Rainfall simulation device and method for soil runoff leaching test

Abstract

The invention discloses a rainfall simulation device and method for a soil runoff leaching test, and relates to the technical field of soil hydrology and environment simulation tests.The rainfall simulation method comprises the following steps that three-dimensional scanning data of the surface of a test soil bed is obtained, and a digital elevation model which covers the surface of the test soil bed and is composed of a plurality of grid units is generated; after the digital elevation model is subjected to depression filling processing, theoretical water flow direction data and theoretical confluence cumulant data of each grid unit are calculated through a multi-flow-direction algorithm. Adjustment is performed based on a dynamic feedback result of actual runoff data, a simulation error source can be accurately positioned in a difference analysis link according to infiltration capacities of different soils and runoff collection characteristics of different earth surfaces, and then the recognition precision of a catchment area and a runoff path is adjusted and optimized through a threshold value so as to adapt to different test conditions, so that the test accuracy is improved. The adaptability to different soil and different earth surface states is improved.

Description

Technical Field

[0001] This invention relates to the field of soil hydrology and environmental simulation test technology, specifically to a rainfall simulation device and method for soil runoff leaching test. Background Technology

[0002] Soil runoff leaching tests are an important tool for studying the migration patterns of soil erosion and agricultural non-point source pollution. They can help researchers evaluate the effectiveness of soil and water conservation measures and improve these measures based on the results. Soil runoff leaching tests are conducted by simulating rainfall on an experimental soil bed.

[0003] Common rainfall simulation methods simulate different raindrop sizes, terminal kinetic energy, and rainfall intensity by controlling nozzle type, water supply pressure, and drop height. Once the rainfall mode of existing simulation devices is set, it remains fixed throughout the entire test. It cannot perform differentiated rainfall simulations based on different terrains, and it lacks the ability to automatically compare and analyze the actual runoff range and runoff path with the expected design target during or after the test, and to dynamically provide feedback and adjust the key parameters in the control model. This makes it impossible for the device's performance to be self-verified and optimized, and its simulation accuracy depends entirely on the initial empirical parameter settings. It also has poor adaptability to different soils and surface conditions. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a rainfall simulation device and method for soil runoff leaching tests.

[0005] To achieve the above objectives, the technical solution of the present invention is as follows:

[0006] A rainfall simulation method for soil runoff leaching experiments includes the following steps:

[0007] Acquire three-dimensional scanning data of the test soil bed surface and generate a digital elevation model covering the test soil bed surface and composed of multiple grid cells;

[0008] After filling depressions in the digital elevation model, a multi-flow direction algorithm is used to calculate the theoretical flow direction data and theoretical cumulative flow data for each grid cell.

[0009] Grid cells with theoretical runoff accumulation data greater than or equal to a preset threshold are marked as catchment area cells;

[0010] Starting from each catchment unit, the theoretical flow direction data of all grid units are calculated in reverse to obtain the potential runoff path data of the corresponding catchment unit.

[0011] Acquire target rainfall parameters and spatial calibration data for each rainfall execution unit;

[0012] Spatial calibration data are input into the fluid dynamics model to determine the effective wetting range projection data formed on the surface of the test soil bed by each rainfall execution unit;

[0013] Based on the target rainfall parameters and potential runoff path data, the target rainfall data for each grid cell is calculated;

[0014] Based on the target rainfall data and effective wet range projection data of all grid cells, a multi-channel control command is generated to drive all rainfall execution units by solving an optimization algorithm.

[0015] The actual water flow direction data and actual flow accumulation data after the execution of multi-channel control commands are obtained, and the difference analysis results are obtained by comparing and analyzing them with the theoretical water flow direction data and theoretical flow accumulation data of the corresponding grid cells.

[0016] The preset threshold is dynamically adjusted based on the difference analysis results.

[0017] Preferably, the preset threshold is dynamically adjusted based on the difference analysis results, specifically including:

[0018] Obtain the actual runoff peak formation time corresponding to the actual water flow direction data And the peak time of the theoretical runoff corresponding to the theoretical flow direction data. Calculate the relative time error;

[0019] Calculate the preset threshold based on relative time error:

[0020] ;

[0021] in, The adjusted preset threshold, The preset threshold before adjustment This is a preset positive calibration coefficient.

[0022] Preferably, the target rainfall data for each grid cell is calculated, specifically including:

[0023] Assign rainfall enhancement weighting coefficients to the grid cells corresponding to potential runoff path data;

[0024] Assign baseline weighting coefficients to the remaining grid cells;

[0025] The target rainfall intensity parameter is multiplied by the rainfall enhancement weighting coefficient and the baseline weighting coefficient to obtain the target rainfall data for each grid cell.

[0026] Preferably, after dynamically adjusting the preset threshold based on the difference analysis results, the method further includes:

[0027] Calculate the centroid of the actual catchment spatial distribution represented by the actual runoff accumulation data and the centroid of the theoretical catchment spatial distribution identified based on the adjusted preset threshold;

[0028] Numerical compensation is performed on the rainfall enhancement weighting coefficient based on the coordinate deviation between the two centroids.

[0029] Preferably, the spatial calibration data is input into the fluid dynamics model to determine the effective wetting range projection data formed on the test soil surface by each rainfall execution unit, specifically including:

[0030] The spatial calibration data includes the installation coordinates, spray angle, and operating flow rate of each rainfall execution unit;

[0031] Spatial calibration data is input into a fluid dynamics model validated by experimental data. The simulation calculation yields the rainfall contribution rate distribution of each rainfall execution unit to the lower grid units under rated operating conditions, which serves as the effective wetting range projection data.

[0032] Preferably, by optimizing the algorithm, multi-channel control commands are generated to drive all rainfall execution units, specifically including:

[0033] A constrained optimization problem is constructed with the activation duration of each rainfall execution unit during the simulation period as the decision variable and the objective being to minimize the sum of squares of the differences between the target rainfall data and the predicted rainfall data of all grid units. The problem is then solved using a linear programming algorithm.

[0034] The predicted rainfall data is obtained by matrix multiplication of the decision variables and the effective wet range projection data.

[0035] Preferably, starting from each catchment unit, a reverse source calculation is performed on the theoretical flow direction data of all grid units, specifically including:

[0036] Starting from each catchment unit;

[0037] Iterate through the theoretical flow direction data of all upstream grid cells adjacent to the catchment unit, and mark the grid cells whose flow direction points to the catchment unit as potential runoff path cells;

[0038] Starting from the marked potential runoff path cell, continue marking the upstream grid cells of that runoff path cell;

[0039] The process is repeated recursively until the watershed is reached, and potential runoff path data is constructed from all marked potential runoff path units.

[0040] A rainfall simulation device for soil runoff leaching tests includes:

[0041] The terrain scanning module is used to collect three-dimensional scanning data of the test soil bed surface;

[0042] The data processing and control module, which is in communication with the terrain scanning module, is configured to perform the following operations:

[0043] Acquire three-dimensional scanning data of the test soil bed surface and generate a digital elevation model covering the test soil bed surface and composed of multiple grid cells;

[0044] After filling depressions in the digital elevation model, a multi-flow direction algorithm is used to calculate the theoretical flow direction data and theoretical cumulative flow data for each grid cell.

[0045] Grid cells with theoretical runoff accumulation data greater than or equal to a preset threshold are marked as catchment area cells;

[0046] Starting from each catchment unit, the theoretical flow direction data of all grid units are calculated in reverse to obtain the potential runoff path data of the corresponding catchment unit.

