Method for calculating water distribution of soil material based on function fitting and mathematical integration

Abstract

This invention relates to the field of machine learning technology, specifically to a method for calculating soil water distribution based on function fitting and mathematical integration. The method includes the following steps: discrete data is generated by collecting voltage and three-dimensional coordinates; a soil state function is fitted via regression; the radial dry density is analyzed and the dominant direction is selected; a directional integral sequence is generated based on normal distance attenuation; the density difference between adjacent coordinates is calculated and connected point sets are extracted to generate structural weights; the soil state, water demand, and weight integration are coupled to obtain the water distribution calculation result; and optimized scheduling sequences are generated by smoothing out boundary regions. In this invention, discrete states are formed by splicing sensor electrical signals and three-dimensional coordinates; the soil state function is fitted via spatial regression; directional attenuation and density connectivity extraction characterize the extension trend and structural weights; and Gaussian integration couples soil state and water demand information. The water distribution result is more suitable for non-uniform distributions, smooths out boundary regions, and improves the targeting of spraying scheduling.

Description

Technical Field

[0001] This invention relates to the field of machine learning technology, and in particular to a method for calculating soil water distribution based on function fitting and mathematical integration. Background Technology

[0002] Machine learning technology primarily involves analyzing, modeling, and predicting data through algorithms and models. Its core aspects include methods such as supervised learning, unsupervised learning, reinforcement learning, and deep learning, as well as techniques such as feature extraction, model training, parameter optimization, loss function design, and data preprocessing. This technology is widely applied in scenarios such as image recognition, natural language processing, intelligent control, and predictive analytics. By constructing mathematical models and training algorithms, computer systems can automatically learn patterns from data and perform reasoning and decision-making.

[0003] Traditional methods for calculating soil water content based on function fitting and mathematical integration involve fitting a functional relationship between the physical properties and moisture content of soil materials during soil engineering and soil preparation, and then using mathematical integration to calculate the water content of various soil types. This method typically establishes a mathematical function model relating the dry and wet density, moisture content, and volume of the soil materials. Based on the soil material proportions and mix design requirements, it integrates and sums the water content of each type of soil material to obtain an estimate of the overall water content.

[0004] Traditional function fitting and mathematical integration for water distribution calculation rely on pre-set physical property relationships and overall mix parameters. When the soil material on site is affected by the accumulation morphology, local water content fluctuations, wet-dry density gradients, and spatial connectivity, a single function can hardly maintain its ability to express non-uniform distribution. The integration process often sums according to regular regions, making it difficult to distinguish between the dominant extension direction and local dense clumps. Water distribution estimation is prone to averaging, causing local water demand deviations to be masked, resulting in excessive or insufficient water distribution in the spraying section. Subsequent scheduling lacks targeting, affecting the uniformity of soil mix and the stability of the project. Summary of the Invention

[0005] To achieve the above objectives, the present invention adopts the following technical solution: a method for calculating soil water distribution based on function fitting and mathematical integration, comprising the following steps:

[0006] S1: Collect and map the dry density electrical signal, moisture content electrical signal and three-dimensional coordinate data of the soil and water material area to be prepared, construct a three-dimensional discrete dataset and input it into the radial basis function neural network to perform spatial distribution regression fitting, and construct the soil state function;

[0007] S2: Call the soil state function to analyze the dry density of the coordinate points of the directional radiation line, accumulate and calculate the directional cumulative scalar and filter the dominant directional line vector, obtain the integral space coordinate points of the soil and water material area to be distributed and calculate the normal distance between the dominant directional line vector and the dominant directional line vector, and inversely attenuate the mapping to generate the directional integral sequence.

[0008] S3: Obtain the discrete spatial coordinates of the soil and water material area to be distributed and input the soil state function mapping to calculate the dry density of the discrete coordinates. Analyze adjacent discrete spatial coordinates to obtain the density difference scalar. Combine the preset connectivity judgment threshold to extract the connected coordinate point set and sum it to generate a structural weight sequence.

[0009] S4: Obtain the spatial micro-element coordinates of the area to be water-distributed soil and input the soil state function to analyze the soil state sequence, and simultaneously analyze the foundation water demand sequence. Combine the directional integral sequence and the structural weight sequence with the Gaussian integral algorithm for coupled integral calculation to generate the water distribution calculation result.

[0010] S5: Compare the elements in the water distribution calculation result with the preset spraying equipment flow threshold and extract the out-of-bounds water distribution distribution area. Based on the out-of-bounds water distribution distribution area, perform numerical smoothing adjustment and optimization on the water distribution calculation result to generate an optimized water distribution scheduling sequence.

[0011] As a further embodiment of the present invention, the soil state function includes dry density distribution parameters, moisture content response parameters, and spatial location index parameters; the directional integral sequence includes directional accumulation results, dominant direction identifiers, and distance attenuation coefficients; the structural weight sequence includes density difference scalars, connectivity identifiers, and structural aggregation weights; the water distribution calculation results include micro-element water demand coefficients, integral coupling results, and spatial water distribution intensity; and the optimized water distribution scheduling sequence includes out-of-bounds distribution identifiers, smoothing correction results, and scheduling sorting indexes.

[0012] As a further aspect of the present invention, the specific steps of S1 are as follows:

[0013] S101: Collect the dry density electrical signal and moisture content electrical signal of the soil and water material area to be prepared and perform time series alignment. Compare the timestamp difference of the electrical signal with the preset time offset limit value point by point, remove data points associated with out-of-bounds timestamps, rearrange the index of the remaining timestamps and normalize the amplitude to obtain synchronous electrical signal pairs.

[0014] S102: Based on the synchronous electrical signal pair, obtain the three-dimensional coordinate data of the water and soil material area to be distributed, perform spatial mapping matching between the three-dimensional coordinate data and the synchronous electrical signal pair, compare the coordinate deviation with the preset adjacent boundary value, remove abnormal deviation items and splice the remaining coordinate items and related items to establish a three-dimensional discrete dataset.

[0015] S103: Call the three-dimensional discrete dataset and input it into the radial basis function neural network, calculate the radial distance between the data item and the center point of the hidden layer and obtain the neuron activation state parameter, calculate the weighted summation of the activation state parameter and the associated weight and perform spatial distribution regression fitting to construct the soil state function.

[0016] As a further aspect of the present invention, the specific steps of S2 are as follows:

[0017] S201: Obtain the coordinate points of the directional radiation lines of the soil and water material area to be distributed and input the soil state function for spatial mapping, solve the associated dry density parameter, and sequentially associate and bind the dry density parameter with the original spatial location points to obtain the radiation line density mapping set.

[0018] S202: Based on the radiation line density mapping set, extract the dry density parameters corresponding to the radiation line direction and perform linear accumulation to obtain multiple radiation direction associated direction accumulation scalars. Compare the accumulation scalars with the preset direction screening extreme values, remove scalar items that have not reached the direction screening extreme values, and generate the dominant direction line vector.

[0019] S203: Collect the integral spatial coordinates of the soil and water material area to be distributed and calculate the normal distance with the dominant direction line vector. Extract the vertical span corresponding to the dominant direction line vector from multiple coordinate points. Call the preset attenuation ratio coefficient to perform inverse proportional attenuation mapping on the vertical span and merge it according to the spatial index to obtain the direction integral sequence.

