Construction method of soil water and salt simulation model in crop root system based on complex network
By constructing a crop root soil water and salt simulation model based on a complex network approach, the problem of obtaining parameters in existing technologies is solved, and accurate water and salt transport simulation and agricultural management support are achieved.
Patent Information
- Application Number
- CN202511777428.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-28
- Publication Date
- 2026-01-09
AI Technical Summary
Existing soil-scale water and salt transport models, when considering root systems, are either difficult to obtain parameters and have a high computational load, or are too simple or too complex, making it difficult to achieve accurate simulation.
A complex network-based approach is adopted to simulate crop root systems by generating regular networks and random edges. Edge weights are set, and a water-salt dynamics module is designed to simulate soil moisture and salt transport. The edge probabilities are adjusted using a probabilistic method to adapt to different growth stages, thus simplifying parameter requirements.
It improves the accuracy of root zone water and salt simulation, adapts to different crop growth stages, reduces computational load, and provides reliable support for agricultural water and salt management.
Smart Images
Figure CN121302987A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of agricultural engineering, and in particular to a method for constructing a soil water and salt simulation model in crop roots based on a complex network. Background Technology
[0002] Existing soil-scale water and salt transport models either oversimplify the root system by using only simple, static root parameters to represent root effects, or overcomplicate it by requiring parameters such as root length, root diameter, and root hair angle, which are not only difficult to obtain but also greatly increase the computational load.
[0003] The rise of complex network technology has provided a solution to root system problems. Generating random edges reflects the randomness of root growth; providing different levels of probability reflects the regularity of root growth; and setting edge weights reflects the hierarchy of the root system. Therefore, a method for constructing a soil water and salt simulation model for crop roots based on complex networks is needed. Summary of the Invention
[0004] The purpose of this invention is to propose a method for constructing a soil water and salt simulation model in crop roots based on complex networks in order to solve the problems of existing technologies.
[0005] To achieve the above objectives, the present invention adopts the following technical solution: This invention includes the following steps: S1: Generate a regular network to construct a network with an adjustable number of nodes. Each node corresponds to a soil body of a set size. Vertical connections are set only between nodes to simulate vertical water transport in the soil. S2: Create random edges based on crop growth characteristics. Use a probabilistic method to add random edges between nodes in the rule network to simulate crop root system. When adding edges, generate edge weights representing root system level at the same time. S3: Design a water-salt dynamics simulation module. Input water is preferentially infiltrated through the matrix flow path controlled by nodes. Water exceeding the infiltration capacity of nodes is transported through the priority flow path controlled by random connections. The priority flow follows the first-in-first-out principle. The total evaporation of soil at surface nodes and transpiration of plants at root nodes is calculated as evapotranspiration. Based on the water difference between evapotranspiration nodes and child nodes, water rebalancing is performed. Soil salinity is set to include solution and solid states. Solution salt migrates with water movement, and solid salt and solution salt are mutually transformed according to a preset relationship as water changes. S4: The model input data includes input moisture, potential evapotranspiration, Kc coefficient reflecting crop growth period, and Kf coefficient reflecting leaf area shading. It initializes the node volume and infiltration rate parameters, edge weights and transmission delay parameters, and outputs soil moisture stratification results and salinity stratification results.
[0006] Furthermore, in step S1, the number of nodes in the rule network is 5×5 to 20×20, and the soil size corresponding to the node is adjusted according to the soil sampling accuracy.
[0007] Furthermore, in step S2, the probability distribution of the probability method is adjusted according to the crop growth stage: a higher probability of connection is set between nodes in the soil surface during the seedling stage, and a higher probability of connection is set between nodes in the deep soil layer during the maturity stage.
[0008] In a further step S2, the edge weight is a dimensionless value. The smaller the value, the higher the root system level and the thicker the root system. The larger the value, the lower the root system level and the thinner the root system.
[0009] Further, in step S31, the infiltration rate I of the matrix flow satisfies the formula: I = Imax × exp(-k × θ), Where Imax is the maximum infiltration rate, θ is the soil moisture content, and k is the reduction coefficient; the matrix flow rate Q satisfies Q=min(input,I), where input is the input moisture.
