An intelligent decision system for precise water distribution in agricultural irrigation areas based on crop water requirement coefficient

Abstract

This invention discloses an intelligent decision-making system for precise water allocation in agricultural irrigation districts based on crop water requirement coefficients. The system comprises three main modules: hardware acquisition and execution, software algorithms and modeling, and digital twin simulation and decision-making. These modules operate in a closed-loop logic of acquisition, processing, simulation, decision-making, execution, and feedback. The hardware module employs a four-dimensional collaborative acquisition, dual-mode transmission, and edge computing scheme. The software module incorporates algorithms and adaptation rules such as dual-factor coupling and dynamic water requirement coefficient correction. The digital twin module completes instruction issuance and parameter correction through full-element modeling, LOD optimization, and multi-objective decision-making. This invention, through dual-path feedback, digital twin collaboration, and multi-algorithm coordination, combined with dynamic water requirement coefficient correction, terraced field spatial adaptation, and soil water-salt coupling modeling, accurately matches plot requirements, balancing water conservation, crop adaptation, and energy consumption, thereby reducing irrigation district management costs and adapting to complex irrigation district operations.

Description

Technical Field

[0001] This invention relates to the field of intelligent agricultural irrigation technology, specifically to an intelligent decision-making system for precise water allocation in agricultural irrigation areas based on crop water requirement coefficients. Background Technology

[0002] Irrigation water distribution in agricultural irrigation districts is a crucial link in ensuring stable agricultural production. With the advancement of agricultural modernization, the scale of irrigation districts is constantly expanding, and the proportion of irrigation demand in complex terrains such as terraced fields and saline-alkali lands is gradually increasing. In recent years, the integration of intelligent technologies with agricultural production has become increasingly profound. Technologies such as digital twins, edge computing, and multi-sensor collaboration are gradually being applied to the operation and management of irrigation districts, driving the transformation of water distribution models from traditional manual control to automation and precision.

[0003] Existing irrigation district water distribution technologies mostly remain at the basic automation level, failing to form mature multi-technology integration solutions. Most systems rely on only a single type of sensor to collect data, resulting in limited data coverage and an inability to comprehensively capture multi-dimensional information such as irrigation district topography, soil conditions, and crop growth. Water distribution control is largely based on fixed formulas or empirical parameters, lacking the ability to adapt to dynamic changes in the field in real time. It cannot flexibly adjust according to differences in terrain and soil water-salt variations in different areas. Furthermore, most technologies lack a complete parameter feedback and correction mechanism, making it difficult to optimize water distribution plans based on actual implementation results. This leads to a disconnect between the water distribution process and actual field needs, failing to meet the refined operational requirements of complex irrigation districts. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of existing technologies and provide an intelligent decision-making system for precise water allocation in agricultural irrigation areas based on crop water requirement coefficients. This system integrates three major modules: hardware acquisition and execution, software algorithms and modeling, and digital twin simulation and decision-making, operating in a closed-loop logic of acquisition-processing-simulation-decision-execution-feedback. The hardware module comprehensively captures multi-dimensional data on irrigation area topography, crops, and soil based on a four-dimensional collaborative acquisition and dual-mode transmission architecture. The software module incorporates physiological-environmental dual-factor coupling, terraced field spatial adaptation, and soil water-salt coupling modeling, accurately calculating dynamic water requirement parameters through a dedicated formula. The digital twin module utilizes full-element three-dimensional simulation and multi-objective optimization algorithms to quickly output adapted irrigation and drainage schemes. Dual-path feedback quantifies and corrects parameters based on deviations, solving the problem of traditional water allocation standardization while adapting to complex scenarios such as terraced fields and saline-alkali land, balancing water-saving effects, crop growth adaptability, and equipment energy consumption, thereby reducing the daily management costs of irrigation areas.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution: an intelligent decision-making system for precise water allocation in agricultural irrigation areas based on crop water requirement coefficient. The system includes a hardware acquisition and execution module, a software algorithm and modeling module, and a digital twin simulation and decision-making module. The three modules are linked step by step according to the complete logic of data acquisition, preprocessing, transmission, algorithm calculation, simulation analysis, decision issuance, execution feedback, and parameter correction.

[0006] The hardware acquisition and execution module design incorporates a four-dimensional collaborative acquisition technology using drones, weather radar, ground weather stations, and satellite remote sensing, coupled with a 5G and LoRa dual-mode transmission architecture and an edge computing node deployment scheme, to achieve comprehensive acquisition of multi-dimensional data in the irrigation area, rapid preliminary preprocessing, precise command execution, and emergency equipment support.

[0007] The software algorithm and modeling module designs a crop physiological-environment dual-factor coupling logic, constructs a dynamic water demand coefficient correction algorithm, sets up a water demand coefficient-soil water and salt coupling submodule, a crop physiological-environment dual-driven water demand coefficient correction submodule, and a terraced vertical space water demand coefficient distribution submodule. Combining the multi-dimensional influencing factors of terraced terrain, spatial adaptation rules are designed to establish a dynamic correlation model between water demand coefficient and soil water and salt, and output dynamic data to support decision-making.

[0008] The digital twin simulation and decision-making module is set up with a digital twin irrigation district full-element mapping submodule, an irrigation and drainage collaborative intelligent decision-making submodule, and a decision output and feedback correction submodule. Based on the Unity3D engine, the module designs a three-dimensional modeling technology for all elements of the irrigation district and uses LOD level of detail optimization technology to achieve multi-element linkage simulation once per second. Through multi-objective optimization algorithm, the module designs irrigation and drainage scheme selection rules and completes instruction issuance, execution data reception, and parameter dynamic correction.