[0047] Acquire target rainfall parameters and spatial calibration data for each rainfall execution unit;

[0048] Spatial calibration data are input into the fluid dynamics model to determine the effective wetting range projection data formed on the surface of the test soil bed by each rainfall execution unit;

[0049] Based on the target rainfall parameters and potential runoff path data, the target rainfall data for each grid cell is calculated;

[0050] Based on the target rainfall data and effective wet range projection data of all grid cells, a multi-channel control command is generated to drive all rainfall execution units by solving an optimization algorithm.

[0051] The actual water flow direction data and actual flow accumulation data after the execution of multi-channel control commands are obtained, and the difference analysis results are obtained by comparing and analyzing them with the theoretical water flow direction data and theoretical flow accumulation data of the corresponding grid cells.

[0052] The preset threshold is dynamically adjusted based on the difference analysis results.

[0053] Preferably, it also includes arrayed rainfall execution units, which are communicatively connected to the data processing and control module. Each rainfall execution unit includes a dripper and a solenoid valve disposed on the dripper. The multi-channel control command is specifically a pulse signal that controls the opening timing and duty cycle of each solenoid valve during the simulation period.

[0054] Preferably, it also includes a sensor feedback module that is communicatively connected to the data processing and control module. This sensor is deployed below the test soil bed or on the critical path to collect actual water flow direction data and actual flow accumulation data, and transmit the data to the data processing and control module.

[0055] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0056] This invention employs a technical chain from 3D scanning modeling to multi-flow-direction algorithm calculation, and then to reverse source tracing to extract runoff paths. It accurately identifies catchment areas and potential runoff paths on the surface of the test soil bed, and allocates target rainfall amounts to the grid based on path characteristics. Furthermore, it generates multi-channel control commands through an optimization algorithm, driving the rainfall execution unit to implement enhanced rainfall in runoff path areas and baseline rainfall in non-path areas. This breaks the traditional model of uniform rainfall across the entire area, allowing artificial rainfall to perfectly match the water flow evolution characteristics of the soil bed topography. The simulation results more closely resemble the runoff leaching process of natural rainfall. By collecting data on the actual water flow direction and cumulative runoff volume of the soil bed after rainfall, and performing precise grid-level comparison with the theoretical data calculated by the multi-flow-direction algorithm, it quantifies core parameters such as runoff path overlap, runoff volume error, and peak time deviation. The system dynamically adjusts the preset threshold for catchment area identification based on the difference analysis results, thereby optimizing the key parameters of the control model through feedback. This closed-loop system upgrades rainfall simulation from a one-time fixed mode to a self-verifiable and continuously optimized intelligent mode. Adjustments are made based on dynamic feedback from actual runoff data. The difference analysis stage accurately identifies the sources of simulation errors, taking into account the infiltration capacity of different soils and the runoff collection characteristics of different surfaces. Threshold adjustments further optimize the identification accuracy of catchment areas and runoff paths, and weight compensation further calibrates rainfall distribution to adapt to different experimental conditions. Without relying on a large number of initial empirical parameters, it can stably output high-precision simulation results on test beds with different soils and surface conditions, improving adaptability to various soil and surface conditions. Attached Figure Description

[0057] The disclosure of this invention is illustrated with reference to the accompanying drawings. It should be understood that the drawings are for illustrative purposes only and are not intended to limit the scope of protection of this invention. In the drawings, the same reference numerals are used to refer to the same parts. Wherein:

[0058] Figure 1 This is a flowchart of the steps of the present invention;

[0059] Figure 2 This is a data flow diagram of the present invention. Detailed Implementation

[0060] It is readily understood that, based on the technical solution of this invention, those skilled in the art can propose various interchangeable structural methods and implementations without altering the essential spirit of the invention. Therefore, the following detailed embodiments and accompanying drawings are merely illustrative examples of the technical solution of this invention and should not be considered as the entirety of the invention or as limitations or restrictions on the technical solution of this invention.

[0061] like Figure 1-2 As shown, a rainfall simulation device and method for soil runoff leaching test includes the following steps:

[0062] Acquire three-dimensional scanning data of the test soil bed surface and generate a digital elevation model covering the test soil bed surface and composed of multiple grid cells;

[0063] Specifically, a test soil bed is a device in a laboratory that simulates a natural soil slope. It is usually a regularly shaped soil trough filled with test soil. The slope, roughness, and soil structure can be filled according to the needs to simulate the real soil surface in the field.

[0064] Furthermore, the test soil bed is scanned using non-contact 3D scanning equipment such as laser 3D scanners or structured light 3D scanners to avoid damaging the soil bed surface during the scanning process. Taking a laser 3D scanner as an example, the laser 3D scanner is set up above or to the side of the test soil bed, and a laser beam is emitted towards the surface of the test soil bed. The laser 3D scanner moves and scans along a preset path to ensure that the entire surface area of ​​the test soil bed is covered. By receiving the time difference or phase difference of the reflected laser, the distance from the laser 3D scanner to each point on the surface of the test soil bed is calculated. Combined with the position and orientation of the laser 3D scanner, the 3D coordinates of each point are finally generated, thus obtaining the 3D scanning data of the test soil bed. This allows for the accurate capture of the micro-topographic features of the test soil bed surface, such as the undulations, small pits, or small protrusions on the surface of the test soil bed.

[0065] Furthermore, due to the presence of dust reflections and equipment error points during scanning, the original 3D scan data contains noise. It is necessary to first remove abnormal data points that deviate from the actual terrain from the original 3D scan data before meshing. According to the experimental accuracy requirements, the planar range of the test soil bed is divided into uniformly sized square grids. Each grid cell has unique planar coordinates. Using an interpolation algorithm, the representative elevation value of the corresponding grid cell is calculated using the 3D scan data around the grid cell. Commonly used interpolation algorithms include the neighborhood averaging method and the inverse distance weighted method. Taking the neighborhood averaging method as an example, the average elevation of all points within a certain range around the grid cell is taken as the elevation of the grid cell. Finally, a digital elevation model is output. The output digital elevation model is a two-dimensional matrix, where each element of the matrix corresponds to a grid cell, and the element value corresponds to the elevation value of the corresponding grid cell.

[0066] After filling depressions in the digital elevation model, a multi-flow direction algorithm is used to calculate the theoretical flow direction data and theoretical cumulative flow data for each grid cell.

[0067] Specifically, minor subsidence on the surface of the test soil bed can create meaningless pseudo-depressions in the digital elevation model (DEM). These pseudo-depressions are typically only a few millimeters deep and are quickly filled by rainwater during natural rainfall. Once full, the water overflows and continues to flow along the terrain, without accumulating for long periods. Without filling these depressions, subsequent flow calculations will be affected. Depression filling essentially involves identifying and filling these minor depressions in the DEM that do not affect the overall runoff pattern, as detailed below:

[0068] Traverse all grid cells in the digital elevation model. Using a single grid cell or multiple adjacent grid cells as units, determine whether the elevation value of the corresponding grid cell is lower than the elevation values ​​of all adjacent grid cells. If so, it is a pseudo depression. Calculate the filling depth and determine the height to be filled for each pseudo depression. The principle is to fill to the lowest outlet elevation around the pseudo depression, so that the pseudo depression is filled to just the height at which water overflows and flows downstream. This avoids overfilling and altering the overall terrain while eliminating the impact of pseudo depressions on water flow calculations. Correct the elevation values ​​of all grid cells within the pseudo depression to the calculated outlet elevation values ​​to generate a digital elevation model without pseudo depressions.