[0020] As a further aspect of the present invention, the specific steps of S3 are as follows:

[0021] S301: Obtain discrete spatial coordinate points of the soil and water material area to be prepared, input the three-dimensional coordinate parameters of the multi-spatial coordinate points into the soil state function for mapping and solving, extract the associated dry density of the multi-coordinate points and arrange and recombine them in sequence according to the coordinate index order to establish a discrete density mapping set.

[0022] S302: Based on the discrete density mapping set, extract the associated density of adjacent discrete spatial coordinate points and perform item-by-item subtraction operation to solve the density difference scalar between adjacent coordinate points and compare it with the preset connectivity judgment threshold. Extract the scalar associated coordinate point index that has not exceeded the boundary and aggregate it to obtain the connected coordinate point set.

[0023] S303: Call the connected coordinate point set and extract the associated density difference scalar of the multi-connected spatial coordinate points. Accumulate and sum all density difference scalars to solve the density change accumulation parameter of the corresponding spatial connected region. Reorganize and encode the sequence according to the corresponding connected index order to generate the structure weight sequence.

[0024] As a further aspect of the present invention, the connectivity determination threshold is obtained by acquiring the soil reference sample attributes, extracting the dry density parameters of adjacent points of the reference sample and performing a subtraction operation to obtain the reference difference sequence, calculating the standard deviation of the elements in the sequence to solve the reference discrete parameters, and multiplying the discrete parameters with a preset confidence coefficient to determine the threshold.

[0025] As a further aspect of the present invention, the specific steps of S4 are as follows:

[0026] S401: Obtain the spatial micro-element coordinates of the soil and water material area to be mixed and input the soil state function for mapping and solution. Extract the corresponding spatial point associated dry density and initial moisture content. Perform serialization combination encoding and data splicing according to the spatial distribution index of the micro-element coordinates to obtain the soil state sequence.

[0027] S402: Based on the soil state sequence, extract the initial moisture content and dry density parameters associated with multiple spatial points, subtract the preset target moisture content scalar from the initial moisture content parameter, and perform algebraic calculations by multiplying the obtained difference and the dry density parameter with the preset infinitesimal volume constant to establish the basic water demand sequence.

[0028] S403: Call the basic water demand sequence, the directional integral sequence, and the structural weight sequence, and perform point-to-point product calculation to obtain multivariate coupling parameters. Assign discrete quadrature elements and integral quadrature coefficients to the multivariate coupling parameters and perform weighted accumulation item by item to generate water distribution calculation results.

[0029] As a further aspect of the present invention, the specific steps of S5 are as follows:

[0030] S501: Compare the elements in the water distribution calculation result with the preset spray equipment flow threshold, filter the water distribution elements whose values ​​exceed the spray equipment flow threshold, extract the spatial distribution coordinate index of the corresponding out-of-bounds elements and perform spatial topology reorganization and union to obtain the out-of-bounds water distribution coordinate set.

[0031] S502: Call the out-of-bounds water allocation coordinate set to extract the water allocation quantity corresponding to the associated spatial location and the water allocation quantity parameter of the adjacent coordinate. Calculate the summation and arithmetic mean of the water allocation quantity elements and the water allocation quantity parameters of the adjacent coordinate, and replace the water allocation quantity corresponding to the original out-of-bounds spatial location to establish a smooth water allocation value set.

[0032] S503: Extract multiple arithmetic averages and non-boundary water allocation elements from the smooth water allocation value set, serialize and merge them according to the order of the multi-element associated spatial coordinate index, perform time-series allocation and scheduling site encoding mapping on the merged element sequence, and generate an optimized water allocation scheduling sequence.

[0033] As a further aspect of the present invention, the flow rate threshold of the spraying equipment is determined by calling the operating parameters of the water supply equipment, extracting the ultimate power parameter of the water pump and the cross-sectional area parameter of the nozzle, multiplying the ultimate power parameter with a preset hydraulic coefficient to obtain the extreme kinetic energy scalar, and performing a ratio calculation between the kinetic energy scalar and the cross-sectional area parameter.

[0034] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0035] In this invention, a discrete state description is formed by splicing sensor electrical signals with three-dimensional coordinates. A soil state function is established by using spatial regression fitting. Combined with directional radiation analysis, normal distance attenuation mapping, and density difference connectivity extraction, the integral space synchronously reflects the dominant extension trend and local structural weights. The infinitesimal state, basic water demand, and spatial weight information are coupled in the Gaussian integral to obtain water distribution results that better fit the non-uniform soil distribution. Furthermore, the water distribution areas that exceed the boundaries are smoothly adjusted to improve the targeting of spraying scheduling, reduce the risk of local over- or under-spraying, and ensure the uniformity of the mixture. Attached Figure Description

[0036] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0037] Figure 1 This is a schematic diagram of the steps of the present invention;

[0038] Figure 2 This is a detailed schematic diagram of S1 of the present invention;

[0039] Figure 3 This is a detailed schematic diagram of S2 of the present invention;

[0040] Figure 4 This is a detailed schematic diagram of S3 of the present invention;

[0041] Figure 5 This is a detailed schematic diagram of S4 of the present invention;

[0042] Figure 6 This is a detailed schematic diagram of S5 of the present invention. Detailed Implementation

[0043] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0044] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0045] Please see Figure 1 This invention provides a method for calculating soil water distribution based on function fitting and mathematical integration, including the following steps:

[0046] S1: Collect and map the dry density electrical signal, moisture content electrical signal and three-dimensional coordinate data of the soil and water material area to be prepared, construct a three-dimensional discrete dataset and input it into the radial basis function neural network to perform spatial distribution regression fitting, and construct the soil state function;

[0047] S2: Call the soil state function to analyze the dry density of the coordinate points of the directional radiation line, accumulate and calculate the directional cumulative scalar and filter the dominant directional line vector, obtain the integral space coordinate points of the soil and water material area to be distributed and calculate the normal distance between the dominant directional line vector and the inverse proportional decay mapping, and generate the directional integral sequence.

[0048] S3: Obtain the discrete spatial coordinates of the soil and water material area to be distributed and input the soil state function mapping to calculate the dry density of the discrete coordinates. Analyze adjacent discrete spatial coordinates to obtain the density difference scalar. Combine the preset connectivity judgment threshold to extract the connected coordinate point set and sum it to generate the structural weight sequence.

[0049] S4: Obtain the spatial micro-element coordinates of the soil and water material area to be distributed and input the soil state function to analyze the soil state sequence. At the same time, analyze the foundation water demand sequence, and combine the directional integral sequence and the structural weight sequence to input the Gaussian integral algorithm for coupled integral calculation to generate the water distribution calculation result.

[0050] S5: Compare the elements in the water distribution calculation result with the preset spray equipment flow threshold and extract the out-of-bounds water distribution distribution area. Based on the out-of-bounds water distribution distribution area, perform numerical smoothing adjustment and optimization on the water distribution calculation result to generate an optimized water distribution scheduling sequence.

[0051] The soil state function includes dry density distribution parameters, moisture content response parameters, and spatial location index parameters. The directional integral sequence includes directional accumulation results, dominant direction identifiers, and distance attenuation coefficients. The structural weight sequence includes density difference scalars, connectivity identifiers, and structural aggregation weights. The water distribution calculation results include micro-element water demand coefficients, integral coupling results, and spatial water distribution intensity. The optimized water distribution scheduling sequence includes out-of-bounds distribution identifiers, smoothing correction results, and scheduling sorting indexes.

