Regional cotton water and fertilizer dynamic decision-making method considering salt stress and freeze-thaw mechanism

Abstract

The invention discloses a regional cotton water and fertilizer dynamic decision-making method considering salt stress and a freeze-thaw mechanism, and the method comprises the steps: obtaining the multi-source historical data of a target region, integrating the multi-source historical data, analyzing the historical rainfall data of the target region, and dividing the historical rainfall data into Fengping low hydroponic years; constructing a crop growth prediction model coupled with salt stress and a freeze-thaw process through the integrated and divided data; performing sensitivity analysis on the crop growth prediction model to obtain an objective function, optimizing the objective function, setting constraint conditions, and specifying a multi-objective optimized water and fertilizer strategy; establishing a farmland information system based on the Internet of Things technology, and storing real-time monitoring data; and establishing a water and fertilizer system dynamic adaptation system through a multi-objective optimized water and fertilizer strategy and the farmland information system based on the Internet of Things technology, and providing an accurate water and fertilizer system according to the water and fertilizer system dynamic adaptation system. According to the invention, a more accurate water and fertilizer system can be provided for farmers.

Description

Technical Field

[0001] The present invention relates to the technical field of agricultural irrigation, and particularly to a dynamic decision-making method for regional cotton water and fertilizer considering salt stress and freeze-thaw mechanism. Background Art

[0002] Traditional agricultural irrigation and fertilization systems usually rely on preset time and dosage, lacking the ability to dynamically adjust according to actual environmental changes and crop needs. This method cannot effectively cope with complex field conditions, easily leading to problems such as water resource waste, soil nutrient loss, environmental pollution, and uneven crop growth, which are particularly prominent in saline-alkali land and freeze-thaw regions.

[0003] In some areas, the climate is dry, the average annual precipitation is extremely low and the evaporation is extremely large, and the contradiction between water resource supply and demand is extremely prominent. In some arable lands that can be utilized, the soil salt content is generally high, and the degree of salinization in some areas is serious, which not only affects the soil permeability, but also has a stress effect on the crop water absorption and nutrient absorption ability. Freeze-thaw cycles will change the water and salt migration process in the soil, making the soil water supply uneven, and affecting root development and nutrient use efficiency.

[0004] Current cotton field irrigation and fertilization indexes are mostly based on soil water content, soil water potential, atmospheric temperature and humidity, etc., with problems such as large spatial variability and delayed response, and it is difficult to timely feedback the actual water and nutrient requirements during the cotton growth process, especially there are obvious deficiencies in coping with salt stress and freeze-thaw changes. Summary of the Invention

[0005] To solve the above technical problems existing in the prior art, the present invention proposes a dynamic decision-making method for regional cotton water and fertilizer considering salt stress and freeze-thaw mechanism, which can optimize the water and fertilizer utilization efficiency while reducing the impact of saline-alkali stress on crop growth, and provide a more scientific and accurate management plan for cotton planting in arid and semi-arid regions.

[0006] To achieve the above object, the present invention provides a dynamic decision-making method for regional cotton water and fertilizer considering salt stress and freeze-thaw mechanism, including:

[0007] Obtain multi-source historical data of the target area, integrate the multi-source historical data, and analyze the historical rainfall data of the target area to divide the wet, normal and dry hydrological years;

[0008] Construct a crop growth prediction model coupling salt stress and freeze-thaw process through the integrated and divided data;

[0009] Conduct a sensitivity analysis on the crop growth prediction model to obtain an objective function, optimize the objective function, set constraint conditions, and specify a water and fertilizer strategy for multi-objective optimization;

[0010] Establish a farmland information system based on Internet of Things technology to store real-time monitoring data;

[0011] Establish a dynamic adaptation system for water and fertilizer regime through the multi-objective optimized water and fertilizer strategy and the farmland information system based on Internet of Things technology, and provide a precise water and fertilizer regime according to the dynamic adaptation system for water and fertilizer regime.

[0012] Preferably, the multi-source historical data includes meteorological data, soil data, crop data, field management data, and fertilizer usage data;

[0013] Dividing the wet, normal, and dry hydrological years includes: classifying river runoff into extremely wet years, relatively wet years, normal years, relatively dry years, and extremely dry years according to national standards, where extremely wet years and relatively wet years are called wet years, and relatively dry years and extremely dry years are called dry years.

[0014] Preferably, the crop growth prediction model coupling salt stress and freeze-thaw process includes:

[0015] Salt module: used to increase the irrigation water volume by considering the leaching coefficient to reduce the soil salt content;

[0016] Freeze-thaw module: used to simulate the snow-water conversion process between surface snow cover and underground frozen soil, and obtain the change of soil moisture content by calculating the snow-water equivalent depth and snow depth parameters;

[0017] Fertilizer module: used to simulate the transport and transformation process of fertilizers in the soil and their promotion effect on crop growth according to fertilizer type, application rate, and fertilization time factors.

