An agricultural water resource optimization method based on crop water requirement and multi-objective optimization.

Abstract

This invention discloses an agricultural water resource optimization method based on crop water requirement and multi-objective optimization. The method includes: S1 Constructing a multi-objective programming model that maximizes crop yield, irrigation water productivity, and minimizes planting water costs; using the Zimmerman method to transform the multi-objective programming model into a deterministic single-objective model; S2 Reading monitoring data collected in the current monitoring period and monitoring data from the previous monitoring period corresponding to periods not collected in the current monitoring period; the monitoring period includes a complete planting cycle for corn and wheat; S3 Inputting all monitoring data into the deterministic single-objective model and solving the single-objective model using Lingo programming to obtain the irrigation amount for corn and wheat at each growth stage within the current monitoring period; S4 Using the irrigation amount obtained in step S3 for the corresponding growth stage of corn and wheat as the guiding irrigation amount for the current growth stage of corn and wheat, based on their current growth stage.

Description

Technical Field

[0001] This invention relates to irrigation water resource optimization technology, specifically to an agricultural water resource optimization method based on crop water requirement and multi-objective optimization. Background Technology

[0002] The North China Plain, as a core agricultural region, is a crucial area for ensuring national food security. However, this high-yield model has long relied on groundwater over-extraction, making groundwater over-extraction and the optimal allocation of farmland water resources a research hotspot in recent years. Although the downward trend of groundwater levels has been somewhat alleviated under the dual impact of climate change and human activities, water scarcity still constrains regional food security and ecological sustainable development. The complementary nature of crop water requirements indicates that winter wheat relies on groundwater to regulate water supply during key growth stages, while summer maize requires supplemental irrigation to compensate for uneven rainfall distribution.

[0003] Meanwhile, in the process of precision agricultural water resource management that is deeply coupled with the reproductive cycle, a complex decision-making environment with multiple uncertainties arises. On the one hand, agricultural water resource management involves multiple conflicting objectives, such as maximizing yield, improving irrigation efficiency, and controlling costs. On the other hand, in the farmland ecosystem, the dynamic circulation and transformation of atmospheric water (including rainfall and farmland evapotranspiration), surface water, soil water, and groundwater, along with vertical and horizontal transport between different water bodies, jointly drive the dynamic allocation of water in the farmland system. Furthermore, the water demand of farmland exhibits significant stage-specific dynamic fluctuations with climate and crop growth.

[0004] Currently, there is considerable research on agricultural water resource management in the winter wheat-summer maize rotation model in the North China Plain. However, previous studies on agricultural water resource management have primarily focused on the static trade-off between economic and ecological benefits, with insufficient attention paid to the comprehensive utilization of precipitation and irrigation, as well as the dynamic water requirements during the crop growth period. This deficiency may lead to decreased water resource utilization efficiency and impact food security. Traditional methods have limitations in balancing objective weights and are ill-equipped to address the complex uncertainties arising from climate change, spatial and temporal mismatches in water demand, and groundwater over-extraction in the North China Plain. Summary of the Invention

[0005] To address the aforementioned shortcomings in existing technologies, the agricultural water resource optimization method based on crop water requirement and multi-objective optimization provided by this invention solves the problem that existing technologies do not pay attention to the comprehensive utilization of precipitation and irrigation and the dynamic water requirement of crops during their growth period, which may lead to a decline in water resource utilization efficiency.

[0006] To achieve the above-mentioned objectives, the technical solution adopted by this invention is as follows:

[0007] A method for optimizing agricultural water resources based on crop water requirement and multi-objective optimization is provided, comprising the following steps:

[0008] S1. Construct a multi-objective programming model that maximizes crop yield and irrigation water productivity and minimizes planting water costs, and use the Zimmerman method to transform the multi-objective programming model into a deterministic single-objective model.

[0009] S2. Read the monitoring data collected in the current monitoring cycle and the monitoring data from the previous monitoring cycle corresponding to the time period not collected in the current monitoring cycle; the monitoring cycle includes a complete planting cycle for corn and wheat;

[0010] S3. Input all monitoring data into a deterministic single-objective model and use Lingo programming to solve the single-objective model to obtain the irrigation amount of corn and wheat in each growth stage during the current monitoring period.

