Power transmission line water and soil loss treatment decision-making method and system

Abstract

The invention belongs to the technical field of water and soil conservation and intelligent power grid crossing, and provides a power transmission line water and soil loss treatment decision-making method and system. Influence factors are extracted from multi-source data, the game theory thought is introduced, the weight of each influence factor determined by a subjective weight analysis method and an objective weight analysis method is subjected to complete compromise solution fusion, and the power transmission line water and soil loss treatment decision-making method is obtained. The optimal comprehensive weight of each influence factor is obtained, and a final comprehensive sensitivity index is obtained through calculation; comparing the final comprehensive sensitivity index with a set grade threshold, determining the sensitivity grade, analyzing the contribution degree of each influence factor to the final comprehensive sensitivity index, and determining a dominant factor; and according to the sensitivity grade and the dominant factor, matching an optimal treatment scheme according to a preset treatment measure knowledge base. According to the invention, power transmission line water and soil loss sensitivity evaluation and treatment with dynamic perception, accurate evaluation and intelligent decision can be realized.

Description

Technical Field

[0001] This invention belongs to the technical field of soil and water conservation and smart grid, specifically relating to a decision-making method and system for soil erosion control in power transmission lines. Background Technology

[0002] The statements in this section are merely background information related to the present invention and do not necessarily constitute prior art.

[0003] Transmission line projects often traverse complex terrains such as mountains and hills. Human activities, such as tower foundation excavation and construction access road building, severely damage existing vegetation and soil structure, easily inducing serious soil erosion and threatening the safe operation of the power grid and the ecological environment. Traditional soil and water conservation schemes for transmission lines mainly rely on manual on-site investigations and historical experience, which have the following significant drawbacks: Existing soil and water conservation schemes for power transmission lines rely on static assessments, primarily using annual average climate data and static vegetation cover, which cannot reflect dynamic risks during construction and under extreme weather conditions such as heavy rain. On the other hand, existing assessment models (such as USLE / RUSLE) mainly target natural slopes and lack effective quantitative indicators for high-intensity, short-duration human disturbances such as engineering construction.

[0004] Moreover, the assessment results often stop at risk level classification, making it difficult to identify the key causes of the risk and failing to provide a direct basis for precise governance. The generation of governance solutions relies on manual design, which is inefficient and lacks optimization. Summary of the Invention

[0005] To address the aforementioned problems, this invention proposes a decision-making method and system for soil and water conservation along power transmission lines. This invention enables dynamic sensing, precise assessment, and intelligent decision-making for the sensitivity assessment and control of soil and water conservation along power transmission lines.

[0006] According to some embodiments, the present invention adopts the following technical solution: A decision-making method for soil and water conservation along power transmission lines includes the following steps: Acquire terrain data, construct a digital elevation model, and extract terrain humidity index and confluence dynamics index from it; Acquire climate and hydrological data and calculate dynamic rainfall erosivity factors; Acquire Earth observation image data, calculate enhanced vegetation index, and retrieve monthly vegetation cover; Intelligent interpretation of UAV orthophotos is performed to identify disturbed patches, calculate their area proportion and spatial distribution, and then calculate the disturbance intensity index. Data extracted or calculated from multiple sources are used as impact factors, and the impact factors are standardized. By introducing game theory, the weights of each influencing factor determined by the subjective weight analysis method and the objective weight analysis method are completely compromised and integrated to obtain the optimal comprehensive weight of each influencing factor. The final comprehensive sensitivity index is calculated by multiplying all standardized influence factors by their corresponding comprehensive weights and then summing the results. The final comprehensive sensitivity index is compared with the set level threshold to determine the sensitivity level. The contribution of each influencing factor to the final comprehensive sensitivity index is analyzed to determine the dominant factor. Based on the sensitivity level and dominant factors, the optimal governance solution is matched according to the pre-set governance measures knowledge base.

[0007] As an alternative implementation method, the process of acquiring terrain data includes using an airborne lidar on a drone to collect data and construct a digital elevation model.

[0008] As an alternative implementation method, the process of acquiring climate and hydrological data and calculating dynamic rainfall erosivity factors includes: using regional meteorological radar gridded rainfall forecast data, dividing the data into grids of fixed size, and calculating the R-value in real time using the rainfall intensity within a set time period during the forecast period to achieve dynamic early warning. The value of R is:

[0009] Em= 0.29×[1-0.72×exp(-0.05×Im)] Where: Em is the unit rainfall kinetic energy during the m-th rainfall period; Im is the average rainfall intensity during the m-th rainfall period; exp() is the natural exponential function; I 30 The rainfall intensity is defined as the rainfall intensity within a given time period; T represents the duration of the rainfall.

[0010] As an alternative implementation method, the process of acquiring Earth observation image data, calculating the enhanced vegetation index, and retrieving monthly vegetation cover includes: using multi-time-series Sentinel-2 imagery to calculate the enhanced vegetation index and retrieve monthly vegetation cover to dynamically reflect seasonal changes in vegetation, wherein the monthly vegetation cover is: fVC = (EVI - EVI) soil ) / (EVI veg -EVI soil ); Among them, EVI is the Enhanced Vegetation Index. soil EVI is the Enhanced Vegetation Index value corresponding to a completely bare, unvegetated soil pixel. veg The enhanced vegetation index value is the value of a pure vegetation pixel.