[0069] Furthermore, after the depression filling process is completed, the theoretical flow direction data and theoretical cumulative runoff data for each grid cell are calculated. Traditional calculation methods use a unidirectional flow algorithm, assuming that water in each grid cell will only flow to the lowest-elevation grid among its eight adjacent grid cells. However, in natural conditions, the soil surface is uneven, and water flow often branches towards multiple depressions, which does not conform to actual runoff patterns. A multidirectional flow algorithm allows water flow to be distributed to multiple downstream depression grids, with the distribution ratio determined based on factors such as the elevation difference and slope between adjacent grids. The calculation results are closer to actual runoff patterns. The specific calculations are as follows:

[0070] Calculate the theoretical flow direction data for each grid cell: Taking the current grid cell as the center, select the 8 neighboring grid cells surrounding it as the judgment objects, compare the elevation values ​​of the current grid cell with the 8 neighboring grid cells, and filter out the neighboring grid cells with lower elevation values ​​than the current grid cell. These filtered grid cells are the downstream grid cells of the current grid cell. Based on the elevation difference and slope between the downstream grid cells and the current grid cell, determine the proportion of water flow allocated to each downstream grid cell of the current grid cell, and record the downstream grid list and the corresponding water flow allocation proportion of each grid cell as the theoretical flow direction data.

[0071] For example, if the elevation of the current grid cell A is 10cm, the elevation of the adjacent grid cell B is 8cm, the elevation of the grid cell C is 9cm, and the elevation of the other neighboring grid cells is higher than 10cm, then the elevation difference between grid cell A and grid cell B is greater, the slope is steeper, and therefore the proportion of water flow allocated to grid cell B is higher, while the proportion allocated to C is lower.

[0072] Calculate the theoretical runoff accumulation data for each grid cell: The calculation process traverses the grid cells in descending order of terrain elevation, calculating upstream grid cells first and downstream grid cells last. Initially, assign a runoff accumulation value of 1 to all grid cells. Starting with the grid cell with the highest elevation value in the digital elevation model, calculate the runoff accumulation for each grid cell sequentially.

[0073] formula:

[0074] Current grid cell cumulative amount = initial value + ∑ (upstream grid cell cumulative amount × proportion flowing into the current grid cell)

[0075] For example: Grid cell B is upstream of grid cell A. Grid cell A allocates 70% of the water flow to grid cell B. If the cumulative amount of grid cell A is 100, then the amount of water flowing into grid cell B is 100 × 70% = 70. The final cumulative amount of grid cell B is 1 + 70 = 71.

[0076] After the traversal is completed, each grid cell has corresponding cumulative flow data. The larger the value, the more water flows through the grid cell, and the easier it is to form a flow.

[0077] Grid cells with theoretical runoff accumulation data greater than or equal to a preset threshold are marked as catchment area cells;

[0078] Specifically, the theoretical cumulative flow data of all grid cells calculated by the multi-flow algorithm after depression filling is retrieved, the cumulative flow value of each grid cell is read one by one, and a judgment is performed on each grid cell:

[0079] If the theoretical cumulative flow of a grid cell is greater than or equal to a preset threshold, the corresponding grid cell will be marked as a catchment area cell.

[0080] If the theoretical cumulative runoff of a grid cell is less than a preset threshold, it is determined to be a normal runoff grid cell and is not marked.

[0081] The final result is a distribution map of all catchment units, including their locations and numbers. These catchment units are the core units for water flow collection on the surface of the test soil bed.

[0082] Starting from each catchment unit, the theoretical flow direction data of all grid units are calculated in reverse to obtain the potential runoff path data of the corresponding catchment unit.

[0083] Specifically, a marked catchment unit is selected as the starting point of this calculation. This catchment unit is used as the endpoint of the reverse source tracing water flow path. All grid units adjacent to this catchment unit are traversed, and the theoretical water flow direction data of the corresponding grid units are consulted. Adjacent grid units whose water flow direction points to this catchment unit are selected to obtain the directly upstream grid units of the catchment unit. The directly upstream grid units of the catchment unit are then marked as first-level potential runoff path units.

[0084] For example, if the adjacent grid cells of catchment unit B are A, C, and D, and 30% of the water flow in grid cell A flows to catchment unit B, and 50% of the water flow in grid cell D flows to catchment unit B, then grid cells A and D are the directly upstream grid cells of catchment unit B and are marked as first-level potential runoff path cells.

[0085] Furthermore, a marked primary potential runoff path unit is selected, and all grid units adjacent to this primary potential runoff path unit are traversed. Grid units whose flow direction points towards this primary potential runoff path unit are selected and marked as secondary potential runoff path units. Then, using the secondary potential runoff path unit as the target, upstream grid units are selected and marked as tertiary potential runoff path units, and so on. When no upstream grid unit pointing towards the current path unit can be found in a certain round of selection, the source tracing stops. Such grid units without upstream water are watershed grid units, i.e., the source and starting point of the runoff path. After rainwater falls on the watershed grid unit, it will gradually flow along the path towards the catchment area unit.

[0086] Furthermore, after completing the full-process source tracing for each catchment unit, all potential runoff path units marked as Level 1, Level 2, Level 3 to Level N potential runoff path units are integrated with the watershed grid units to obtain the potential runoff path data for that catchment unit. This potential runoff path unit data includes the coordinates, levels, and water distribution ratio information of all grid units along the path, which can be intuitively presented as a path map from the watershed to the catchment area. Compared with the uniform watering of traditional rainfall simulation and the lack of distinction between runoff paths and ordinary areas, reverse source tracing accurately identifies the water inflow channels of the catchment area, allowing subsequent rainfall control to have a clear target area.

[0087] Acquire target rainfall parameters and spatial calibration data for each rainfall execution unit;

[0088] Specifically, the target rainfall parameters are pre-set characteristic indicators of natural rainfall that need to be simulated based on the experimental objectives. The target rainfall parameters determine the total amount, intensity, and process of rainfall to be applied to the surface of the test soil bed. The target rainfall parameters typically include the following parameters:

[0089] Rainfall intensity: The amount of rainfall per unit time is the core indicator for simulating the magnitude of rainfall. It is usually expressed as light rain (1 mm / h), moderate rain (3 mm / h), and heavy rain (10 mm / h).

[0090] Rainfall duration: The duration of a single simulated rainfall event, such as 30 minutes, 1 hour, or 2 hours;

[0091] Total rainfall: The total rainfall required to reach the surface of the test soil bed can be calculated by multiplying the rainfall intensity and the duration of rainfall;

[0092] Rainfall event type: that is, the temporal distribution characteristics of rainfall, such as: uniform rainfall, intermittent rainfall, increasing or decreasing rainfall;

[0093] Spatial distribution characteristics of rainfall: Requirements for the distribution of rainfall on the surface of the test soil bed, such as uniform rainfall across the entire area or localized enhanced rainfall;

[0094] Spatial calibration data are input into the fluid dynamics model to determine the effective wetting range projection data formed on the surface of the test soil bed by each rainfall execution unit;

[0095] The target rainfall parameters are not randomly set, but determined based on the experimental objectives:

[0096] If the experiment simulates the natural rainfall runoff characteristics of a certain region, then refer to the historical rainfall monitoring data of that region, such as the rainfall intensity and duration records of the local meteorological station;

[0097] If the experiment is to study the effect of different rainfall intensities on soil leaching, multiple sets of gradient rainfall intensities, such as 1 mm / h, 3 mm / h, and 5 mm / h, are artificially set.

