Self-adaptive drip irrigation control system and method fused with soil solution optimal model

Abstract

The invention discloses a self-adaptive drip irrigation control system and method fused with a soil solution optimal model, and belongs to the field of agricultural irrigation. The system integrates a layered soil and soil solution monitoring module, a weather and crop growth monitoring module, a data acquisition and management module, an irrigation and fertilization execution module and an operation feedback module. Through deep-buried multi-layer soil sensing and solution collection, real-time monitoring of key parameters such as moisture and nutrients of different soil layers is achieved, and a dynamic target is set in combination with crop growth requirements and meteorological data. The system adopts a soil solution optimal model, performs dynamic balance analysis on monitoring data and target parameters, introduces a multi-factor linkage and layered regulation strategy, intelligently generates an optimal water and fertilizer scheme and accurately executes the optimal water and fertilizer scheme. Automatic recording and quality feedback are achieved in the whole process, model self-learning and scheme optimization are supported, the crop nutrient utilization rate is increased, resource waste is reduced, and efficient and intelligent precision agricultural water and fertilizer management is achieved.

Description

Technical Field

[0001] This invention belongs to the field of agricultural irrigation, and more specifically relates to an adaptive drip irrigation control system and method that integrates an optimal soil solution model. Background Technology

[0002] With the increasing pursuit of high-yield, high-quality, and resource-efficient crop utilization in modern agriculture, refined water and fertilizer management has become a core element in improving farmland productivity and sustainable environmental development. Traditional irrigation and fertilization methods rely heavily on manual experience or single soil moisture indicators, lacking systematic monitoring of soil solution nutrient dynamics and soil layer differentiation. This easily leads to low nutrient utilization, crop supply and demand imbalances, and water and fertilizer waste, and may even exacerbate secondary soil salinization and other environmental problems. In recent years, the rapid development of sensor and Internet of Things (IoT) technologies has provided a technological foundation for the intelligent and automated management of field water and fertilizer.

[0003] However, existing intelligent drip irrigation systems often focus on feedback regulation of surface soil moisture or simple meteorological factors at a single point, lacking comprehensive linkage capabilities across multiple soil layers and factors. This makes it difficult to achieve precise responses to the heterogeneity of different crops, growth stages, and soil layers. Meanwhile, as the medium through which plants directly absorb nutrients, real-time, stratified monitoring and dynamic regulation of parameters such as soil solution (moisture, concentration of major nutrient ions, pH, and salinity) are crucial for ensuring efficient crop absorption, reducing nutrient leaching, and mitigating environmental risks. Therefore, there is an urgent need to develop a novel intelligent drip irrigation control system that integrates an optimal model of stratified soil solution, considers meteorological and crop growth conditions, and possesses adaptive feedback optimization capabilities. This system would provide strong support for efficient agricultural production and the coordinated development of resources and the environment. Summary of the Invention

[0004] This invention aims to address the problems of existing irrigation and fertilization systems, such as the lack of precise monitoring and dynamic control of soil solution states in different soil layers, inability to take into account meteorological changes and crop growth needs, resulting in low water and fertilizer utilization efficiency and resource waste. It provides an intelligent drip irrigation control technology that can integrate the optimal model of layered soil solution, adaptively adjust irrigation and fertilization schemes, achieve precise and efficient supply of water and nutrients in the crop root zone, significantly improve agricultural resource utilization and reduce environmental risks.

[0005] To achieve the above objectives, the present invention employs the following technical solution: The system includes: The soil solution monitoring unit uses a combination of deep-buried multi-parameter soil sensors and soil solution collectors, which are deployed in different soil layers of the field. The sensor depth is set according to the distribution of crop roots. The meteorological and crop growth monitoring unit uses an integrated field meteorological station and portable crop physiological parameter monitoring equipment, including temperature and humidity sensors, rainfall meters, wind speed and direction instruments, and solar radiation sensors. It automatically collects climate data and synchronizes it to the data acquisition and management platform wirelessly. The data acquisition and management platform, based on the Internet of Things architecture, integrates the data outputs of various sensor terminals and collectors, aggregates and pre-processes the data using an edge computing gateway, and then transmits it to the cloud server via 4G / 5G or LoRa wireless network. The irrigation and fertilization execution system consists of a control host, distributed solenoid valves, a main / branch pipeline network, and an automatic fertilizer dispenser. The control host automatically calculates the required water and fertilizer input based on the latest sensing data and decision-making models, and transmits the data to the zone actuators via RS485, Modbus, or wireless means. The main and branch pipelines are equipped with flow meters and pressure sensors to ensure uniform and accurate irrigation in each zone. The operation monitoring and quality feedback module tracks the implementation effects of irrigation and fertilization operations and the crop growth status throughout the entire process.