[0052] Please see Figure 2 The specific steps of S1 are as follows:

[0053] S101: Collect the dry density electrical signal and moisture content electrical signal of the soil and water material area to be prepared and perform time series alignment. Compare the timestamp difference of the electrical signal with the preset time offset limit value point by point, remove data points associated with out-of-bounds timestamps, rearrange the index of the remaining timestamps and normalize the amplitude to obtain synchronous electrical signal pairs.

[0054] Dry density and moisture content electrical signals were acquired using a frequency domain reflection probe array buried 500 mm below the surface of the soil and water material area. The sampling frequency was set to 100 Hz. These two sets of dynamic voltage signals, along with hardware clock timestamps, were synchronously extracted and time-series aligned. The timestamps of the dry density and moisture content electrical signal data frames were extracted and subtracted to obtain the timestamp difference. This timestamp difference was then compared point-by-point with a preset time offset limit of 5 milliseconds, determined experimentally based on the minimum response delay of soil moisture migration. When the absolute value of the timestamp difference exceeded 5 milliseconds, the data points with out-of-bounds timestamps were discarded. The remaining data points were then re-indexed in ascending order of time sequence. Subsequently, the global maximum and minimum amplitudes of the dry density and moisture content electrical signals are extracted. The minimum amplitude is subtracted from the current signal amplitude to obtain the absolute increment. This absolute increment is then divided by the difference between the maximum and minimum amplitudes to obtain the range quotient. Amplitude normalization is then performed to obtain the synchronization electrical signal pair. For example, the current amplitude of the dry density electrical signal is 1.5 volts, corresponding to a timestamp of 10000 milliseconds, and the current amplitude of the moisture content electrical signal is 2.0 volts, with a timestamp of 10002 milliseconds. The timestamp deviation is 2 milliseconds, which is less than the 5 millisecond limit, so this point is retained. The dry density signal has a global maximum value of 5.0 volts and a minimum value of 1.0 volt. Subtracting 1.0 volt from 1.5 volts yields 0.5 volts. Dividing this by the range of 4.0 volts gives a normalized amplitude of 0.125 for the dry density. Similarly, the moisture content signal has a maximum value of 6.0 volts and a minimum value of 1.0 volt. Subtracting 1.0 volt from 2.0 volts yields 1.0 volt. Dividing this by the range of 5.0 volts gives a normalized value of 0.2. These values ​​are combined to form a synchronization signal pair containing the values ​​0.125 and 0.2.

[0055] S102: Based on the synchronous electrical signal pair, obtain the three-dimensional coordinate data of the water and soil material area to be distributed, perform spatial mapping matching between the three-dimensional coordinate data and the synchronous electrical signal pair, compare the coordinate deviation with the preset adjacent boundary value, remove abnormal deviation items and splice the remaining coordinate items and related items to establish a three-dimensional discrete dataset.

[0056] Based on the spatial distribution of the acquired synchronous electrical signals, a laser differential positioning receiver is used to acquire three-dimensional coordinate data of the soil and water material distribution area. Positioning information, including eastward, northward, and elevation parameters, is extracted from the dataset. Spatial mapping and matching are performed between the three-dimensional coordinate data and the synchronous electrical signals. The coordinates of the physical sensor calibration positions bound to the synchronous electrical signals are extracted. The eastward, northward, and elevation parameters of the sensor calibration positions are subtracted from their corresponding values ​​in the acquired three-dimensional coordinate data to obtain three spatial direction differences. These three spatial direction differences are squared and summed. The square root of the sum is then extracted to obtain the coordinate deviation. This coordinate deviation is compared with a preset proximity limit value, which is set to 50 mm based on the fixed error range of the positioning device. When the coordinate deviation is greater than 50 mm, abnormal deviation items are removed. For data nodes with deviation values ​​less than or equal to this limit, the remaining coordinate items and associated items are concatenated through a sequence merging operation to establish a three-dimensional discrete dataset containing three spatial attributes and two signal attributes.

[0057] Table 1. Structural Elements of a 3D Discrete Dataset

[0058]

[0059] Table 1 shows the structural elements of the three-dimensional discrete dataset. As shown in Table 1, substituting the normalized dry density value of 0.125 and the normalized moisture content value of 0.2 obtained from the previous node, the associated sensor calibration location has an east coordinate of 1000 mm, a north coordinate of 2000 mm, and an elevation of 500 mm. The measured three-dimensional coordinate data are 1020 mm east, 1990 mm north, and an elevation of 510 mm. Subtracting the measured coordinates from the calibration coordinates yields an east difference of 20 mm, a north difference of -10 mm, and an elevation difference of 10 mm. Squaring these differences yields 400, 100, and 100 respectively. Summing these values ​​gives 600. Calculating the square root gives a coordinate deviation of approximately 24.49 mm. Since the coordinate deviation of 24.49 mm does not exceed the 50 mm limit, 1020 mm, 1990 mm, and 510 mm are extracted along with 0.125 and 0.2 to form a 5-dimensional vector, which is then merged into the 3D discrete dataset. This execution node involves calculating the Euclidean deviation based on the coordinate parameters, thereby triggering subsequent operations such as data item retention and matrix concatenation.

[0060] S103: Call the three-dimensional discrete dataset and input it into the radial basis function neural network. Calculate the radial distance between the data item and the center point of the hidden layer and obtain the neuron activation state parameter. Calculate the weighted sum of the activation state parameter and the associated weight and perform spatial distribution regression fitting to construct the soil state function.

[0061] A three-dimensional discrete dataset is input into a radial basis function neural network. This network structure includes an input layer composed of the aforementioned 5-dimensional vector parameters, a hidden layer consisting of 256 nodes using Gaussian kernel activation functions, and an output layer that outputs a single predicted value using a linear superposition mechanism. The specific values ​​of the 5-dimensional vector parameters in the input data items are extracted, and the corresponding 5-dimensional coordinate parameters of the hidden layer center point are obtained. The center point coordinates are pre-extracted from historical soil calibration samples using a grid search algorithm. The radial distance between the data item and the hidden layer center point is calculated by subtracting each parameter of the input data from the corresponding parameters of the hidden layer center point point by point. The resulting 5 feature differences are summed by squaring, and the square root of the sum is extracted to generate the radial distance parameter. For example, the previously obtained eastward coordinates (1020 mm), northward coordinates (1990 mm), elevation (510 mm), normalized dry density (0.125), and normalized moisture content (0.2) are used. For a pre-clustered center point, its coordinates are 1018 mm east, 1992 mm north, elevation 511 mm, normalized dry density 0.125, and normalized moisture content 0.2. Subtracting these points one by one yields a difference of 2 mm east, -2 mm north, -1 mm elevation, and 0 for two signal attributes. Squaring these differences yields 4, 4, 1, 0, and 0 respectively. Summing these values ​​gives 9, and extracting the square root yields a radial distance parameter of 3 mm. Dividing this 3 mm radial distance parameter by a smoothing control parameter (set to 3.0 based on data divergence) and squaring the resulting quotient 1.0 yields 1.0. Then, the negative 1 power of the natural constant is used to calculate the neuron activation state parameter, approximately 0.368. The local weight coefficient leading to the output layer from the hidden node is extracted and set to 0.4. The local response value of 0.1472 is obtained by multiplying the single node activation state parameter of 0.368 with the corresponding local weight coefficient of 0.4. The response values ​​of 256 hidden nodes are summed. The cumulative sum of the weak activation responses of all other nodes is set to 1.2. A bias constant of 0.5 is added to the sum for spatial distribution regression fitting. Finally, the total evaluation value of the soil state function is 1.8472.