[0009] Furthermore, the hardware acquisition and execution module comprises five functional sub-modules, each working collaboratively along a link, specifically designed as follows: The integrated air-ground-space multi-dimensional data acquisition sub-module scans the irrigation area at a fixed cycle of 30 minutes, simultaneously acquiring topographic data, meteorological data, and crop growth data; the UAV is equipped with a hyperspectral camera and GPS positioning module; the crop physiology and soil multi-parameter sensing sub-module divides the terraced fields into acquisition units every 5 meters according to the vertical elevation gradient zones; in saline-alkali land, sensors are evenly distributed according to plot area, and the sensors include leaf water potential sensors, stem flow rate sensors, and soil... Soil moisture sensors, soil water and salt content sensors, and soil temperature sensors complete data acquisition every five minutes, and all sensors have daily automatic calibration functions; the irrigation and drainage equipment status monitoring and intelligent execution submodule collects the operating parameters of water pumps, valves, and drainage outlets in real time, and responds to commands within seconds and drives the equipment to act; the end-to-end data transmission and edge processing submodule is deployed according to the rule of one edge computing node per 1,000 mu of irrigation area, and the initial data preprocessing time is strictly controlled within 0.3 seconds; the backup power supply and emergency protection submodule is designed with automatic switching in case of power failure, real-time fault alarm and emergency operation plan.

[0010] Furthermore, the software algorithm and modeling module includes an algorithm and hardware adaptation optimization sub-module, specifically designed as follows: A comprehensive hardware device parameter database is constructed, including the operating parameter ranges, data output formats, and communication protocol standards for different types of sensors and actuators; real-time hardware operation status monitoring logic is designed to identify operational fluctuations by comparing actual hardware operating parameters with database standard parameters; and dynamic adjustment rules for algorithm parameters are established to adaptively adjust data processing thresholds, calculation step sizes, and other algorithm parameters according to the degree of hardware operational fluctuations.

[0011] Furthermore, the dynamic water demand coefficient correction algorithm of the software algorithm and modeling module is implemented through a crop physiological-environmental factor dual-driven water demand coefficient correction submodule. This submodule is designed with weighted coupling logic of physiological and environmental factors, and uses the constructed dynamic water demand coefficient correction formula to calculate the plot-level dynamic water demand coefficient. The formula is as follows:

[0012]

[0013] in, This refers to the crop water requirement coefficient after dynamic correction at the plot level. This is the basic water requirement coefficient for the corresponding crop growth period; The weighting coefficient for physiological indicators ranges from 0.4 to 0.6, and is set according to the differences in crop variety characteristics. To measure the water potential of crop leaves; The appropriate leaf water potential threshold for crops at the corresponding growth stage; To measure the stem flow rate of crops; The appropriate stem flow rate threshold for crops at the corresponding growth stage; This is a comprehensive correction coefficient for environmental factors; all parameters were trained and calibrated using field trials of similar crops at multiple growth stages, covering over 100,000 sets of historical data from different climate scenarios. The calculation is derived from three environmental factors: temperature, humidity, and soil moisture. The calculation formula is as follows:

[0014]

[0015] Where Σ is the summation symbol; The values ​​are 1, 2, and 3, which correspond to air temperature, humidity, and soil moisture, respectively. For the first The influence coefficient of environmental factors ranges from -0.2 to 0.3; For the first Measured values ​​of environmental factors; is the optimal crop growth threshold for the i-th type of environmental factor.

[0016] Furthermore, the software algorithm and modeling module completes the spatial adaptation of water demand coefficients through the terraced field vertical space water demand coefficient distribution submodule. This submodule designs multi-factor comprehensive adaptation logic for terraced field terrain and adopts the constructed terraced field vertical space water demand coefficient correction formula to achieve differentiated customization of water demand coefficients for different gradient zones. The formula is as follows:

[0017]

[0018] in, Customized crop water requirement coefficients for the corresponding gradient zones of terraced fields; This refers to the crop water requirement coefficient after dynamic correction at the plot level. The elevation influence coefficient ranges from 0.02 to 0.05 per meter and is calibrated using comparative test data on water transport losses in different elevation gradient zones. The average elevation of the gradient zone corresponding to the terraced fields; The lowest elevation in the irrigation area is used as a benchmark; This is the slope correction factor, with a value ranging from 0.003 to 0.008 per degree, calibrated based on the results of irrigation uniformity tests on terraced fields with different slopes; The slope of the gradient zone corresponding to the terraced fields; It is a sine function; The weighting of soil texture influence ranges from 0.1 to 0.15, and is calibrated through water demand characteristic tests in irrigation areas with multiple soil textures. To determine the soil texture difference coefficient corresponding to the gradient zone, the coefficients were set according to the soil classification standards: 1.2 for sandy soil, 1.0 for loam, and 0.8 for clay. This is the average soil texture coefficient of the irrigation area, with a default value of 1.0.

[0019] Furthermore, the digital twin irrigation district full-element mapping submodule serves as the simulation unit of the digital twin simulation and decision-making module. It designs a collaborative simulation logic across four element layers, with each layer functioning as follows: The geospatial layer recreates geographical elements such as irrigation district topography, canal network, roads, and water sources, with an accuracy of ±0.1 meters; the crop growth layer correlates with dynamic water demand coefficient data to simulate the response of crop growth to irrigation measures; the soil water and salt layer is linked to the water demand coefficient-soil water and salt coupling submodule. The coupling between the water demand coefficient and soil water and salt is achieved through a dynamic correlation model designed in the water demand coefficient-soil water and salt coupling submodule. The soil water and salt layer adopts a constructed coupling correlation formula:

[0020]

[0021] in, The dynamic soil electrical conductivity is denoted by λ, which is the influence coefficient of the water demand coefficient. The value of λ ranges from 0.3 to 0.5 and is calibrated through training data from the coupling test of soil water and salt with the water demand coefficient. is a natural exponential function; μ is the irrigation dilution coefficient, with a value ranging from 0.002 to 0.005, calibrated through comparative experiments on the soil water and salt dilution effects under different irrigation amounts; The initial soil electrical conductivity of the irrigation area is the measured value, determined by the average value of multiple sampling points in different irrigation areas. This refers to the crop water requirement coefficient after dynamic correction at the plot level. The maximum water requirement coefficient for the corresponding crop growth period was determined through field trials with similar crops; This refers to the amount of irrigation water.