[0011] As an alternative implementation method, the process of identifying disturbed patches, calculating their area proportion and spatial distribution, and then calculating the disturbance intensity index includes: the disturbed patches include the excavation face, the spoil heap, and the construction access road; using a pre-trained neural network model, the disturbed patches are identified, their area proportion and spatial distribution are calculated, and the disturbance intensity index is defined as: DII = f(Perturbation patch type weight, perturbation area ratio, perturbation duration).

[0012] As an alternative implementation method, the process of standardizing the data for impact factors includes: using range standardization; for positive factors, the standardization formula is: F i = (X i - X min ) / (X max - X min ) For negative factors, the standardization formula is: F i = (X max - X i ) / (X max - X min ) Among them, F i Let X be the value of the i-th region. i The standardized value, X min X is the minimum value of the entire region. max This represents the maximum value for the entire region.

[0013] As an alternative implementation method, the process of introducing game theory ideas and completely compromising and integrating the weights of each influencing factor determined by the subjective weight analysis method and the objective weight analysis method includes: using the AHP (Analytic Hierarchy Process) to calculate the eigenvectors of the matrix through mathematical methods, ensuring the logical consistency of expert judgment through consistency checks, and obtaining a set of weights representing expert experience. The entropy weight method is used to assign weights based on the degree of dispersion of the data. The higher the degree of dispersion of the data, the higher its weight, resulting in a set of weights that are completely driven by the data and reflect the data patterns. The weights obtained by the subjective weight analysis method and the weights obtained by the entropy weight method are linearly weighted to obtain the final optimized weights.

[0014] As an alternative implementation method, the process of multiplying all standardized influence factors by their corresponding comprehensive weights and then summing the results to calculate the final comprehensive sensitivity index includes: S = ∑(W) opti ×F i ) Wherein, S is the comprehensive sensitivity index, representing the level of soil and water loss sensitivity at a certain point. The higher the S value, the stronger the soil and water loss sensitivity at that point. ∑ is the summation symbol, which means summing the results of all factors within the parentheses, where i ranges from 1 to n, and n is the total number of selected factors; W opti For the optimal comprehensive weight of the i-th factor, the sum of the weights of all factors equals 1; F i is the standardized value of the i-th factor.

[0015] As an alternative implementation method, the process of comparing the final comprehensive sensitivity index with the set level threshold to determine the sensitivity level includes: using the natural breakpoint method combined with engineering experience to divide the sensitivity into multiple levels, ensuring that the intra-class difference is minimized and the inter-class difference is maximized, and determining the set level threshold for dividing adjacent levels into different levels.

[0016] As an alternative implementation method, the process of analyzing the contribution of each influencing factor to the final comprehensive sensitivity index and determining the dominant factor includes: quantifying the contribution of each factor to the final sensitivity assessment result using the SHAP method, and selecting the influencing factor with the highest contribution as the dominant factor.

[0017] As an alternative implementation method, the process of matching the optimal governance scheme according to the preset governance measures knowledge base is replaced by: constructing a decision model, wherein the decision model aims to minimize the total cost function, the expected soil erosion function, and the ecological negative effect function, and is solved with several of the following constraints as constraints: total cost cannot exceed the budget, drainage capacity constraint, mutual exclusion constraint of measures, material balance constraint, slope stability constraint, vegetation restoration rate constraint, mandatory measures constraint, soil loss control constraint, and investment intensity constraint, to obtain the optimal governance scheme.

[0018] A decision-making system for soil and water conservation along power transmission lines includes: The multi-source data acquisition and factor calculation module is configured to acquire topographic data, construct a digital elevation model, and extract topographic humidity index and confluence dynamics index from it; acquire climate and hydrological data and calculate dynamic rainfall erosivity factor; acquire Earth observation image data, calculate enhanced vegetation index, and invert monthly vegetation cover; intelligently interpret UAV orthophotos, identify disturbed patches, calculate their area proportion and spatial distribution, and then calculate disturbance intensity index; and use the data extracted or calculated from multi-source data as influencing factors and standardize the data of influencing factors. The optimal comprehensive weight calculation module is configured to introduce game theory ideas, and to fully compromise and merge the weights of each influencing factor determined by the subjective weight analysis method and the objective weight analysis method to obtain the optimal comprehensive weight of each influencing factor. The comprehensive sensitivity index calculation module is configured to multiply all standardized influence factors by their corresponding comprehensive weights and then sum them to calculate the final comprehensive sensitivity index. The sensitivity level determination module is configured to compare the final comprehensive sensitivity index with a set level threshold, determine the sensitivity level, analyze the contribution of each influencing factor to the final comprehensive sensitivity index, and determine the dominant factor. The governance solution decision module is configured to match the optimal governance solution based on the sensitivity level and dominant factors, according to a pre-defined governance measures knowledge base.