[0062] Please see Figure 3 The specific steps of S2 are as follows:

[0063] S201: Obtain the coordinates of the directional radiation lines of the soil and water material area to be distributed and input the soil state function for spatial mapping. Solve the associated dry density parameter, and sequentially associate and bind the dry density parameter with the original spatial location points to obtain the radiation line density mapping set.

[0064] Multiple radially arranged spatial location data points were collected on the surface of the soil and water material area using a laser scanner at a 5-degree angular resolution and a 100-millimeter step size. These data points generated directional radial coordinate points containing east, north, and elevation coordinates. The acquired directional radial coordinate points were then sequentially input into a previously constructed soil state function for spatial mapping calculation. The east, north, and elevation coordinate values ​​of the directional radial coordinate points were extracted and constructed as a three-dimensional input feature vector. This vector was then input into the previously constructed soil state function based on a radial basis function (RBF) neural network architecture for nonlinear spatial mapping. The RBF activation values ​​between the input vector and the center points of neurons in each hidden layer were calculated using hidden layers, and a nonlinear weighted sum was performed using the output layer weight matrix to directly output a preliminary density fluctuation value with the physical dimension of "kilograms per cubic meter". For example, the east coordinate of a certain directional radial coordinate point was extracted as 1050 mm, the north coordinate as 2020 mm, and the elevation as 520 mm. After inputting the data into the soil state function (RBF neural network) for forward propagation calculation, the network output directly maps and generates a preliminary density fluctuation value of approximately 80.50 kg / m³. This preliminary density fluctuation value is then summed with the basic dry density constant, which is determined to be 1400.00 kg / m³ using an on-site sampling and drying method. Finally, the associated dry density parameter corresponding to the current coordinate point is calculated to be 1480.50 kg / m³. Next, the original spatial location data of the radiation line coordinates in the current direction are extracted. The east coordinates (1050 mm), north coordinates (2020 mm), elevation coordinates (520 mm), and the calculated associated dry density parameter (1480.50 kg / m³) are combined into a one-dimensional array. This array is then serialized and bound according to the sampling time sequence. Multiple one-dimensional arrays are stored in a preset database, ultimately obtaining a radiation line density mapping set that completely covers the sampling area.

[0065] S202: Based on the radiation line density mapping set, extract the dry density parameters corresponding to the radiation line direction and perform linear accumulation to obtain the cumulative scalar of multiple radiation direction related directions. Compare the cumulative scalar with the preset direction screening extreme value, remove the scalar terms that have not reached the direction screening extreme value, and generate the dominant direction line vector.

[0066] The generated radiation line density map set is called, and the data is classified and extracted according to the radiation angle. All one-dimensional arrays under the same radiation angle are clustered into a single directional subset. For each directional subset, the dry density parameters corresponding to all radiation directions contained therein are extracted, and these dry density parameters are linearly summed to obtain the cumulative scalar of multiple radiation direction associations covering all radiation directions. For example, if a radiation direction angle is set to 45 degrees, the dry density parameters corresponding to 3 sampling points in this directional subset are extracted. These 3 parameters are substituted into the previous solution logic to obtain the specific values ​​of 1480.50 kg / m³, 1510.20 kg / m³, and 1495.30 kg / m³, respectively. The three dry density parameters mentioned above are continuously added together, i.e., 1480.50, 1510.20, and 1495.30 are summed to obtain a cumulative scalar value associated with the 45-degree radiation direction, which is 4486.00 kg / m³. Subsequently, the calculated cumulative scalar values ​​for each direction are compared one by one with preset directional screening extreme values. The preset directional screening extreme values ​​are set based on the minimum cumulative bearing capacity required for the target compacted area, and are determined and set at 4000.00 kg / m³ through multiple geotechnical compressive strength tests. The cumulative scalar value of 4486.00 kg / m³ obtained in the above example is judged. Since its value is greater than the preset extreme value of 4000.00 kg / m³, it is judged to meet the standard and retained. If the cumulative scalar value in a certain direction is 3800.00 kg / m³, since it is less than the screening extreme value, a rejection operation is performed to completely delete the scalar item and its associated coordinate data that did not meet the screening extreme value of the direction. After rejecting the non-compliant directions, the radial direction with the largest cumulative scalar value among all the remaining retained directions is extracted. The 3D spatial direction vector corresponding to the largest scalar value is extracted as the dominant direction line vector representing the densest direction of the soil. The value of 4486.00 kg / m³ corresponding to the 45-degree direction in this example is set as the global maximum value. The geometric vector corresponding to the 45-degree direction is then generated as the dominant direction line vector.

[0067] S203: Collect the spatial coordinates of the water and soil material area to be distributed and calculate the normal distance with the dominant direction line vector. Extract the vertical span corresponding to the dominant direction line vector from multiple coordinate points. Call the preset attenuation ratio coefficient to map the vertical span attenuation inversely and merge it according to the spatial index to obtain the direction integral sequence.

[0068] The integrated spatial coordinates of the soil and water material distribution area were acquired using a distributed lidar system with a spatial resolution of 100 mm. The absolute east-west, absolute north-south, and absolute elevation coordinates of the target spatial location were extracted. The normal distance between these integrated spatial coordinates and the dominant direction line vector generated in the previous steps was calculated. Specifically, the spatial vector connecting the integrated spatial coordinates to the starting point of the dominant direction line vector was extracted. The cross product matrix of this spatial vector and the dominant direction line vector was calculated. The magnitude of this cross product matrix was then calculated and divided by the magnitude of the dominant direction line vector itself, thereby extracting the vertical span from multiple coordinates to the dominant direction line vector. For example, the absolute east-west coordinates of a certain integrated spatial coordinates point were extracted as 1100 mm, the absolute north-south coordinates as 2050 mm, and the absolute elevation coordinates as 530 mm. Combining the specific coordinate parameters of the 45-degree dominant direction line vector mentioned earlier, after the cross product and magnitude division operations, the vertical span from this integrated spatial coordinates point to the dominant direction line vector was calculated to be 40 mm. After obtaining the vertical span, a preset attenuation ratio coefficient is used to perform inverse proportional attenuation mapping on the vertical span. The preset attenuation ratio coefficient is set based on the diffusion resistance model of water lateral infiltration in the soil, and its specific value is determined to be 120 square millimeters. The preset attenuation ratio coefficient is divided by the calculated vertical span, and a constant 1 is added to obtain the attenuation weight value. Substituting the example data, the attenuation ratio coefficient of 120 square millimeters is divided by the vertical span of 40 millimeters, resulting in a quotient of 3. This quotient is then added to the constant 1 to obtain an attenuation weight value of 4. This attenuation weight value is multiplied by the initial moisture content parameter of the integral space coordinate point to complete the attenuation mapping of the coordinate point. Subsequently, the 3D index label of the integral space coordinate point is extracted, and the attenuated data is appended and merged layer by layer according to the spatial index label to finally construct a directional integral sequence reflecting the moisture distribution state around the dominant direction.