[0022] Furthermore, the multi-objective optimization algorithm of the digital twin simulation and decision-making module is implemented through the irrigation and drainage collaborative intelligent decision-making sub-module. This sub-module is specifically designed as follows: An improved multi-objective particle swarm optimization algorithm is adopted, setting three major optimization objectives: water saving rate, crop growth suitability, and equipment energy consumption. An objective function is constructed, and weight allocation rules are determined. A three-level control threshold system is designed, using the critical values ​​of soil conductivity and crop water requirement coefficient as key indicators to divide the trigger intervals for three irrigation and drainage modes: conventional precision irrigation mode is triggered when soil conductivity < critical value and water requirement coefficient ≥ critical value; precision salt leaching irrigation mode is triggered when soil conductivity ≥ critical value and water requirement coefficient ≥ critical value; and drainage followed by supplementary irrigation mode is triggered when soil conductivity ≥ critical value and water requirement coefficient < critical value. The algorithm decision-making time is strictly controlled within 0.5 seconds, and the time for instructions to be sent from the decision output sub-module to the hardware execution layer does not exceed 0.2 seconds.

[0023] Furthermore, the digital twin simulation and decision-making module achieves closed-loop correction through a decision output and feedback correction sub-module. This sub-module is designed with a dual-path feedback correction logic: first, the equipment execution data feedback path, which receives actual equipment execution data such as pump speed, valve opening, and drainage flow rate from the hardware execution layer, compares it with preset command values ​​to calculate the deviation, and feeds it back to the digital twin model to correct simulation parameters; second, the irrigation area status data feedback path, which receives real-time monitoring data such as water demand coefficient, soil water and salt, and crop physiological indicators from the irrigation area, compares it with algorithm prediction values, and feeds it back to the software algorithm module to correct the water demand coefficient threshold, terrace adaptation parameters, and decision model weights; at the same time, it designs parameter correction amplitude rules, which quantify and adjust according to the degree of data deviation, fine-tuning parameters when the deviation is ≤5%, and triggering deep correction when the deviation is >5%.

[0024] Compared with existing technologies, this intelligent decision-making system for precise water allocation in agricultural irrigation areas based on crop water requirement coefficient has the following beneficial effects:

[0025] I. This invention establishes a decision output and feedback correction submodule with dual-path feedback correction logic, combined with a digital twin irrigation district full-element mapping submodule and multi-algorithm collaborative design, enabling irrigation district water allocation to form a complete closed-loop control from data acquisition to parameter correction. The dynamic water demand coefficient correction algorithm, combined with physiological-environment dual-factor weighted coupling logic and terrace vertical space adaptation rules, can specifically adjust water demand parameters for areas with different elevations, slopes, and soil textures, solving the water resource waste problem caused by the use of uniform standards in traditional water allocation; the dual-path feedback verifies the data from equipment execution and irrigation district status data, quantifies and adjusts parameters according to the degree of deviation, and accurately matches the actual needs of each plot during water allocation, making the simulation results more consistent with the actual situation on site.

[0026] II. This invention establishes a dynamic correlation model between water demand coefficient and soil water-salt ratio through deep integration of software algorithms and modeling modules, digital twin simulation, and decision-making modules. It also incorporates an improved multi-objective particle swarm optimization algorithm and a three-level control threshold system. Compared to traditional water distribution technologies, this invention balances water conservation, crop growth suitability, and equipment energy consumption, while shortening control time through second-level linkage simulation and rapid decision response. The hardware module adopts a four-dimensional collaborative acquisition and dual-mode transmission architecture, coupled with optimized design for algorithm and hardware adaptation, ensuring stable data acquisition efficiency, reducing malfunctions, and lowering the manpower and material costs of daily irrigation area management.

[0027] Other advantages, objectives and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination or study, or may be learned from the practice of the invention. Attached Figure Description

[0028] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0029] Figure 1 This is a schematic diagram of the overall system architecture and module relationships of the present invention;

[0030] Figure 2 This is a flowchart of the dynamic water demand coefficient correction algorithm of the present invention;

[0031] Figure 3 This is a flowchart of the digital twin simulation and decision-making closed loop of the present invention. Detailed Implementation

[0032] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.

[0033] Example 1

[0034] This embodiment details the complete operation process and technical implementation of an intelligent decision-making system for precision water allocation in agricultural irrigation areas based on crop water requirement coefficients. The system is highly adaptable to the irrigation needs of special terrains such as terraced fields and saline-alkali land, solving the problems of insufficient targeting and delayed response of traditional water allocation technologies.

[0035] The specific execution process is as follows:

[0036] System initialization and hardware acquisition and execution module startup

[0037] After the system is powered on and initialized, the communication links between modules are first verified to ensure smooth data transmission between the hardware acquisition and execution module, the software algorithm and modeling module, and the digital twin simulation and decision-making module. For example... Figure 1 As shown, the three-layer module operates in a sequential manner, following a complete logic of data acquisition, preprocessing, transmission, algorithm calculation, simulation analysis, decision issuance, execution feedback, and parameter correction, forming a closed-loop control system. Subsequently, the hardware acquisition and execution module activates the five functional sub-modules in the order of "acquisition sub-module startup - transmission sub-module readiness - emergency support sub-module standby." Each sub-module operates based on preset link coordination rules, with the core objective of comprehensively, accurately, and stably acquiring multi-dimensional data from the irrigation area.