[0098] Furthermore, a rainfall actuator refers to the hardware device that enables artificial rainfall, typically an array of nozzles or drippers. Spatial calibration data describes the installation location, operating posture, and water output capacity of each rainfall actuator, specifically including:

[0099] Installation coordinates: The position of the rainfall execution unit in three-dimensional space, usually with a corner point of the test soil bed as the origin to establish a coordinate system;

[0100] Jet tilt angle: The angle between the direction of water discharge from the nozzle and the horizontal or vertical plane, which determines the coverage area of ​​the rainfall;

[0101] Operating flow rate: The amount of water discharged per unit time by the rainfall actuator under rated operating conditions.

[0102] The spatial calibration data is obtained in the following ways:

[0103] The working flow rate, maximum spray angle of the nozzle, etc. can be directly found in the equipment manual as the factory parameters. The installation coordinates are measured with a tape measure or laser rangefinder, and the spray tilt angle is calibrated with an angle meter to ensure that the parameters of each nozzle or dripper are consistent with the actual installation status.

[0104] Based on the target rainfall parameters and potential runoff path data, the target rainfall data for each grid cell is calculated;

[0105] Specifically, based on potential runoff path data, all grid cells of the test soil bed are divided into two categories: potential runoff path grid cells and non-potential runoff path grid cells, forming a grid type zoning map. Potential runoff path grid cells represent the main flow areas of water under natural rainfall, requiring a higher rainfall allocation to simulate the effect of water flow convergence. Non-potential runoff path grid cells are all grid cells excluding potential runoff path grid cells, including watershed grid cells and non-channel slope grid cells. Water flow in non-potential runoff path grid cells does not directly flow into the catchment area, and allocating a baseline rainfall amount is sufficient to meet the rainfall coverage requirements of the experiment.

[0106] Weighting coefficients are the core control parameters for achieving differentiated rainfall, and specifically include:

[0107] Rainfall enhancement weighting coefficient : This is used to ensure that the rainfall in the potential runoff path grid cells is higher than the target rainfall intensity, thereby enhancing the potential runoff collection effect. This is particularly useful when simulating storm runoff that emphasizes water collection and scouring. A value of 1.3 to 1.5 is acceptable; if the simulation focuses on slow infiltration of light rain runoff, A value of 1.1 to 1.2 is acceptable.

[0108] Benchmark weighting coefficient : The rainfall in non-potential runoff path grid cells is equal to the target rainfall intensity to satisfy basic coverage.

[0109] Furthermore, the target rainfall for each grid cell is calculated using a formula, with separate calculations for the two types of grid cells:

[0110] The target rainfall per unit area grid cell = target rainfall intensity × weighting coefficient × rainfall duration, which can be converted into a mathematical formula:

[0111] ;

[0112] in:

[0113] The target rainfall for the i-th grid cell is expressed in mm, representing thickness, which can be converted to volume.

[0114] The target rainfall intensity is a preset value for the test, such as 3 mm / h;

[0115] The weight coefficient for the i-th grid cell is taken as the potential runoff path grid cell. Non-potential runoff path grid cells are taken ;

[0116] The duration of rainfall is a preset value for the experiment, such as 1 hour.

[0117] Example calculation is as follows:

[0118] Assumptions for the experiment:

[0119] Target rainfall intensity ;

[0120] Rainfall duration ;

[0121] Rainfall Enhancement Weight ;

[0122] Benchmark weight ;

[0123] Calculation results:

[0124] Target rainfall of potential runoff path grid cells ;

[0125] Target rainfall for non-potential runoff path grid cells: .

[0126] Compared to the uniform watering of the entire area in traditional artificial rainfall, the method uses differentiated weights to allow rainfall to accurately match the water flow patterns of the terrain, thus significantly improving the simulation accuracy.

[0127] Based on the target rainfall data and effective wet range projection data of all grid cells, a multi-channel control command is generated to drive all rainfall execution units by solving an optimization algorithm.

[0128] Specifically, the effective wet range projection data represents the rainfall contribution rate of each rainfall execution unit j to each grid i. This value is used to determine the rainfall capacity of the sprinkler head, and its range is 0-1. This indicates that nozzle j cannot pour water onto grid i.

[0129] Construct a two-dimensional matrix (number of nozzles × number of grids):

[0130] (m is the total number of nozzles).

[0131] Rainfall contribution rate The physical meaning is the proportion of rainfall that sprinkler j can provide to grid cell i per working unit time to the total water output of the sprinkler. Its value is calculated by the fluid dynamics model combined with the sprinkler spatial calibration data, and already includes the influence of sprinkler installation position, spray angle and working flow rate.

[0132] Furthermore, we first construct a constrained optimization problem, which is to minimize the error between the actual rainfall and the target rainfall in the grid cells by adjusting the operating parameters of the sprinkler, as follows:

[0133] Determine the decision variables: that is, determine the activation duration of each rainfall execution unit. The operating flow rate of the rainfall execution unit is the rated value, i.e., the factory calibration or field measurement. The amount of rainfall it delivers to the grid unit is proportional to the duration of operation. , The total rainfall duration set for the test, i.e. the sprinkler head opening time, cannot be negative and cannot exceed the total rainfall duration.

[0134] Define the objective function as follows: Minimize the sum of squared errors between the predicted rainfall and the target rainfall for all grid cells. The formula is:

[0135] ;

[0136] in:

[0137] The target rainfall for grid cell i;

[0138] The predicted rainfall for grid cell i is calculated from the sprinkler opening duration and contribution rate. The core calculation formula is as follows:

[0139] That is, the predicted rainfall for each grid cell is equal to the sum of the product of the contribution rate of each sprinkler that can reach that grid cell and its operating duration.

[0140] In order to solve To meet the actual operating requirements of the hardware, the following constraints need to be added:

[0141] Nonnegativity constraint: That is, the nozzle opening time cannot be negative;

[0142] Duration constraints: This means that the duration of operation of a single sprinkler head cannot exceed the total duration of rainfall;

[0143] Flow constraints: , The rated flow rate of nozzle j is... This represents the total water consumption allowed for the experiment.

[0144] Furthermore, the quadratic objective function that minimizes the sum of squared errors is transformed into a linear objective function through linearization. The constraint on sprinkler duration is transformed into a system of linear inequalities. Given the input grid, sprinkler contribution rate matrix, and target rainfall array, a linear programming algorithm is run to solve for the optimal operating duration of each sprinkler. The details are as follows:

[0145] Data preprocessing: The contribution rate matrix Target rainfall vector The parameters of the constraints are organized into a format that can be recognized by the linear programming solver.

[0146] Objective function linearization: Transform the quadratic objective function into a linear function. If the solver only supports linear objective functions, it can be approximated by Taylor expansion or replaced by the absolute value of the error.

[0147] The preprocessed model is input into a linear programming solver, such as the linprog function in MATLAB. The solver iteratively calculates the optimal solution for the decision variables using the simplex method or interior point method. This refers to the optimal opening time of the nozzle. Check whether the optimal solution satisfies all constraints. If not, adjust the constraint parameters or contribution rate matrix and solve again.

[0148] Then calculate the optimal start time. The commands are converted into multi-channel control instructions that the rainfall execution unit can recognize. The specific conversion rules are as follows:

[0149] Command format: Pulse signal, which controls the opening and closing of the solenoid valve;

[0150] Timing allocation: Determine the start time of each sprinkler head and convert the start time into a time interval to avoid water pressure fluctuations caused by simultaneous start of sprinklers;

[0151] Duty cycle calculation: If an intermittent rainfall pattern is adopted, Converted into a cyclic duty cycle of on and off;

[0152] Channel division: Each rainfall execution unit corresponds to an independent control channel, ensuring that commands can drive each nozzle individually without interference;

[0153] Signal conversion: The time interval is converted into pulse control signals, such as a high level 1 representing on and a low level 0 representing off, ultimately forming a multi-channel control command.