[0069] Please see Figure 4 The specific steps of S3 are as follows:

[0070] S301: Obtain the discrete spatial coordinates of the soil and water material area to be prepared, input the three-dimensional coordinate parameters of the multi-spatial coordinates into the soil state function for mapping and solution, extract the associated dry density of the multi-coordinates and arrange and recombine them in sequence according to the coordinate index order to establish a discrete density mapping set.

[0071] Multiple discrete spatial coordinate points were collected in the soil and water material preparation area using a high-precision laser rangefinder with a grid spacing of 200 mm. The absolute coordinates (east, north, and elevation) of each sampling location were extracted. The extracted 3D coordinate parameters of the multiple spatial coordinate points were used as input vectors, which were then sequentially input into a soil state function based on a radial basis function (RBF) neural network for nonlinear mapping. Through nonlinear activation of the hidden layer neurons and fully connected mapping of the output layer weights, a fluctuation density correction with correct physical dimensions (kg / m³) was directly output. This fluctuation density correction was then summed with the locally determined basic dry density constant to obtain the associated dry density of the multiple coordinate points. Substituting the previously associated data into the calculation, the absolute coordinates (east, 1060 mm, north, 2030 mm, and elevation) of a certain discrete spatial coordinate point were extracted. After nonlinear mapping processing using the soil state function (RBF neural network), the fluctuation density correction was directly and analytically calculated to be approximately 106.12 kg / m³. This 106.12 kg / m³ was then added to the aforementioned basic dry density constant of 1400.00 kg / m³, yielding a correlated dry density of 1506.12 kg / m³ for the given coordinate point. After obtaining the correlated dry density, the hardware acquisition timestamp and the device's built-in 3D spatial array sorting number were extracted from the current discrete spatial coordinate point. The correlated dry density value was then bound to this sorting number as a key-value pair. Strictly following the ascending order of coordinate index numbers, all bound data nodes were sequentially rearranged and stored in a pre-defined continuous storage block, ultimately establishing a discrete density mapping set covering the entire survey area.

[0072] S302: Based on the discrete density mapping set, extract the associated density of adjacent discrete spatial coordinate points and perform item-by-item subtraction operation to solve the density difference scalar between adjacent coordinate points and compare it with the preset connectivity judgment threshold. Extract the index of the scalar associated coordinate points that have not exceeded the boundary and aggregate them to obtain the connected coordinate point set.

[0073] Based on the generated discrete density mapping set, adjacent discrete spatial coordinate points within a preset physical distance range are extracted according to the sorting number of the 3D spatial array. The associated dry density values ​​of the current reference node and its surrounding adjacent nodes are read. The associated dry densities of the acquired adjacent discrete spatial coordinate points are subtracted item by item, and the absolute value parameter of the calculation result is extracted to obtain the density difference scalar between adjacent coordinate points. The density difference scalar between adjacent coordinate points is compared with the preset connectivity judgment threshold. The preset connectivity judgment threshold is set based on the limiting density gradient limit for maintaining continuous capillary permeability of water inside the soil, and is determined and set to 25.00 kg / m³ through multiple water migration and permeability physical model tests. Substituting into the actual calculation example, a discrete spatial coordinate point calculated above is extracted as the current reference node, and its associated dry density is 1506.12 kg / m³. The first adjacent discrete spatial coordinate point is extracted along the north direction of the survey area, and the associated dry density of this adjacent point is 1490.50 kg / m³. Subtracting 1506.12 kg / m³ from 1490.50 kg / m³ and extracting the absolute value yields a density difference scalar of 15.62 kg / m³ between adjacent coordinate points. Data comparison shows this density difference scalar of 15.62 kg / m³ is less than the preset connectivity threshold of 25.00 kg / m³, indicating a good hydraulic connection between the two spatial nodes, and this scalar is considered to be within the bounds. For nodes deemed to be outside the bounds, their network connection with the reference node is directly severed. For scalars within the bounds, the 3D spatial array sorting index corresponding to the associated coordinate points is extracted. Using a disjoint-set data structure path compression algorithm, the index of the current reference node and the indices of adjacent nodes satisfying the connectivity condition are merged into the same cluster root node. Data merging and classification aggregation operations are performed, traversing all coordinate points to remove duplicate isolated node features, ultimately establishing a set of connected coordinate points with similar density features and spatial interconnection.

[0074] S303: Call the connected coordinate point set and extract the associated density difference scalar of the multi-connected spatial coordinate points. Accumulate and sum all density difference scalars to solve the density change accumulation parameter of the corresponding spatial connected region. Reorganize and encode the sequence according to the corresponding connected index order to generate the structure weight sequence.

[0075] The generated set of connected coordinate points is invoked. For each independent connected component block within the set, all interconnected multi-connected spatial coordinate points within that block are traversed and read. The pre-calculated correlation density difference scalar between each pair of adjacent nodes is extracted. The extracted density difference scalars within the current connected block are continuously summed to obtain the density variation accumulation parameter characterizing the overall non-uniformity of the corresponding spatial connected region. Substituting the data obtained in the preceding steps, actual business calculations are performed. It is set that a specific spatial connected region contains four pairs of connected node relationships. The correlation density difference scalars are extracted sequentially. The first scalar is 15.62 kg / m³, obtained earlier. The remaining three scalars are derived through the same underlying calculation logic, with specific values ​​of 12.30 kg / m³, 8.50 kg / m³, and 20.10 kg / m³, respectively. The four values ​​mentioned above are continuously added together, i.e., the four floating-point numbers 15.62, 12.30, 8.50, and 20.10 are summed to calculate the cumulative density variation parameter of the corresponding spatially connected region, which is 56.52 kg / m³. This parameter directly reflects the degree of structural loosening and permeability within the current connected region. After solving, the connectivity index number corresponding to each connected node in the original spatial dataset is extracted. Using the calculated cumulative density variation parameter as the local weight base, the underlying data nodes are sequentially rearranged and reorganized according to the topological extension traversal order from the local nodes to the global network architecture based on the corresponding connectivity index number. The calculated cumulative density variation parameter is accurately distributed to each mapping node according to the network index path, generating a 1D feature array carrying spatial distribution sorting and density variation weight attributes. Finally, the numerical conversion and encoding of the entire target compaction area is completed, generating a structural weight sequence for subsequent precise water allocation path planning guidance.

[0076] Please see Figure 5 The specific steps of S4 are as follows:

[0077] S401: Obtain the spatial micro-element coordinates of the soil and water material area to be mixed and input the soil state function for mapping and solution. Extract the corresponding spatial point associated dry density and initial moisture content. Perform serialization combination encoding and data splicing according to the spatial distribution index of the micro-element coordinates to obtain the soil state sequence.