[0038] Once activated, the integrated air-space-ground data acquisition submodule performs full-area scanning at fixed intervals. This submodule employs a four-dimensional collaborative acquisition technology utilizing drones, weather radar, ground weather stations, and satellite remote sensing. Drones, equipped with hyperspectral cameras and GPS positioning modules, focus on collecting crop growth data and micro-topographic details in specific areas of the irrigation district. Ground weather stations collect meteorological environmental data such as temperature, humidity, and wind speed at fixed points within the irrigation district. Satellite remote sensing provides macro-level topographic data for the entire irrigation district. Weather radar supplements this with data on precipitation, cloud cover, and other meteorological warnings. The data acquisition actions of these four types of equipment are synchronized to ensure coverage of core dimensions such as irrigation district topography, meteorology, and crop growth, avoiding data blind spots.

[0039] The crop physiology and soil multi-parameter sensing submodule is designed with layout logic tailored to the terrain and soil characteristics of terraced fields and saline-alkali land. The terraced areas are divided into data acquisition units according to vertical elevation gradient zones, while the saline-alkali land is divided into evenly spaced plots. Sensor selection covers leaf water potential sensors, stem flow rate sensors, soil moisture sensors, soil water and salt content sensors, and soil temperature sensors, comprehensively capturing crop physiological status and soil environmental data. This submodule continuously collects data at a set frequency, and all sensors have built-in daily automatic calibration logic. By comparing with standard reference values, it corrects acquisition deviations, ensuring data accuracy meets the requirements of subsequent algorithm calculations.

[0040] The irrigation and drainage equipment status monitoring and intelligent execution submodule scans the operating parameters of terminal execution equipment such as pumps, valves, and drain outlets in real time, including pump speed, valve opening, drainage flow rate, and equipment operating current and voltage. This data is used to determine whether the equipment is in normal working condition and to provide a reference for the accurate issuance of subsequent decision-making instructions. When receiving irrigation and drainage instructions from the digital twin simulation and decision-making module, this submodule can achieve a response time of seconds, driving the corresponding equipment to act according to the instructions, ensuring that the decision-making plan is implemented quickly.

[0041] The end-to-end data transmission and edge processing submodule deploys edge computing nodes according to the irrigation district scale. Its core function is to solve the problem of rapid transmission and preliminary processing of massive amounts of collected data. This submodule adopts a dual-mode transmission architecture of 5G and LoRa. 5G transmission is used for high-priority, high-capacity data transmission, while LoRa transmission is used for low-power, wide-coverage, routine data transmission. The two transmission methods serve as backups for each other, ensuring uninterrupted data transmission. After receiving data, the edge computing nodes immediately perform preliminary preprocessing operations, including removing outliers, standardizing data formats, and filtering key data. This reduces the processing pressure of invalid data on subsequent modules, and the preprocessing time is strictly controlled within a set range to ensure the high efficiency of data transmission and processing.

[0042] The backup power and emergency backup submodule monitors the power supply status and operating conditions of the hardware modules throughout the process, and has a built-in automatic power failure switching circuit and fault diagnosis logic. When an external power interruption is detected, it immediately switches to backup power to avoid interruption of data acquisition and execution. When an abnormality is detected in the equipment operation, it triggers an alarm signal in real time and activates the emergency operation plan, such as shutting down the faulty equipment, activating the backup equipment, or adjusting the acquisition frequency to ensure that core data is not lost, thus ensuring the overall stability of the hardware module operation.

[0043] Deep data processing and modeling in the software algorithm and modeling module

[0044] After receiving the preprocessed data transmitted by the hardware module, the software algorithm and modeling module starts a multi-submodule collaborative working mode. The core objective is to transform the raw data into dynamic core data that supports decision-making. Through multi-algorithm coupling and modeling, it achieves accurate calculation and differentiated adaptation of the water demand coefficient.

[0045] The algorithm and hardware adaptation optimization submodule first initiates the hardware status verification process. This submodule pre-builds a comprehensive hardware device parameter database, which includes the operating parameter ranges, data output formats, and communication protocol standards for different models of sensors and actuators, essentially creating a "standard profile" for each hardware device. By comparing the actual hardware operating parameters with the standard parameters in the database in real time, this submodule can quickly identify the amplitude of hardware operation fluctuations, such as decreased sensor acquisition accuracy or transmission module speed fluctuations. For these fluctuations, the submodule dynamically adjusts algorithm parameters according to preset algorithm parameter adjustment rules, adaptively adjusting data processing thresholds, calculation step sizes, and other algorithm parameters—for example, when sensor acquisition fluctuations are large, the data processing threshold range is appropriately expanded to avoid data processing distortion caused by hardware fluctuations, ensuring that the algorithm processing accuracy and hardware operating status remain consistent.

[0046] The reason why the crop physiological-environmental factor dual-driven water requirement coefficient correction submodule adopts dual-factor coupling is that crop water requirement is not only affected by its own physiological state, but also closely related to external environmental conditions, and single-factor calculation is difficult to reflect the actual water requirement. For example Figure 2 As shown, this submodule first extracts physiological indicators such as crop leaf water potential and stem flow rate, as well as environmental indicators such as air temperature, humidity, and soil moisture from the hardware-acquired data. Then, it calls the constructed dynamic water demand coefficient correction formula for calculation. The formula is as follows:

[0047]

[0048] in, As the crop water requirement coefficient dynamically adjusted at the plot level, it serves as the core basis for subsequent water allocation decisions. The basic water requirement coefficient corresponding to the crop's growth period is determined through field trials of similar crops at multiple growth periods to ensure the basic water requirement pattern of the crop's growth. As the weighting coefficients of physiological indicators, they are set differently according to the characteristics of crop varieties, because different crops have different sensitivities to their own physiological state and environmental conditions. For example, the physiological indicators of drought-resistant crops have relatively higher weights. To measure the water potential of crop leaves, The ratio of the two values ​​represents the difference between the current physiological water requirement of the crop and the optimal water requirement, corresponding to the appropriate leaf water potential threshold for crops during their growth period. To measure crop stem flow rate, The appropriate stem flow rate threshold for crops during their growth period is also used to quantify the physiological water requirements of crops. This is a comprehensive correction coefficient for environmental factors, used to adjust for the impact of environmental conditions on crop water requirements. All parameters have been calibrated through field trials of similar crops at multiple growth stages, using historical data covering a large number of different climatic scenarios, to ensure the reliability of the calculation results.