[0154] Unlike the traditional, crude approach of simultaneously switching all sprinklers on and off and uniformly controlling rainfall intensity, this method uses optimized algorithms to differentiate the operation of sprinklers, allowing rainfall distribution to accurately match the water flow patterns of the terrain, thus significantly improving the realism of runoff simulation.

[0155] The actual water flow direction data and actual flow accumulation data after the execution of multi-channel control commands are obtained, and the difference analysis results are obtained by comparing and analyzing them with the theoretical water flow direction data and theoretical flow accumulation data of the corresponding grid cells.

[0156] Specifically, starting from the start of the rainfall execution unit, data on the actual water flow direction and the actual cumulative runoff volume are continuously collected at preset time intervals until the runoff completely recedes after the rainfall ends. The collected data are then compared and analyzed, as follows:

[0157] The comparative analysis of water flow direction data aims to verify whether the simulated runoff path matches the actual runoff path.

[0158] Calculate the flow direction deviation angle, which measures the degree of deviation between the actual flow direction and the theoretical flow direction in a single grid cell:

[0159] ;

[0160] in, The deviation angle of the water flow direction in the i-th grid cell;

[0161] The actual water flow direction angle of the i-th grid cell;

[0162] The theoretical flow direction angle of the i-th grid cell;

[0163] This indicates that the direction simulation is accurate;

[0164] This indicates a slight deviation in the direction simulation;

[0165] This indicates that there is a significant deviation in the directional simulation, which requires focused optimization.

[0166] Calculate runoff path overlap to measure the degree of actual-to-theoretical match between overall potential runoff paths:

[0167] ;

[0168] in, This represents the number of grid cells where the actual path overlaps with the theoretical path.

[0169] This represents the total number of grid cells for the theoretical path.

[0170] This indicates that the runoff path simulation is excellent.

[0171] This indicates that there are systematic errors in the runoff path simulation;

[0172] This indicates that the runoff path simulation is successful.

[0173] The comparative analysis of cumulative runoff data aims to verify whether the simulated runoff volume matches the actual runoff volume, as detailed below:

[0174] The absolute error of a single grid cell directly reflects the difference between the actual and theoretical flow rates of that cell.

[0175] ;

[0176] in, The absolute error of the cumulative flow in the i-th grid cell;

[0177] This represents the actual cumulative flow of the i-th grid cell;

[0178] Let be the theoretical cumulative flow of the i-th grid.

[0179] This indicates that the actual runoff volume is greater than the theoretical value, indicating either more rainfall or a stronger runoff collection capacity.

[0180] : This indicates an exact match;

[0181] This indicates that the actual runoff is less than the theoretical value, resulting in less rainfall or more soil infiltration.

[0182] Calculate the relative error of individual grid cells, a standardized error metric used to eliminate grid size differences:

[0183] ;

[0184] in, The relative error of the cumulative flow in the i-th grid;

[0185] This indicates that the water volume simulation is highly accurate.

[0186] This indicates that the accuracy of the water volume simulation is moderate.

[0187] This indicates that the water volume simulation has a large error.

[0188] Calculate the peak runoff time deviation across the entire region to measure the synchronicity of runoff processes and reflect the deviation between the actual peak runoff time and the theoretical time:

[0189] ;

[0190] in, This refers to the time deviation of peak runoff.

[0191] This refers to the peak time of actual runoff formation;

[0192] This represents the peak time of theoretical runoff formation.

[0193] The smaller the value, the better the synchronization of the process.

[0194] Based on the analysis of the experimental process, common causes of error include:

[0195] Terrain modeling error: Insufficient 3D scanning accuracy leads to deviations between the digital elevation model and the actual soil surface;

[0196] Sprinkler control error: The actual working flow rate and spray angle of the rainfall actuator are inconsistent with the calibration data, resulting in a deviation in the effective wetting range;

[0197] Changes in soil properties: During rainfall, the soil surface may crust, erode, or collapse, altering the actual runoff patterns;

[0198] Parameter setting error: The initial values ​​of the preset threshold and rainfall weight coefficient are unreasonable, which leads to the deviation in the identification of theoretical catchment area and path.

[0199] The final output of the difference analysis results includes:

[0200] Grid-level difference data table: Specific values ​​for the water flow direction deviation angle, relative error of the cumulative flow, and peak time deviation for each grid;

[0201] Visualized difference map:

[0202] Runoff path difference map: Theoretical paths, actual paths, and overlapping paths are marked with different colors;

[0203] Convergence cumulative error heatmap: The relative error is represented by the intensity of the color, for example, red is the high error area and green is the low error area;

[0204] Error Source Analysis Report: Sorted by error contribution rate, this report lists the main causes of simulation deviations.

[0205] Unlike traditional rainfall simulations that only look at the results, this method uses precise grid-level comparisons to quantify the magnitude and distribution of simulation errors, providing a clear evaluation standard for simulation accuracy.

[0206] The preset threshold is dynamically adjusted based on the difference analysis results.

[0207] The aforementioned technology, through a technical chain from 3D scanning modeling to multi-flow direction algorithm calculation, and then to reverse tracing to extract runoff paths, accurately identifies the catchment units and potential runoff paths on the surface of the test soil bed. Based on the path characteristics, it allocates grid target rainfall differentially, and then generates multi-channel control commands through optimization algorithms to drive the rainfall execution unit to implement enhanced rainfall in the runoff path area and baseline rainfall in the non-path area. This breaks the traditional pattern of uniform rainfall across the entire area, allowing artificial rainfall to perfectly match the water flow evolution characteristics of the soil bed topography, and the simulation results are closer to the runoff leaching process of natural rainfall.

[0208] By collecting data on the actual water flow direction and cumulative runoff volume of the soil bed after rainfall, and comparing it with the theoretical data calculated by the multi-flow direction algorithm at the grid level, core indicators such as runoff path overlap, runoff volume error, and peak time deviation are quantified. Based on the difference analysis results, the preset threshold for catchment area identification is dynamically adjusted to achieve feedback optimization of key parameters of the control model. Through the closed-loop system, rainfall simulation is upgraded from a one-time fixed mode to an intelligent mode that can be self-verified and continuously optimized.

[0209] By adjusting the simulation results based on dynamic feedback from actual runoff data, and considering the infiltration capacity of different soils and the runoff collection characteristics of different surfaces, the simulation error sources are accurately located in the difference analysis stage. Then, the identification accuracy of the catchment area and runoff path is optimized by threshold adjustment, and the rainfall distribution is further calibrated by weight compensation to adapt to different experimental conditions. Without relying on a large number of initial empirical parameters, it can stably output high-precision simulation results on test soil beds with different soils and surface conditions, thus improving the adaptability to different soils and surface conditions.

[0210] The preset thresholds are dynamically adjusted based on the results of the difference analysis, specifically including:

[0211] Obtain the actual runoff peak formation time corresponding to the actual water flow direction data And the peak time of the theoretical runoff corresponding to the theoretical flow direction data. Calculate the relative time error;

[0212] Calculate the preset threshold based on relative time error:

[0213] ;

[0214] in, The adjusted preset threshold, The preset threshold before adjustment This is a preset positive calibration coefficient.