[0006] In one aspect, the method includes: Active monitoring of soil and solution conditions at different depths, including water content, main nutrient molecules, pH value, and salinity, to ensure coverage of the main root zone and leaching layer; Set target parameters for crop growth period, based on crop variety and growth stage, and clarify the most suitable comprehensive threshold range of soil solution (concentration of water, nitrogen, phosphorus, potassium, and pH) for each period. Combine meteorological conditions and historical experience in increasing yield to formulate phased nutrient supply and soil moisture targets. The system generates dynamic equilibrium analysis and optimal control schemes. It adopts the optimal soil solution model, introduces a hierarchical dynamic weighting mechanism and the linkage of meteorological-crop-soil ternary factors, and performs weighted control on the actual water and fertilizer requirements of different soil root zones. It also combines meteorological forecasts and corrects soil water and fertilizer scheduling parameters to improve crop nutrient utilization and reduce nutrient leaching and resource waste. Precise execution and full-process recording of operations; control results are sent to the intelligent control host to drive the irrigation valves to start in zones and the water and fertilizer machine to accurately mix; the amount of fertilizer used in each area is controlled in layers according to the target threshold. Feedback on results and further optimization of the plan; continuous tracking of crop growth indicators and comparison with initial targets; for plots that do not meet the standards or fluctuate significantly, adjusting solution parameters and management plans.

[0007] In one scheme, the active monitoring of the stratified soil and solution state includes: selecting monitoring points in different representative areas of the field based on crop root distribution and soil heterogeneity, and burying multi-parameter soil sensors and soil solution collectors in the main root zone, the infiltration layer and the necessary deeper soil layers. Using a high-precision integrated sensor for soil moisture, pH, conductivity, and temperature, and a negative pressure ceramic cup or polymer membrane infiltration device, soil moisture, pH, salinity, and major nutrient molecule data of different soil layers are collected on a regular basis and transmitted to a cloud platform for analysis via the Internet of Things.

[0008] In one approach, setting target parameters for crop growth period includes: retrieving relevant scientific research literature, soil fertility test results, and local planting experience based on the characteristics of the crop variety and the needs of each growth stage, combined with real-time and historical meteorological data. Using crop water and fertilizer requirement models and soil-crop-meteorological ternary balance analysis, the comprehensive threshold ranges of soil moisture, nitrogen, phosphorus, potassium, pH and electrical conductivity at each stage were dynamically set.

[0009] In one scheme, the generation of dynamic equilibrium analysis and optimal control scheme includes: receiving and integrating real-time soil solution monitoring data from different depths, and automatically comparing them with the target parameter range to obtain the current dynamic gap; Using the optimal soil solution model, the soil profile is stratified and a stratified dynamic weighting mechanism is introduced. Combined with meteorological input and crop growth data, a ternary regulation equation of meteorology-crop-soil is established, which couples soil moisture transport and dynamic mass balance of ion concentration. Through an information feedback mechanism, the weight parameters and leaching penalty are dynamically optimized based on historical data and crop growth response to achieve adaptive regulation. Using the improved particle swarm optimization algorithm, the optimal hierarchical input is determined, generating the most economical and efficient water and fertilizer scheduling scheme in batches, which is then distributed to the execution system for implementation.

[0010] In one approach, the precise execution and full-process recording of operations include: the optimized water and fertilizer input instructions are automatically sent to the intelligent control host in each area through the Internet of Things platform; the host intelligently calls each group of irrigation valves to achieve zoned and layered irrigation control; and the optimal fertilizer solution is precisely input to the designated soil layer and area through the automatic proportioning system of the integrated water and fertilizer machine.

[0011] In one approach, the result feedback and approach optimization include: comprehensively tracking the crop growth status, final yield, and quality of the target area across multiple indicators, and comparing and analyzing them with the target parameters; for plots with deviations, actively correcting the solution model parameters and dynamically adjusting the strategy.

[0012] In one embodiment, the improved particle swarm optimization algorithm includes: encoding the water and fertilizer input at each layer and at each time step into the dimension of the particles, and introducing weight distribution and gradient transfer domain knowledge into an adaptive update mechanism for particle velocity and inertial weights. It combines individual and global optimal guidance with domain gradient influence correction to achieve dynamic optimization under multi-objective conditions; The objective function considers the requirements of sufficient nutrients, minimum leaching, and lowest cost, and uses a constraint penalty function to handle interval thresholds and maximum input agronomic constraints to prevent the objective parameters from exceeding limits and resources from being wasted. Each round automatically selects the current optimal solution, using random perturbation and dynamic parameter adjustment to improve search accuracy and convergence speed.

[0013] Beneficial effects of this invention: This invention achieves precise perception and decentralized control of the dynamic state of moisture and nutrients in different soil layers by organically integrating a layered soil solution optimal model with an intelligent drip irrigation control system, breaking through the limitations of traditional single monitoring and programmed management.

[0014] The system can dynamically generate and optimize water and fertilizer input plans based on crop growth stage, soil stratification characteristics, and meteorological factors, ensuring that the crop root zone is always in an optimal state of water and nutrient supply. This significantly improves water and fertilizer utilization efficiency and reduces the risk of nutrient leaching and environmental pollution. Furthermore, this invention supports fully automated operation and feedback correction, possessing adaptive learning and regional adaptability capabilities. This not only ensures a steady increase in crop yield and quality but also significantly reduces labor input and management costs, promoting the green and efficient development of modern agriculture. Attached Figure Description

[0015] Figure 1 This is a system block diagram of the present invention; Figure 2 This is a flowchart of the method of the present invention; Figure 3 This is a flowchart of the dynamic balance analysis and optimal control process of this invention. Detailed Implementation

[0016] To facilitate understanding of the present invention, a more complete description will be given below with reference to the accompanying drawings. Typical embodiments of the invention are shown in the drawings. However, the invention can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete.