[0011] S4. Based on the current growth stage of corn and wheat, use the irrigation amount corresponding to the growth stage obtained in step S3 as the guiding irrigation amount for the current growth stage of corn and wheat.

[0012] The beneficial effects of this invention are as follows: This scheme takes winter wheat and summer maize as research objects. It uses monitoring data collected within the monitoring period and monitoring data from the previous monitoring period corresponding to the non-occurring period to form a complete monitoring period data. By combining Lingo programming to solve the single-objective model, the irrigation amounts for each crop at each growth stage can be obtained under the conditions of maximizing crop yield, irrigation water productivity, and minimizing planting water costs within the monitoring period. The irrigation amount at the current growth stage of the crop is used as the guiding irrigation amount. This scheme determines the irrigation amount for each growth stage in this way, deeply integrating the complementary water requirements of winter wheat and summer maize, which can reduce groundwater over-extraction and improve water resource utilization efficiency.

[0013] This scheme accurately determines the water resource allocation process under the winter wheat-summer maize rotation pattern, taking into account the water requirements at different growth stages. This enables the multi-objective programming model of this scheme to provide a water resource allocation plan that conforms to the actual situation and provides a reliable basis for the optimal management of farmland water resources. Unlike traditional static irrigation allocation methods, this study quantifies the dynamic process of the water cycle and key parameters, such as crop water requirements, soil moisture changes, and evaporation, through the model, and outputs irrigation amounts to help decision-makers understand and manage water resource allocation more scientifically. Specifically, it accurately determines the water requirements at each growth stage.

[0014] Furthermore, the multi-objective programming model includes objectives for maximizing crop yield, maximizing irrigation water productivity, and minimizing planting water costs;

[0015] The expression for the crop yield maximization objective is:

[0016]

[0017] in, Total crop yield, in kg; The yield per unit area of ​​crop c, in kg / hm² 2 ; Let hm be the planting area of ​​crop c. 2 C represents the total number of crops, with a value of 2. The potential yield per unit area of ​​crop c, kg / hm² 2 ; Let be the crop coefficient of crop c during its growth period t; The potential evapotranspiration of crop c during its growth period t, in mm; The water sensitivity index of crop c during its growth period t; The soil moisture content of crop c during its growth period t; Field water holding capacity; The wilting coefficient; This represents the reference crop evapotranspiration rate of crop c during its growth period t.

[0018] The expression for the irrigation water productivity maximization objective is:

[0019]

[0020] in, For irrigation water productivity, kg / m 3 ; Let C be the total yield of crop c, in kg; Let be the irrigation amount (in mm) for crop c during its growth period t.

[0021] The expression for minimizing the cost of water used in planting is:

[0022]

[0023] in, Total cost, CNY; The planting cost of crop c, CNY / hm 2 ; Water cost for crop c, CNY / m 3 .

[0024] The beneficial effects of the above technical solution are: this multi-objective programming model maximizes crop yield ( ), Maximizing the productivity of irrigation water ( ) and minimizing the cost of water used for planting ( The collaborative optimization mechanism significantly improves the resource utilization efficiency and economic sustainability of agricultural systems. The goal is to quantify the relationship between crop water quantity and yield by incorporating the crop water production function (Jensen model); The goal is to drive a preference for high-value crops through "unit irrigation water output" (increased yield in the numerator / water-saving in the denominator); The goal is to utilize water price parameters ( It automatically reduces the amount of irrigation needed for water-intensive crops.

[0025] Furthermore, the constraints of the multi-objective programming model include water balance constraints, water demand constraints, soil moisture content constraints, and non-negativity constraints; the expression for the water balance constraint is:

[0026]

[0027] in, The soil moisture content of crop c during its growth period t; and , respectively, represent the soil moisture content (in mm) of crop c at growth stage t and growth stage t+1; The effective rainfall for crop c during its growth period t, in mm; Let be the amount of exchange between the root zone and buffer zone of crop c during growth period t, in mm; The root zone depth of crop c at growth stage t is 1m; Field holding capacity; denoted as the critical water storage capacity of the root zone, in mm; a and b are dimensionless empirical parameters related to soil texture.

