and reducing resource waste and negative impact on the environment. Not only soil and crop data are included, but also meteorological factors such as light, rainfall and wind speed are focused on. The loss law of fertilizer solution in different environments (such as rainfall leaching and strong wind drift) is depicted through a liquid fertilizer loss dynamics model, and the model parameters are dynamically optimized using a deep deterministic policy gradient reinforcement learning algorithm, solving the limitations of single data dimension and poor adaptability of static models in existing systems, so that the fertilization scheme can accurately match the complex and variable actual conditions in the field. Through the real-time feedback mechanism, the fertilization control module can monitor the device state and environmental parameter abnormalities of the pump group, valve and pipeline pressure in the instruction execution unit, solving the problem of imperfect abnormal response mechanism in existing systems, avoiding the influence of fertilization effect or equipment damage caused by faults, and improving the stability and reliability of system operation. Thus, the problems of low precision, serious resource waste, single data dimension, poor model adaptability and imperfect abnormal response mechanism in the prior art are solved.

[0070] A specific embodiment of a fertilization control system for liquid fertilizer will be described below. Taking the flowering and fruit setting period of greenhouse tomatoes as an example, as shown in FIG. 1, the fertilization control system includes: Figure 2 In the interactive interface unit of the interactive module, the user inputs the crop type: tomato, the planting area: E-002, and the growth cycle: flowering and fruit setting period through the touch screen. ​

[0071] The crop knowledge graph loading unit loads the wheat seedling stage knowledge graph to obtain key parameters: nutrient demand: N: P: K = 3: 1: 4, and the fertilization response curve before the nutrient reaches the critical value is in the shape of a parabola, and it decreases rapidly after exceeding.

[0072] The data acquisition module collects the following data in real time: air temperature: 25°C, relative humidity: 62%, wind speed: 1.8 m / s, photosynthetic photon flux density: 800 μmol / m²・s, rainfall: 0 mm, sensor unit real-time acquisition: soil EC value: 2.5 mS / cm, soil NPK content: N = 150 mg / kg, P = 50 mg / kg, K = 200 mg / kg, fertilizer solution concentration: N = 150 mg / L, P = 50 mg / L, K = 100 mg / L, pipeline pressure: 0 MPa, crop growth status: fruit diameter is about 5 cm on average, and the data processing module performs noise reduction and standardization processing on the data.

[0073] The loss model unit in the instruction generation module constructs a liquid fertilizer loss kinetics model,

[0074] The laboratory experiment obtains the following parameters: soil runoff critical rainfall 15 mm, light compensation point density = 100 μmol / m²・s, light saturation point density = 800 μmol / m²・s, transpiration promoting critical wind speed = 1.5 m / s, drift dominant critical wind speed = 3.0 m / s, fertilizer solution concentration 200 mg / L, photosynthetic photon flux density I = 600 μmol / m²・s, wind speed V = 1 m / s, a = 0.003, b = 0.02, c = 0.06, d = 0.89, g = 0.2, h = 1.1, j = 0.7, k = 1.5, m = 0.002; After optimization by deep deterministic policy gradient reinforcement learning algorithm, a = 0.002, b = 0.01, c = 0.05, = 100 μmol / m²・s, = 800 μmol / m²・s, = 1.5 m / s, = 3.0 m / s, d = 0.85, g = 0.3, h = 1.2, j = 0.6, k = 1.5, m = 0.002.