[0018] Preferably, the salt module calculates the irrigation quota according to the field water holding capacity, measured soil moisture content, irrigation wetting ratio, planned wetting layer depth, and salt leaching coefficient parameters to control the soil salt content;

[0019] The calculation method of the irrigation quota is:

[0020] ;

[0021] In the formula, is the field water holding capacity; is the measured soil moisture content; is the irrigation wetting ratio; is the planned wetting layer depth; is the salt leaching coefficient; is the irrigation quota;

[0022] The freeze-thaw module divides the snow-water equivalent depth by the density of snow to convert the snow-water equivalent depth into the actual snow depth, specifically:

[0023] ;

[0024] ;

[0025] Wherein, is the snow water equivalent depth; is the daily snow accumulation rate; is the daily snow sublimation rate; is the daily snow melting rate; is the snow depth; is the snow density.

[0026] Preferably, a sensitivity analysis is performed on the crop growth prediction model, and the objective function obtained includes:

[0027] Through the Sobol index method in the SALib library, a sensitivity analysis is jointly performed on the crop growth model. By analyzing the first-order Sobol index and the second-order Sobol index, the parameters sensitive to the crop growth index are screened out, and the sensitivity parameters are calibrated, while the non-sensitivity parameters adopt the default values; among them, the first-order Sobol index is used to measure the contribution of a single input variable to the variance of the output variable; the second-order index is used to measure the contribution of the input variable and its combination to the variance of the output variable;

[0028] The NSGA-II algorithm is used for solving, and the objective function is set to minimize the error between the simulated values and the observed values of the leaf area index, above-ground biomass, and final yield at different growth stages.

[0029] Preferably, the objective function is:

[0030] ;

[0031] Wherein, and are respectively the observed value and the simulated value of the leaf area index, and are respectively the observed value and the simulated value of the above-ground biomass, and respectively represent the observed value and the simulated value of the yield, n is the sampling times of the leaf area index and the above-ground biomass in the experiment, and m is the sampling times of the yield.

[0032] Preferably, the optimization of the objective function is:

[0033] ;

[0034] ;

[0035] ;

[0036] Wherein, TWSO is the yield, and TWSO sim is the simulated yield, I is the total irrigation water volume, and I sim is the simulated total irrigation water volume, N is the total fertilization amount, and N sim is the simulated total fertilization amount;

[0037] The constraint conditions are set as:

[0038] ;

[0039] ;

[0040] Wherein, R is the available water volume, P is the rainfall, and ET is the crop water consumption;

[0041] ;

[0042] Wherein, is the average annual yield over the years;

[0043] The decision variables are non - negative:

[0044] ;

[0045] ;

[0046] Output the Pareto - optimal solution set, that is, the combination of the optimal irrigation water volume, fertilization amount and yield.

[0047] Preferably, the farmland information system based on the Internet of Things technology stores real - time monitoring data including:

[0048] Crop data such as leaf area index and above - ground biomass obtained by remote sensing technology inversion;

[0049] Soil moisture content, soil water permeability and soil temperature are transmitted in real - time through the intelligent entropy meters laid in the field;

[0050] Field management data manually input by the field manager, where the field management data includes crop variety, planting date, harvest date, fertilization status, soil moisture content thresholds for starting irrigation and soil nutrient thresholds for fertilization respectively;

[0051] Rainfall, maximum temperature, minimum temperature, wind speed, solar radiation, vapor pressure and snow depth data are monitored in real - time through small meteorological stations in the field.

[0052] Preferably, the water - fertilizer regime dynamic adaptation system includes:

[0053] Data reading module: used to read the real - time data stored in the farmland information system based on the Internet of Things technology, including relevant parameters of soil salt content and freeze - thaw process;

[0054] Data analysis and storage module: used to analyze and store the read data, and evaluate the current soil salinity status and freeze-thaw state by combining the simulation results of the salinity module and the freeze-thaw module;

[0055] Reference water and fertilizer regime selection module: used to select an optimized water and fertilizer regime as the reference water and fertilizer regime for dynamic correction according to requirements;

[0056] Crop growth status prediction module: used to obtain the crop growth status by using the reference water and fertilizer regime and the real-time data-driven model, compare it with the observed value, and decide whether to correct the water and fertilizer regime;

[0057] Dynamic adaptation module: used to re-determine the optimal water and fertilizer regime through coupling the crop model with the particle swarm optimization algorithm;

[0058] User interface and operation module: used to provide a visual interface for operations;

[0059] System management and maintenance module: used to manage, repair and improve the system.