[0078] A mobile detection platform equipped with high-resolution laser scanning hardware and a soil moisture dielectric sensor is used to perform gridded scanning sampling in the area to be prepared with soil and water, with equidistant spatial steps of 50 mm, to acquire absolute coordinate data of multiple spatial micro-elements covering the entire area. The three-dimensional coordinate parameters of the current micro-element location are extracted, including an absolute coordinate of 1060 mm to the east, 2030 mm to the north, and an absolute elevation coordinate of 520 mm. These three-dimensional coordinate parameters are then input into a pre-configured soil state function for nonlinear spatial mapping solution. This soil state function is constructed using a ternary polynomial interpolation algorithm. The three-dimensional positional deviations are obtained by subtracting the input micro-element's three-dimensional coordinates from the coordinates of the spatial reference point. The deviations in each of the three dimensions are then multiplied by their corresponding spatial heterogeneity attenuation coefficients and summed. The resulting compensation constant is directly added to the reference soil property value, thereby accurately extracting the dry density and initial moisture content associated with the corresponding spatial point location. Substituting the actual coordinate system data extracted in the previous steps, calculations were performed. The absolute coordinates to the east (1060 mm), to the north (2030 mm), and the absolute elevation coordinates (520 mm) were solved using the soil state function. The associated dry density of this micro-element coordinate was obtained as 1506.12 kg / m³, and the initial moisture content was 12.50%. After obtaining the attribute data, the three-dimensional spatial distribution index number of this micro-element coordinate in the global survey area was extracted and converted into a one-dimensional linear digital code 152005. Following the sequential traversal order of the micro-element coordinate spatial distribution index from smallest to largest, the distribution index number, associated dry density value, and initial moisture content value were combined and encoded, and stored in a contiguous memory segment to complete data splicing. Finally, a soil state sequence covering the micro-element information of the entire area was obtained.

[0079] S402: Based on the soil state sequence, extract the initial moisture content and dry density parameters associated with multiple spatial points, subtract the preset target moisture content scalar from the initial moisture content parameter, and perform algebraic calculations by multiplying the obtained difference and the dry density parameter with the preset infinitesimal volume constant to establish the basic water demand sequence.

[0080] Based on the generated soil state sequence, each independent micro-element data frame is traversed using a memory addressing pointer to accurately extract the initial moisture content and dry density parameters associated with the corresponding multi-spatial points. The preset target moisture content scalar corresponding to the current soil type is read from the preset standard construction process database. This preset target moisture content scalar is determined by a standard compaction test, with the optimal moisture content value set at 18.00%. The preset target moisture content scalar is directly subtracted from the extracted initial moisture content parameter to calculate the moisture content gap of the current micro-element soil. Subsequently, the dry density parameter of the corresponding micro-element and the preset micro-element volume constant are called. The preset micro-element volume constant is calculated based on three consecutive multiplications at a previous grid scanning step size of 50 mm, and its value is 0.000125 cubic meters. The calculated moisture content difference, the associated dry density parameter of the micro-element, and the preset micro-element volume constant are multiplied algebraically to obtain the absolute water mass required to reach the target state for the current spatial single micro-element. Substituting the data obtained earlier into a logical deduction, the initial moisture content of a certain spatial micro-element is extracted to be 12.50%, with a corresponding dry density of 1506.12 kg / m³. Subtracting the preset target moisture content of 18.00% from the initial moisture content of 12.50% yields a moisture content difference of 5.50%, which is converted to a floating-point number of 0.055 for calculation. Multiplying the floating-point difference of 0.055, the dry density of 1506.12 kg / m³, and the preset micro-element volume constant of 0.000125 m³ directly, the absolute water demand is calculated to be 0.01035 kg. Following the existing spatial distribution index order in the soil state sequence, the absolute water demand values ​​calculated individually for each micro-element are arranged and written in an orderly manner to establish a basic water demand sequence reflecting the distribution of water demand. The execution node of this operation logic is to calculate the absolute water demand at a single point based on the difference in water content and the infinitesimal parameters, thereby triggering the subsequent operation of constructing the water demand sequence.

[0081] S403: Call the basic water demand sequence, directional integral sequence and structural weight sequence and perform point-to-point product calculation to obtain multivariable coupling parameters. Assign discrete quadrature elements and integral quadrature coefficients to the multivariable coupling parameters and perform weighted accumulation item by item to generate water distribution calculation results.

[0082] The generated basic water demand sequence is invoked, and the stored directional integral sequence and the structural weight sequence constructed in the previous steps are read simultaneously. The directional integral sequence is generated by calculating the gradient of the three-dimensional gravitational potential energy and capillary seepage potential energy within the soil, representing the transport loss of water in a specific spatial orientation, and the current micro-element directional integral constant is extracted as 1.05. The structural weight sequence generated earlier includes the density variation weight attribute of the current micro-element node, with a corresponding value of 1.12. The water demand value of the current micro-element node in the basic water demand sequence is multiplied point-to-point with the parameter value at its aligned position in the directional integral sequence and the structural weight sequence to obtain a multivariate coupling parameter that comprehensively considers seepage loss and structural compactness. The basic water demand of 0.01035 kg obtained in the previous example is multiplied term by term by the directional integral constant 1.05 and the structural weight value 1.12 to calculate the multivariate coupling parameter of the micro-element node as 0.01217 kg. Based on the principle of numerical integration, discrete quadrature elements and integral quadrature coefficients are assigned to the extracted multivariate coupled parameters. The discrete quadrature elements are set to a normalized spatial distance of 1.0 between adjacent data sampling points, and the integral quadrature coefficients are uniformly set to a weighted multiplier of 1.0 for internal nodes according to Simpson's integral rule. The multivariate coupled parameters are multiplied by the assigned discrete quadrature elements and integral quadrature coefficients to obtain the nodal integral components, which are then continuously and weightedly accumulated within a specified connected region according to the integration operation rules. The current local connected region is set to contain only three element nodes, with their respective nodal integral components being 0.01217 kg, 0.01150 kg, and 0.01320 kg. These three floating-point values ​​are continuously summed to generate the final water allocation calculation result for this local region, which is 0.03687 kg. The execution node of this operation logic is to derive the overall regional water replenishment result based on the multi-sequence parameters and integral factors, thereby triggering subsequent operations for final water replenishment control.

[0083] Please see Figure 6 The specific steps of S5 are as follows:

[0084] S501: Compare the elements in the water distribution calculation result with the preset spray equipment flow threshold, filter the water distribution elements whose values ​​exceed the spray equipment flow threshold, extract the spatial distribution coordinate index of the corresponding out-of-bounds elements, and perform spatial topology reorganization and union to obtain the out-of-bounds water distribution coordinate set.

[0085] By establishing a low-level data cursor, the sequence of water distribution calculation results generated in the previous steps is read byte by byte in a contiguous memory segment. A preset spray equipment flow threshold is obtained. This preset flow threshold is determined based on the physical flow cross-sectional area of ​​the centrifugal pump configured on-site under standard rated operating conditions, with the maximum single-point discharge volume. By reading the pump nameplate parameters and verifying them with a measuring cup on-site, the preset spray equipment flow threshold is determined to be 0.03500 kg. Elements within the water distribution calculation results are extracted one by one. The water distribution element 0.03687 kg obtained from the previous steps is substituted into the preset flow threshold 0.03500 kg and compared logically. After comparison, 0.03687 kg is greater than 0.03500 kg, resulting in an out-of-bounds difference of 0.00187 kg exceeding the physical limit. Based on this comparison result, the water distribution element is determined to have triggered an out-of-bounds condition, and water distribution elements with values ​​exceeding the spray equipment flow threshold are filtered out. For the filtered outbound data, the spatial distribution coordinate index of the corresponding outbound element is extracted. Here, the extracted distribution coordinate index number is 152005. To integrate the outbound situations within locally connected regions, spatial topology reorganization and union operations are performed on the distribution coordinate indices of all marked outbound elements. A hash set space is initialized, and the first extracted spatial distribution coordinate index 152005 is stored in the hash set as the reference cluster center. Then, the remaining outbound indices are traversed, and the Manhattan distance calculation logic is used to verify the absolute spatial distance between the three-dimensional Cartesian coordinates of the remaining indices and the coordinates of the reference cluster center. If the distance value is equal to or less than the preset 50 mm grid side length, it is determined that the two are connected in spatial topology. The connected and overlapping coordinate indices are then subjected to a union operation, and duplicate coordinate point records are simultaneously removed to obtain the outbound water distribution coordinate set. The execution node of this operation logic is to obtain the outbound identification result based on the comparison between the water distribution calculation result and the equipment flow limit, thereby triggering the subsequent operations of abnormal point coordinate extraction and topology reorganization.