[0017] Unless otherwise defined, all technical and scientific terms used in this invention have the same meaning as understood by one of ordinary skill in the art to which this invention pertains. The terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. To facilitate understanding, the invention will now be described more fully with reference to the accompanying drawings. Typical embodiments of the invention are shown in the drawings. However, the invention can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided to make the disclosure of the invention more thorough and complete.

[0018] like Figure 1 As shown, an adaptive drip irrigation control system that incorporates an optimal soil solution model is presented. The system consists of the following components: The soil solution monitoring unit employs a combination of deeply buried multi-parameter soil sensors and soil solution collectors deployed in different soil layers (such as the main root zone and the infiltration layer) within the field. The sensor depth can be precisely set according to the crop root distribution, typically divided into the surface layer (10-20cm), the root active layer (30-50cm), and the leaching layer (50-100cm). Each monitoring point is equipped with a highly sensitive soil moisture sensor, conductivity sensor, pH sensor, and a collector capable of automatically collecting soil solution samples. The collected solution is periodically extracted to a surface sampling port for on-site or laboratory analysis of key nutrient ions (such as NH4). + NO3 - PO4³ - K + The concentration of soil water and fertilizer is measured. All sensors are synchronously driven by a low-power data acquisition unit and have automatic calibration and anti-contamination functions to ensure data accuracy and long-term reliability. The monitoring frequency and time interval can be remotely set as needed to achieve dynamic high-resolution sensing of the soil water and fertilizer environment.

[0019] The meteorological and crop growth monitoring unit utilizes an integrated field weather station and portable crop physiological parameter monitoring equipment. The weather station deployment includes temperature and humidity sensors, rainfall meters, anemometers, and solar radiation (sunlight) sensors, automatically collecting climate data and wirelessly synchronizing it to a data platform. Crop growth monitoring uses non-contact laser distance meters or ultrasonic distance meters to record plant height, and imagers or multispectral cameras to analyze leaf area index and canopy leaf color, while infrared thermometers detect canopy temperature and transpiration. Some parameters, such as leaf nitrogen content and chlorophyll index, can be periodically measured using portable foliar instruments. Each device automatically collects data at a set frequency. The system possesses AI image recognition and data preprocessing capabilities, which can initially filter out outliers, ensuring real-time synchronization and accurate feedback of climate and crop status.

[0020] The data acquisition and management platform, based on an IoT architecture, integrates data outputs from various sensor terminals and data collectors. After aggregation and preliminary processing via an edge computing gateway, the data is transmitted to a cloud server through 4G / 5G or LoRa wireless networks. The platform includes a highly structured database for storing heterogeneous data from multiple sources, such as soil, meteorology, and crops, and supports real-time and historical data retrieval, analysis, and visualization. The embedded data management software features automatic anomaly labeling, data integrity checks, and redundant backup functions, facilitating backend model calls and long-term trend analysis. The user interface is user-friendly, enabling remote data access, chart display, and system parameter settings, while also providing open API interfaces to support seamless integration with third-party intelligent analysis modules.

[0021] The irrigation and fertilization execution system consists of a control host, distributed solenoid valves, a main / branch pipeline network, and an automatic fertilizer dispenser. Based on the latest sensor data and decision-making models, the control host automatically calculates the required water and fertilizer input and transmits this information to the zone actuators via RS485, Modbus, or wirelessly. The main and branch pipelines are equipped with flow meters and pressure sensors to ensure uniform and precise irrigation in each zone. The multi-functional fertilizer dispenser automatically mixes fertilizer solutions of different concentrations according to crop stage and deficiencies, and then injects them into the water flow through electric valves, achieving true integrated water and fertilizer management. The system has task scheduling, execution feedback, and manual intervention interfaces. It can automatically alarm and shut off the relevant pipeline areas in case of abnormal flow or equipment failure, minimizing resource waste and operational risks.

[0022] The operation monitoring and quality feedback module is responsible for tracking the implementation effects of irrigation and fertilization operations and crop growth throughout the entire process. The system automatically compares predetermined thresholds with actual monitoring results, setting early warning thresholds for key indicators such as water, nutrient utilization, crop growth dynamics, and equipment operating conditions. The platform has a built-in quality tracking and traceability mechanism; all operation data and effect evaluations are compiled into electronic archives, providing decision-making references for managers and farmers. Upon detecting anomalies, the system generates new control plans and distributes them to the execution layer through automated suggestions or remote expert diagnosis, forming a closed loop of "monitoring-analysis-adjustment-feedback." Data accumulation can be used for subsequent model optimization and regional experience summarization, achieving intelligent, refined, and sustainable improvement in farmland management.