[0060] Preferably, when the crop growth status prediction module finds that the simulated above-ground biomass is lower than the observed value or the water demand decision error does not meet the constraints, the dynamic adaptation module re-determines the optimal water and fertilizer regime through coupling the crop growth model and using the particle swarm optimization algorithm; among them, the salinity module reduces the soil salinity content by adjusting the irrigation quota, and the freeze-thaw module optimizes the soil water management to ensure the water supply and root growth of cotton during the freeze-thaw process.

[0061] Compared with the prior art, the present invention has the following advantages and technical effects:

[0062] (1) The present invention accurately controls the water demand through real-time monitoring and data analysis, optimizes the utilization of water resources, and improves agricultural production efficiency;

[0063] (2) The present invention uses Internet of Things technology to collect real-time meteorological, soil, crop and field management data, improving the timeliness and accuracy of irrigation decision-making;

[0064] (3) Embedding the salinity and freeze-thaw modules into the existing crop growth model can more comprehensively simulate the crop growth conditions, improve the applicability and prediction accuracy of the model. The salinity module effectively reduces the soil salinity content by adjusting the irrigation quota, improving the water absorption and nutrient status of crops; the freeze-thaw module can accurately simulate the snow water conversion process between surface snow cover and underground frozen soil, optimize the soil water management, ensure the water supply and root growth of crops during the freeze-thaw process, thereby improving the irrigation efficiency and yield of crops;

[0065] (4) The present invention constructs a dynamic correction system for crop water and fertilizer regimes, which can dynamically adjust the water and fertilizer regimes according to real-time data and crop growth conditions, reducing the damage caused to crops by irrigation delay. Description of the Drawings

[0066] The drawings forming a part of this application are used to provide a further understanding of this application. The schematic embodiments and descriptions thereof of this application are used to explain this application and do not constitute an improper limitation to this application. In the drawings:

[0067] Figure 1 It is a flow chart of a regional cotton water and fertilizer dynamic decision-making method considering salt stress and freeze-thaw mechanism according to an embodiment of the present invention. Specific Embodiments

[0068] It should be noted that, without conflict, the embodiments in this application and the features in the embodiments can be combined with each other. The following will refer to the drawings and combine with the embodiments to detail this application.

[0069] It should be noted that the steps shown in the flowchart of the drawings can be executed in a computer system such as a set of computer-executable instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than here.

[0070] This embodiment further details the technical solution by listing the cotton planting in a demonstration area of a certain place.

[0071] A regional cotton water and fertilizer dynamic decision-making method considering salt stress and freeze-thaw mechanism, as Figure 1 , includes the following steps:

[0072] S1: Multi-source data integration and hydrological year division;

[0073] Collect meteorological, soil, crop, field management data and fertilizer usage as crop model driving data, analyze historical rainfall data to divide wet, normal and dry hydrological years, and simulate respectively under different hydrological year scenarios. Data such as fertilizer type, application rate, and fertilization time provide basic information for subsequent fertilizer decision-making.

[0074] The method for dividing hydrological years is as follows:

[0075] In the national standard "Standard for Basic Hydrological Terms and Symbols" (GB / T 50095-98), the high, normal, and low river runoff is divided into five categories: extremely wet year, relatively wet year, normal year, relatively dry year, and extremely dry year. In water resources analysis, extremely wet year and relatively wet year are often referred to as wet years; relatively dry year and extremely dry year are called dry years. Extremely wet year: The year when the river runoff is the maximum or close to the maximum value in previous years; relatively wet year: The year when the river runoff is significantly greater than the average value; normal year: The year when the river runoff is closest to the average value; relatively dry year: The year when the river runoff is significantly less than the average value; extremely dry year: The year when the river runoff is the minimum or close to the minimum value in previous years.

[0076] The runoff series generally follows the P-Ⅲ type probability distribution, and the frequency analysis method is used to determine the statistical parameters and design values of each frequency as the standard for dividing runoff and high, normal, and low hydrological years. The division method is shown in Table 1.

[0077] Table 1

[0078] Level of abundant, normal and dry years Classification standard (P) Abundant flow year P ≦ 37.5% Normal flow year 37.5% < P ≦ 62.5% Dry flow year P > 62.5%

[0079] S2: Construction of a crop growth model coupling salt stress and freeze-thaw process;

[0080] Based on the existing crop growth model, a salt module, a freeze-thaw module, and a fertilizer module are embedded. The salt module mainly considers the leaching coefficient to increase the irrigation water volume to achieve the purpose of reducing the soil salt content; the freeze-thaw module simulates the snow-water conversion process between surface snow cover and subsurface frozen soil, and accurately reflects the change of soil moisture content by calculating parameters such as snow-water equivalent depth and snow depth; the fertilizer module simulates the transport and transformation process of fertilizers in the soil and their promoting effect on crop growth according to factors such as fertilizer type, application rate, and fertilization time, so as to act more comprehensively on the crop growth model.