[0075] The instruction generation engine unit generates a fertilization scheme by calculation: taking P as an example, according to the crop knowledge graph, the flowering and fruit setting period, the regional total P required is 5.8 kg, the fertilizer concentration is 200 mg / L, the total amount of fertilization during the flowering and fruit setting period is about 9.74 L per day, the air temperature and humidity, rainfall, light and photon flux density, wind speed every hour in the next 24 hours are obtained through the LSTM time series model, the optimal fertilization time window is 8 to 12 o'clock by substituting the liquid fertilizer loss kinetics model and combining the air temperature and humidity, the actual amount of fertilizer to be applied is 12.48 L, the fertilizer solution flow rate is 0.05 L / min, the fertilization scheme E-002, P fertilizer, 12.48 L, 8 to 12 o'clock is generated, the user modifies the time to 8 to 10 o'clock, the scheme is changed to E-002, P fertilizer, 11.19 L, 8 to 10 o'clock, and the fertilization equipment control instruction E-002, 003 (P fertilizer tank), 11.19 L, 7.58 open, 10.02 close, 30 Hz, 0, 3 MPa) is generated.

[0076] In the fertilization equipment control module, the instruction execution unit obtains and executes the fertilization equipment control instruction, 7.58 open, the fertilization is started, and the fertilization process is detected, no abnormality occurs in the middle, and 10.02 is normally closed.

[0077] After the fertilization is completed, the actual fertilizer loss is 0.19 kg, the predicted fertilizer loss is 0.29 kg, and the difference is -0.1 kg according to the real-time weather condition calculation, and the daily fertilization amount is reduced from 9.74 L to 9.72 L.

[0078] In summary, the embodiments of the present application obtain the nutrient requirements and nutrient ratios of crops at different growth stages from the crop knowledge graph, provide a basis for calculating the amount of fertilizer to be applied, deploy field weather stations and sensors to collect data, denoise and standardize the data, denoising can increase the reliability of the data, and standardizing the data facilitates system identification, build a liquid fertilizer loss kinetics model and optimize it, predict future weather parameters, generate a fertilization scheme based on the optimized liquid fertilizer loss kinetics model and the crop knowledge graph, transmit the scheme to the interactive interface for user parameter modification, finally determine the scheme and generate the fertilization equipment control instruction, which can improve the accuracy of fertilization, reduce fertilizer waste and pollution, and also avoid yield reduction due to insufficient fertilizer amount, thereby solving the problems of lack of consideration of complex field weather conditions and deep modeling of fertilizer loss.

[0079] Next, a fertilization control method for liquid fertilizer according to an embodiment of the present application is described with reference to the accompanying drawings.

[0080] As shown in Figure 3 , the fertilization control method for liquid fertilizer includes the following steps: In step S101, the crop knowledge graph, the crop growth state and the generation period are obtained.

[0081] It can be understood that the embodiments of the present application can accurately anchor the nutrient demand patterns and growth characteristics of crops at different stages by obtaining crop knowledge maps, crop growth status and growth cycles, and provide a scientific basis for fertilization decisions based on crop biological characteristics.

[0082] In step S102, the crop growth stage is obtained according to the crop growth status and growth cycle, the NPK amount required for the growth stage is obtained according to the crop knowledge map, and the total fertilizer amount and daily fertilizer amount required by the crop are calculated based on the soil NPK content and the concentration of NPK fertilizer.

[0083] It can be understood that the embodiment of the present application determines the growth stage by combining the crop growth status and cycle, obtains the NPK amount required for this stage based on the crop knowledge map, and then calculates the total fertilization amount and daily fertilization amount by associating the existing NPK content in the soil and the fertilizer concentration. This not only accurately fills the gap between soil nutrients and crop needs, avoids excess or insufficient nutrients caused by blind fertilization, but also realizes on-demand supply by refining the daily fertilization amount, so that the fertilization rhythm is highly matched with the crop growth rhythm, and effectively improves the fertilizer utilization rate.