[0019] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention improves the accuracy and objectivity of the sensitivity assessment process by constructing an evaluation index system, standardizing factor data, determining factor weights, and comprehensively calculating the sensitivity assessment results. It also quantifies the contribution of each factor to the final sensitivity assessment results, providing a direct basis for precise governance.

[0020] This invention, through the combination of dominant factors, matches the optimal technical measures package from the knowledge base, which can generate effective governance solutions in a timely manner. It can also use the model to modify relevant factors according to the governance solution to simulate and predict the amount of erosion after governance, quantitatively evaluate the governance benefits, and support the comparison of solutions.

[0021] This invention is a dynamic system that updates parameters and evaluation results periodically (e.g., quarterly or annually) as new inspection data, monitoring data, and meteorological data are input, enabling rolling forecasting and adaptive management.

[0022] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0023] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.

[0024] Figure 1 This is a schematic diagram of a decision-making process for soil and water conservation along power transmission lines, based on one embodiment. Detailed Implementation

[0025] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0026] It should be noted that the following detailed description is illustrative and intended to provide further explanation of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0027] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0028] Where there is no conflict, the embodiments and features described in this application may be combined with each other.

[0029] Example 1 A decision-making method for soil and water conservation along power transmission lines, such as Figure 1 As shown, it includes the following steps: Acquire terrain data, construct a digital elevation model, and extract terrain humidity index and confluence dynamics index from it; Acquire climate and hydrological data and calculate dynamic rainfall erosivity factors; Acquire Earth observation image data, calculate enhanced vegetation index, and retrieve monthly vegetation cover; Intelligent interpretation of UAV orthophotos is performed to identify disturbed patches, calculate their area proportion and spatial distribution, and then calculate the disturbance intensity index. Data extracted or calculated from multiple sources are used as impact factors, and the impact factors are standardized. By introducing game theory, the weights of each influencing factor determined by the subjective weight analysis method and the objective weight analysis method are completely compromised and integrated to obtain the optimal comprehensive weight of each influencing factor. The final comprehensive sensitivity index is calculated by multiplying all standardized influence factors by their corresponding comprehensive weights and then summing the results. The final comprehensive sensitivity index is compared with the set level threshold to determine the sensitivity level. The contribution of each influencing factor to the final comprehensive sensitivity index is analyzed to determine the dominant factor. Based on the sensitivity level and dominant factors, the optimal governance solution is matched according to the pre-set governance measures knowledge base.

[0030] In other schemes, the process of matching the optimal governance scheme according to the preset governance measures knowledge base can be replaced by constructing a decision model. The decision model aims to minimize the total cost function, the expected soil erosion function, and the ecological negative effect function. It is solved with several constraints, including total cost not exceeding the budget, drainage capacity constraints, mutual exclusion constraints of measures, material balance constraints, slope stability constraints, vegetation restoration rate constraints, mandatory measures constraints, soil loss control constraints, and investment intensity constraints, to obtain the optimal governance scheme.

[0031] The following is a detailed description of each step.

[0032] First, data acquisition and preprocessing: Topographic data A high-precision (0.1m) DEM was acquired using an airborne LiDAR on a drone, replacing traditional surveying methods and capable of penetrating vegetation to obtain accurate topographic data. Based on the DEM, the Terrain Moisture Index (TWI) and Spiral Dynamics Index (SPI) were automatically extracted as proxy indicators for assessing soil stability and runoff erosion capacity.

[0033] The system uses gridded rainfall forecast data from regional meteorological radar with a spatial resolution of 1km×1km, providing forecasts for the next 72 hours.

[0034] Dynamic rainfall erosivity (R) factor. Instead of using the annual average, the R value is calculated in real time based on the maximum 30-minute rainfall intensity (I30) within the forecast period, enabling dynamic early warning.

[0035] Em= 0.29×[1-0.72×exp(-0.05×Im)] in: Em represents the kinetic energy of rainfall per unit volume during the m-th rainfall period, in MJ·ha⁻¹·mm⁻¹. Im represents the average rainfall intensity during the m-th rainfall period, in millimeters per hour (mm / h). exp() is the natural exponential function. T represents the duration of rainfall.

[0036] Vegetation cover and management factors (C) Using multi-time-series Sentinel-2 imagery (10m resolution), the Enhanced Vegetation Index (EVI) is calculated via Google Earth Engine or PIE-Engine platform to retrieve monthly vegetation cover (fVC), dynamically reflecting seasonal changes in vegetation and making the assessment more realistic.

[0037] fVC = (EVI - EVI) soil ) / (EVI veg -EVI soil ) EVI stands for Enhanced Vegetation Index. soil EVI represents the Enhanced Vegetation Index (EVI) value corresponding to a completely bare, unvegetated soil pixel. It signifies the "zero state" of vegetation cover within the monitoring area. When the EVI value equals EVI_soil, it means that there is almost no vegetation there; it is bare land, i.e., the surface completely destroyed during engineering construction. veg It refers to the EVI value of a pure vegetation pixel. It represents the typical EVI value that can be observed when a pixel (e.g., 10m x 10m) in a remote sensing image is completely covered by healthy, dense vegetation. It is the upper limit of the EVI value that can be achieved in a specific area and at a specific time, like a "perfect score" standard.