[0023] like Figure 2 As shown, an adaptive drip irrigation method incorporating an optimal soil solution model is proposed. The implementation steps are as follows: Step 1: Active monitoring of stratified soil and solution conditions Deep-buried soil sensors and solution samplers are used in key areas of the field to monitor the water content, main nutrient molecules, pH value, and salinity of soil layers at various depths regularly or in real time, ensuring coverage of the main root zone and leaching layer.

[0024] In representative areas of the field, appropriate monitoring points were selected based on crop root distribution and soil heterogeneity. Multi-parameter soil sensors and soil solution collectors were installed in the main root zone (20-40 cm), the infiltration layer (40-80 cm), and, where necessary, deeper soil layers. The soil sensors employed high-precision integrated sensors for soil moisture, pH, conductivity, and temperature, accurately capturing the dynamics of water content, pH, and salinity at different soil layers. The solution collectors used negative-pressure ceramic cups or polymer membrane permeation devices to actively extract soil solution samples from the target soil layer at set time intervals. The samples were then directly used after simple filtration for analysis with ion-selective electrodes or integrated nutrient analysis modules to detect ammonium nitrogen (NH4+). + ), nitrate nitrogen (NO3) - ), phosphorus (PO4³) - ), potassium (K) + Rapid detection of key nutrient molecules such as nitrogen, phosphorus, and nitrogen. Each sensor and acquisition device automatically powers on / off, collects data at set intervals, and performs self-diagnostics via a data acquisition terminal. The system has a reserved communication interface to wirelessly transmit real-time or periodic monitoring data to a field gateway via the Internet of Things (such as LoRa, NB-IoT, or 4G / 5G), and then uploads it to a cloud platform. To ensure data continuity and representativeness, the monitoring frequency can be dynamically adjusted according to the crop growth stage, and can be increased to minute-level real-time monitoring during critical growth periods and extreme weather. Data from all soil layers are synchronously labeled, and the dynamic profile can be intuitively queried on the platform, forming a continuous water and fertilizer environment change map. This stratified, high-resolution active monitoring method provides a solid data foundation and scientific basis for subsequent precise regulation.

[0025] Step 2: Set target parameters for crop growth period Based on crop varieties and growth stages, the most suitable comprehensive threshold range of soil solution (such as water, nitrogen, phosphorus, potassium concentration, pH, etc.) for each stage is determined. Combined with meteorological conditions and historical experience in increasing yield, phased nutrient supply and soil moisture targets are formulated.

[0026] In the stage of setting target parameters for crop growth, it is first necessary to systematically compile the optimal soil moisture and nitrogen (NH4+) levels for each growth stage of the crop, based on the variety characteristics and physiological requirements of each growth stage (such as emergence, jointing, heading, and grain filling), by referring to relevant scientific research literature, soil fertility test results, and past local planting experience. + / NO3 - ), phosphorus (PO4³) - ), potassium (K) +The system comprehensively assesses threshold ranges for parameters such as pH and conductivity. Combining real-time and historical meteorological data (e.g., rainfall, temperature, evaporation), it utilizes crop water and fertilizer requirement models and soil-crop-meteorological ternary balance analysis to dynamically correct and optimize the reasonable ranges of various indicators. For example, nutrient supply thresholds are appropriately increased during periods of rapid growth, while inputs are reduced during the slow-growth period and before harvest to prevent nitrogen leaching and quality decline. Furthermore, through remote sensing, environmental, and field data mining of historically high-yielding fields, yield-increasing factors are summarized, such as optimal water management models and key fertilization strategies. Big data analysis is used to assist in setting stratified soil moisture (water ranges), phased nutrient concentration targets, and safe pH ranges for the current field and growth stage. Ultimately, these target parameters serve as constraints for full-cycle intelligent monitoring and dynamic regulation, written into the system database to support automatic early warning and adjustment. If new monitoring data deviates from the target range, the system can quickly issue early warnings, diagnose the issue, and provide precise targeting for subsequent regulation models. This entire process not only ensures the scientific nature and regional adaptability of decision-making but also provides a crucial foundation for improving resource utilization efficiency and crop yield stability.

[0027] Step 3: Generate dynamic equilibrium analysis and optimal control scheme Based on the real-time soil solution data monitored in step 1, a dynamic comparison is made with the target parameters in step 2. An optimal soil solution model is employed, incorporating a stratified dynamic weighting mechanism and a linkage between meteorological, crop, and soil ternary factors. This allows for decentralized regulation of the actual water and fertilizer requirements of different soil root zones. Combined with meteorological forecasts and autonomous learning feedback, soil water and fertilizer scheduling parameters are corrected to precisely improve crop nutrient utilization and significantly reduce nutrient leaching and resource waste. The actual deficit in the current field (i.e., the gap between the target and the current situation) is calculated. Through model derivation, the most economical and effective water and fertilizer input schemes are designed in stratified and phased stages, ensuring sufficient nutrient and water supply to the root zone while avoiding excessive nutrient leaching or salt accumulation. This process highlights the core regulatory role of the optimal soil solution model.