[0084] For example, if it is monitored that tomatoes are in the fruit swelling stage (judged by fruit diameter and growth cycle), the crop knowledge map shows that 1.2kg / mu of nitrogen, 0.8kg / mu of phosphorus, and 1.5kg / mu of potassium are needed at this stage; according to sensor detection, the current nitrogen, phosphorus, and potassium contents in the soil are 0.5kg / mu, 0.6kg / mu, and 0.7kg / mu respectively, which means that 0.7kg / mu of nitrogen, 0.2kg / mu of phosphorus, and 0.8kg / mu of potassium need to be supplemented; combined with The NPK concentration in the liquid fertilizer applied (such as 100g / L nitrogen, 80g / L phosphorus, and 120g / L potassium) can be used to calculate the total fertilizer requirement (7L / mu of nitrogen liquid, 2.5L / mu of phosphorus liquid, and 6.7L / mu of potassium liquid), and then evenly distributed over the 20-day fruit expansion period to obtain the daily fertilizer amount (0.35L / mu of nitrogen liquid, 0.125L / mu of phosphorus liquid, and 0.335L / mu of potassium liquid), which not only accurately meets the nutrient needs of tomatoes during the fruit expansion period, but also avoids fertilizer waste.

[0085] In step S103, a liquid fertilizer loss dynamics model is constructed and optimized, and the air temperature and humidity, rainfall, photosynthetic photon flux density, and wind speed in the next 24 hours are predicted using the LSTM time series model. The predicted rainfall, photosynthetic photon flux density, and wind speed are input into the optimized liquid fertilizer loss dynamics model to obtain the estimated solute loss per unit volume. Combined with the predicted air temperature and humidity, the optimal fertilization time window is obtained, and the corresponding fertilization time and estimated fertilizer loss are calculated based on the optimal fertilization time window.

[0086] It can be understood that, by constructing and optimizing the liquid fertilizer loss kinetic model, combining the prediction of future 24-hour meteorological parameters by the LSTM time series model, and incorporating key meteorological factors into the fertilizer solution loss calculation, the expected unit volume solute loss amount is accurately obtained; at the same time, the best fertilization time window is determined by combining the air temperature and humidity prediction, and the fertilization duration and the expected fertilizer loss amount are calculated, which not only realizes the scientific prediction of the loss of fertilizer solution under different meteorological conditions, but also maximizes the loss reduction by selecting an appropriate period for fertilization, effectively making up for the fertilizer waste caused by the traditional fertilization ignoring the environmental impact, and further improving the accuracy of the fertilization scheme and the fertilizer utilization rate.

[0087] It should be noted that the prediction of future 24-hour meteorological parameters by the LSTM time series model is realized by constructing a neural network model including an input layer, a hidden layer (including LSTM units) and an output layer, and the core formula involved in the prediction process can be summarized as follows: for the input historical meteorological parameter sequence (such as air temperature and humidity, rainfall, etc. time step data), information filtering and updating are performed through the forget gate, input gate and output gate in the LSTM unit, wherein the forget gate determines to discard the historical state information through the sigmoid function, the input gate updates the cell state by combining the sigmoid function and the tanh function, and the output gate outputs the current hidden state based on the sigmoid function and the tanh function. Finally, the predicted values of each meteorological parameter in the next 24 hours are obtained through the mapping of the output layer, which provides reliable future environmental parameter input for the liquid fertilizer loss kinetic model and improves the accuracy of the prediction of fertilizer solution loss.

[0088] In step S104, the actual amount of fertilizer to be applied on the day is calculated according to the daily amount of fertilizer and the expected fertilizer loss amount, the fertilizer solution flow rate is calculated according to the actual amount of fertilizer to be applied on the day and the fertilization duration, and the variable frequency pump frequency is mapped, and the best spraying pressure is determined according to the adjustable nozzle model, which is converted into the pump group pressure valve control value to generate the fertilization scheme. The fertilization equipment control instruction is generated according to the fertilization scheme, and the fertilization is performed according to the fertilization equipment control instruction.