[0038] Construction disturbance factor (D) Based on AI models such as YOLOv8 or SAM, the system intelligently interprets UAV orthophotos, automatically identifies disturbed features such as excavation faces, soil dumps, and construction access roads, and calculates their area proportion and spatial distribution.

[0039] Define the Disturbance Intensity Index (DII): DII = f(Disturbance Type Weight, Disturbance Area Ratio, Disturbance Duration). For example, Excavation Face Weight (1.0) > Construction Access Road (0.6) > Temporary Soil Stockpile (0.3).

[0040] A comprehensive index method is adopted, but game theory is introduced into the weight calculation. The AHP (subjective) and CRITIC (objective) weights are completely compromised to obtain the optimal comprehensive weight Wopt, overcoming the arbitrariness of traditional additive integration. S = ∑(W opti ×F i (F) i (These are the standardized factor values) S: Comprehensive Sensitivity Index. This is the final result to be calculated, a numerical value representing the level of soil erosion sensitivity at a certain point (such as a tower location). The higher the S value, the stronger the soil erosion sensitivity at that point.

[0041] ∑: Summation symbol. It represents the summation of the calculation results of all factors within the parentheses (i from 1 to n, where n is the total number of selected factors).

[0042] W opti : The optimized comprehensive weight of the i-th factor. It represents the relative importance of this factor to the sensitivity of soil erosion. The sum of all weights equals 1.

[0043] Fi : The standardized factor value of the i-th factor. It removes the dimensions of the original data (such as slope, rainfall) and maps it to a comparable numerical range (usually 0-1 or 0-100).

[0044] To calculate S, the following four key steps need to be completed: Step 1: Constructing an evaluation index system (F i ) First, determine which factors to use to evaluate sensitivity. For example, select the following five key factors: X1 is the slope (°); X2 is the rainfall erosivity (R value); X3 is the soil erodibility (K value); X4 is the vegetation cover; X5 is the construction disturbance intensity (DII).

[0045] Step 2: Factor data standardization (calculate F) i ) The original data for each factor have different dimensions and ranges (e.g., slope 0-90°, rainfall intensity may be in the thousands), making direct comparison and calculation impossible. They must be standardized to the same scale (e.g., 0-1).

[0046] Range standardization For positive factors (the larger the factor value, the higher the sensitivity, such as slope and rainfall intensity): F i = (X) i - X min ) / (X max - X min ); For negative factors (the larger the factor value, the lower the sensitivity, such as vegetation cover): F i = (X) max - X i ) / (X max -X min ); Example: Suppose the original vegetation cover value of a certain tower base is X4 = 0.6, and the maximum vegetation cover of the entire area is 0.9, and the minimum is 0.1. Vegetation cover is a negative factor. F4 = (0.9 - 0.6) / (0.9 - 0.1) = 0.3 / 0.8 = 0.375.

[0047] After this step, the values ​​F of all factors are... i All values ​​are converted to the range [0, 1].

[0048] Step 3: Determine factor weights (calculate W) _opt_i ) This is the most crucial step, determining the scientific validity of the model. We employ a combined subjective and objective weighting method to balance expert experience with data patterns.

[0049] Subjective weight (W) _subjective_i - AHP (Analytic Hierarchy Process) Method: Several experts were invited to compare the relative importance of each factor in pairs (e.g., "How much more important is slope than rainfall force?") and a judgment matrix was constructed.

[0050] Calculation: The eigenvectors of the matrix are calculated using mathematical methods (such as the sum-product method) to obtain a set of weights. A consistency check (CR < 0.1) is then performed to ensure the logical consistency of the expert judgment.

[0051] Output: A set of weights [W_a, W_b, W_c, W_d, W_e] representing expert experience, which sum to 1.

[0052] Objective weight (W) _objective_i - Entropy Weight Method Principle: Weights are assigned based on the dispersion of the data. The greater the data variation (the more dispersed) of a factor, the stronger its ability to distinguish the sensitivity of different points, and its weight should also be greater.

[0053] calculate: a. Data standardization (already approved by F) i matrix); b. Calculate the weight of the j-th sample value under the i-th factor: p ij = F ij / ∑(F ij ) c. Calculate the information entropy of the i-th factor: e i = -k ∑(p ij × ln(p) ij (where k = 1 / ln(n)); d. Calculate the coefficient of variation: g i = 1 - e i ; e. Calculate the weights: W_objective_i = g i / ∑(g i ).

[0054] Output: A set of weights that are entirely driven by data and reflect the patterns in the data.