[0215] Specifically, the preset threshold is the watershed for identifying catchment units, but the initial preset threshold is an empirical value set manually, which deviates from the actual runoff characteristics of the soil bed.

[0216] If the threshold is too high: the number of identified catchment areas is too small, the theoretical runoff path is not covered enough, and the actual runoff peak will occur later than the theoretical value.

[0217] If the threshold is too low: the number of identified catchment areas is too large, the theoretical runoff path is excessively extended, and the actual runoff peak occurs earlier than the theoretical value.

[0218] Therefore, it is necessary to adjust the preset threshold in reverse based on the difference between the actual and theoretical peak times, as follows:

[0219] Calculating the relative time error is a core indicator for quantifying the deviation between the actual and theoretical peak runoff times. It involves obtaining the actual peak runoff formation time corresponding to the actual flow direction data. And the peak time of the theoretical runoff corresponding to the theoretical flow direction data. Calculate the relative time error:

[0220] ;

[0221] : The actual peak occurred earlier, indicating that the initial threshold was too low and the catchment area was too large;

[0222] like : The actual peak occurred later, indicating that the initial threshold was too high and the catchment area was too small;

[0223] like The actual peak value is completely synchronized with the theoretical peak value, and the threshold value does not need to be adjusted.

[0224] Furthermore, by substituting the relative time error into the formula, a new preset threshold is calculated:

[0225] ;

[0226] in, The sensitivity of threshold adjustment is controlled by β. The larger β is, the more pronounced the threshold change with error. To avoid excessive threshold adjustment leading to system oscillation, the range of β is [value missing]. .

[0227] The following is an example calculation:

[0228] when At that time, the known parameters are: , , ;

[0229] Calculate the relative time error: ;

[0230] Calculate the new threshold: ;

[0231] When the threshold is reduced from 10 L to 9 L, more grid cells will be marked as catchment areas, the theoretical runoff path will be more complete, and the theoretical peak time in the next simulation will be delayed to synchronize with the actual peak.

[0232] when At that time, the known parameters are: , , ;

[0233] Calculate the relative time error: ;

[0234] Calculate the new threshold: ;

[0235] Raising the threshold from 10 L to 11 L will mark more grid cells as catchment areas, resulting in a more complete theoretical runoff path. The theoretical peak time in the next simulation will be earlier and synchronized with the actual peak time.

[0236] By utilizing the relative error of runoff peak time, the identification threshold of catchment units is dynamically corrected, making subsequent catchment area and runoff path identification closer to actual runoff patterns, ultimately improving the accuracy of rainfall simulation.

[0237] Calculate the target rainfall data for each grid cell, specifically including:

[0238] Assign rainfall enhancement weighting coefficients to the grid cells corresponding to potential runoff path data;

[0239] Assign baseline weighting coefficients to the remaining grid cells;

[0240] The target rainfall intensity parameter is multiplied by the rainfall enhancement weighting coefficient and the baseline weighting coefficient to obtain the target rainfall data for each grid cell.

[0241] After dynamically adjusting the preset threshold based on the difference analysis results, it also includes:

[0242] Calculate the centroid of the actual catchment spatial distribution represented by the actual runoff accumulation data and the centroid of the theoretical catchment spatial distribution identified based on the adjusted preset threshold;

[0243] Numerical compensation is performed on the rainfall enhancement weighting coefficient based on the coordinate deviation between the two centroids.

[0244] Specifically, the centroid of the spatial distribution of the catchment area is used as a quantitative indicator to describe the spatial location of the catchment area on the surface of the test soil bed. The cumulative runoff of each catchment unit is used as a weight to calculate the weighted average coordinates of all catchment units. By using a single coordinate point to accurately characterize the spatial center location of the catchment area, the ambiguity of describing the location using scattered grid coordinates is avoided.

[0245] The spatial centroids of the actual catchment area and the adjusted theoretical catchment area were calculated separately, yielding two centroid coordinates:

[0246] Calculate the centroid of the actual catchment area spatial distribution :

[0247] enter:

[0248] Coordinates of actual catchment units: Planar coordinates of each actual catchment unit (i=1,2,..., ), This represents the actual number of catchment units;

[0249] Actual cumulative runoff: The actual cumulative runoff volume of each actual catchment area unit. ;

[0250] The centroid coordinates are calculated using a weighted average method, with the actual cumulative flow as the weight:

[0251] ;

[0252] This center of gravity is the location of the area where the actual runoff is most concentrated.

[0253] Calculate the centroid of the spatial distribution of the theoretical catchment area after adjustment. :

[0254] Use the adjusted new preset threshold Re-identify theoretical catchment units

[0255] enter:

[0256] Adjusted coordinates of theoretical catchment units: Planar coordinates of each theoretical catchment unit ,

[0257] ( ), This represents the theoretical number of catchment units;

[0258] Adjusted theoretical runoff accumulation: Theoretical runoff accumulation for each theoretical catchment area unit. ;

[0259] A weighted average method is used, with the theoretical cumulative flow as the weight:

[0260] ;

[0261] This center of gravity is the central position of the area where the theoretical runoff water volume is most concentrated.

[0262] Furthermore, by the coordinate differences, the spatial position deviation between the actual catchment area and the theoretical catchment area is quantified to obtain two indicators of the deviation in the X direction and the deviation in the Y direction:

[0263] ;

[0264] If it indicates that the center of gravity of the actual catchment area is on the right side of the theoretical center of gravity, then it is on the left side;

[0265] If it indicates that the center of gravity of the actual catchment area is above the theoretical center of gravity, then it is below;

[0266] If and the centers of gravity completely coincide and the spatial position matching degree is 100%, and there is no need to compensate the weight coefficient.

[0267] Furthermore, according to the coordinate deviation, numerical compensation is performed on the rainfall enhancement weight coefficient, and the potential runoff path grids in the direction of the deviation of the actual center of gravity are increased in the rainfall enhancement weight coefficient, and vice versa, so that the water volume center of the theoretical rainfall distribution approaches the actual center of gravity. The compensation principle is as follows:

[0268] Do not change the weight coefficients of all runoff path grids, and only compensate the grids close to the direction of the center of gravity deviation to avoid new errors caused by excessive adjustment;

[0269] The compensation amplitude is proportional to the absolute value of the center of gravity deviation: the greater the deviation, the greater the compensation amplitude, and the smaller the deviation, the smaller the compensation amplitude.

[0270] The specific compensation formula is:

[0271] Let the original rainfall enhancement weight coefficient be , and the compensated weight coefficient be , and the compensation coefficient be k (0 < k ≤ 0.2, to avoid excessive compensation amplitude):

[0272] For the deviation in the X direction :

[0273] For the potential runoff path grids on the right side of the center of gravity of the theoretical catchment area, the weight coefficient compensation formula:

[0274] ;

[0275] For the potential runoff path grid to the left of the centroid of the theoretical catchment area, the weighting coefficients remain unchanged or are slightly reduced, whereby... The total length of the test soil bed in the X direction is used to normalize the deviation value and avoid imbalance in the compensation range due to different soil bed sizes.

[0276] For X-direction deviation :

[0277] For the potential runoff path grid to the left of the centroid of the theoretical catchment area, the weighting coefficient compensation formula is as follows:

[0278] ;

[0279] For the potential runoff path grid to the right of the centroid of the theoretical catchment area, the weighting coefficient remains unchanged or is slightly reduced.

[0280] Y-direction compensation: The logic is completely consistent with the X-direction compensation, and the weighting coefficient compensation formula is as follows:

[0281] .