[0028] The expression for the required water volume constraint is:

[0029]

[0030] in, This refers to the irrigation water utilization coefficient. Let be the water requirement of crop c during its growth period t, in mm; The target water allocation for crop c during its growth period t, in mm; , The confidence level; It serves as a credibility measure, used to quantify the probability of an ambiguous event occurring.

[0031] The expression for soil moisture content constraint is:

[0032] ;

[0033] The expression for the nonnegativity constraint is:

[0034] .

[0035] The beneficial effects of the above technical solution are as follows: through a dynamic synergistic mechanism of water balance constraints, water demand constraints, soil moisture content constraints, and non-negative constraints, the precision and adaptability of agricultural water cycle management are significantly improved. Its core innovation lies in upgrading traditional static constraints into a dynamic closed-loop water cycle system with multi-variable linkage: the water balance constraint couples soil moisture content, root zone depth, and groundwater exchange to form a feedback loop of "precipitation-irrigation-evapotranspiration-groundwater recharge," accurately quantifying the impact of soil texture and deep infiltration; the water demand constraint introduces fuzzy confidence (Cr) to regulate precipitation uncertainty, dynamically responding to drought risk through a dual-threshold elastic range (minimum water demand ensures survival, target water allocation inhibits waste); the soil moisture content constraint, through dual-boundary control of the wilting coefficient and field capacity, collaboratively ensures the physiological water needs of crops. The non-negative constraint eliminates non-physical interpretations, ensuring algorithm stability. This constraint system breaks through the fragmented limitations of traditional models, achieving global synergistic optimization of water-food-ecological risks.

[0036] Furthermore, the objective function of the deterministic single-objective model is:

[0037]

[0038] in, For satisfaction;

[0039] The constraints of the objective function are:

[0040]

[0041]

[0042] in, and They are respectively The lower and upper limits; These are auxiliary parameters used to describe the nonlinearity of the membership function, and are specified by the decision-maker. ; and They are respectively The lower and upper limits; and They are respectively The lower and upper limits; and for The quartile values ​​and median; and They are respectively The quartile values ​​and the median; For conversion symbols;

[0043] In deterministic single-objective models Using Lingo programming as the variable, the single-objective model is solved and the output is... .

[0044] The beneficial effects of the above technical solution are: this solution improves satisfaction ( The innovative design that maximizes the objective achieves a systematic breakthrough in multi-objective collaborative optimization. The multi-objective membership integration mechanism utilizes nonlinear parameters. Dynamically adjust the membership functions of output, water productivity, and cost targets (e.g., using a structured approach for output targets). This allows for a gain bias towards high-efficiency objectives; fuzzy constraint deterministic transformation: converting the credibility measure (Cr) constraint into a constraint based on quartiles ( The deterministic inequality of precipitation is used to quantify the randomness of precipitation in order to ensure water needs for survival, while avoiding conservative irrigation.

[0045] Furthermore, the monitoring data includes meteorological data, soil parameters, crop growth parameters, and economic parameters. The meteorological data includes daily average temperature, minimum and maximum temperature, relative humidity, average wind speed, sunshine duration, and precipitation. The soil parameters include soil bulk density, soil moisture content, saturated moisture content, wilting coefficient, and field capacity. The crop growth parameters include growth period duration, yield, plant height, biomass, and leaf area. The economic parameters include seed cost, fertilizer cost, labor cost, pesticide cost, agricultural machinery cost, water cost, and crop unit price.

[0046] Furthermore, the confidence level , The value of is 0.5. Attached Figure Description

[0047] Figure 1 This is a flowchart of an agricultural water resource optimization method based on crop water requirement and multi-objective optimization.

[0048] Figure 2 This diagram illustrates the changes in temperature, effective rainfall, and crop water requirements during the growing season under a winter wheat-summer maize rotation model.

[0049] Figure 3 The diagram shows the trade-off results of the model objective function under six different confidence levels, where (a) is the trade-off result in 2023 and (b) is the trade-off result in 2024.

[0050] Figure 4 This diagram illustrates the range of different target values.

[0051] Figure 5 This figure shows the optimization results of various water elements at different growth stages under different scenarios in 2023 under the winter wheat-summer maize rotation model.