[0081] When simulating cotton irrigation and fertilization, the influence of salt stress and freeze-thaw on irrigation and fertilization amount is considered. When the soil volumetric water content is lower than the lower limit of the irrigation threshold, irrigation is started; when the soil nutrient content is lower than the lower limit of the fertilization threshold, fertilization is started.

[0082] In order to regulate the influence of salt stress on cotton, a salt module is set in this embodiment, and the irrigation quota is adjusted by considering the leaching coefficient to achieve the purpose of reducing the soil salt content. Specifically, the salt module calculates the irrigation quota according to parameters such as field water holding capacity, measured soil water content, irrigation wetting ratio, planned wetting layer depth, and salt leaching coefficient, so as to effectively control the soil salt content and improve the water absorption and nutritional status of cotton. This module has good adaptability and application effect under different hydrological year scenarios, and can significantly improve the growth quality and yield of cotton.

[0083] The calculation formula of the irrigation quota is as follows:

[0084] ;

[0085] In the formula, (cm 3 ∙cm -3 ) is the field water holding capacity; (cm 3 ∙cm -3 ) is the measured soil water content; when > , let > ; is the irrigation wetting ratio; is the planned wetting layer depth; is the salt leaching coefficient.

[0086] In this embodiment, the value range of the leaching coefficient of the salt module is 1.0 - 2.0. According to the monitoring data of soil salt content, the irrigation quota is dynamically adjusted to ensure that the soil salt content is within an appropriate range.

[0087] Before cotton planting, the soil undergoes freeze-thaw, which affects the initial soil moisture content. Therefore, the source code of the crop growth model is modified to call the snow module under the soil water balance module to simulate the freeze-thaw scenario before cotton sowing. This module describes the accumulation and melting of snow due to precipitation, snowmelt, and sublimation. It aims to track the thickness of the snow layer present on the surface, such as the snow water equivalent depth. By dividing the snow water equivalent depth by the density of snow, the snow water equivalent depth can be converted into the actual snow depth.

[0088] The conversion formula between snow water equivalent and snow depth is as follows:

[0089] ;

[0090] ;

[0091] In the formula, (cm) is the snow water equivalent depth; (cm∙d -1 ) is the daily accumulation rate of snow; (cm∙d -1 ) is the daily sublimation rate of snow; (cm∙d -1 ) is the daily melting rate of snow; (cm) is the depth of snow; (cm∙cm -1 ) is the density of snow.

[0092] By real-time monitoring of meteorological data such as air temperature and snowfall, accurately calculate the snow water equivalent depth and snow depth, providing a reliable basis for the water and fertilizer decision-making of cotton.

[0093] In addition, the methods for calculating the snow accumulation rate, sublimation rate, and melting rate are as follows:

[0094] Snow accumulation rate: If the minimum temperature on the day is less than or equal to the lower limit of the minimum critical temperature of the snow cover, it is considered that all rainfall will be converted into snow accumulation. If the minimum temperature is greater than or equal to the upper limit of the minimum critical temperature of the snow cover, there is no snow accumulation. If the minimum temperature is between these two critical temperatures, the snow accumulation rate is a part of the rainfall, and the proportion of this part is as follows:

[0095] ;

[0096] In the formula, T min (°C) is the minimum temperature, T minc1 (°C) is the upper limit of the critical minimum temperature of the snow cover, T minc2 (°C) is the lower limit of the critical minimum temperature of the snow cover.

[0097] Snow sublimation rate: If the snow water equivalent depth is greater than the snow water equivalent threshold and there is no snow accumulation, the snow sublimation rate is equal to the evaporation rate. Otherwise, the snow sublimation rate is 0. The sublimation rate cannot exceed the existing snow water equivalent depth, so a limiting function is used to ensure that the sublimation rate does not exceed the existing snow amount.

[0098] Snow melting rate: If the minimum temperature is less than the critical minimum temperature for snow melting, there is no snow melting. If the minimum temperature is greater than or equal to 0 °C and the maximum temperature is less than the critical maximum temperature for snow melting, there is also no snow melting. Otherwise, the snow melting rate is calculated as follows:

[0099] ;

[0100] In the formula, R mi (cm∙d -1 ) is the snow melting rate, T min (°C) is the minimum temperature, T minc (°C) represents the critical minimum temperature for snow melting, R mt (cm∙°C -1 ∙d -1 ) is the melting rate per degree Celsius above the critical minimum temperature per day.

[0101] Limit the melting rate: The melting rate cannot exceed the existing snow amount minus the sublimated snow amount. Use a limiting function to ensure that the melting rate does not exceed , in the formula, H w (cm) is the snow water equivalent depth, R s(cm∙d -1 ) is the sublimation rate of snow.