[0282] Furthermore, after compensation, the spatial centroid of the theoretical catchment area is recalculated. If the deviation between the new theoretical centroid and the actual centroid is smaller than the deviation before compensation, the compensation is effective. If the deviation is still large, the above steps can be repeated for iterative compensation. By further correcting the precipitation enhancement weighting coefficient through the deviation of the spatial centroid of the catchment area, the precipitation distribution not only matches the temporal characteristics of runoff but also the spatial characteristics of runoff, ultimately achieving precise optimization of precipitation simulation in both spatiotemporal dimensions.

[0283] Spatial calibration data are input into the fluid dynamics model to determine the effective wetting range projection data formed on the test soil surface by each rainfall execution unit, specifically including:

[0284] Spatial calibration data includes the installation coordinates, spray angle, and operating flow rate of each rainfall execution unit;

[0285] Spatial calibration data is input into a fluid dynamics model validated by experimental data. The simulation calculation yields the rainfall contribution rate distribution of each rainfall execution unit to the lower grid units under rated operating conditions, which serves as the effective wetting range projection data.

[0286] By optimizing the algorithm, multi-channel control commands are generated to drive all rainfall execution units, specifically including:

[0287] A constrained optimization problem is constructed with the activation duration of each rainfall execution unit during the simulation period as the decision variable and the objective being to minimize the sum of squares of the differences between the target rainfall data and the predicted rainfall data of all grid units. The problem is then solved using a linear programming algorithm.

[0288] The predicted rainfall data is obtained by matrix multiplication of the decision variables and the effective wet range projection data.

[0289] Specifically, let's take a small-scale experimental soil bed scenario as an example for calculation:

[0290] Test scenario preset:

[0291] Grid division: The surface of the test soil bed was divided into 4 grid cells, numbered as follows: ;

[0292] Rainfall execution unit: 3 sprinklers (rainfall execution units) are arranged, numbered as follows ;

[0293] Experimental constraints: Total rainfall duration Duration of operation of a single nozzle .

[0294] Input data:

[0295] Target rainfall data for all grids: Based on target rainfall parameters and potential runoff path data, the target rainfall for the four grids (unit: mm) was calculated:

[0296] ;

[0297] illustrate: and It belongs to the potential runoff path grid cell and is assigned a rainfall enhancement weight (target rainfall 3.6 mm). and For non-potential runoff path grid cells, assign a baseline weight (target rainfall 3.0 mm).

[0298] Effective wetting range projection data: The spatial calibration data of the three nozzles were input into the hydrodynamic model to obtain the rainfall contribution rate matrix of the nozzle grid. express right Rainfall contribution rate:

[0299] ;

[0300] illustrate: represent For every minute of work, it can provide Provides 0.4mm of rainfall. represent Unable to water .

[0301] Furthermore, let's assume The optimal opening duration is , , That is, the array of decision variables:

[0302] .

[0303] Predicted rainfall for grid i It equals the sum of the products of the contribution rate of all nozzles to it and their corresponding operating duration, expressed as a matrix:

[0304] ;

[0305] The specific calculation formula is as follows:

[0306] .

[0307] Furthermore, the optimization objective is to minimize the sum of squared errors between the predicted and target rainfall amounts for all grid cells:

[0308] ;

[0309] The constraints are:

[0310] ;

[0311] Furthermore, the above optimization problem is input into a linear programming solver, and the optimal solution is obtained through iterative calculation:

[0312] ;

[0313] Substituting the optimal duration into the rainfall prediction formula, we get:

[0314] ;

[0315] The optimal activation duration is converted into a pulse control command that the rainfall execution unit can recognize. The command includes three core elements: channel number, activation sequence, and duty cycle, as shown in the example below:

[0316] Control Channel Number Corresponding nozzle Activation time Time allocation (total duration 10 minutes) Duty cycle (optional in intermittent mode) Command format (pulse signal) Channel 1 Nozzle1 6min Turn on in 0-6 minutes, turn off in 6-10 minutes 60% (6-second cycle on / 4-second cycle off) High level (1) lasts for 6 minutes, low level (0) lasts for 4 minutes. Channel2 Nozzle2 6min Turn on in 0-6 minutes, turn off in 6-10 minutes 60% (6-second cycle on / 4-second cycle off) High level (1) lasts for 6 minutes, low level (0) lasts for 4 minutes. Channel 3 Nozzle3 3min Turn on in 0-3 minutes, turn off in 3-10 minutes 30% (3s cycle on / 7s cycle off) High level (1) lasts for 3 minutes, low level (0) lasts for 7 minutes.

[0317] Starting from each catchment unit, reverse source tracing calculations are performed on the theoretical flow direction data of all grid units, specifically including:

[0318] Starting from each catchment unit;

[0319] Iterate through the theoretical flow direction data of all upstream grid cells adjacent to the catchment unit, and mark the grid cells whose flow direction points to the catchment unit as potential runoff path cells;

[0320] Starting from the marked potential runoff path cell, continue marking the upstream grid cells of that runoff path cell;

[0321] The process is repeated recursively until the watershed is reached, and potential runoff path data is constructed from all marked potential runoff path units.

[0322] A rainfall simulation device for soil runoff leaching tests includes:

[0323] The terrain scanning module is used to collect three-dimensional scanning data of the test soil bed surface;

[0324] The data processing and control module, which is in communication with the terrain scanning module, is configured to perform the following operations:

[0325] Acquire three-dimensional scanning data of the test soil bed surface and generate a digital elevation model covering the test soil bed surface and composed of multiple grid cells;

[0326] After filling depressions in the digital elevation model, a multi-flow direction algorithm is used to calculate the theoretical flow direction data and theoretical cumulative flow data for each grid cell.

[0327] Grid cells with theoretical runoff accumulation data greater than or equal to a preset threshold are marked as catchment area cells;

[0328] Starting from each catchment unit, the theoretical flow direction data of all grid units are calculated in reverse to obtain the potential runoff path data of the corresponding catchment unit.

[0329] Acquire target rainfall parameters and spatial calibration data for each rainfall execution unit;

[0330] Spatial calibration data are input into the fluid dynamics model to determine the effective wetting range projection data formed on the surface of the test soil bed by each rainfall execution unit;

[0331] Based on the target rainfall parameters and potential runoff path data, the target rainfall data for each grid cell is calculated;

[0332] Based on the target rainfall data and effective wet range projection data of all grid cells, a multi-channel control command is generated to drive all rainfall execution units by solving an optimization algorithm.

[0333] The actual water flow direction data and actual flow accumulation data after the execution of multi-channel control commands are obtained, and the difference analysis results are obtained by comparing and analyzing them with the theoretical water flow direction data and theoretical flow accumulation data of the corresponding grid cells.

[0334] The preset threshold is dynamically adjusted based on the difference analysis results.

[0335] It also includes array-configured rainfall execution units, which are communicatively connected to the data processing and control module. Each rainfall execution unit includes a dripper and a solenoid valve mounted on the dripper. The multi-channel control command specifically refers to pulse signals that control the opening timing and duty cycle of each solenoid valve during the simulation period.

[0336] It also includes a sensor feedback module that communicates with the data processing and control module. This module is deployed below the test soil bed or on the critical path to collect data on the actual water flow direction and the actual cumulative flow, and then transmits the data to the data processing and control module.

[0337] Specifically, the sensor placement principle is as follows: sensors should be placed at key locations on the test soil bed, including:

[0338] Theoretical catchment unit, the endpoint of runoff convergence;

[0339] Key nodes of potential runoff paths, such as path bifurcations and watershed locations;

[0340] A standard grid area is used as a control.