[0052] Figure 6 This figure shows the optimization results of various water elements at different growth stages under different scenarios in 2024 under the winter wheat-summer maize rotation model. Detailed Implementation

[0053] The specific embodiments of the present invention are described below to enable those skilled in the art to understand the present invention. However, it should be understood that the present invention is not limited to the scope of the specific embodiments. For those skilled in the art, various changes are obvious as long as they are within the spirit and scope of the present invention as defined and determined by the appended claims. All inventions utilizing the concept of the present invention are protected.

[0054] refer to Figure 1 , Figure 1 A flowchart illustrating an agricultural water resource optimization method based on crop water requirement and multi-objective optimization is shown; for example... Figure 1 As shown, the method includes steps S1 to S4.

[0055] In step S1, a multi-objective programming model is constructed to measure crop yield, irrigation water productivity, and planting water cost. The Zimmerman method is then used to transform the multi-objective programming model into a deterministic single-objective model.

[0056] In one embodiment of the present invention, the multi-objective programming model includes a crop yield maximization objective, an irrigation water productivity maximization objective, and a planting water cost minimization objective;

[0057] The expression for the crop yield maximization objective is:

[0058]

[0059] in, Total crop yield, in kg; The yield per unit area of ​​crop c, in kg / hm² 2 ; Let hm be the planting area of ​​crop c. 2 C represents the total number of crops, with a value of 2. The potential yield per unit area of ​​crop c, kg / hm² 2 ; Let be the crop coefficient of crop c during its growth period t; The potential evapotranspiration of crop c during its growth period t, in mm; The water sensitivity index of crop c during its growth period t; The soil moisture content of crop c during its growth period t; Field water holding capacity; The wilting coefficient; This represents the reference crop evapotranspiration rate of crop c during its growth period t.

[0060] The expression for the irrigation water productivity maximization objective is:

[0061]

[0062] in, For irrigation water productivity, kg / m 3 ; Let C be the total yield of crop c, in kg; Let be the irrigation amount (in mm) for crop c during its growth period t.

[0063] In this scheme, irrigation water productivity Water productivity is an indicator that measures the agricultural output generated per unit of water resources. A higher water productivity value indicates higher irrigation water utilization efficiency and more effective conversion into crop yield.

[0064] The expression for minimizing the cost of water used in planting is:

[0065]

[0066] in, Total cost, CNY; The planting cost of crop c, CNY / hm 2 ; Water cost for crop c, CNY / m 3 .

[0067] The constraints of the multi-objective programming model include water balance constraints, water demand constraints, soil moisture content constraints, and non-negativity constraints; the expression for the water balance constraint is:

[0068]

[0069] in, The soil moisture content of crop c during its growth period t; and , respectively, represent the soil moisture content (in mm) of crop c at growth stage t and growth stage t+1; The effective rainfall for crop c during its growth period t, in mm; Let be the amount of exchange between the root zone and buffer zone of crop c during growth period t, in mm; The root zone depth of crop c at growth stage t is 1m; Field holding capacity; denoted as the critical water storage capacity of the root zone (mm); a and b are dimensionless empirical parameters related to soil texture.

[0070] The expression for the required water volume constraint is:

[0071]

[0072] in, This refers to the irrigation water utilization coefficient. Let be the water requirement of crop c during its growth period t, in mm; The target water allocation for crop c during its growth period t, in mm; , The confidence level; It is a credibility measure used to quantify the probability of an ambiguous event occurring.

[0073] The expression for soil moisture content constraint is:

[0074] ;

[0075] The expression for the nonnegativity constraint is:

[0076] .

[0077] In step S2, the monitoring data collected in the current monitoring cycle and the monitoring data in the previous monitoring cycle corresponding to the period when no data was collected in the current monitoring cycle are read; the monitoring cycle includes a complete planting cycle of corn and wheat.

[0078] To facilitate understanding, a small example is provided below. In North China, the growing season for winter wheat is from October to June of the following year, and for summer maize, it is from June to October. Therefore, the monitoring period is from June of the current year to June of the following year. Assuming that six months have been monitored in the current period, and the monitoring data for these six months has been collected in real time, the monitoring data from June to November is the data read in the current monitoring period. Since this scheme requires monitoring data for the complete monitoring period to accurately predict irrigation amounts for each growing season, this scheme uses data from the previous monitoring period to supplement the data for the periods not yet occurring in the current monitoring period. That is, data from December to June of the following year from the previous monitoring period is added to the current monitoring period. Then, all of these data are input into a deterministic single-objective model and solved using Lingo programming.