[0028] like Figure 3 As shown, S301 first receives and integrates real-time soil solution monitoring data from different depths (such as the main root zone, leaching layer, etc.). (Represents the real-time value of the i-th layer, index j such as moisture / nitrogen / phosphorus / potassium, etc., at time t), and the range of target parameters set in step 2. Automatic comparison and calculation are performed. The difference represents the current dynamic gap. The system employs a stratified soil solution optimization model, dividing the soil profile into n layers, and introducing a stratified dynamic weighting mechanism based on the actual contribution of each layer to crop absorption. ( ,and The primary root zone is assigned a high weight, while the leaching layer is assigned a low weight. This is based on a comprehensive consideration of soil water and fertilizer dynamics and meteorological inputs. Rainfall, evaporation, temperature), crop growth data ( (Leaf area, plant height, water consumption rate), establish a ternary regulation equation coupling meteorology, crop, and soil: in: Let be the amount of water and fertilizer input that needs to be replenished at the current moment for the i-th level, index j (a decision variable to be optimized). Simulate the soil state at each time step after input I; This indicates penalties such as excessive leaching / nutrient loss / salt accumulation. Its weighting factor; The goal is to ensure that the water and nutrients in each root zone are close to the optimal threshold, while avoiding excessive irrigation that would waste resources.

[0029] S302. Coupled with soil moisture transport and ion mass balance equations, the migration and dissipation of different interlayer recharge inputs are predicted and corrected: 1. Water transport: in: Soil moisture content of the i-th layer; Current layer supply amount; Evaporation / Crop water consumption; Infiltration / leakage.

[0030] 2. Dynamic mass balance of ion concentration: in: The i-th layer j-th index (e.g., NO3) - K + )concentration; Crop root absorption; Nutrient loss due to water displacement.

[0031] S303. Based on the information feedback mechanism, the system autonomously learns from historical data and crop growth responses (the coupling relationship between yield, quality, and water and fertilizer changes), and continuously optimizes the weight parameters. , Loss of penal To achieve dynamic self-adaptation: (Where Y represents recent production volume / product quality,) (for learning rate) S304. Through model derivation, under the premise of satisfying the interval threshold, not exceeding the maximum input, and minimizing resource waste, the particle swarm optimization algorithm is improved. The optimal stratified input amount is solved by the improved particle swarm optimization algorithm with complex soil-crop-meteorological characteristics. It generates the most economical and effective water and fertilizer scheduling plans in batches (layered and timed) and distributes them to the execution system for implementation.

[0032] In the specific solution of the optimal hierarchical input quantity Furthermore, when generating cost-effective batch water and fertilizer scheduling schemes, an improved particle swarm optimization algorithm tailored to the complex characteristics of soil, crops, and weather is employed. This algorithm combines the search and constraint handling capabilities for optimal input quantities with domain knowledge to dynamically optimize the water and fertilizer supply amount at each layer and at each time step under multi-objective conditions (such as sufficient nutrients, minimal leaching, and lowest cost).

[0033] The improved particle swarm optimization process is as follows: First, for each optimization cycle, the variables to be optimized, such as water and fertilizer inputs, are... Encoded as the dimension of the particles, the position of each particle This represents the current scheme (where n is the number of layers, m is the number of water and fertilizer types, and p is the particle number). To reflect the coupling effects between soil layers and the actual scheduling constraints, knowledge from areas such as weight distribution and gradient migration was introduced, and the following adjustments were made to the inertial weights and velocity adaptation: The formulas for updating particle velocity and position are: in: The velocity vector of the p-th particle in the k-th iteration; The inertia weight is dynamically adjusted with each iteration, specifically as follows: in , To preset the maximum and minimum weights, The maximum number of iterations; , Individual / group learning factors; Uniformly distributed random numbers in the interval [0,1] improve search diversity; The current best historical position (local extremum) of the p-th particle; The globally optimal position in the iteration history of all particles; Introducing a domain-specific adjustment coefficient to account for the influence of soil physical response gradient; The gradient of the objective function at the current position expresses the system's sensitivity to changes in soil water and fertilizer—approximately calculated using a domain model.

[0034] objective function For multi-objective weighted penalty type: in: enter The resulting model predicts soil conditions; Target value (midpoint of the interval); Losses such as excessive leaching / salt accumulation (calculated based on soil migration model); Comprehensive cost function (measured cost, resource input). As weight; , , This is a dynamic weighting factor.

[0035] To meet agronomic requirements such as "interval threshold", not exceeding the maximum input, and minimizing waste, a constraint penalty function is adopted: in: Maximum allowed input for each layer and each nutrient type; Minimum soil parameter requirements (lower threshold); High-weighted penalty coefficient.

[0036] During each search round, the particle swarm optimization (PSO) is guided towards the input region that satisfies the objective and constraints through velocity and displacement updates, history-population optimal guidance, and neighborhood gradient fine-tuning. At the end of each iteration, the PSO is automatically selected based on the lowest... The value of the particle is taken as the current optimal solution. The algorithm also introduces random perturbation and dynamic parameter adjustment mechanisms to prevent getting trapped in local optima and improve optimization accuracy and convergence speed.

[0037] Finally, when the convergence condition is met or the set number of iterations is reached, the current value is extracted. The corresponding input vector is the optimal input for different soil layers within different time periods. Based on this, the system automatically generates water and fertilizer supply operation plans for each zone and each time period, achieving precise input at different levels and times, fully taking into account the actual needs of crops, efficient use of resources, and environmental safety, and then distributing the plans to the intelligent execution system for precise irrigation and fertilization.