[0010] In further step S33, the conversion relationship between the solid salt and the solution salt satisfies: Safter=Sbefore-Sbefore×(1-θ), Where Safter represents the solid salt content after transformation, Sbefore represents the solid salt content before transformation, and θ represents the soil moisture content.
[0011] Furthermore, the formula for calculating the potential evapotranspiration data is as follows: Where ET0 is the daytime evapotranspiration, Δ is the slope of the vapor pressure curve, Rn is the net radiation on the crop surface, G is the soil heat flux density (G=0 if the daily scale is used as a reference), γ is the hygrometer constant, T is the daily average temperature, u2 is the wind speed at a height of 2 meters, es is the average saturated water vapor pressure at temperature T, ea is the actual water vapor pressure difference at the dew point temperature, and es-ea is the water vapor pressure difference value.
[0012] Furthermore, the node volume is 10mm~50mm, the maximum infiltration rate is 5mm / timestep~20mm / timestep, and the edge transmission delay parameter is 1~5timestep.
[0013] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention simulates root systems by combining a rule network with probabilistic random edge connections. It eliminates the need for difficult-to-measure parameters such as root length and diameter, flexibly adapts to different crop growth stages, significantly improves the accuracy of root zone water and salt simulation, provides reliable support for agricultural water and salt management, and has a wide range of applications and practicality. Attached Figure Description
[0014] Figure 1 This is a flowchart illustrating the method for constructing a soil water and salt simulation model in crop roots based on a complex network, as proposed in this invention. Figure 2 This is a schematic diagram of the root network structure of a method for constructing a soil water and salt simulation model in crop roots based on a complex network, as proposed in this invention. Figure 3 This invention presents a method for constructing a soil water and salt simulation model of crop roots based on complex networks. The diagram shows the root network at different growth stages. Figure 4 The results of running the algorithm of this invention in the Matlab 2018a interface; Detailed Implementation
[0015] Reference Figure 1 The present invention includes the following steps: S1: Generate a regular network to construct a network with an adjustable number of nodes. Each node corresponds to a soil body of a set size. Vertical connections are set only between nodes to simulate vertical water transport in the soil. S2: Create random edges based on crop growth characteristics. Use a probabilistic method to add random edges between nodes in the rule network to simulate crop root system. When adding edges, generate edge weights representing root system level at the same time. S3: Design a water-salt dynamics simulation module. Input water is preferentially infiltrated through the matrix flow path controlled by nodes. Water exceeding the infiltration capacity of nodes is transferred through the priority flow path controlled by random connections. The priority flow follows the first-in-first-out principle. The total evaporation of soil at surface nodes and the transpiration of plants at root nodes are calculated as evapotranspiration. Based on the water difference between evapotranspiration nodes and child nodes, water rebalancing is performed. Soil salinity is set to include solution and solid states. Solution salt migrates with water movement, and solid salt and solution salt are mutually transformed according to a preset relationship as water changes. S4: The model input data includes input moisture, potential evapotranspiration, Kc coefficient reflecting crop growth period, and Kf coefficient reflecting leaf area shading. It initializes the node volume and infiltration rate parameters, edge weights and transmission delay parameters, and outputs soil moisture stratification results and salinity stratification results.
[0016] Step 1: Generate a rule network like Figure 2As shown, a 10×10 regular network (the number of nodes can be varied as needed) is generated, where each node represents a soil mass of a certain area and depth. Currently, only the vertical water transport process is considered, therefore the nodes are only connected vertically.
[0017] Step 2: Create random edges like Figure 3 As shown, random edges are added between nodes to represent crop roots. The model considers crop type and root characteristics at different growth stages, using a probabilistic approach to add random edges. For example, a higher probability is set for edges in the top 10-20 cm soil layer during the seedling stage, and a higher probability is set for edges in the deeper soil layer (30-40 cm) during the maturity stage. Simultaneously, edge weights are randomly generated, from 1 to n, representing root levels, with level 1 being the thickest root and level n being the finest root.
[0018] Step 3: Dynamic Process The water-salt dynamics process is divided into three modules: water infiltration, evapotranspiration, and salt transport.