[0079] In this scheme, the monitoring data includes meteorological data, soil parameters, crop growth parameters, and economic parameters. The meteorological data includes daily average temperature, minimum and maximum temperature, relative humidity, average wind speed (2m), sunshine duration, and precipitation. Soil parameters include soil bulk density, soil moisture content, saturated moisture content, wilting coefficient, and field capacity. Crop growth parameters include growth period duration, yield, plant height, biomass, and leaf area. Economic parameters include seed cost, fertilizer cost, labor cost, pesticide cost, agricultural machinery cost, water cost, and crop unit price.

[0080] In step S3, all monitoring data are input into a deterministic single-objective model, and the single-objective model is solved using Lingo programming to obtain the irrigation amount of corn and wheat in each growth stage during the current monitoring period.

[0081] In one embodiment of the present invention, the objective function of the deterministic single-objective model is:

[0082]

[0083] in, For satisfaction;

[0084] The constraints of the objective function are:

[0085]

[0086]

[0087] in, and They are respectively The lower and upper limits; These are auxiliary parameters used to describe the nonlinearity of the membership function, and are specified by the decision-maker. ; and They are respectively The lower and upper limits; and They are respectively The lower and upper limits; and for The quartile values ​​and median; and They are respectively The quartile values ​​and the median; For conversion symbols;

[0088] In deterministic single-objective models Using Lingo programming as the variable, the single-objective model is solved and the output is... .

[0089] In step S4, based on the current growth stage of corn and wheat, the irrigation amount corresponding to the growth stage obtained in step S3 is used as the guiding irrigation amount for the current growth stage of corn and wheat.

[0090] The multi-objective programming model constructed in this scheme solves the problem of low adaptability of traditional models to spatiotemporal heterogeneous parameters. By deeply integrating the water demand complementarity of winter wheat and summer maize, and combining the multi-objective programming model with the Zimmerman algorithm, it generates optimized irrigation strategies for sensitive stages such as the jointing and tasseling stages, improves irrigation efficiency and reduces groundwater over-extraction, and simultaneously optimizes grain yield, groundwater extraction-replenishment balance and ecological sustainability goals, providing scientific support for farmland water resource management.

[0091] The feasibility of the agricultural water resource optimization method proposed in this scheme will be described in detail below with specific examples:

[0092] 1. Determine the study area

[0093] North China (32°~40°N, 114°~121°E) boasts fertile soil and a suitable climate, making it an important agricultural production base in my country. The region belongs to the warm temperate semi-humid continental monsoon climate zone, with an average temperature of 12.1℃ and annual precipitation of 647 mm. Rainfall is concentrated from June to September, exhibiting significant seasonal variation, particularly in July and August, accounting for 60% of the annual precipitation. The actual annual evaporation is approximately 423 mm, with a potential evaporation of 1164.4 mm. Annual sunshine hours are 2580 hours, and the frost-free period is approximately 190 days. Winter wheat and summer maize are the main crops grown in this region, accounting for 51.4% and 35.1% of the total grain planting area, respectively. The winter wheat-summer maize rotation is a unique planting pattern in this region, with winter wheat growing from October to June of the following year, and summer maize growing from June to October. Winter wheat requires approximately 450-600 mm of water during its growing season, while the average annual rainfall is only 139 mm (concentrated during the winter wheat growing season), making it difficult to meet its water needs. Summer maize's growing season coincides with both rain and heat, with abundant rainfall, ranging from approximately 334-458 mm during the growing season, while its water requirement is only about 350-450 mm. This mismatch between water use for key crops and rainfall timing in North China constitutes a contradictory system of groundwater over-extraction and rainfall resource waste, resulting in a decline in the overall water resource utilization efficiency of the region.

[0094] 2. Sources of monitoring data

[0095] A. Field experiment data

[0096] The basic experimental data in this embodiment comes from the Daxing Water-Saving Irrigation Experimental Station of the National Water-Saving Irrigation Technology Research Center in Beijing (39°37′25″N, 116°25′51″E), a winter wheat-summer maize rotation irrigation experiment from 2022 to 2024. The main data types include meteorological data (including daily average temperature, minimum and maximum temperature, relative humidity, average wind speed (2 m), sunshine duration and precipitation, etc.), soil parameters (including soil bulk density, soil moisture content, saturated moisture content, wilting coefficient and field capacity, etc.), and crop growth parameters (including growth period duration, yield, plant height, biomass and leaf area, etc.).