[0102] By introducing the salt module and the freeze-thaw module, the irrigation efficiency and yield of cotton in the region have been significantly improved. The salt module effectively reduces the soil salt content and improves the water absorption and nutritional status of cotton; the freeze-thaw module optimizes the soil water management and ensures the water supply and root growth of cotton during the freeze-thaw process. The combined application of these two modules provides strong support for the efficient water use and sustainable development of cotton in the region.

[0103] S3: Sensitivity analysis and calibration of model parameters under multiple factors;

[0104] Through the Sobol index method in SALib (Sensitivity Analysis Library), a sensitivity analysis is jointly conducted with the crop growth model. The first-order Sobol index and the second-order Sobol index are analyzed to screen out the parameters sensitive to the crop growth indicators, and the sensitivity parameters are calibrated, while the non-sensitive parameters adopt the default values.

[0105] SALib is a Python library for performing global sensitivity analysis. It supports multiple sensitivity analysis methods, and the common method is the analysis method based on the Sobol index.

[0106] The Sobol index can be divided into the first-order (first-order), second-order (second-order), and higher-order Sobol indices.

[0107] First-order Sobol index: The first-order Sobol index (S i ) measures the contribution of a single input variable to the variance of the output variable, and the calculation formula is:

[0108] ;

[0109] In the formula, V represents variance, E represents expectation, Y is the output variable, and X i is the i-th input variable.

[0110] Second-order Sobol index: The second-order Sobol index (S ij ) measures the contribution of two input variables and their combination to the variance of the output variable, and the calculation formula is:

[0111] ;

[0112] In the formula, X i and X j are two different input variables.

[0113] SALib uses the Monte Carlo simulation method to estimate Sobol indices. It approximates the expected value and variance by generating random samples (sampling). By sampling multiple times, the estimated values of the Sobol indices can be obtained.

[0114] Specifically, the steps for using SALib include:

[0115] Define the problem: Define the distributions of the input parameters and the output function.

[0116] Generate samples: Use the Monte Carlo method to generate random samples of the input parameters.

[0117] Calculate the output: For each sample, calculate the value of the output variable.

[0118] Calculate the Sobol indices: Use these samples to calculate the estimated values of the Sobol indices.

[0119] Analyze the results: Analyze the results of the Sobol indices to understand the influence of each input variable on the output.

[0120] Result interpretation:

[0121] If the first-order Sobol index of a certain parameter is large, it indicates that this parameter has a greater influence on the output. If the second-order Sobol index of a certain parameter is large, it indicates that the interaction between this parameter and other parameters has a greater influence on the output.

[0122] This global sensitivity analysis helps to understand the influence of input variables on the output, optimize the model, and make decisions. Through sensitivity analysis, the behavior of the model can be better understood, and it can be determined which factors are decisive, thus providing useful information when the model needs to be adjusted or optimized.

[0123] Couple the crop growth model and the multi-objective optimization model for parameter calibration. The solution method uses the NSGA-II algorithm, and the objective function is set to minimize the errors between the simulated values and the observed values of the leaf area index, above-ground biomass, and final yield at different growth stages.

[0124] The specific implementation method is as follows:

[0125] Set the objective function:

[0126] ;

[0127] In the formula, and are the observed value and the simulated value of the leaf area index respectively, and are the observed value and the simulated value of the above-ground biomass respectively, and They respectively represent the observed and simulated values of yield. n is the number of samplings of leaf area index and aboveground biomass in the experiment, and m is the number of samplings of yield.

[0128] Generate the initial population: Set the initial parameters of the algorithm, including population size, number of iterations, crossover and mutation probabilities. Generate a random initial population, where each individual represents a combination of parameters;

[0129] Model simulation: Input the model driving data and the parameters to be calibrated to run the model and obtain the simulation results;

[0130] Evaluation: Calculate the error between the observed and simulated values, and evaluate the fitness of each individual according to the objective function, that is, the comprehensive index of the minimum irrigation amount, minimum fertilization amount and maximum yield;

[0131] Selection: Use the NSGA-II algorithm to perform the screening of multi-objective optimization. Select excellent individuals from the population through the crowding algorithm to ensure the balance of multiple objectives during the screening process;

[0132] Crossover: Perform crossover on the selected individuals to generate new individuals. Decide whether to perform crossover according to the set crossover probability, and apply the crossover operator to the individuals for crossover to generate new individuals;

[0133] Mutation: Perform mutation on the individuals after crossover to increase the population diversity. Decide whether to perform mutation according to the mutation probability, and use the mutation operator to perform mutation operations on the individuals to generate new individuals;

[0134] Produce a new population: Use the non-dominated sorting and crowding degree algorithms to replace the original population and the newly generated individuals to generate a new population;

[0135] Termination condition: The conditions for terminating the optimization, such as reaching the maximum number of iterations or obtaining the optimal solution set;

[0136] Output the result: Output the optimal solution set, that is, the optimal combination of a set of parameters, which minimizes the error between the simulated value and the observed value.