[0049] Comprehensive correction coefficient for environmental factors The calculation formula is derived from three core environmental factors:

[0050]

[0051] Where Σ is the summation symbol, used to summarize the corrective effects of the three types of environmental factors; i is the environmental factor number, corresponding to air temperature, humidity, and soil moisture, which are the most critical environmental factors affecting crop water requirements. The value range of the i-th type of environmental factor has been calibrated through multiple sets of experiments to ensure that it can accurately reflect the degree of influence of different environmental factors on water demand. For the measured value of the i-th type of environmental factor, Let be the crop's suitable growth threshold for the i-th type of environmental factor. The ratio of the difference between the current environment and the suitable environment to this threshold is used to quantify the deviation between the current environment and the suitable environment, and then... The weighted adjustment magnitude of this environmental factor is obtained, and finally summarized as follows: This enables dynamic correction of the water demand coefficient by the environment.

[0052] The terraced field vertical spatial water demand coefficient distribution submodule is designed with adaptation logic tailored to the specific characteristics of terraced terrain. Significant differences in elevation, slope, and soil texture exist between different gradient zones of terraces, leading to substantial variations in crop water demand across different areas of the same irrigation district. Traditional uniform water allocation methods cannot meet these demands. This submodule extracts average elevation, slope, and soil texture data for each gradient zone from hardware-collected data, initiates multi-factor comprehensive adaptation logic for terraced terrain, and employs a constructed terraced field vertical spatial water demand coefficient correction formula to achieve differentiated customization of water demand coefficients for different gradient zones. The formula is as follows:

[0053]

[0054] in, The customized crop water requirement coefficient for the corresponding gradient zone of the terraced fields is the final water requirement reference value for a specific gradient zone; This represents the average elevation of the gradient zone corresponding to the terraced fields. The lowest elevation in the irrigation area is used as the benchmark. The ratio of the difference between the two to the benchmark reflects the elevation difference. γ is the elevation influence coefficient. Because different elevations will lead to changes in water loss, temperature and humidity, etc., which will affect water demand, this coefficient is trained and calibrated through comparative test data of water loss in different elevation gradient zones. The slope of the gradient zone corresponding to the terraced fields. Used to quantify the impact of slope on irrigation water distribution The slope correction coefficient is determined based on the test results of irrigation uniformity of terraced fields with different slopes. The steeper the slope, the easier it is for irrigation water to be lost, and the water demand coefficient needs to be adjusted accordingly. To correspond to the soil texture difference coefficient in the gradient zone, the coefficients were set according to soil classification standards. Sandy soil, loam, and clay soil have different water retention capacities and significantly different water demand coefficients. The weighting of soil texture was determined through water demand characteristics tests in irrigation areas with multiple soil textures. This is the average soil texture coefficient for the irrigation area. The default value is a unified benchmark value, which is used to eliminate the impact of overall differences in soil texture among different irrigation areas and ensure the universality of the adaptation logic.

[0055] The water demand coefficient-soil water-salt coupling submodule focuses on the water allocation needs of special soil environments such as saline-alkali land, establishing a dynamic relationship between the water demand coefficient and soil water and salt levels. Crop water requirements and soil water and salt conditions influence each other; excessive soil water and salt levels affect crop water absorption, while irrigation water allocation alters soil water and salt concentrations. Therefore, a coupling model between the two is needed. This submodule uses the constructed coupling relationship:

[0056]

[0057] in, Dynamic soil electrical conductivity is used to reflect the soil water and salt status in real time. The initial soil electrical conductivity of the irrigation area was determined by the average value of multiple sampling points in different irrigation areas to ensure the representativeness of the initial data. The influence coefficient of water demand coefficient is trained and calibrated using experimental data coupling soil water and salt with water demand coefficient. It is used to quantify the influence of water demand coefficient on soil water and salt. The higher the water demand coefficient, the more water the crop absorbs and the higher the soil water and salt concentration. This refers to the crop water requirement coefficient after dynamic correction at the plot level. The maximum water requirement coefficient for the corresponding crop growth period is determined through field trials of similar crops and serves as an upper limit reference for the water requirement coefficient. It is a natural exponential function used to simulate the dilution effect of irrigation on soil water and salt; The irrigation dilution coefficient was determined through a comparative experiment on the soil water and salt dilution effect under different irrigation amounts. The greater the irrigation amount, the more obvious the soil water and salt dilution. Irrigation volume directly affects changes in soil water and salt concentration. This coupled model enables dynamic calculation of the linkage between water demand coefficient and soil water and salt content, providing a core basis for subsequent irrigation and drainage pattern decisions.

[0058] After completing the above data processing and modeling, the software algorithm and modeling module transmits core data such as plot-level dynamic water demand coefficient, customized water demand coefficient of each gradient zone of the terrace, and dynamic soil electrical conductivity to the digital twin simulation and decision-making module in a fixed format, providing data support for simulation analysis and decision-making.

[0059] Simulation analysis and command issuance of digital twin simulation and decision-making modules

[0060] After receiving the core data transmitted by the software algorithm and modeling module, the digital twin simulation and decision-making module starts the full-element simulation and multi-objective decision-making process. The core objective is to build a virtual image of the irrigation area based on real-time data, simulate the effects of different water distribution schemes, and finally output the optimal irrigation and drainage instructions.