[0038] Step 4: Precise Execution and Complete Work Recording The control results are sent to the intelligent control host, driving the irrigation valves to start in zones and the integrated water and fertilizer machine to precisely mix the fertilizer, adjusting the dosage in each area according to target thresholds. The entire operation is automatically recorded, and alarms can be triggered in a timely manner when irrigation anomalies, equipment malfunctions, or abnormal water and fertilizer feedback occur. Through synchronous response between the soil layer and crop, real-time response to the actual needs of the crop is achieved, rather than programmed management.

[0039] During the precise execution and full-process recording phase, optimized water and fertilizer input instructions are automatically sent to the corresponding smart control hosts in each area of ​​the field via an IoT platform. Based on the dynamic replenishment needs of each soil layer and crop root zone, the hosts intelligently activate each set of irrigation valves to achieve zoned and layered irrigation control. Through the integrated water and fertilizer machine's automatic mixing system, the calculated optimal fertilizer solution (a proportionally mixed blend of nitrogen, phosphorus, potassium, and other nutrients, with concentrations accurate to the target range) is precisely input into different pipeline networks and irrigation units. The entire process follows the optimal input volume. The spatiotemporal distribution is automatically executed, and each execution link includes fertilizer pump speed, valve opening and closing, and real-time flow rate. ) and the actual injection volume per layer ( All of these processes are meticulously controlled and recorded to ensure that every drop of irrigation water and fertilizer is accurately measured to the designated layer and area.

[0040] Meanwhile, the system automatically records all key data throughout the entire operation process digitally, including the actual amount of irrigation water used, execution time, duration, current valve status, equipment operating parameters, and real-time feedback on soil and crop status for each operation. The central management platform continuously monitors the integrity and safety of the operation chain. When abnormal irrigation flow, equipment failure, communication interruption, or significant discrepancies between soil / crop monitoring results and model predictions are detected, the system automatically sends real-time alarm information to the management terminal and can trigger automatic shutdown, overload protection, or user intervention to minimize the risk of anomalies.

[0041] During the execution period, the system continuously and synchronously collects feedback data from soil profiles and crops (such as soil moisture content, nutrient concentration, crop leaf color, and other physiological indicators), and dynamically calculates the deviation between the current response and the target. This allows for real-time adjustments to subsequent water and fertilizer inputs, forming a three-dimensional closed-loop control system encompassing soil, crops, and operations. This enables dynamic and precise management guided by the actual needs of crops, completely replacing the previous static and formulaic irrigation and fertilization operations, and greatly improving nutrient utilization and operational intelligence.

[0042] Step 5: Results Feedback and Solution Optimization After the operation is completed, crop growth indicators (soil solution, crop growth, yield and quality, etc.) are continuously tracked and compared with the initial targets. For plots that do not meet the standards or fluctuate significantly, the solution parameters and management plan are adjusted to trace the causes, and data is gradually accumulated to improve regional adaptability and scalability.

[0043] During the results feedback and scheme optimization phase, after the operation is completed, the system will continuously track the crop growth status in the target area in all aspects. The main monitored indicators include soil solution status in different soil layers (such as moisture, nitrogen, phosphorus, potassium ion concentration and pH value), crop growth (such as leaf area index, plant height, leaf color, stem diameter and biomass and other physiological and ecological indicators), and final yield and quality data (such as grain content, protein content, marketable rate, etc.). This data is collected regularly through multiple means such as field IoT sensors, automated image analysis systems and manual sampling and testing, and uploaded to the central database in real time.

[0044] The system automatically summarizes the actual crop performance throughout the entire growth cycle and compares it with the target parameters set in the initial step 2, calculating the deviation of each key indicator. For plots with significant deviations or large process fluctuations, the system will activate the retrospective diagnostic module, using methods such as time-series data mining, causal analysis, and anomaly detection to locate the main factors affecting crop performance (such as extreme weather events, equipment failures, local fertility imbalances, etc.), and actively correct the core parameters of the solution model (stratification weights, migration loss coefficients, crop absorption efficiency, etc.) based on the feedback results.

[0045] During the optimization process, the system utilizes historically accumulated data to conduct model self-learning, continuously improving the solution model's adaptability to different crops, varieties, and habitats through machine learning algorithms (such as Bayesian regression, random forests, and deep neural networks). Based on feedback reports, management dynamically adjusts subsequent management strategies, such as altering the spatiotemporal distribution of fertilizer and water inputs, providing targeted remedial measures for abnormal areas, and solidifying and promoting high-performing technical elements. All optimization iterations and parameter corrections are meticulously recorded for long-term traceability and continuous improvement.

[0046] This closed-loop feedback system enables the entire system not only to achieve precise one-time management, but also to continuously eliminate system biases and improve the model's ability to generalize to regional ecological, climatic and crop diversity conditions, laying a solid data foundation and decision support for subsequent large-scale promotion.