[0137] Three groups of sensitivity parameters and the corresponding three localized crop growth models are obtained for three typical hydrological years.

[0138] S4: Formulation of water and fertilizer strategies based on multi-objective optimization;

[0139] When optimizing the reference water and fertilizer regime, the corresponding crop growth model is selected according to the hydrological year type in the planned year. The salt module, freeze-thaw module, and fertilizer module embedded in the model can more accurately reflect the soil salinity, freeze-thaw process, and the impact of fertilizers on cotton growth under different hydrological year scenarios. By coupling the NSGA-II algorithm, the objective function is set to maximize yield, minimize irrigation water volume, and maximize fertilizer use efficiency. At the same time, considering the control of soil salinity content by the salt module, the optimization of soil water management by the freeze-thaw module, and the promotion of crop growth by the fertilizer module, a more scientific and reasonable water and fertilizer management plan can be obtained. The same optimization model as in S3 is used, but its objective function needs to be modified and some constraint conditions are added.

[0140] The objective function in this optimization is as follows:

[0141] ;

[0142] ;

[0143] ;

[0144] In the formula, TWSO (kg∙ha -1 ) is the yield, TWSO sim (kg∙ha -1 ) is the simulated yield, I (mm) is the total irrigation water volume, I sim (mm) is the simulated total irrigation water volume, N (kg∙ha -1 ) is the total fertilizer application amount, N sim (kg∙ha -1 ) is the simulated total fertilizer application amount.

[0145] The constraint conditions are set as follows:

[0146] ;

[0147] ;

[0148] In the formula, R (mm) is the available water volume, P (mm) is the rainfall, and ET (mm) is the crop water consumption.

[0149] In addition, in this example, another constraint is added as follows:

[0150] ;

[0151] In the formula, (kg∙ha -1 ) is the average annual yield over the years.

[0152] The decision variables are non-negative:

[0153] ;

[0154] ;

[0155] Output result: Output the Pareto optimal solution set, that is, the combination of the optimal irrigation water volume, fertilization amount and yield. Combinations that meet the requirements can be selected from it, namely three different combinations of high water use efficiency, high yield and water saving.

[0156] Due to the introduction of the salt module and the freeze-thaw module, the optimized water-fertilizer regime can better adapt to the soil salinization and freeze-thaw phenomena in the area, effectively reduce the soil salt content, ensure the water supply and root growth of cotton during the freeze-thaw process, and thus improve the yield and irrigation efficiency of cotton.

[0157] S5: Establish a farmland information system based on Internet of Things technology;

[0158] To store all real-time monitoring data and provide data input for the dynamic correction system.

[0159] The farmland information system based on Internet of Things technology includes real-time crop data, real-time soil data, real-time field management data, real-time meteorological data and real-time fertilizer data. Crop data such as leaf area index and aboveground biomass are obtained by remote sensing technology inversion; the intelligent entropy meter laid in the field transmits real-time soil moisture content, soil water permeability, soil temperature and other related soil data; the field management data manually input by field managers include but are not limited to crop varieties, planting dates, harvest dates, fertilization status (including fertilizer types, application rates, fertilization times, etc.), the soil moisture content thresholds for starting irrigation and the soil nutrient thresholds for fertilization respectively; the small meteorological station in the field monitors the must-have data driven by the model such as rainfall, maximum temperature, minimum temperature, wind speed, solar radiation, vapor pressure, snow depth, etc.

[0160] S6: Build a dynamic adaptation system for cotton water-fertilizer regime;

[0161] The construction of a dynamic correction system for cotton's water and fertilizer regime consists of a data reading module, a data analysis and storage module, a reference water and fertilizer regime selection module, a crop growth status prediction module, a dynamic adaptation module, a user interface and operation module, and a system management and maintenance module, jointly forming a precise water and fertilizer correction system. The salt module in the dynamic correction system can adjust the water and fertilizer strategy in a timely manner according to the real-time monitored soil salt content data. For example, in the dry season, the soil salt content may rise rapidly, and the module will immediately recommend increasing the irrigation volume to quickly reduce the soil salt content and prevent cotton from being stressed by salt. This dynamic adjustment can effectively reduce the growth retardation and yield decline of cotton caused by excessive soil salt, and improve the timeliness and effectiveness of irrigation. The freeze-thaw module can monitor meteorological data such as air temperature and snowfall in real time, accurately predict the soil water change during the freeze-thaw process, and thus optimize the irrigation timing. For example, when the air temperature gradually rises in spring, the module predicts that the snow is about to melt, and it will recommend reducing the irrigation volume appropriately before the snow melts to avoid water waste; when the air temperature drops suddenly and the soil freezes, the module will recommend irrigating in advance to ensure that there is enough water in the soil for cotton to absorb and avoid insufficient water supply caused by soil freezing. The fertilizer module can monitor the nutrient situation in the soil in real time. If the nutrients are insufficient, it will immediately send a fertilization signal.