[0097] The growth process of winter wheat and summer maize can be divided into the following stages: The growth stages of winter wheat include sowing-overwintering period (S-OW), overwintering-greening period (OW-GU), greening-jointing period (GU-J), jointing-tasseling period (JB), tasseling-grain filling period (BF), and grain filling-harvest period (F-H1); The growth stages of summer maize include sowing-jointing period (SJ), jointing-tasseling period (JT), tasseling-grain filling period (TF), and grain filling-harvest period (F-H2).

[0098] B. Meteorological data

[0099] Changes in temperature, effective rainfall, and crop water requirements during the growth period under the winter wheat-summer maize rotation model are as follows: Figure 1 As shown, the data represents climate data for 2022-2023 and 2023-2024, respectively. mean The upper and lower limits of the shaded area represent the average temperature during the winter wheat-summer maize growing season; the upper and lower limits of the shaded area represent the highest and lowest temperatures, respectively; P represents the effective rainfall; 2023 and 2024 represent the years 2022-2023 and 2023-2024, respectively, and so on.

[0100] 3. Performance Evaluation of Multi-Objective Programming Model

[0101] Based on the data obtained above, step S3 of this scheme was executed. During the execution process, six different confidence levels were selected, and the trade-off results are referenced. Figure 3 . Figure 3 Different schemes in the text represent different confidence levels. (From 0.5 to 1.0). The area of ​​the triangle represents the degree of coordinated development; the larger the area, the better the coordinated development under a specific situation. Figure 3 The vertices of F1-AY, F2-IWUE, and F3-TC correspond to the optimal values ​​of three different objectives. The closer the index value is to the vertex, the better the index is.

[0102] In each option, the value of the objective function is constrained between its corresponding maximum and minimum values. By setting upper and lower limits for each objective, the range of objective values ​​is determined, and normalization methods are applied for standardization. The closer the indicator value is to 1, the better it meets the decision-maker's needs. For example, At that time, the maximum output target was 1.708 × 10 4 The maximum irrigation water productivity is 5.13 kg / m³, while the minimum cost target is 1.183 × 10⁻⁶ kg / m³. 4 CNY. If these three targets are considered simultaneously, production could reach 1.655 × 10⁻⁶. 4 The irrigation water productivity is 4.87 kg / m³, and the cost is 1.193 × 10⁻⁶ kg / m³. 4 CNY.

[0103] The above demonstrates that there are certain trade-offs among the multiple objectives, and the results obtained through multi-objective collaborative optimization are reasonable. Numerically, the cost index shows relatively small changes across different schemes, remaining consistently close to 1, indicating its relative stability. In contrast, the yield and irrigation water productivity indices show significant variations. This suggests that in water resource optimization, cost is relatively easy to control, while yield and water efficiency require appropriate trade-offs in multi-objective optimization.

[0104] The model's satisfaction level decreased as the confidence level increased. Specifically, the satisfaction levels for 2023 were 0.72, 0.68, 0.65, 0.61, 0.57, and 0.54, respectively; and for 2024, they were 0.71, 0.67, 0.63, 0.60, 0.57, and 0.54, respectively. This indicates that models with lower confidence levels provide higher satisfaction levels compared to models with higher confidence levels, which is why this approach will be used in this study. Setting all confidence levels to 0.5 enables more efficient water resource optimization, i.e., improving water resource utilization. Furthermore, it is noteworthy that the results from the low-confidence model cover a wider area than those from the high-confidence model, suggesting that the low-confidence approach may offer more flexible solutions that contribute to the sustainable allocation of agricultural water and food resources.