[0047] Example: I. Experimental Background and Site Setup This embodiment was conducted in a smart greenhouse in a modern agricultural park, with tomatoes as the crop and a planting density of 20,000 plants per hectare. The greenhouse area is 1 hectare, the soil is light loam, and there are 20 typical monitoring units (monitoring points). The system equipment is fully equipped as described in the "System Composition" section, and sensors and solution sampling devices are embedded in the three soil layers: the main root zone (0~20cm), the secondary root zone (20~40cm), and the leaching zone (40~60cm).

[0048] II. Monitoring and Target Parameter Setting Based on the water and fertilizer requirements of tomatoes at different growth stages, and in conjunction with meteorological and crop growth monitoring, target parameters for key stages are set (see table below).

[0049] III. Monitoring Data Collection and Anomaly Identification (Taking Typical Data from a Monitoring Point as an Example) During the early flowering-fruit setting stage (May 10th), the following data were collected by the stratified soil and solution sensors: Compared with the target parameters, it was found that the water and nutrients (especially NO3-N and K) in the main root zone (0-20cm) were significantly lower than the requirements, indicating a large water and fertilizer deficit.

[0050] IV. Model Analysis and Operational Decision Making By calling the optimal soil solution model and inputting the differences between the actual and target values ​​for each layer, combined with the greenhouse weather forecast for the next 3 days (expected temperature 24-32℃, no rainfall, strong sunshine), the model automatically assigns dynamic weights (60% for the primary root zone, 30% for the secondary root zone, and 10% for the leaching zone) to derive the optimal control scheme. The system calculates the following additional irrigation and fertilizer required at this monitoring point: Add 8L / m² of water-fertilizer mixture to the 0-20cm layer, with the following ratio: NO3-N 80mg / L, K 250mg / L; Add 4L / ㎡ of water to the 20-40cm layer through seepage irrigation, and appropriately increase K by 60mg / L; Adjust the pH of the fertilizer solution to around 6.5 to ensure it meets the suitable environment for the root zone.

[0051] Typical decision parameters and execution configuration table V. Precise Execution and Record Keeping Based on the model output, the intelligent control host activates the zoned drip irrigation valves, and the integrated water and fertilizer machine automatically mixes the fertilizer according to the specified ratio. The entire execution process is recorded: valve opening and closing, flow rate in each zone, EC / pH monitoring of the fertilizer solution, operation time, etc. If any process parameters are abnormal (such as insufficient water flow or deviation in fertilizer mixing ratio), the system will immediately alarm, and manual intervention will be required in a timely manner.

[0052] VI. Post-operation feedback on crops and soil Samples were taken again at 24 hours and 72 hours after the operation, and the results are as follows: Crop growth monitoring revealed an improvement in leaf color index, an average increase in plant height of 1.6 cm, and a decrease in canopy temperature of 0.7 °C, indicating a significant improvement in water and nutrient supply.

[0053] During the growing season, the system continuously tracks data and, based on subsequent data collection and yield and quality assessments, detects slight potassium accumulation in the leaching layer at localized points, automatically reducing subsequent potassium fertilizer input. The accumulated monitoring-decision-execution-feedback big data is used for model retraining, optimizing weight allocation and dynamic adjustment capabilities.

[0054] This embodiment demonstrates the specific process and effectiveness of the system in greenhouse tomato soil stratification monitoring, precise response to crop demand, optimal water and fertilizer input decision-making, and closed-loop management of the entire operation. It verifies the application value and promotion prospects of this invention in improving crop nutrient utilization, reducing leaching and resource waste, and promoting intelligent drip irrigation.

[0055] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM).

[0056] It should be understood that the above detailed description of the technical solutions of the present invention with reference to preferred embodiments is illustrative and not restrictive. Those skilled in the art can modify the technical solutions described in the embodiments or substitute some of the technical features based on reading this specification; however, these modifications or substitutions do not cause the essence of the corresponding technical solutions to depart from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An adaptive drip irrigation control system integrating an optimal soil solution model, characterized in that: The system includes: The soil solution monitoring unit uses a combination of deep-buried multi-parameter soil sensors and soil solution collectors, which are deployed in different soil layers of the field. The sensor depth is set according to the distribution of crop roots. The meteorological and crop growth monitoring unit uses an integrated field meteorological station and portable crop physiological parameter monitoring equipment, including temperature and humidity sensors, rainfall meters, wind speed and direction instruments, and solar radiation sensors. It automatically collects climate data and synchronizes it to the data acquisition and management platform wirelessly. The data acquisition and management platform, based on the Internet of Things architecture, integrates the data outputs of various sensor terminals and collectors, aggregates and pre-processes the data using an edge computing gateway, and then transmits it to the cloud server via 4G / 5G or LoRa wireless network. The irrigation and fertilization execution system consists of a control host, distributed solenoid valves, a main / branch pipeline network, and an automatic fertilizer dispenser. The control host automatically calculates the required water and fertilizer input based on the latest sensing data and decision-making models, and transmits the data to the zone actuators via RS485, Modbus, or wireless means. The main and branch pipelines are equipped with flow meters and pressure sensors to ensure uniform and accurate irrigation in each zone. The operation monitoring and quality feedback module tracks the implementation effects of irrigation and fertilization operations and the crop growth status throughout the entire process.