[0055] Assigning combined weights (W_opt_i): By combining subjective and objective weights, the final optimal weights are obtained. A commonly used method is linear weighting: W _opt_i = α ×W _subjective_i+ β × W _objective_i Where α + β = 1, for example, taking α = 0.7, β = 0.3 indicates a greater emphasis on expert experience.

[0056] Step 4: Calculate the overall sensitivity index (S) All standardized factor values ​​F i The product is multiplied by its corresponding comprehensive weight W_opt_i and then summed.

[0057] S = (W_opt1× F1) + (W_opt2× F2) + (W_opt3× F3) + (W_opt4× F4) + (W_opt5× F5).

[0058] Sensitivity level classification In this embodiment, the natural breakpoint method (Jenks) combined with engineering experience is used to divide the results into 5 levels, as shown in Table 1, to ensure that the intra-class differences are minimized and the inter-class differences are maximized.

[0059] Table 1

[0060] Key Sensitive Factor Analysis and Acquisition To implement precise measures, it is essential to identify the dominant factors leading to sensitivity. The following is a detailed analysis of each factor and a strengthened explanation of how to obtain them, with particular emphasis on innovative considerations for construction-related factors, as shown in Table 2.

[0061] Table 2

[0062] Factor Contribution Analysis: Using Explainable Artificial Intelligence (XAI) technology, the contribution of each factor to the final sensitivity assessment result can be quantified through the SHAP algorithm. For example, the analysis results may show that "in the current extremely sensitive area, the contribution of construction disturbance factor accounts for 35%, followed by slope factor at 28%", which provides a direct basis for precise governance.

[0063] Based on sensitivity levels and dominant factors, as shown in Table 3, and combined with a pre-built knowledge base, prevention and control plans can be automatically generated and optimized.

[0064] Table 3

[0065] Based on the combination of dominant factors, the optimal technical measures package is matched from the knowledge base. For example, for a tower base dominated by "steep slope + disturbance", the measures package of "slope cutting and gradation + ecological retaining wall + slope revegetation + interception and drainage" is automatically matched.

[0066] Generate a "one tower, one policy" governance plan design document. In some embodiments, this is achieved using a model, including: Design schematic diagram: Automatically generated 2D / 3D renderings.

[0067] Bill of Quantities: Automatically calculated quantities of building materials, earthwork, and seedlings.

[0068] Budget estimate: A preliminary estimate generated based on the built-in price information database.

[0069] Construction timing recommendations: The optimal construction time window is recommended based on weather forecasts.

[0070] Effect simulation: The model can also be used to correct various factors after treatment, simulate and predict the amount of erosion after treatment (A_post), quantitatively evaluate the treatment benefits, and support the selection of alternatives.

[0071] In the above process, the assessment of soil erosion is based on the soil erosion equation, the complete mathematical expression of which is: A=R dynamic ×K×LS×C current ×P×DII; Dynamic rainfall erosivity factor (R dynamic ):

[0072] Among them, I 30 The maximum rainfall intensity over 30 minutes (mm / h); t i is the duration of rainfall (h); n is the number of time periods for calculation.

[0073] Data sources: meteorological station observation data, GPM / TRMM satellite precipitation products, and weather forecast data.

[0074] Soil erodibility factor (K):

[0075] in SAN represents sand content (%); SIL represents dust content (%); CLA represents clay content (%); C represents organic carbon content (%). Data sources: field soil sampling and laboratory analysis, World Soil Database (HWSD), and multispectral remote sensing inversion.

[0076] Topographic factor (LS)

[0077]

[0078] Where λ is the slope length (m); M is the slope length exponent (usually taken as 0.5); and θ is the slope angle (°).

[0079] Data source: High-precision DEM acquired by UAV LiDAR, terrain information extracted from remote sensing stereo image pairs.

[0080] Vegetation cover factor (C_current)

[0081] NDVI is the Normalized Difference Vegetation Index; α and β are empirical parameters (usually α=2, β=1).

[0082] Data source: Sentinel-2 / Landsat multispectral imagery, UAV multispectral camera.

[0083] Construction Disturbance Intensity Index (DII)

[0084] A i W represents the area (m²) of the i-th type of disturbance. i Perturbation type weights (excavation = 1.0, dumping = 0.8, access road = 0.6, temporary = 0.4); DF i For duration factor, (t) i (where n is the duration in weeks); n is the number of disturbance types; Data sources: AI recognition of drone orthophotos, construction logs, and on-site surveys.

[0085] XAI Driving Factor Diagnostic Model Factor contribution analysis based on SHAP values:

[0086] N represents the set of all features; S represents a subset of features; f represents the trained prediction model; and x represents the input sample.

[0087] In some embodiments, a multi-objective optimization decision-making model can be constructed to solve for the optimal decision scheme.

[0088] The goal of the decision model is:

[0089] The constraints are satisfied:

[0090] Where x is a decision vector, representing a specific governance plan. It is an m-dimensional vector: x = (x_1, x_2, ..., x_m). Each dimension x... jIt can be expressed as: Binary selection: x j ∈ {0, 1} indicates whether the j-th measure is adopted (for example, 1 indicates that a retaining wall is built, and 0 indicates that it is not built).