[0105] 4. Objective Analysis of Multi-Objective Programming Models

[0106] The multi-objective programming model in this scheme includes three objectives: yield, irrigation water productivity, and cost, and incorporates uncertain constraint variables. Also known as the confidence level, (the variable affects the allocation of agricultural water and soil resources) multiple scenarios are generated, thus affecting the target value. Statistical analysis is performed on the changes in the target value under different scenarios, and the results are as follows: Figure 4 As shown. In Under different combinations of scenarios, the output in 2023 ranged from 1.43 to 1.71 × 10⁻⁶. 4 kg, with an optimized value of 1.61 × 10 4 kg, representing a 4.91% increase compared to actual scenarios; the irrigation water productivity ranged from 4.13 to 5.26 kg / m² over 23 years. 3 The optimized value is 4.68 kg / m³ 3 This represents a 5.84% improvement compared to the actual scenario; the cost ranges from 1.18 to 1.21 × 10⁻⁶. 4 CNY, optimized value is 1.195 × 10 4 The CNY is 5.94% lower than the actual scenario. Production in 2024 is projected to range from 1.50 to 1.75 × 10⁻⁶. 4 kg, with an optimized value of 1.61 × 10 4 kg, representing a 3.13% increase compared to actual scenarios; irrigation water productivity ranged from 4.29 to 5.36 kg / m². 3 The optimized value is 4.79 kg / m³ 3 This represents a 5.74% improvement compared to the actual scenario; the cost ranges from 1.175 to 1.198 × 10⁻⁶. 4 CNY, optimized value is 1.185×10 4 The CNY has decreased by 8.40% compared to the actual situation.

[0107] The data above shows that the multi-objective programming model constructed in this scheme is used to predict irrigation volume, and then crop irrigation is guided based on this data. This can significantly improve crop yield and irrigation water productivity, and reduce costs by a large proportion.

[0108] 5. Water resource allocation during the crop growing season

[0109] Irrigation and rainfall are the main sources of water replenishment for crops in the field. Figure 5 and Figure 6 The optimization results of various water elements at different growth stages under different scenarios in 2023 and 2024 under the winter wheat-summer maize rotation model are presented. The optimization is based on the number of days in the growth period, crop water requirement, precipitation, water shortage sensitivity, and water allocation target. Different scenarios in the figure represent different confidence levels (…). (From 0.5 to 1.0). Analysis shows that for winter wheat, the total water allocation is largest during the jointing-heading stage and the heading-grain-filling stage, followed by the greening-jointing stage, while the water allocation is roughly equal during the sowing-overwintering, overwintering-greening, and grain-filling-harvest stages. For summer maize, the total water allocation is largest during the jointing-heading stage, followed by the tasseling-grain-filling stage and the sowing-jointing stage, while the water allocation is smallest during the grain-filling-harvest stage.

[0110] When the confidence level ranges from 0.5 to 1, the proportion of crop water allocation at different growth stages does not change significantly, and the total water allocation increases with the increase of confidence level. The proportion of water allocation for winter wheat at different growth stages in 2023 is [3.48%±0.17%], [6.57%±0.16%], [6.64%±0.31%], [8.83%±0.40%], [6.51%±0.32%], [3.92%±0.13%]; the proportion of water allocation for summer maize at different growth stages in 2023 is [10.75%±0.18%], [37.80%±0.14%], [12.79%±0.46%], [2.71%±0.14%]. The proportion of water allocation for different growth stages of winter wheat in 2024 was [2.78%±0.17%], [4.68%±0.20%], [5.53%±0.32%], [7.79%±0.38%], [6.55%±0.37%], and [6.26%±0.16%]; the proportion of water allocation for different growth stages of summer maize in 2024 was [17.37%±0.45%], [39.65%±0.10%], [5.65%±0.31%], and [3.75%±0.09%].

[0111] pass Figure 5 and Figure 6 Analysis of water distribution at different confidence levels shows that... The water allocation efficiency is highest when the value is 0.5. Compared with the actual treatment, the total average irrigation water consumption for winter wheat was reduced by 14.76%, and the total average irrigation water consumption for summer maize was reduced by 16.62%. It can be seen that the agricultural water resource optimization method adopted in this scheme can significantly reduce irrigation water consumption, thereby achieving the goal of water conservation.