[0086] S502: Call the out-of-bounds water allocation coordinate set to extract the water allocation quantity corresponding to the associated spatial location and the water allocation quantity parameter of the adjacent coordinate. Perform item-by-item summation and arithmetic mean calculation on the water allocation quantity element and the water allocation quantity parameter of the adjacent coordinate. Replace the water allocation quantity corresponding to the original out-of-bounds spatial location to establish a smooth water allocation value set.

[0087] The generated out-of-bounds water allocation coordinate set is invoked, and a reading channel is established to traverse the node data within this coordinate set. Using a hash mapping mechanism, with spatial distribution coordinate index 152005 as the search key, the water allocation corresponding to the associated spatial location is located and extracted. Substituting this into the previous value yields 0.03687 kg. Based on a 50 mm spatial grid step size constant, the distribution coordinate matrix is ​​offset by 1 grid unit in four orthogonal directions (east, west, south, and north) to retrieve and extract the water allocation parameters of adjacent coordinates. The adjacent coordinate indices are obtained as 152004, 152006, 151005, and 153005, and the extracted water allocation parameter values ​​for these four adjacent coordinates are 0.03100 kg, 0.03200 kg, 0.02900 kg, and 0.03050 kg, respectively. A smoothing and noise reduction operation is then initiated, performing a summation operation on each water allocation element and its adjacent coordinate water allocation parameters. The total water allocation for the local cluster is calculated as 0.15937 kg by performing consecutive additions with 0.03100 kg, 0.03200 kg, 0.02900 kg, and 0.03050 kg. Next, an arithmetic mean calculation is performed, resulting in 5 valid elements participating in the calculation. The total water allocation of 0.15937 kg is then divided by the total number of elements (5), yielding a smoothed arithmetic mean of 0.03187 kg. A memory overwrite instruction is invoked to directly write the calculated 0.03187 kg, replacing the original water allocation of 0.03687 kg at location 152005 in the out-of-bounds space. The aforementioned neighborhood extraction and calculation processes are then iteratively performed on all out-of-bounds nodes to eliminate local extreme values ​​and establish a smoothed water allocation numerical set. The execution node of this operation logic is to calculate the local arithmetic mean result based on the data of the out-of-bounds point and adjacent points, thereby triggering the memory overwrite operation of the original out-of-bounds data.

[0088] S503: Extract the arithmetic mean of multiple elements and the water distribution elements that have not exceeded the limit from the smooth water distribution data set, and serialize and merge them according to the order of the spatial coordinate index of the multi-element association. Then, perform time-series allocation and scheduling site encoding mapping on the merged element sequence to generate an optimized water distribution scheduling sequence.

[0089] For the established smooth water distribution numerical set, the repaired memory block data is read via the data bus. The arithmetic mean of multiple terms obtained after neighborhood averaging is extracted, including the smooth water distribution of 0.03187 kg corresponding to spatial index 152005 generated by the previous replacement. Simultaneously, a doubly linked list traversal mechanism is used to completely extract the non-boundary water distribution elements that have not triggered boundary value markers, extracting the originally normal non-boundary water distribution of 0.03100 kg at index 152004. The associated spatial coordinate indices and their corresponding three-dimensional absolute coordinates of all elements are extracted. Abandoning simple numerical sorting, a Boustrophedon reciprocating continuous trajectory planning algorithm is used for spatial topological sorting. Based on this algorithm, an "S"-shaped continuous traversal trajectory conforming to the actual physical path of the spraying equipment is generated, ensuring that adjacent nodes in the one-dimensional sequence are also closely connected in the real three-dimensional physical space. Following the order of this continuous physical trajectory, the corresponding smooth water allocation and non-boundary water allocation elements are stored in a continuous physical memory segment for serialization, splicing, and merging to form a one-dimensional continuous water allocation data stream. Subsequently, time-series allocation and scheduling point encoding mapping operations are performed on the spliced ​​and merged element sequence. A preset driving speed parameter of 0.50 meters per second is obtained, and the actual three-dimensional Euclidean physical distance between adjacent trajectory nodes is read. This actual physical distance (such as a straight-line distance of 0.05 meters within the grid or a turning distance) is divided by the driving speed of 0.50 meters per second to calculate the actual driving time interval between corresponding points. The timestamp of the operation's starting node is set to 0.00 seconds. Following the order of the physical trajectory nodes, the corresponding driving time intervals are continuously accumulated and summed for subsequent nodes to complete the time-series allocation. The first node's timestamp is 0.00 seconds. If index 152004 is the 105th node and the current interval is 0.10 seconds, its timestamp is allocated to 10.40 seconds. If index 152005 (located at a newline position, with a calculated actual turning trajectory distance of 0.08 meters) is the 106th node, the turning time is increased by 0.16 seconds, and the allocated timestamp is 10.56 seconds. The register address book of the control hardware is read synchronously, and the coordinate indices are mapped and bound to the scheduling position codes of the control valve opening in chronological order, generating an optimized water distribution scheduling sequence containing timestamps and absolute position information. The execution node of this operation logic is to obtain the absolute timing result based on the real spatial physical trajectory planning and equipment travel parameter calculation, thereby triggering the issuance of the work instruction sequence.

[0090] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for calculating soil water distribution based on function fitting and mathematical integration, characterized in that, Includes the following steps: S1: Collect and map the dry density electrical signal, moisture content electrical signal and three-dimensional coordinate data of the soil and water material area to be prepared, construct a three-dimensional discrete dataset and input it into the radial basis function neural network to perform spatial distribution regression fitting, and construct the soil state function; S2: Call the soil state function to analyze the dry density of the coordinate points of the directional radiation line, accumulate and calculate the directional cumulative scalar and filter the dominant directional line vector, obtain the integral space coordinate points of the soil and water material area to be distributed and calculate the normal distance between the dominant directional line vector and the dominant directional line vector, and inversely attenuate the mapping to generate the directional integral sequence. S3: Obtain the discrete spatial coordinates of the soil and water material area to be distributed and input the soil state function mapping to calculate the dry density of the discrete coordinates. Analyze adjacent discrete spatial coordinates to obtain the density difference scalar. Combine the preset connectivity judgment threshold to extract the connected coordinate point set and sum it to generate a structural weight sequence. S4: Obtain the spatial micro-element coordinates of the soil and water material area to be distributed and input the soil state function to analyze the soil state sequence. At the same time, analyze the foundation water demand sequence, and combine the directional integral sequence and the structural weight sequence to input the Gaussian integral algorithm for coupled integral calculation to generate the water distribution calculation result.