[0019] 3.1 Moisture Infiltration Module Water enters the soil via two pathways: the matrix flow pathway and the preferred flow pathway. Matrix flow is controlled by nodes, which are generalized as buckets where water accumulates. When a node reaches its maximum volume Cmax, downward matrix flow occurs. Matrix flow is controlled by the input intensity and infiltration rate.
[0020] Q = min(input, I) Where Q is the matrix flow of Q infiltration, input is the input water, and I is the infiltration rate (unit: mm / time step).
[0021] Infiltration rate is a function of soil moisture content.
[0022] I = Imax×exp(-k ×θ) Where Imax is the maximum infiltration rate (unit: mm / time step), θ is the soil moisture content (unit: m3 / m3), and k is the reduction coefficient (dimensionless).
[0023] Priority flow is controlled by random edges. Random edges are generalized as a queue, following the FIFO (First In, First Out) principle. Input moisture enters the queue from the parent node, accumulates in the queue, and when the queue length is reached, the moisture stored at the front of the queue is released to the child node.
[0024] Intake water is first allocated to the matrix flow, and water with an infiltration rate greater than that at the nodes is allocated to the preferential flow.
[0025] 3.2 Evaporation Module STEP 1: Water Loss. Evapotranspiration is divided into soil evaporation (E) and plant transpiration (T). Soil evaporation occurs only at surface nodes, while plant transpiration occurs at root nodes (nodes reached by random edges). The evapotranspiration rate is ET = E + T.
[0026] Where E or T = min(C, E or T), and C is the minimum water storage of the node.
[0027] STEP 2: Moisture Rebalancing. Compare the storage moisture difference between the evapotranspiration node and its child nodes. If the moisture of the evapotranspiration node is less than that of its child nodes, then allocate half of the moisture difference to the evapotranspiration node.
[0028] 3.3 Salt Transport Module Soil salts exist in both solution and solid states. Solution salts shift upwards or downwards proportionally with the influx and evaporation of water. When water enters a node, a corresponding proportion of solid salts are converted into solution salts. Safter=Sbefore - Sbefore×(1-θ) Where Safter and Sbefore are the solid salts before and after dissolution, respectively, and θ is the soil moisture content (unit: m3 / m3).
[0029] When evaporation causes a decrease in water content, a corresponding proportion of the solution-state salt is converted into solid-state salt.
[0030] Step 4: Parameter setting, input data and initialization The model input data includes input water (rainfall + irrigation) (unit: mm / time step), potential evapotranspiration (unit: mm / time step), Kc coefficient (unit: dimensionless) reflecting crop growth period, and Kf coefficient (unit: dimensionless) reflecting leaf area shading.
[0031] Potential evapotranspiration is calculated using a formula.
[0032] Where ET0 is the daytime evapotranspiration (mm / day), Δ is the slope of the vapor pressure curve (kPa / ℃), Rn is the net radiation on the crop surface (MJ / m2 / day), G is the soil heat flux density (MJ / m2 / day), if the daily scale is taken as the reference, G=0, γ is the hygrometer constant (kPa / ℃), T is the daily average temperature (℃), u2 is the wind speed at a height of 2 meters (m / s), es is the average saturated water vapor pressure at temperature T (kPa), ea is the actual water vapor pressure difference at the dew point temperature (kPa), and es-ea is the water vapor pressure difference value (VPD, kPa).
[0033] The parameter settings include node parameter settings and edge parameter settings.
[0034] a. Abstract each soil block node as a bucket, with bucket parameters including maximum volume Cmax (unit: mm) and maximum infiltration rate Imax (unit: mm / time step).
[0035] b. Abstract each random root system edge as a queue. The queue parameters include the level weight W (unit: dimensionless), representing the root system level, and the queue length D (unit: time step), representing the rate of water movement and the delay time in the macropores.
[0036] Initialize the moisture and salt content (including solution salt and solid salt) in the nodes.
[0037] The algorithm is written in Matlab.
[0038] like Figure 4 As shown, STEP 1: Open the Matlab software. Since the network module is only included in Matlab versions 2018a and later, a Matlab version 2018a or later must be used.
[0039] STEP 2: Open the main program (Mymain.m) in the Matlab main page. Also, place the function file generate_Edges.m, which generates random edges, in the same folder as the main program.