[0061] The digital twin irrigation district full-element mapping submodule serves as the core simulation unit, designing a collaborative simulation logic across four element layers: geospatial layer, crop growth layer, soil water and salinity layer, and irrigation and drainage system layer. The geospatial layer constructs a 3D model of all irrigation district elements using the Unity3D engine, recreating geographical features such as topography, canal networks, roads, and water sources. Accuracy indicators are rigorously calibrated to ensure consistency between the virtual terrain and the actual irrigation district. The crop growth layer, linked to dynamic water demand coefficient data output by the software module, simulates the crop's growth response to different irrigation measures under current water demand conditions, such as growth rate when irrigation is sufficient and wilting state when water demand is insufficient. The soil water and salinity layer, linked to the water demand coefficient-soil water and salinity coupling submodule, updates the virtual soil water and salinity status in real time based on coupling correlation, simulating the dilution process of soil water and salinity after irrigation and the concentration process after drainage. The irrigation and drainage system layer, linked to equipment operation data collected by the hardware module, recreates the real-time operating status of equipment such as pumps and valves in the virtual model. This submodule employs Level of Detail (LOD) optimization technology to reduce model accuracy in non-critical areas, ensuring simulation efficiency and enabling multi-element linkage simulation once per second, thus ensuring real-time synchronization between the virtual image and the actual irrigation area status.

[0062] The intelligent decision-making submodule for coordinated irrigation and drainage utilizes an improved multi-objective particle swarm optimization algorithm. Compared to traditional algorithms, this algorithm offers advantages such as faster convergence and better uniformity of solution distribution in multi-objective optimization problems, making it suitable for balancing multiple demands in the water allocation process. The algorithm sets three optimization objectives: water-saving rate, crop growth suitability, and equipment energy consumption. It constructs an objective function and determines weight allocation rules—the weight allocation is pre-set based on the operational needs of the irrigation district; for example, the weight of water-saving rate can be increased in water-scarce areas, while the weight of equipment energy consumption can be increased in energy-constrained areas. Simultaneously, this submodule designs a three-level control threshold system, using critical values ​​for soil electrical conductivity and crop water requirement coefficient as key indicators. These two critical values ​​are calibrated using long-term experimental data from similar irrigation districts, corresponding to the suitable soil water-salt range and water requirement range for crop growth, respectively.

[0063] Based on the real-time soil conductivity and crop water requirement coefficient obtained from simulation, the submodule divides the trigger intervals for three irrigation and drainage modes: When the soil conductivity is less than the critical value and the water requirement coefficient is greater than or equal to the critical value, it indicates that the soil water and salt conditions are suitable but the crop is short of water, triggering the conventional precision irrigation mode, which only needs to supplement the water required for crop growth; when the soil conductivity is greater than or equal to the critical value and the water requirement coefficient is greater than or equal to the critical value, it indicates that the soil water and salt levels are excessive and the crop is short of water, triggering the precision salt leaching irrigation mode, which dilutes the soil salt while supplementing water; when the soil conductivity is greater than or equal to the critical value and the water requirement coefficient is less than the critical value, it indicates that the soil water and salt levels are excessive but the crop is not short of water, triggering the first-drain-then-replenish irrigation mode, which first drains the saline water and then supplements a small amount of suitable water. The algorithm decision-making time is strictly controlled within the set range to ensure rapid response to dynamic changes in the irrigation area. The instructions are sent to the hardware execution layer through the decision output submodule, and the transmission time does not exceed the set upper limit, ensuring the timely implementation of the decision plan.

[0064] The decision output and feedback correction submodules simultaneously initiate dual-path feedback correction logic. The core objective is to optimize system parameters based on actual execution results, forming a closed-loop control. For example... Figure 3 As shown, this closed-loop process encompasses two paths: equipment execution data feedback and irrigation district status data feedback. By comparing preset values ​​with actual values ​​to calculate the deviation, and quantifying and correcting system parameters based on the degree of deviation, continuous optimization of "simulation-decision-execution-feedback-correction" is achieved. The first path is the equipment execution data feedback path: receiving actual equipment execution data such as pump speed, valve opening, and drainage flow rate from the hardware execution layer, comparing it with preset values ​​in the decision command to calculate the deviation—for example, if the command requires a valve opening of 80%, and the actual opening is 75%, then a 5% deviation is calculated. This deviation is fed back to the digital twin model to correct the virtual equipment's operating parameters, making subsequent simulations more closely resemble actual equipment performance. The second path is the irrigation district status data feedback path: receiving real-time monitoring data such as water demand coefficient, soil water and salt content, and crop physiological indicators from the hardware module, comparing it with the predicted values ​​from the software algorithm module. For example, if the algorithm predicts a water demand coefficient of 0.8, and the actual monitored value is 0.7, then the deviation is calculated and fed back to the software module to correct the water demand coefficient threshold, terrace adaptation parameters, and decision model weights, improving the accuracy of subsequent calculations and decisions. Meanwhile, the submodule is designed with parameter correction rules that are quantitatively adjusted according to the degree of data deviation. When the deviation is small, the parameters are fine-tuned to avoid system fluctuations, and when the deviation is large, deep correction is triggered to ensure system accuracy. The system parameters are continuously optimized through dual-path feedback.

[0065] System closed-loop control and continuous operation

[0066] After receiving irrigation and drainage instructions from the digital twin simulation and decision-making module, the hardware acquisition and execution module immediately drives the corresponding equipment to take action: In the conventional precision irrigation mode, the water pump is started and the valve is opened at the set opening degree to supply water to the target area according to the calculated irrigation volume; in the precision salt washing irrigation mode, the water pump output power is adjusted to increase the water supply and extend the irrigation time to ensure that the salt is fully diluted; in the drain-then-replenish irrigation mode, the drain outlet is opened first to discharge the saline water, and after the set drainage time is reached, the drain outlet is closed, and then the irrigation equipment is started to replenish the appropriate amount of water.