[0162] The following are the functions implemented by each module:

[0163] Data reading module: Read the real-time data stored in S5, including the soil salt content and relevant parameters of the freeze-thaw process.

[0164] Data analysis and storage module: Analyze and store the read data, and accurately evaluate the current soil salt status and freeze-thaw state in combination with the simulation results of the salt module and the freeze-thaw module.

[0165] Reference water and fertilizer regime selection module: Select an optimized water and fertilizer regime in S4 as the reference water and fertilizer regime for dynamic correction according to the requirements.

[0166] Crop growth status prediction module: Use the reference water and fertilizer regime and the real-time data-driven model to obtain the crop growth status, compare it with the observed values, and decide whether to correct the water and fertilizer regime.

[0167] Dynamic adaptation module: If the water demand decision error does not meet the constraints or the simulated aboveground biomass is less than the observed value, re-decide the optimal water and fertilizer regime through the coupling of the crop model and the particle swarm algorithm.

[0168] User interface and operation module: A visual interface that facilitates user operation.

[0169] System management and maintenance module: User management and the repair and improvement of system functions.

[0170] When the crop growth status prediction module finds that the simulated above-ground biomass is lower than the observed value or the water requirement decision error does not meet the constraints, the dynamic adaptation module will couple with the crop growth model and use the particle swarm algorithm to re-determine the optimal water and fertilizer regime. The salinity module reduces the soil salinity content by adjusting the irrigation quota, while the freeze-thaw module optimizes the soil water management to ensure the water supply and root growth of cotton during the freeze-thaw process, thereby reducing the damage caused by irrigation delay to the crop and improving the dynamic correction effect of the water and fertilizer regime.

[0171] Through the above functions, a more accurate water and fertilizer regime can be provided for farmers, clearly presenting the irrigation and fertilization dates and dosages, saving water and fertilizer while ensuring crop growth.

[0172] The above is only a preferred specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by those skilled in the art within the technical scope disclosed by the present application should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A regional cotton water and fertilizer dynamic decision-making method considering salt stress and freeze-thaw mechanism, characterized in that: include: Acquire multi-source historical data of the target area, integrate the multi-source historical data, and analyze the historical rainfall data of the target area to divide the wet, normal and dry years; A crop growth prediction model coupling salt stress and freeze-thaw process was constructed through integrated and divided data; Conducting sensitivity analysis on the crop growth prediction model to obtain an objective function, optimizing the objective function, setting constraints, and specifying a multi-objective optimization water and fertilizer strategy; Establish a farmland information system based on Internet of Things technology to store real-time monitoring data; A dynamic adaptation system for water and fertilizer system is established through the multi-objective optimized water and fertilizer strategy and the farmland information system based on the Internet of Things technology, and a precise water and fertilizer system is provided according to the dynamic adaptation system for water and fertilizer system.

2. The regional cotton water and fertilizer dynamic decision-making method considering salt stress and freeze-thaw mechanism according to claim 1 is characterized in that: The multi-source historical data includes meteorological data, soil data, crop data, field management data and fertilizer usage data; The classification of the wet, normal and dry years includes: according to national standards, the wet, normal and dry river runoff are divided into extremely wet years, relatively wet years, normal years, relatively dry years and extremely dry years, among which the extremely wet years and relatively wet years are called wet years, and the relatively dry years and extremely dry years are called dry years.

3. The regional cotton water and fertilizer dynamic decision-making method considering salt stress and freeze-thaw mechanism according to claim 1 is characterized in that: The crop growth prediction model coupled with salt stress and freeze-thaw process includes: Salt module: used to increase irrigation water and reduce soil salinity by taking into account the leaching coefficient; Freeze-thaw module: used to simulate the snow-water conversion process between surface snow and underground frozen soil, and obtain the change of soil moisture content by calculating the snow water equivalent depth and snow depth parameters; Fertilizer module: used to simulate the movement and transformation of fertilizers in the soil, as well as its effect on crop growth based on fertilizer type, application amount and application time.

4. The regional cotton water and fertilizer dynamic decision-making method considering salt stress and freeze-thaw mechanism according to claim 3 is characterized in that: The salt module calculates the irrigation quota and controls the soil salt content according to the field water holding capacity, the measured soil moisture content, the irrigation wetting ratio, the planned wetting layer depth and the salt leaching coefficient parameters; The calculation method of the irrigation quota is: In the formula, θ f is the field water holding capacity; θ(z) is the measured soil moisture content; β is the irrigation wetting ratio; D W is the planned wet layer depth; R is the salt leaching coefficient; M is the irrigation quota; The freeze-thaw module converts the snow water equivalent depth into the actual snow depth by dividing the snow density, specifically: In the formula, H w is the snow water equivalent depth; R ai is the daily accumulation rate of snow; R si is the daily sublimation rate of snow; R mi is the daily melting rate of snow; H s is the depth of snow; V s is the density of snow.