[0341] The physical installation location of the sensor is bound to the grid cell coordinates of the digital elevation model in advance to ensure that each set of data collected corresponds to a specific grid cell.

[0342] Actual water flow direction data are collected using soil moisture sensor arrays, electromagnetic flow meters, or staining tracer methods (such as brilliant blue solution);

[0343] Actual runoff accumulation data is collected using weighing runoff buckets (placed at the outlet of the catchment area), ultrasonic flow meters, or grid water volume monitoring sensors.

[0344] The technical scope of this invention is not limited to the content described above. Those skilled in the art can make various modifications and variations to the above embodiments without departing from the technical concept of this invention, and all such modifications and variations should fall within the protection scope of this invention.

Claims

1. A rainfall simulation method for soil runoff leaching experiments, characterized in that, Includes the following steps: Acquire three-dimensional scanning data of the test soil bed surface and generate a digital elevation model covering the test soil bed surface and composed of multiple grid cells; After filling depressions in the digital elevation model, a multi-flow direction algorithm is used to calculate the theoretical flow direction data and theoretical cumulative flow data for each grid cell. Grid cells with theoretical runoff accumulation data greater than or equal to a preset threshold are marked as catchment area cells; Starting from each catchment unit, the theoretical flow direction data of all grid units are calculated in reverse to obtain the potential runoff path data of the corresponding catchment unit. Acquire target rainfall parameters and spatial calibration data for each rainfall execution unit; Spatial calibration data are input into the fluid dynamics model to determine the effective wetting range projection data formed on the surface of the test soil bed by each rainfall execution unit; Based on the target rainfall parameters and potential runoff path data, the target rainfall data for each grid cell is calculated; Based on the target rainfall data and effective wet range projection data of all grid cells, a multi-channel control command is generated to drive all rainfall execution units by solving an optimization algorithm. The actual water flow direction data and actual flow accumulation data after the execution of multi-channel control commands are obtained, and the difference analysis results are obtained by comparing and analyzing them with the theoretical water flow direction data and theoretical flow accumulation data of the corresponding grid cells. The preset threshold is dynamically adjusted based on the difference analysis results.

2. The rainfall simulation method for soil runoff leaching test according to claim 1, characterized in that: The preset thresholds are dynamically adjusted based on the results of the difference analysis, specifically including: Obtain the actual runoff peak formation time corresponding to the actual water flow direction data And the peak time of the theoretical runoff corresponding to the theoretical flow direction data. Calculate the relative time error; Calculate the preset threshold based on relative time error: ； in, The adjusted preset threshold, The preset threshold before adjustment This is a preset positive calibration coefficient.

3. The rainfall simulation method for soil runoff leaching test according to claim 1, characterized in that: Calculate the target rainfall data for each grid cell, specifically including: Assign rainfall enhancement weighting coefficients to the grid cells corresponding to potential runoff path data; Assign baseline weighting coefficients to the remaining grid cells; The target rainfall intensity parameter is multiplied by the rainfall enhancement weighting coefficient and the baseline weighting coefficient to obtain the target rainfall data for each grid cell.

4. The rainfall simulation method for soil runoff leaching test according to claim 3, characterized in that: After dynamically adjusting the preset threshold based on the difference analysis results, it also includes: Calculate the centroid of the actual catchment spatial distribution represented by the actual runoff accumulation data and the centroid of the theoretical catchment spatial distribution identified based on the adjusted preset threshold; Numerical compensation is performed on the rainfall enhancement weighting coefficient based on the coordinate deviation between the two centroids.

5. The rainfall simulation method for soil runoff leaching test according to claim 1, characterized in that: Spatial calibration data are input into the fluid dynamics model to determine the effective wetting range projection data formed on the test soil surface by each rainfall execution unit, specifically including: The spatial calibration data includes the installation coordinates, spray angle, and operating flow rate of each rainfall execution unit; Spatial calibration data is input into a fluid dynamics model validated by experimental data. The simulation calculation yields the rainfall contribution rate distribution of each rainfall execution unit to the lower grid units under rated operating conditions, which serves as the effective wetting range projection data.

6. The rainfall simulation method for soil runoff leaching test according to claim 1, characterized in that: By optimizing the algorithm, multi-channel control commands are generated to drive all rainfall execution units, specifically including: A constrained optimization problem is constructed with the activation duration of each rainfall execution unit during the simulation period as the decision variable and the objective being to minimize the sum of squares of the differences between the target rainfall data and the predicted rainfall data of all grid units. The problem is then solved using a linear programming algorithm. The predicted rainfall data is obtained by matrix multiplication of the decision variables and the effective wet range projection data.

7. The rainfall simulation method for soil runoff leaching test according to claim 1, characterized in that: Starting from each catchment unit, reverse source tracing calculations are performed on the theoretical flow direction data of all grid units, specifically including: Starting from each catchment unit; Iterate through the theoretical flow direction data of all upstream grid cells adjacent to the catchment unit, and mark the grid cells whose flow direction points to the catchment unit as potential runoff path cells; Starting from the marked potential runoff path cell, continue marking the upstream grid cells of that runoff path cell; The process is repeated recursively until the watershed is reached, and potential runoff path data is constructed from all marked potential runoff path units.

8. A rainfall simulation device for soil runoff leaching tests, comprising a rainfall simulation method for soil runoff leaching tests according to any one of claims 1-7, characterized in that, include: The terrain scanning module is used to collect three-dimensional scanning data of the test soil bed surface; The data processing and control module, which is in communication with the terrain scanning module, is configured to perform the following operations: Acquire three-dimensional scanning data of the test soil bed surface and generate a digital elevation model covering the test soil bed surface and composed of multiple grid cells; After filling depressions in the digital elevation model, a multi-flow direction algorithm is used to calculate the theoretical flow direction data and theoretical cumulative flow data for each grid cell. Grid cells with theoretical runoff accumulation data greater than or equal to a preset threshold are marked as catchment area cells; Starting from each catchment unit, the theoretical flow direction data of all grid units are calculated in reverse to obtain the potential runoff path data of the corresponding catchment unit. Acquire target rainfall parameters and spatial calibration data for each rainfall execution unit; Spatial calibration data are input into the fluid dynamics model to determine the effective wetting range projection data formed on the surface of the test soil bed by each rainfall execution unit; Based on the target rainfall parameters and potential runoff path data, the target rainfall data for each grid cell is calculated; Based on the target rainfall data and effective wet range projection data of all grid cells, a multi-channel control command is generated to drive all rainfall execution units by solving an optimization algorithm. The actual water flow direction data and actual flow accumulation data after the execution of multi-channel control commands are obtained, and the difference analysis results are obtained by comparing and analyzing them with the theoretical water flow direction data and theoretical flow accumulation data of the corresponding grid cells. The preset threshold is dynamically adjusted based on the difference analysis results.

9. A rainfall simulation device for soil runoff leaching test according to claim 8, characterized in that: It also includes an array of rainfall execution units, which are communicatively connected to the data processing and control module. Each rainfall execution unit includes a dripper and a solenoid valve mounted on the dripper. The multi-channel control command is specifically a pulse signal that controls the opening sequence and duty cycle of each solenoid valve during the simulation period.

10. A rainfall simulation device for soil runoff leaching test according to claim 8, characterized in that: It also includes a sensor feedback module that communicates with the data processing and control module. This module is installed below the test soil bed or on the critical path to collect data on the actual water flow direction and the actual cumulative flow, and then transmits the data to the data processing and control module.

Images