[0067] During equipment operation, each acquisition submodule continuously collects data at a set frequency: the integrated air-space-ground acquisition submodule monitors crop growth changes after irrigation; the crop physiology and soil multi-parameter sensing submodule collects real-time data on soil moisture and water-salt content; and the irrigation and drainage equipment status monitoring submodule records the actual operating parameters of the equipment. This data is transmitted to the edge computing node for preprocessing and then simultaneously sent to the software algorithm and modeling module and the digital twin simulation and decision-making module. The software module recalculates the water demand coefficient and soil water-salt coupling value based on the new data; the digital twin module updates the virtual simulation model; and the decision-making module compares the new data with the expected results. If there are discrepancies, the parameters are adjusted through the feedback correction submodule, generating new irrigation and drainage instructions for execution, forming a complete closed loop of "data acquisition-processing-simulation-decision-execution-feedback-correction".

[0068] The system operates continuously according to the above closed-loop logic, responding in real time to the dynamic changes of various factors such as the topography, soil, crops, and weather in the irrigation area, and always outputting the optimal irrigation and drainage plan to ensure that the water distribution in the irrigation area meets the needs of crop growth while avoiding water waste and soil environmental degradation.

[0069] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A smart decision-making system for precise water allocation in agricultural irrigation areas based on crop water requirement coefficient, characterized in that, The system includes a hardware acquisition and execution module, a software algorithm and modeling module, and a digital twin simulation and decision-making module. The three modules are linked step by step according to the complete logic of data acquisition, preprocessing, transmission, algorithm calculation, simulation analysis, decision issuance, execution feedback, and parameter correction. The hardware acquisition and execution module is designed to utilize a four-dimensional collaborative acquisition technology involving drones, weather radar, ground weather stations, and satellite remote sensing, combined with a 5G and LoRa dual-mode transmission architecture and an edge computing node deployment scheme. This is used for the comprehensive acquisition of multi-dimensional data in irrigation areas, rapid preliminary preprocessing, precise execution of instructions, and emergency support for equipment. The software algorithm and modeling module designs a crop physiological-environment dual-factor coupling logic, constructs a dynamic water demand coefficient correction algorithm, sets up a water demand coefficient-soil water and salt coupling submodule, a crop physiological-environment dual-driven water demand coefficient correction submodule, and a terraced vertical space water demand coefficient distribution submodule. Combining the multi-dimensional influencing factors of terraced terrain, spatial adaptation rules are designed to establish a dynamic correlation model between water demand coefficient and soil water and salt, and output dynamic data to support decision-making. The digital twin simulation and decision-making module is set up with a digital twin irrigation district full-element mapping submodule, an irrigation and drainage collaborative intelligent decision-making submodule, and a decision output and feedback correction submodule. Based on the Unity3D engine, the module designs a three-dimensional modeling technology for all elements of the irrigation district and uses LOD level of detail optimization technology to achieve multi-element linkage simulation once per second. Through multi-objective optimization algorithm, the module designs irrigation and drainage scheme selection rules and completes instruction issuance, execution data reception, and parameter dynamic correction.

2. The intelligent decision-making system for precise water allocation in agricultural irrigation areas based on crop water requirement coefficient as described in claim 1, characterized in that, The hardware acquisition and execution module comprises five functional sub-modules, each working collaboratively along a link. The specific design is as follows: The integrated air-ground-space multi-dimensional data acquisition sub-module scans the irrigation area at a fixed cycle of 30 minutes, simultaneously collecting topographic data, meteorological data, and crop growth data. The drone is equipped with a hyperspectral camera and a GPS positioning module. The crop physiology and soil multi-parameter sensing sub-module divides the terraced fields into acquisition units every 5 meters according to the vertical elevation gradient. Sensors are evenly distributed across saline-alkali land plots, including leaf water potential sensors, stem flow rate sensors, and soil moisture sensors. The sensors, including soil water and salt content sensors and soil temperature sensors, complete data acquisition every five minutes, and all sensors have a daily automatic calibration function; the irrigation and drainage equipment status monitoring and intelligent execution submodule collects the operating parameters of water pumps, valves, and drainage outlets in real time, and responds to commands within seconds and drives the equipment to act; the end-to-end data transmission and edge processing submodule is deployed according to the rule of one edge computing node per 1,000 mu of irrigation area, and the initial data preprocessing time is strictly controlled within 0.3 seconds; the backup power supply and emergency protection submodule is designed with automatic switching in case of power failure, real-time fault alarm and emergency operation plan.

3. The intelligent decision-making system for precise water allocation in agricultural irrigation areas based on crop water requirement coefficient as described in claim 1, characterized in that, The software algorithm and modeling module includes an algorithm and hardware adaptation optimization sub-module, specifically designed as follows: A comprehensive hardware device parameter database is constructed, including the operating parameter ranges, data output formats, and communication protocol standards of different types of sensors and actuators; real-time hardware operation status monitoring logic is designed to identify the amplitude of operational fluctuations by comparing the actual hardware operating parameters with the database standard parameters. Establish dynamic adjustment rules for algorithm parameters, and adaptively adjust data processing thresholds and calculation step size algorithm parameters according to the degree of hardware operation fluctuations.