Claims

1. An agricultural water resource optimization method based on crop water requirement and multi-objective optimization, characterized in that, The method comprises the steps of: S1, constructing a multi-objective programming model for maximizing crop yield, irrigation water productivity and minimizing planting water cost, and converting the multi-objective programming model into a deterministic single-objective model by using the Zimmerman method; S2, reading monitoring data collected in a current monitoring period and monitoring data corresponding to a period not collected in the current monitoring period in a previous monitoring period; the monitoring period comprises a complete planting period of corn and wheat; S3, inputting all the monitoring data into the deterministic single-objective model, and solving the single-objective model by using Lingo programming to obtain irrigation amounts of corn and wheat in each growth period in the current monitoring period; S4, according to the growth period of corn and wheat, using the irrigation amount of the corresponding growth period obtained in step S3 as the guiding irrigation amount of the current growth period of corn and wheat; The multi-objective programming model comprises a crop yield maximization objective, an irrigation water productivity maximization objective and a planting water cost minimization objective; The expression of the crop yield maximization objective is: wherein, is the total yield of the crop, kg; is the unit area yield of the crop c, kg / hm 2 ; is the planting area of the crop c, hm 2 ; C is the total number of crops, taking a value of 2; is the potential unit area yield of the crop c, kg / hm 2 ; is the crop coefficient of the crop c at growth period t; is the potential evapotranspiration of the crop c at growth period t, mm; is the water sensitivity index of the crop c at growth period t; is the soil moisture content of the crop c at growth period t; is the field water holding rate; is the wilting coefficient; is the reference crop evapotranspiration of the crop c at growth period t; The expression of the irrigation water productivity maximization objective is: wherein, is the irrigation water productivity, kg / m 3 ; is the total yield of crop c, kg; is the irrigation amount of crop c at growth stage t, mm; The expression of the planting water cost minimization objective is: wherein, is the total cost, CNY; is the planting cost of crop c, CNY / hm 2 ; is the water cost of crop c, CNY / m 3 .

2. The crop water requirement and multi-objective optimization based agricultural water resources optimization method as claimed in claim 1 wherein, Constraint conditions of the multi-objective programming model comprise water balance constraint, required water allocation constraint, soil water content constraint and non-negative constraint; the expression of the water balance constraint is: wherein, is the soil water content for crop c at growth stage t; and are the soil water content for crop c at growth stage t and growth stage t+1, respectively, mm; is the effective rainfall for crop c at growth stage t, mm; is the exchange amount for crop c between the root zone and the buffer zone at growth stage t, mm; is the root zone depth for crop c at growth stage t, 1 m; is the field capacity; is the critical water storage for the root zone, mm; a and b are dimensionless empirical parameters related to soil texture. The expression of the required water allocation constraint is: wherein, is the irrigation water use efficiency; is the water requirement of crop c at growth stage t, mm; is the target water allocation of crop c at growth stage t, mm; , is the confidence level; is the believability measure, which quantifies the likelihood of the fuzzy event to occur; The expression of the soil water content constraint is: ； The expression of the non-negative constraint is: 。 3. The crop water requirement and multi-objective optimization based agricultural water resources optimization method as claimed in claim 2, wherein, The objective function of the deterministic single-objective model is: wherein, satisfaction; Constraint conditions of the objective function are: wherein, and are the lower and upper limits of ; is a helper parameter to describe the nonlinearity of the membership function, specified by the decision maker, ; and are the lower and upper limits of ; and are the lower and upper limits of ; and are the lower and upper limits of ; and are the lower and upper limits of ; is a conversion symbol; In the deterministic single objective model For the variable, the Lingo programming is used to solve the single objective model and output .

4. The crop water requirement and multi-objective optimization based agricultural water resources optimization method, according to any one of claims 1 to 3, characterized in that, The monitoring data comprise meteorological data, soil parameters, crop growth parameters and economic parameters; the meteorological data comprise daily average air temperature, minimum air temperature and maximum air temperature, relative humidity, average wind speed, sunshine duration and precipitation; the soil parameters comprise soil bulk density, soil water content, saturated water content, wilting coefficient and field water holding capacity; the crop growth parameters comprise growth period duration, yield, plant height, biomass and leaf area; and the economic parameters comprise seed cost, fertilizer cost, labor cost, pesticide cost, agricultural machinery cost, water cost and crop unit price.

5. The crop evapotranspiration and multi-objective optimization based agricultural water resources optimization method according to claim 3, wherein, the confidence level , are each 0.5.

Images