[0089] It can be understood that, by calculating the actual amount of fertilizer to be applied on the day according to the daily amount of fertilizer and the expected fertilizer loss amount, the application can ensure that sufficient nutrients are provided for crops under the premise of considering the loss of fertilizer solution, and avoid insufficient fertilization due to loss; secondly, the fertilizer solution flow rate is calculated according to the actual amount of fertilizer to be applied and the fertilization duration, and the variable frequency pump frequency is mapped, which realizes the accurate correspondence between the amount of fertilizer and the equipment operation parameters, and ensures the accurate execution of the amount of fertilizer; thirdly, the best spraying pressure is determined according to the adjustable nozzle model and converted into the pump group pressure valve control value, which takes into account the fertilization effect and equipment safety, and finally generates the fertilization scheme and control instruction and executes it.

[0090] According to a fertilization control method for liquid fertilizer proposed in an embodiment of the present application, an acquisition module collects multi-dimensional data such as air temperature and humidity, wind speed, soil EC value, and crop growth status, and combines it with the nutrient requirements of different growth stages in the crop knowledge graph. This avoids the problem of excessive or insufficient fertilizer caused by traditional fertilization relying on fixed cycles or empirical values, significantly improves fertilizer utilization, and reduces resource waste and negative impacts on the environment. It not only incorporates soil and crop data, but also focuses on integrating meteorological factors such as light, rainfall, and wind speed. The liquid fertilizer loss dynamics model is used to characterize the loss pattern of fertilizer liquid in different environments (such as rainfall leaching and strong wind drift), and the deep deterministic policy gradient reinforcement learning algorithm is used to dynamically optimize the model parameters. This solves the limitations of the existing system's single data dimension and poor adaptability of static models, so that the fertilization plan can accurately match the complex and changeable actual conditions in the field. The fertilization control module uses a real-time feedback mechanism, combined with the abnormality detection function in the instruction execution unit, to promptly monitor abnormalities in equipment status and environmental parameters such as pumps, valves, and pipeline pressure. This addresses the existing system's imperfect abnormality response mechanism, preventing fertilization effects or equipment damage caused by faults, and improving the stability and reliability of system operation. This addresses existing agricultural fertilization issues such as low accuracy, severe resource waste, single-dimensional data, poor model adaptability, and imperfect abnormality response mechanisms.