2. The method for calculating soil water distribution based on function fitting and mathematical integration according to claim 1, characterized in that, The soil state function includes dry density distribution parameters, moisture content response parameters, and spatial location index parameters. The directional integral sequence includes directional accumulation results, dominant direction identifiers, and distance attenuation coefficients. The structural weight sequence includes density difference scalars, connectivity identifiers, and structural aggregation weights. The water distribution calculation results include micro-element water demand coefficients, integral coupling results, and spatial water distribution intensity.

3. The method for calculating soil water distribution based on function fitting and mathematical integration according to claim 1, characterized in that, The specific steps of S1 are as follows: S101: Collect the dry density electrical signal and moisture content electrical signal of the soil and water material area to be prepared and perform time series alignment. Compare the timestamp difference of the electrical signal with the preset time offset limit value point by point, remove data points associated with out-of-bounds timestamps, rearrange the index of the remaining timestamps and normalize the amplitude to obtain the synchronous electrical signal pair. S102: Based on the synchronous electrical signal pair, obtain the three-dimensional coordinate data of the water and soil material area to be distributed, perform spatial mapping matching between the three-dimensional coordinate data and the synchronous electrical signal pair, compare the coordinate deviation with the preset adjacent boundary value, remove abnormal deviation items and splice the remaining coordinate items and related items to establish a three-dimensional discrete dataset. S103: Call the three-dimensional discrete dataset and input it into the radial basis function neural network, calculate the radial distance between the data item and the center point of the hidden layer and obtain the neuron activation state parameter, calculate the weighted summation of the activation state parameter and the associated weight and perform spatial distribution regression fitting to construct the soil state function.

4. The method for calculating soil water distribution based on function fitting and mathematical integration according to claim 1, characterized in that, The specific steps of S2 are as follows: S201: Obtain the coordinate points of the directional radiation lines of the soil and water material area to be distributed and input the soil state function for spatial mapping, solve the associated dry density parameter, and sequentially associate and bind the dry density parameter with the original spatial location points to obtain the radiation line density mapping set. S202: Based on the radiation line density mapping set, extract the dry density parameters corresponding to the radiation line direction and perform linear accumulation to obtain multiple radiation direction associated direction accumulation scalars. Compare the accumulation scalars with the preset direction screening extreme values, remove scalar items that have not reached the direction screening extreme values, and generate the dominant direction line vector. S203: Collect the integral spatial coordinates of the soil and water material area to be distributed and calculate the normal distance with the dominant direction line vector. Extract the vertical span corresponding to the dominant direction line vector from multiple coordinate points. Call the preset attenuation ratio coefficient to perform inverse proportional attenuation mapping on the vertical span and merge it according to the spatial index to obtain the direction integral sequence.

5. The method for calculating soil water distribution based on function fitting and mathematical integration according to claim 1, characterized in that, The specific steps for S3 are as follows: S301: Obtain discrete spatial coordinate points of the soil and water material area to be prepared, input the three-dimensional coordinate parameters of the multi-spatial coordinate points into the soil state function for mapping and solving, extract the associated dry density of the multi-coordinate points and arrange and recombine them in sequence according to the coordinate index order to establish a discrete density mapping set. S302: Based on the discrete density mapping set, extract the associated density of adjacent discrete spatial coordinate points and perform item-by-item subtraction operation to solve the density difference scalar between adjacent coordinate points and compare it with the preset connectivity judgment threshold. Extract the scalar associated coordinate point index that has not exceeded the boundary and aggregate it to obtain the connected coordinate point set. S303: Call the connected coordinate point set and extract the associated density difference scalar of the multi-connected spatial coordinate points. Accumulate and sum all density difference scalars to solve the density change accumulation parameter of the corresponding spatial connected region. Reorganize and encode the sequence according to the corresponding connected index order to generate the structure weight sequence.

6. The method for calculating soil water distribution based on function fitting and mathematical integration according to claim 5, characterized in that, The connectivity determination threshold is obtained by acquiring the soil reference sample attributes, extracting the dry density parameters of adjacent points of the reference sample and performing a subtraction operation to obtain the reference difference sequence, calculating the standard deviation of the elements in the sequence to solve the reference discrete parameters, and multiplying the discrete parameters with a preset confidence coefficient to determine the threshold.

7. The method for calculating soil water distribution based on function fitting and mathematical integration according to claim 1, characterized in that, The specific steps of S4 are as follows: S401: Obtain the spatial micro-element coordinates of the area to be mixed with water and soil and input the soil state function for mapping and solution. Extract the corresponding spatial point associated dry density and initial moisture content. Perform serialization combination encoding and data splicing according to the spatial distribution index of the micro-element coordinates to obtain the soil state sequence. S402: Based on the soil state sequence, extract the initial moisture content and dry density parameters associated with multiple spatial points, subtract the preset target moisture content scalar from the initial moisture content parameter, and perform algebraic calculations by multiplying the obtained difference and the dry density parameter with the preset infinitesimal volume constant to establish the basic water demand sequence. S403: Call the basic water demand sequence, the directional integral sequence, and the structural weight sequence, and perform point-to-point product calculation to obtain multivariate coupling parameters. Assign discrete quadrature elements and integral quadrature coefficients to the multivariate coupling parameters and perform weighted accumulation item by item to generate water distribution calculation results.

8. The method for calculating soil water distribution based on function fitting and mathematical integration according to claim 1, characterized in that, The method further includes: S5: Compare the elements in the water distribution calculation result with the preset spray equipment flow threshold and extract the out-of-bounds water distribution distribution area. Based on the out-of-bounds water distribution distribution area, perform numerical smoothing adjustment and optimization on the water distribution calculation result to generate an optimized water distribution scheduling sequence. The optimized water allocation scheduling sequence includes out-of-bounds distribution identifiers, smoothing correction results, and scheduling sorting indexes.

9. The method for calculating soil water distribution based on function fitting and mathematical integration according to claim 8, characterized in that, The specific steps of S5 are as follows: S501: Compare the elements in the water distribution calculation result with the preset spray equipment flow threshold, filter the water distribution elements whose values ​​exceed the spray equipment flow threshold, extract the spatial distribution coordinate index of the corresponding out-of-bounds elements and perform spatial topology reorganization and union to obtain the out-of-bounds water distribution coordinate set. S502: Call the out-of-bounds water distribution coordinate set to extract the water distribution volume corresponding to the associated spatial location and the water distribution volume parameters of adjacent coordinates, perform item-by-item summation and arithmetic mean calculation on the water distribution volume elements and the water distribution volume parameters of adjacent coordinates, and replace the water distribution volume corresponding to the original out-of-bounds spatial location to establish a smooth water distribution value set. S503: Extract multiple arithmetic averages and non-boundary water allocation elements from the smooth water allocation value set, serialize and merge them according to the order of the multi-element associated spatial coordinate index, perform time-series allocation and scheduling site encoding mapping on the merged element sequence, and generate an optimized water allocation scheduling sequence.

10. The method for calculating soil water distribution based on function fitting and mathematical integration according to claim 9, characterized in that, The flow rate threshold of the spraying equipment is determined by calling the operating parameters of the water supply equipment, extracting the ultimate power parameter of the water pump and the cross-sectional area parameter of the nozzle, multiplying the ultimate power parameter with the preset hydraulic coefficient to obtain the extreme kinetic energy scalar, and then performing a ratio calculation between the kinetic energy scalar and the cross-sectional area parameter.

Images