[0091] Continuous variable: x j ∈ {0, 1}, representing the intensity or scale of a certain measure (e.g., the proportion of area for vegetation restoration).

[0092] Integer variable: x j ∈ {Z} represents the quantity of a certain measure (e.g., the number of drainage ditches).

[0093] F(x) is a vector of objective functions containing three objectives that need to be minimized simultaneously: f1(x): Total cost function (unit: ten thousand yuan). Where c j k is the unit cost of the j-th measure. j This is its fixed cost. This function calculates the total construction investment for scheme x.

[0094] f2(x): Expected soil erosion function (unit: t / km²·a); f2(x) = A post = A pre (1 - η1×x1) (1 - η2×x2) ... (1 - η m ×x m ).

[0095] Where A pre This is the predicted erosion amount before measures are taken (calculated by the dynamic RUSLE model), η j is the erosion reduction efficiency of the j-th measure (between 0 and 1). This function predicts the remaining erosion after implementing treatment plan x.

[0096] f3(x): Ecological negative effect function (dimensionless, taken as negative to minimize), i.e.: f3(x) = ; Where e j It is the ecological benefit score of the j-th measure (e.g., the e-value of the vegetation measure). j The value is high, while the e value of concrete hardening measures is high. j (The value may be negative). Minimizing f3(x) is equivalent to maximizing ecological benefits.

[0097] ω is the feasible solution space, defined by various constraints: (a) Budget constraint: f1(x) ≤ B (total cost cannot exceed budget B)

[0098] Among them, c j x is the unit cost of measure j; j For the implementation scale of the j-th measure, k j Let B be the fixed cost of measure j, and let B be the total budget.

[0099] Constraint Interpretation: The total cost of the governance solution must not exceed the project budget.

[0100] Data source: The cost of the measures comes from engineering quotas and market prices, and the budget is determined by the project investment plan.

[0101] (ii) Engineering constraints: (1) Drainage capacity constraints Objective: To ensure that the designed drainage system is capable of handling runoff generated by heavy rainfall.

[0102] Formula: (Estimated total runoff / (Number of drainage ditches × Design drainage capacity of a single drainage ditch)) -1 ≤ 0

[0103] in, To design the total runoff generated by rainfall, ; Design drainage capacity of drainage ditch i .

[0104] Calculated value: Total runoff / Total drainage capacity (this ratio must be ≤ 1); Allowed value: 1; Explanation: If the calculation result is greater than 0, it indicates that the drainage capacity is insufficient and the plan is not feasible.

[0105] (2) Mutually exclusive constraints of measures Objective: To prevent conflicting measures (such as both full paving and ecological grass planting) from being deployed on the same plot of land.

[0106] formula: ; The sum of the implementation intensity of measure j and the sum of the implementation intensity of measure k minus 1 ≤ 0; Calculated value: The sum of the implementation strengths of the two mutually exclusive measures.

[0107] Allowed value: 1 Explanation: If both measures are fully implemented (both with an intensity of 1), then 1 + 1 - 1 = 1 > 0, violating the constraint. This formula mandates that the sum of the intensities of the two measures cannot exceed 1.

[0108] (3) Material balance constraints Objective: To control the difference in earthwork excavation and filling within the project and reduce transportation and environmental impact.

[0109] Formula: Absolute value (total excavation volume - total fill volume) - allowable earthwork imbalance margin ≤ 0; ; V cut Total excavation volume (m³) 3 ); V fill Total fill volume (m) 3 ); δ: Allowable earthwork balance margin (m) 3 ).

[0110] Calculated value: The absolute difference between the excavation and filling volumes.

[0111] Allowable value: The allowable earthwork imbalance margin (a preset constant, such as 500 cubic meters). Explanation: The difference between excavation and filling costs must be controlled within the allowable margin.

[0112] (4) Slope stability constraints Mathematical expression:

[0113]

[0114] Effective stress calculation:

[0115] Anti-skid force calculation:

[0116] Calculation of glide force:

[0117] Rainfall-backflow influencing factors

[0118] Engineering measure effect coefficient

[0119] xi represents the intensity of the measure implementation (0-1), and η represents the coefficient of the measure implementation effect; Parameter description: FS(t): Safety factor, the ratio of slope resistance to sliding force, dimensionless, >1.0.

[0120] The parameters are explained in Table 4.

[0121] Table 4

[0122] (5) Vegetation restoration rate constraint

[0123] Formula for calculating vegetation restoration degree: Recovery rate constraint:

[0124] NDVI calculation:

[0125] NIR represents the reflectance in the near-infrared band, Red represents the reflectance in the red band, and NDVI data can be obtained through satellite remote sensing, UAV multispectral imagery, etc.

[0126] Logistic growth model: k is a growth parameter, which can be obtained by fitting a time series NDVI, and can be referenced to the growth characteristics of similar vegetation in the local area.

[0127] Seasonal adjustment factor:

[0128] Where t peak The peak growing season months Site condition coefficient:

[0129] Where S represents the score (0-1) for soil, slope, and moisture conditions.