[0091] A fertilization control method for liquid fertilizer will be described below through a specific embodiment, taking cucumber planting as an example, including the following steps: In step 1, the crop type is input as "greenhouse cucumber" through the interactive module, and the corresponding crop knowledge map is loaded. The map contains the nutrient requirements of cucumbers in the germination period, seedling period, flowering and fruiting period (for example, the amount of nitrogen and potassium required during the flowering and fruiting period increases significantly) and the fertilization response curve; at the same time, the image sensor of the acquisition module collects growth status data such as cucumber leaf color, plant height, and fruit number, and combines it with planting records to determine that its growth cycle is 90 days. In step 2, based on leaf expansion (average leaf area of ​​25 cm²) and growth cycle (45 days of planting), the cucumbers are determined to be in the flowering and fruiting stage. The knowledge graph extracts that 2.0 kg / mu of nitrogen, 0.8 kg / mu of phosphorus, and 2.5 kg / mu of potassium are required during this stage. Sensor detection shows that the current nitrogen, phosphorus, and potassium contents in the soil are 1.2 kg / mu, 0.6 kg / mu, and 1.0 kg / mu, respectively, requiring additional nitrogen of 0.8 kg / mu, phosphorus of 0.2 kg / mu, and potassium of 1.5 kg / mu. Combined with the concentration of liquid fertilizer applied (150 g / L nitrogen, 100 g / L phosphorus, and 200 g / L potassium), the total fertilizer requirement is calculated to be 5.3 L / mu of liquid nitrogen, 2.0 L / mu of liquid phosphorus, and 7.5 L / mu of liquid potassium. Evenly distributed over the 30-day flowering and fruiting period, the daily fertilizer requirement is 0.18 L / mu of liquid nitrogen, 0.07 L / mu of liquid phosphorus, and 0.25 L / mu of liquid potassium. In step three, a liquid fertilizer loss kinetics model was constructed and parameters were optimized using the DDPG algorithm. An LSTM model was used to predict meteorological parameters for the next 24 hours: air temperature 18-28°C, humidity 60-70%, no rainfall, photosynthetic photon flux density 800-1200 μmol / m²·s, and wind speed 0.5-1.2 m / s. These meteorological parameters were input into the model, yielding an estimated solute loss per unit volume of 8%. Combined with the temperature and humidity forecasts, the optimal fertilization window was determined to be 9:00-11:00 the next day (when temperatures are suitable and wind speeds are low), corresponding to a fertilization duration of two hours. The estimated fertilizer loss was 8% of the daily fertilizer application rate. In step 4, the actual amount of fertilizer to be applied on that day is calculated: 0.196 L / mu of nitrogen liquid (0.18 L ÷ 92%), 0.076 L / mu of phosphorus liquid, and 0.272 L / mu of potassium liquid. Based on a 2-hour fertilization duration, the total flow rate of the fertilizer liquid is calculated to be 0.272 L / h (total fertilizer requirement ÷ 2h), which is mapped to a variable frequency pump frequency of 35 Hz. Based on the parameters of the adjustable nozzle (model PT-15), the optimal spraying pressure is determined to be 0.25 MPa, which is converted into a pump group pressure valve control value of 4.2 V. A fertilization plan and control instructions are generated, including "Area A - Cucumber - Flowering and Fruiting Period - 0.196 L / mu of nitrogen liquid - 9:00 on - 11:00 off - pump frequency 35 Hz - pressure 0.25 MPa", and executed by the fertilization control module. During the application process, the abnormality detection unit monitored in real time: the pump current remained stable at 3.2A (within the rated current range of 3.0±15%), the valve opened and closed within 2 seconds, the pipeline pressure remained at 0.25MPa, and the fertilizer flow rate remained stable at 0.136L / h. All parameters met the preset thresholds, indicating normal system operation. After this fertilization, the chlorophyll content of cucumber leaves increased by 12%, the fruit expansion rate accelerated, and the fertilizer utilization rate increased by 23% compared to traditional fertilization methods. There was no leaf curling caused by excess nutrients.

[0092] In summary, the embodiment of the present application obtains crop knowledge maps, growth status and cycles, determines the growth stage and calculates the total fertilizer application amount and daily fertilizer application amount, and then uses the model to predict meteorological parameters and fertilizer liquid loss to determine the optimal fertilization window. Finally, the actual fertilizer application amount is calculated, the equipment parameters are converted and fertilization is executed. The entire process combines multi-dimensional data and intelligent models to achieve precise fertilization, and abnormal monitoring ensures the stable operation of the system, which effectively improves fertilizer utilization and crop growth effects in practical applications.

[0093] In the description of the present specification, the description of the terms "one embodiment", "some embodiments", "an example", "a specific example", or "some examples" and the like means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. In the present specification, the illustrative expressions of the above terms are not necessarily directed to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any appropriate manner in any one or more embodiments or examples. Furthermore, the person skilled in the art can combine and combine the different embodiments or examples described in the present specification and the features of the different embodiments or examples, without contradiction.

[0094] In addition, the terms "first", "second" are used only for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the technical features indicated. Thus, the features defined with "first", "second" can include at least one of the features, explicitly or implicitly. In the description of the present application, the meaning of "a plurality of" is at least two, for example, two, three, etc., unless otherwise explicitly specified.

[0095] Any process or method descriptions in flow charts or described elsewhere herein can be understood as representing one or more steps of a method or process, including a set of executable instructions for implementing custom logic functions or processes, and the scope of preferred embodiments of the present application includes additional implementation involving other steps, which can be performed at substantially the same time or in reverse order or in other order, depending on the functionality involved, as will be understood by those skilled in the art of the embodiments described herein.

[0096] It should be understood that parts of the present application can be implemented in hardware, software, firmware or a combination thereof. In the above-described embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. As in another embodiment, if implemented in hardware, any of the following technologies known in the art or their combinations can be used: discrete logic circuit with logic gate circuit for implementing logic functions on data signals, application specific integrated circuit with suitable combination logic gate circuit, programmable gate array (PGA), field programmable gate array (FPGA) and the like.