[0040] STEP 3: Click the "Run" button on the "Editor" page. Wait until the double arrow ">>" appears in the command line window to indicate that the code execution has finished.
[0041] STEP 4: The variables named WS_t and SS_t are the moisture stratification results and salinity stratification results, respectively. Double-clicking the variable names in the workspace will display the results in the variable window. Moisture is expressed as a percentage (%), and salinity is expressed as μs / mm.
[0042] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.
Claims
1. A method for constructing a soil water and salt simulation model in crop roots based on complex networks, characterized in that, Includes the following steps: S1: Generate a regular network to construct a network with an adjustable number of nodes. Each node corresponds to a soil body of a set size. Vertical connections are set only between nodes to simulate vertical water transport in the soil. S2: Create random edges based on crop growth characteristics. Use a probabilistic method to add random edges between nodes in the rule network to simulate crop root system. When adding edges, generate edge weights representing root system level at the same time. S3: Design a water-salt dynamics simulation module. Input water is preferentially infiltrated through the matrix flow path controlled by nodes. Water exceeding the infiltration capacity of nodes is transported through the priority flow path controlled by random connections. The priority flow follows the first-in-first-out principle. The total evaporation of soil at surface nodes and transpiration of plants at root nodes is calculated as evapotranspiration. Based on the water difference between evapotranspiration nodes and child nodes, water rebalancing is performed. Soil salinity is set to include solution and solid states. Solution salt migrates with water movement, and solid salt and solution salt are mutually transformed according to a preset relationship as water changes. S4: The model input data includes input moisture, potential evapotranspiration, Kc coefficient reflecting crop growth period, and Kf coefficient reflecting leaf area shading. It initializes the node volume and infiltration rate parameters, edge weights and transmission delay parameters, and outputs soil moisture stratification results and salinity stratification results.
2. The method for constructing a soil water and salt simulation model in crop roots based on complex networks according to claim 1, characterized in that, In step S1, the number of nodes in the rule network is 5×5 to 20×20, and the soil size corresponding to the node is adjusted according to the soil sampling accuracy.
3. The method for constructing a soil water and salt simulation model in crop roots based on complex networks according to claim 1, characterized in that, In step S2, the probability distribution of the probability method is adjusted according to the crop growth stage. During the seedling stage, a higher probability of connecting edges is set between nodes in the soil surface, and during the maturity stage, a higher probability of connecting edges is set between nodes in the deep soil layer.
4. The method for constructing a soil water and salt simulation model in crop roots based on complex networks according to claim 1, characterized in that, In step S31, the infiltration rate I of the matrix flow satisfies the formula: I = Imax × exp(-k × θ), Where Imax is the maximum infiltration rate, θ is the soil moisture content, and k is the reduction coefficient; the matrix flow rate Q satisfies Q=min(input,I), where input is the input moisture.
5. The method for constructing a soil water and salt simulation model in crop roots based on complex networks according to claim 1, characterized in that, In step S3, the conversion relationship between the solid salt and the solution salt satisfies: Safter=Sbefore-Sbefore×(1-θ), Where Safter represents the solid salt content after transformation, Sbefore represents the solid salt content before transformation, and θ represents the soil moisture content.
6. The method for constructing a soil water and salt simulation model in crop roots based on complex networks according to claim 1, characterized in that, The formula for calculating the potential evapotranspiration data is as follows: Where ET0 is the daytime evapotranspiration, Δ is the slope of the vapor pressure curve, Rn is the net radiation on the crop surface, G is the soil heat flux density (G=0 if the daily scale is used as a reference), γ is the hygrometer constant, T is the daily average temperature, u2 is the wind speed at a height of 2 meters, es is the average saturated water vapor pressure at temperature T, ea is the actual water vapor pressure difference at the dew point temperature, and es-ea is the water vapor pressure difference value.
7. The method for constructing a soil water and salt simulation model in crop roots based on complex networks according to claim 1, characterized in that, The node has a volume of 10mm to 50mm, a maximum infiltration rate of 5mm / timestep to 20mm / timestep, and a connection transmission delay parameter of 1 to 5timestep.