[0130] The parameters involved are shown in Table 5.

[0131] Table 5

[0132] (iv) Policy constraints: (certain mandatory regulations must be met, for example, specific measures must be taken in extremely sensitive areas) (1) Mandatory measures X mandatory -1≤0 X mandatory Decision variables for coercive measures (0 or 1) Purpose: To mandate the use of a certain measure in specific areas (such as highly sensitive areas).

[0133] Example: Whether or not a slag trap dam is adopted - 1 = 0; Actual value: Whether or not a slag trap is used (this is a variable; use = 1, no use = 0); Required value: 1 (meaning "must be used"); Explanation: This is a hard equation. If the variable's value is 0, then 0 - 1 = -1 ≠ 0, and the proposal is rejected because it is non-compliant. This formula mandates that the variable must equal 1.

[0134] (2) Constraints on soil loss control A post -T=0; Parameter description: A post Predicted soil erosion after treatment (t / (km²·a)); T: Permissible soil loss (t / (km²·a)); Explanation of constraint: The amount of soil loss after treatment must be controlled within the allowable loss range.

[0135] Data source: Permissible soil loss standards in the "Classification and Grading Standards for Soil Erosion" (SL190-2007).

[0136] Objective: To mandate that the amount of soil loss after remediation must equal the allowable amount of loss.

[0137] Formula: Predicted soil erosion after remediation - Allowable soil loss = 0 Actual value: Predicted soil erosion after treatment (calculated from an erosion model). Required value: Permissible soil erosion (a legally mandated constant threshold) Explanation: This is one of the strictest constraints. It requires that the optimized solution control the erosion amount to exactly equal to the allowable value, not exceeding it, and is also encouraged not to fall below it (because too low a value may mean excessive cost).

[0138] (3) Investment intensity constraints Mathematical expression: I actual ≤I standard Parameter description: I actual Actual investment intensity (ten thousand yuan / km²) I standard Standard investment intensity (ten thousand yuan / km²) Explanation of constraints: Investment per unit area must not exceed industry standards to ensure investment efficiency.

[0139] Data source: Industry investment quota standards and engineering cost database.

[0140] The update / collection frequency of various data is shown in Table 7.

[0141] Table 7

[0142] The model prediction results are compared with field measurements (such as the volume of erosion gullies calculated by UAV 3D reality models) and high-precision monitoring data (such as BeiDou displacement monitoring data) to calculate the relative error or coefficient of determination (R²) and verify the model accuracy.

[0143] Dynamic updates: The model should be a dynamic system. With the input of new inspection data, monitoring data, and meteorological data, the model parameters and evaluation results should be updated regularly (e.g., quarterly or annually) to achieve rolling forecasts and adaptive management.

[0144] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made by those skilled in the art without creative effort within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A decision-making method for soil and water conservation along power transmission lines, characterized in that, Includes the following steps: Acquire terrain data, construct a digital elevation model, and extract terrain humidity index and confluence dynamics index from it; Acquire climate and hydrological data and calculate dynamic rainfall erosivity factors; Acquire Earth observation image data, calculate enhanced vegetation index, and retrieve monthly vegetation cover; Intelligent interpretation of UAV orthophotos is performed to identify disturbed patches, calculate their area proportion and spatial distribution, and then calculate the disturbance intensity index. Data extracted or calculated from multiple sources are used as impact factors, and the impact factors are standardized. By introducing game theory, the weights of each influencing factor determined by the subjective weight analysis method and the objective weight analysis method are completely compromised and integrated to obtain the optimal comprehensive weight of each influencing factor. The final comprehensive sensitivity index is calculated by multiplying all standardized influence factors by their corresponding comprehensive weights and then summing the results. The final comprehensive sensitivity index is compared with the set level threshold to determine the sensitivity level. The contribution of each influencing factor to the final comprehensive sensitivity index is analyzed to determine the dominant factor. Based on the sensitivity level and dominant factors, the optimal governance solution is matched according to the pre-set governance measures knowledge base.

2. The decision-making method for soil and water conservation along power transmission lines as described in claim 1, characterized in that, The process of acquiring terrain data includes using airborne lidar on drones to collect data and constructing a digital elevation model.

3. The decision-making method for soil and water conservation along power transmission lines as described in claim 1, characterized in that, The process of acquiring climate and hydrological data and calculating dynamic rainfall erosivity factors includes: using regional meteorological radar to grid rainfall forecast data, dividing the data into grids of fixed size, and calculating the R-value in real time using rainfall intensity within a set time period during the forecast period to achieve dynamic early warning. The value of R is: Em= 0.29×[1-0.72×exp(-0.05×Im)] Where: Em is the unit rainfall kinetic energy during the m-th rainfall period; Im is the average rainfall intensity during the m-th rainfall period; exp() is the natural exponential function; I 30 The rainfall intensity is defined as the rainfall intensity within a given time period; T represents the duration of the rainfall.