[0097] Those skilled in the art of the present technology can understand that all or part of the steps carried out by the above-described embodiment method can be completed by a program instructing the relevant hardware. The program can be stored in a computer readable storage medium, and when executed, includes one or a combination of steps of the method embodiment.

[0098] Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limitations on the present application. Ordinary technicians in this field can change, modify, replace and modify the above embodiments within the scope of the present application.

Claims

1. A fertilization control system for liquid fertilizer, characterized in that: include: Interaction module, acquisition module, instruction generation unit and fertilization control module; wherein, The interactive module is used to input the crop type and load the corresponding crop knowledge graph; The acquisition module is used to collect air temperature and humidity, wind speed, photosynthetic photon flux density, and rainfall, and deploys sensor modules to collect soil EC value, NPK content, fertilizer solution concentration, fertilizer solution flow, pipeline pressure, and crop growth status data; The instruction generation unit is used to construct a liquid fertilizer loss dynamics model, and use a deep deterministic policy gradient reinforcement learning algorithm for dynamic optimization, use an LSTM time series model to predict future meteorological parameters, and generate a fertilization plan based on the optimized liquid fertilizer loss dynamics model, the crop knowledge graph and the predicted future meteorological parameters, and generate a fertilization control instruction according to the fertilization plan, wherein the liquid fertilizer loss dynamics model includes a rainfall correction function and a photosynthetic photon flux density-wind speed joint correction function, which is used to quantify the fertilizer liquid loss rate under different meteorological conditions, wherein the formula of the liquid fertilizer loss dynamics model is: ； in, is the solute loss per unit volume, For time, is the rainfall correction function value, is the correction function value of photosynthetic photon flux density and wind speed, is the concentration of fertilizer solution, is the model error, is the photosynthetic photon flux density, is the wind speed; The fertilization control module is used to receive the fertilization equipment control instruction, execute the fertilization equipment control instruction and provide real-time feedback.

2. The fertilization control system for liquid fertilizer according to claim 1, characterized in that: The interactive module includes: an interactive interface unit and a crop knowledge graph loading unit, wherein the interactive interface unit is used to support users to interactively modify the generated fertilization plan; the crop knowledge graph loading unit is used to load the corresponding crop knowledge graph according to the information, and the crop knowledge graph includes the nutrient requirements and fertilization response curves of crops at different growth stages.

3. The fertilization control system for liquid fertilizer according to claim 1, characterized in that: The acquisition module includes: a field weather station unit, a sensor unit, and a data processing unit, wherein: The field weather station unit is used to collect air temperature and humidity, wind speed, photosynthetic photon flux density, and rainfall; The sensor unit is used to collect soil EC value, NPK content, fertilizer concentration, fertilizer flow, pipeline pressure and crop growth status; The data processing unit is used for noise reduction and standardization processing of the data acquired by the field weather station unit and the sensor unit.

4. The fertilization control system for liquid fertilizer according to claim 1, characterized in that: The instruction generation unit includes a loss model unit and an instruction generation engine unit, wherein the loss model unit is used to construct and continuously update the liquid fertilizer loss dynamics model based on the actual loss data after fertilization; the instruction generation engine unit is used to generate a fertilization plan, and use a deep deterministic policy gradient algorithm to dynamically optimize the parameters of the liquid fertilizer loss dynamics model with the optimization goal of minimizing the deviation between the predicted fertilizer loss and the actual loss. The instruction generation engine unit supports users to interactively adjust the optimized fertilization plan and generate fertilization control instructions based on the adjusted fertilization plan.