5. The regional cotton water and fertilizer dynamic decision-making method considering salt stress and freeze-thaw mechanism according to claim 1 is characterized in that: The sensitivity analysis of the crop growth prediction model is performed to obtain the objective function including: The sensitivity analysis is conducted by combining the crop growth model with the Sobol index method in the SALib library. By analyzing the first-order Sobol index and the second-order Sobol index, the parameters sensitive to the crop growth index are screened out, the sensitive parameters are calibrated, and the non-sensitive parameters use the default values; wherein the first-order Sobol index is used to measure the contribution of a single input variable to the variance of the output variable; the second-order index is used to measure the contribution of the input variable and its combination to the variance of the output variable; The NSGA-Ⅱ algorithm was used to solve the problem, and the objective function was set to minimize the error between the simulated and observed values ​​of leaf area index, aboveground biomass and final yield at different growth stages.

6. The regional cotton water and fertilizer dynamic decision-making method considering salt stress and freeze-thaw mechanism according to claim 5 is characterized in that: The objective function is: In the formula, and are the observed and simulated values ​​of leaf area index, TAGP i obs and TAGP i sim are the observed and simulated values ​​of aboveground biomass, and represent the observed and simulated values ​​of yield respectively, n is the sampling times of leaf area index and aboveground biomass in the experiment, and m is the sampling times of yield.

7. The regional cotton water and fertilizer dynamic decision-making method considering salt stress and freeze-thaw mechanism according to claim 6 is characterized in that: The objective function is optimized as follows: maxTWSO=maxTWSO sim ; minI=minI sim ; myN=myN sim ; In the formula, TWSO is the output, TWSO sim is the simulated yield, I is the total irrigation amount, I sim is the total amount of simulated irrigation, N is the total amount of fertilizer, and N sim To simulate the total amount of fertilizer applied; The constraints are set as: I<R; In the formula, R is the available water, P is the rainfall, and ET is the crop water consumption; In the formula, is the average output over the years; The decision variable is not negative: I sim ≥0； TWSO sim ≥0; Output the Pareto optimal solution set, that is, the combination of optimal irrigation amount, fertilizer amount and yield.

8. The regional cotton water and fertilizer dynamic decision-making method considering salt stress and freeze-thaw mechanism according to claim 1 is characterized in that: The farmland information system based on the Internet of Things technology stores real-time monitoring data including: Leaf area index and aboveground biomass crop data were obtained through remote sensing technology inversion; The soil moisture content, soil permeability and soil temperature are transmitted in real time through the Zhientropy instrument laid in the field; Field management data manually input by a field manager, wherein the field management data includes crop variety, planting date, harvest date, fertilization status, soil moisture threshold for initiating irrigation, and soil nutrient threshold for fertilization; Real-time monitoring of rainfall, maximum temperature, minimum temperature, wind speed, solar radiation, vapor pressure and snow depth data is carried out through small weather stations in the fields.

9. The regional cotton water and fertilizer dynamic decision-making method considering salt stress and freeze-thaw mechanism according to claim 1 is characterized in that: The water-fertilizer system dynamic adaptation system comprises: Data reading module: used to read the real-time data stored in the farmland information system based on the Internet of Things technology, including soil salt content and relevant parameters of the freeze-thaw process; Data analysis and storage module: used to analyze and store the read data, and evaluate the current soil salinity and freeze-thaw status by combining the simulation results of the salt module and freeze-thaw module; Reference water-fertilizer system selection module: used to select the optimized water-fertilizer system as the reference water-fertilizer system for dynamic correction according to demand; Crop growth status prediction module: used to use the reference water-fertilizer system and the real-time data-driven model to obtain the crop growth status, and compare it with the observed value to decide whether to modify the water-fertilizer system; Dynamic adaptation module: used to re-determine the optimal water and fertilizer system through the crop model coupled with the particle swarm algorithm; User interface and operation module: used to provide a visual interface for operation; System management and maintenance module: used to manage, repair and improve the system.

10. The regional cotton water and fertilizer dynamic decision-making method considering salt stress and freeze-thaw mechanism according to claim 9 is characterized in that: When the crop growth status prediction module finds that the simulated aboveground biomass is lower than the observed value or the water demand decision error does not meet the constraints, the dynamic adaptation module uses the particle swarm algorithm to re-determine the optimal water and fertilizer system by coupling the crop growth model; among them, the salt module reduces the soil salt content by adjusting the irrigation quota, and the freeze-thaw module optimizes soil moisture management to ensure the moisture supply and root growth of cotton during the freeze-thaw process.

Images