4. The intelligent decision-making system for precise water allocation in agricultural irrigation areas based on crop water requirement coefficient as described in claim 1, characterized in that, The dynamic water demand coefficient correction algorithm of the software algorithm and modeling module is implemented through a crop physiological-environmental factor dual-driven water demand coefficient correction submodule. This submodule is designed with weighted coupling logic of physiological and environmental factors, and uses the constructed dynamic water demand coefficient correction formula to calculate the plot-level dynamic water demand coefficient. The formula is as follows: ； in, This refers to the crop water requirement coefficient after dynamic correction at the plot level. This is the basic water requirement coefficient for the corresponding crop growth period; These are the weighting coefficients for physiological indicators; To measure the water potential of crop leaves; The appropriate leaf water potential threshold for crops at the corresponding growth stage; To measure the stem flow rate of crops; The appropriate stem flow rate threshold for crops at the corresponding growth stage; This is the comprehensive correction coefficient for environmental factors; The calculation is derived from three environmental factors: temperature, humidity, and soil moisture. The calculation formula is as follows: ； Where Σ is the summation symbol; The values ​​are 1, 2, and 3, which correspond to air temperature, humidity, and soil moisture, respectively. For the first Environmental factor influence coefficient; For the first Measured values ​​of environmental factors; is the optimal crop growth threshold for the i-th type of environmental factor.

5. The intelligent decision-making system for precise water allocation in agricultural irrigation areas based on crop water requirement coefficient as described in claim 1, characterized in that, The software algorithm and modeling module completes the spatial adaptation of water demand coefficient through the terraced field vertical space water demand coefficient distribution submodule. This submodule designs multi-factor comprehensive adaptation logic for terraced field terrain and adopts the constructed terraced field vertical space water demand coefficient correction formula. The calculation formula is as follows: ； in, Customized crop water requirement coefficients for the corresponding gradient zones of terraced fields; This refers to the crop water requirement coefficient after dynamic correction at the plot level. Elevation influence coefficient; The average elevation of the gradient zone corresponding to the terraced fields; This is the elevation value at the lowest point in the irrigation area; This is the slope correction factor; The slope of the gradient zone corresponding to the terraced fields; It is a sine function; Soil texture affects the weighting; This represents the soil texture difference coefficient corresponding to the gradient zone; This represents the average soil texture coefficient of the irrigation area.

6. The intelligent decision-making system for precise water allocation in agricultural irrigation areas based on crop water requirement coefficient as described in claim 1, characterized in that, The digital twin irrigation district full-element mapping submodule is the simulation unit of the digital twin simulation and decision-making module. It designs a collaborative simulation logic across four element layers, with the following functions for each layer: The geospatial layer recreates the irrigation district's topography, canal network, roads, and water source geographical elements, with an accuracy of ±0.1 meters; the crop growth layer correlates with dynamic water demand coefficient data to simulate the crop's response to irrigation measures; the soil water and salt layer is linked to the water demand coefficient-soil water and salt coupling submodule. The coupling between the water demand coefficient and soil water and salt is achieved through a dynamic correlation model designed in the water demand coefficient-soil water and salt coupling submodule. The coupling correlation formula for the soil water and salt layer is as follows: ； in, The dynamic soil electrical conductivity is denoted by λ, which is the influence coefficient of the water demand coefficient. The value of λ ranges from 0.3 to 0.5 and is calibrated through training data from the coupling test of soil water and salt with the water demand coefficient. is a natural exponential function; μ is the irrigation dilution coefficient, with a value ranging from 0.002 to 0.005, calibrated through comparative experiments on the soil water and salt dilution effects under different irrigation amounts; The initial soil electrical conductivity of the irrigation area is the measured value, determined by the average value of multiple sampling points in different irrigation areas. This refers to the crop water requirement coefficient after dynamic correction at the plot level. The maximum water requirement coefficient for the corresponding crop growth period was determined through field trials with similar crops; This refers to the amount of irrigation water.

7. The intelligent decision-making system for precise water allocation in agricultural irrigation areas based on crop water requirement coefficient as described in claim 1, characterized in that, The multi-objective optimization algorithm of the digital twin simulation and decision-making module is implemented through the irrigation and drainage collaborative intelligent decision-making sub-module. The sub-module is specifically designed as follows: an improved multi-objective particle swarm optimization algorithm is adopted, setting three major optimization objectives: water saving rate, crop growth suitability, and equipment energy consumption. An objective function is constructed and a weight allocation rule is determined. A three-level control threshold system is designed, with the critical values ​​of soil conductivity and crop water requirement coefficient as key indicators. The trigger intervals for three irrigation and drainage modes are divided: when soil conductivity < critical value and water requirement coefficient ≥ critical value, the conventional precision irrigation mode is triggered; when soil conductivity ≥ critical value and water requirement coefficient ≥ critical value, the precision salt leaching irrigation mode is triggered; and when soil conductivity ≥ critical value and water requirement coefficient < critical value, the first-drainage-then-replenishment irrigation mode is triggered. The algorithm decision-making time is strictly controlled within 0.5 seconds, and the time for the instruction to be sent from the decision output sub-module to the hardware execution layer does not exceed 0.2 seconds.

8. The intelligent decision-making system for precise water allocation in agricultural irrigation areas based on crop water requirement coefficient as described in claim 1, characterized in that, The digital twin simulation and decision-making module achieves closed-loop correction through a decision output and feedback correction sub-module. This sub-module is designed with a dual-path feedback correction logic: First, the equipment execution data feedback path receives actual execution data of the equipment, such as pump speed, valve opening, and drainage flow rate, from the hardware execution layer. This data is compared with preset command values ​​to calculate the deviation and fed back to the digital twin model to correct the simulation parameters. Second, the irrigation area status data feedback path receives real-time monitoring data of water demand coefficient, soil water and salt content, and crop physiological indicators from the irrigation area. This data is compared with algorithm predictions and fed back to the software algorithm module to correct the water demand coefficient threshold, terrace adaptation parameters, and decision model weights. Simultaneously, a parameter correction amplitude rule is designed, quantifying adjustments based on the degree of data deviation. Parameters are fine-tuned when the deviation is ≤5%, and deep correction is triggered when the deviation is >5%.

Images