5. The fertilization control system for liquid fertilizer according to claim 1, characterized in that: The formula for the rainfall correction function value is: ； in, is the critical rainfall for runoff generation, is the actual rainfall, is the nonlinear influence intensity, is the linear slope, is the basic correction value; The photosynthetic photon flux density and wind speed correction function expressions are as follows: ； in, is the light compensation point density, is the light saturation point density, The critical wind speed for transpiration promotion is is the drift-dominated critical wind speed, is the basic absorption efficiency coefficient, is the low wind speed transpiration promotion coefficient, is the drift loss coefficient at medium wind speed, is the high wind speed absorption reference coefficient, is the high wind speed drift attenuation coefficient, is the high light suppression coefficient, and e is a natural constant.

6. The fertilization control system for liquid fertilizer according to claim 1, characterized in that: The fertilization control module includes: a fertilization unit and an instruction execution unit, wherein: The fertilization unit is used to arrange field fertilization equipment, which includes a fertilizer storage tank, a water storage tank, a pipeline network, an adjustable nozzle, and a variable frequency pump; The instruction execution unit is used to perform fertilization according to the fertilization equipment control instruction.

7. The fertilization control system for liquid fertilizer according to claim 6, characterized in that: The instruction execution unit also includes: an abnormality determination unit, wherein the abnormality determination unit is used to collect the pump group operation status and valve opening and closing data in real time, receive the fertilizer liquid flow, pipeline pressure, air temperature and humidity, and soil EC value data from the data acquisition module, compare the collected data with the preset threshold, and when the preset conditions are met, determine it as an abnormality and trigger an alarm.

8. The fertilization control system for liquid fertilizer according to claim 7, characterized in that: The preset conditions include: the vibration amplitude, temperature or current value of the pump group exceeds the normal operating range of the equipment; the valve fails to complete the opening and closing action within the specified time, or an abnormal opening and closing state occurs during non-instruction periods; the fertilizer liquid flow rate is continuously lower than the minimum design flow rate or continuously higher than the maximum design flow rate; the pipeline pressure continuously exceeds the safety pressure range or is lower than the minimum pressure to maintain normal operation of the system; the air temperature and humidity or soil EC value continuously deviates from the suitable range for crop growth or the system operation safety threshold.

9. The fertilization control system for liquid fertilizer according to claim 1, characterized in that: The fertilization plan includes the area number, fertilizer type, fertilizer amount and fertilization time; the fertilization control instruction includes the area number, fertilizer storage tank number, opening time, closing time, variable frequency pump frequency and pump group pressure valve control value.

10. A method for a fertilization control system for liquid fertilizers according to any one of claims 1 to 9, characterized in that: The method comprises: Obtain crop knowledge graph, crop growth status and production cycle; The crop growth stage is obtained according to the crop growth state and growth cycle, the NPK amount required for the growth stage is obtained according to the crop knowledge map, and the total amount of NPK fertilizer required for the crop and the daily amount of NPK fertilizer are calculated based on the soil NPK content and the concentration of NPK fertilizer; Construct and optimize a liquid fertilizer loss kinetics model, using an LSTM time series model to predict air temperature and humidity, rainfall, photosynthetic photon flux density, and wind speed for the next 24 hours. The predicted rainfall, photosynthetic photon flux density, and wind speed are input into the optimized liquid fertilizer loss kinetics model to obtain the estimated solute loss per unit volume. Combined with the predicted air temperature and humidity, the optimal fertilization time window is obtained, and the corresponding fertilization duration and estimated fertilizer loss are calculated based on the optimal fertilization time window. The actual amount of fertilizer to be applied on the day is calculated based on the daily fertilizer amount and the expected fertilizer loss amount. The fertilizer liquid flow rate is calculated based on the actual amount of fertilizer to be applied on the day and the fertilization duration, and mapped to the frequency of the variable frequency pump. Based on the adjustable nozzle model, the optimal spraying pressure is determined and converted into a pump group pressure valve control value to generate a fertilization plan. A fertilization equipment control instruction is generated based on the fertilization plan, and fertilization is performed according to the fertilization equipment control instruction.

Images