4. The decision-making method for soil and water conservation along power transmission lines as described in claim 1, characterized in that, The process of identifying disturbed features, calculating their area proportion and spatial distribution, and then calculating the disturbance intensity index includes: disturbed features include excavation faces, spoil heaps, and construction access roads; using a pre-trained neural network model, disturbed features are identified, their area proportion and spatial distribution are calculated, and the disturbance intensity index is defined as: DII = f(Perturbation patch type weight, perturbation area ratio, perturbation duration).

5. The decision-making method for soil and water conservation along power transmission lines as described in claim 1, characterized in that, The process of standardizing impact factors includes: using range standardization; for positive factors, the standardization formula is: F i =(X i - X min ) / (X max - X min ) For negative factors, the standardization formula is: F i =(X max - X i ) / (X max - X min ) Among them, F i Let X be the value of the i-th region. i The standardized value, X min X is the minimum value of the entire region. max This represents the maximum value for the entire region.

6. The decision-making method for soil and water conservation along power transmission lines as described in claim 1, characterized in that it introduces... The game theory approach involves a complete compromise and fusion of the weights of various influencing factors determined by subjective and objective weight analysis methods. This process includes: using the Analytic Hierarchy Process (AHP), calculating the eigenvectors of the matrix using mathematical methods, ensuring the logical consistency of expert judgments through consistency checks, and obtaining a set of weights representing expert experience. The entropy weight method is used to assign weights based on the degree of dispersion of the data. The higher the degree of dispersion of the data, the higher its weight, resulting in a set of weights that are completely driven by the data and reflect the data patterns. The weights obtained by the subjective weight analysis method and the weights obtained by the entropy weight method are linearly weighted to obtain the final optimized weights.

7. The decision-making method for soil and water conservation along power transmission lines as described in claim 1, characterized in that, The process of calculating the final comprehensive sensitivity index by multiplying all standardized influence factors by their corresponding comprehensive weights and then summing the results includes: S = ∑(W opti ×F i ) Wherein, S is the comprehensive sensitivity index, representing the level of soil and water loss sensitivity at a certain point. The higher the S value, the stronger the soil and water loss sensitivity at that point. ∑ is the summation symbol, which means summing the results of all factors within the parentheses, where i ranges from 1 to n, and n is the total number of selected factors; W opti For the optimal comprehensive weight of the i-th factor, the sum of the weights of all factors equals 1; F i is the standardized value of the i-th factor.

8. The decision-making method for soil and water conservation along power transmission lines as described in claim 1, characterized in that, The process of comparing the final comprehensive sensitivity index with the set level threshold to determine the sensitivity level includes: using the natural breakpoint method combined with engineering experience to divide the sensitivity into multiple levels, ensuring that the intra-class difference is minimized and the inter-class difference is maximized, and determining the set level threshold for dividing adjacent levels into different levels. Alternatively, the process of analyzing the contribution of each influencing factor to the final comprehensive sensitivity index and determining the dominant factor includes: quantifying the contribution of each factor to the final sensitivity assessment result using the SHAP method, and selecting the influencing factor with the highest contribution as the dominant factor.

9. The decision-making method for soil and water conservation along power transmission lines as described in claim 1, characterized in that, According to the pre-set knowledge base of governance measures, the process of matching the optimal governance solution is replaced by: constructing a decision model. The decision model aims to minimize the total cost function, the expected soil erosion function, and the ecological negative effect function. It is solved by considering several constraints, including total cost not exceeding the budget, drainage capacity constraints, mutual exclusion constraints of measures, material balance constraints, slope stability constraints, vegetation restoration rate constraints, mandatory measures constraints, soil loss control constraints, and investment intensity constraints, to obtain the optimal governance solution.

10. A decision-making system for soil and water conservation along power transmission lines, characterized in that, include: The multi-source data acquisition and factor calculation module is configured to acquire topographic data, construct a digital elevation model, and extract topographic humidity index and confluence dynamics index from it; acquire climate and hydrological data and calculate dynamic rainfall erosivity factor; acquire Earth observation image data, calculate enhanced vegetation index, and invert monthly vegetation cover; intelligently interpret UAV orthophotos, identify disturbed patches, calculate their area proportion and spatial distribution, and then calculate disturbance intensity index; and use the data extracted or calculated from multi-source data as influencing factors and standardize the data of influencing factors. The optimal comprehensive weight calculation module is configured to introduce game theory ideas, and to fully compromise and merge the weights of each influencing factor determined by the subjective weight analysis method and the objective weight analysis method to obtain the optimal comprehensive weight of each influencing factor. The comprehensive sensitivity index calculation module is configured to multiply all standardized influence factors by their corresponding comprehensive weights and then sum them to calculate the final comprehensive sensitivity index. The sensitivity level determination module is configured to compare the final comprehensive sensitivity index with a set level threshold, determine the sensitivity level, analyze the contribution of each influencing factor to the final comprehensive sensitivity index, and determine the dominant factor. The governance solution decision module is configured to match the optimal governance solution based on the sensitivity level and dominant factors, according to a pre-defined governance measures knowledge base.

Images