2. An adaptive drip irrigation method incorporating an optimal soil solution model, wherein the method is applicable to the system described in claim 1, characterized in that: The method includes: Active monitoring of soil and solution conditions at different depths, including water content, main nutrient molecules, pH value, and salinity, to ensure coverage of the main root zone and leaching layer; Set target parameters for crop growth period, based on crop variety and growth stage, and clarify the most suitable comprehensive threshold range of soil solution (concentration of water, nitrogen, phosphorus, potassium, and pH) for each period. Combine meteorological conditions and historical experience in increasing yield to formulate phased nutrient supply and soil moisture targets. The system generates dynamic equilibrium analysis and optimal control schemes. It adopts the optimal soil solution model, introduces a hierarchical dynamic weighting mechanism and the linkage of meteorological-crop-soil ternary factors, and performs weighted control on the actual water and fertilizer requirements of different soil root zones. It also combines meteorological forecasts and corrects soil water and fertilizer scheduling parameters to improve crop nutrient utilization and reduce nutrient leaching and resource waste. Precise execution and full-process recording of operations; control results are sent to the intelligent control host to drive the irrigation valves to start in zones and the water and fertilizer machine to accurately mix; the amount of fertilizer used in each area is controlled in layers according to the target threshold. Feedback on results and further optimization of the plan; continuous tracking of crop growth indicators and comparison with initial targets; for plots that do not meet the standards, adjustment of solution parameters and management plan.

3. The adaptive drip irrigation method based on an optimal soil solution model according to claim 2, characterized in that: The aforementioned active monitoring of stratified soil and solution states includes: selecting monitoring points in different representative areas of the field based on crop root distribution and soil heterogeneity, and burying multi-parameter soil sensors and soil solution collectors in the main root zone, infiltration layer, and necessary deeper soil layers. Using a high-precision integrated sensor for soil moisture, pH, conductivity, and temperature, and a negative pressure ceramic cup or polymer membrane infiltration device, soil moisture, pH, salinity, and major nutrient molecule data of different soil layers are collected on a regular basis and transmitted to a cloud platform for analysis via the Internet of Things.

4. The adaptive drip irrigation method based on an optimal soil solution model according to claim 2, characterized in that: The setting of target parameters for crop growth period includes: retrieving relevant scientific research literature, soil fertility test results and local planting experience based on the characteristics of the crop variety and the needs of each growth stage, combined with real-time and historical meteorological data; Using crop water and fertilizer requirement models and soil-crop-meteorological ternary balance analysis, the comprehensive threshold ranges of soil moisture, nitrogen, phosphorus, potassium, pH and electrical conductivity at each stage were dynamically set.

5. The adaptive drip irrigation method based on an optimal soil solution model according to claim 2, characterized in that: The aforementioned dynamic equilibrium analysis and optimal control scheme includes: receiving and integrating real-time soil solution monitoring data from different depths, and automatically comparing them with the target parameter range to obtain the current dynamic gap; Using the optimal soil solution model, the soil profile is stratified and a stratified dynamic weighting mechanism is introduced. Combined with meteorological input and crop growth data, a ternary regulation equation of meteorology-crop-soil is established, which couples soil moisture transport and dynamic mass balance of ion concentration. Through an information feedback mechanism, the weight parameters and leaching penalty are dynamically optimized based on historical data and crop growth response to achieve adaptive regulation. Using the improved particle swarm optimization algorithm, the optimal hierarchical input is determined, generating the most economical and efficient water and fertilizer scheduling scheme in batches, which is then distributed to the execution system for implementation.

6. The adaptive drip irrigation method based on an optimal soil solution model according to claim 2, characterized in that: The precise execution and full-process recording of the operation include: the optimized water and fertilizer input instructions are automatically sent to the intelligent control host in each area through the Internet of Things platform, the host intelligently calls each group of irrigation valves to realize zoned and layered irrigation control, and the optimal fertilizer solution is accurately input into the designated soil layer and area through the automatic proportioning system of the integrated water and fertilizer machine.

7. The adaptive drip irrigation method based on an optimal soil solution model according to claim 2, characterized in that: The results feedback and scheme optimization include: comprehensively tracking the crop growth status, final yield, and quality of the target area across multiple indicators, and comparing and analyzing them with the target parameters; actively correcting the solution model parameters and dynamically adjusting the strategy for plots with deviations.

8. The adaptive drip irrigation method based on an optimal soil solution model according to claim 5, characterized in that: The improved particle swarm optimization algorithm includes: encoding the water and fertilizer input at each layer and at each time step into the dimension of the particles, and introducing the weight distribution and gradient transfer domain knowledge into the adaptive update mechanism of particle velocity and inertia weight. It combines individual and global optimal guidance with domain gradient influence correction to achieve dynamic optimization under multi-objective conditions; The objective function considers the requirements of sufficient nutrients, minimum leaching, and lowest cost, and uses a constraint penalty function to handle interval thresholds and maximum input agronomic constraints to prevent the objective parameters from exceeding limits and resources from being wasted. Each round automatically selects the current optimal solution, using random perturbation and dynamic parameter adjustment to improve search accuracy and convergence speed.

Images