Dam site comparison and selection method fused with dam break scenario simulation

Abstract

The invention belongs to the technical field of water conservancy projects, provides a dam site comparison and selection method fused with dam break scenario simulation, and aims to solve the problems that traditional dam site comparison and selection depend on static parameters and neglect extreme dam break risks and multi-dimensional disaster influences. The method comprises the following steps: obtaining multi-source geographic space data of a watershed, and primarily screening candidate dam sites; a three-dimensional geomechanical model is constructed and coupled with a fluid dynamic equation, and the dam break flood routing and sediment transportation process is simulated; quantifying a population evacuation time threshold value, infrastructure inundation loss and an ecological damage index; and combining the engineering cost and the power generation income, constructing a comprehensive risk income scoring model by adopting an entropy weight method, and sorting and preferably selecting the first three dam sites. The system correspondingly realizes the modularization function. According to the method, through fusion of dynamic simulation and multi-objective decision, scientificity, disaster resistance toughness and ecological economy collaboration of dam site comparison and selection are remarkably improved, and quantifiable and reusable technical support is provided for hydraulic engineering toughness planning.

Description

Technical Field

[0001] This invention belongs to the technical field of dam site selection for water conservancy projects. Specifically, it relates to a dam site selection method that integrates dam break scenario simulation. Background Technology

[0002] Dam site selection, as a core step in the early decision-making process of a project, directly determines the safety, economic efficiency, and ecological sustainability of the project throughout its entire life cycle. Traditional dam site selection methods largely rely on static spatial data such as topography, geological structure, and hydrological characteristics, ranking multiple options through empirical formulas or weighted scoring systems. Essentially, these methods are static evaluation frameworks based on historical data and static models. However, the complex external environment faced by modern large-scale water conservancy projects, including frequent extreme weather events, active geological activity, and dense populations and economies, makes it difficult for static evaluation models to capture the dynamic failure paths of the dam body under extreme conditions and their cascading disaster effects. This results in systemic risk blind spots in site selection schemes when dealing with sudden dam failure events.

[0003] Among these, dam break scenario simulation, as a key technical direction for assessing the extreme safety boundary of dams, aims to recreate the instability process of dams under inducing factors such as super-standard floods, seismic disturbances, or structural aging through numerical modeling, and to quantify its spatiotemporal impact on downstream river morphology, settlements, infrastructure, and ecosystems. This direction requires that the selection method not only identify the "optimal terrain node" but also predict the "weakest disaster link," thereby avoiding high-risk site selection at the source. However, current mainstream selection processes generally regard dam break simulation as a post-safety verification step independent of site selection decisions, lacking a dynamic coupling mechanism with core selection dimensions such as terrain selection, economic assessment, and ecological constraints. This makes it difficult for risk assessment results to back-optimize site selection parameters, creating a technical gap of "post-selection disaster verification, disaster-induced selection."

[0004] While some existing technologies incorporate DEM data processing or CFD flow field simulation to improve terrain identification accuracy or local structure optimization capabilities, they generally avoid integrated modeling of dam break dynamics and the overall impact field. Their risk assessments often remain at the level of static safety factor calculation or local scour depth estimation, and they have not established a quantitative mapping relationship between dam break probability and site selection parameters. Summary of the Invention

[0005] To address the problems in existing technologies, this invention provides a dam site selection method that integrates dam break scenario simulation. By constructing a candidate dam site preliminary screening model driven by multi-source geospatial data, and combining a dynamic simulation engine for dam break physical processes with a multi-dimensional disaster impact assessment system, it achieves the quantification and comprehensive comparison and ranking of the systematic risks of candidate dam sites under extreme conditions, thereby improving the scientific nature, foresight, and disaster resilience of dam site decisions.

[0006] According to one aspect of the present invention, a dam site selection method integrating dam break scenario simulation is provided, comprising: Acquire digital elevation model data for the entire target watershed, hydrogeological map data, historical rainfall intensity distribution data, spatial distribution data of densely populated areas, spatial coordinate data of important infrastructure, and boundary vector data of ecologically sensitive areas; Based on the digital elevation model data, the topology of the watershed network is extracted, potential confluence paths and topographic convergence areas are identified, and the distribution of rock permeability coefficient and fault orientation information in hydrogeological map data are combined to select an initial set of candidate dam sites that meet the basic geological stability requirements. For each candidate dam site in the initial set of candidate dam sites, a three-dimensional geomechanical finite element model is constructed. The structural material parameters of the dam body are set as follows: concrete compressive strength grade 50, elastic modulus 30,000 MPa, Poisson's ratio 0.2, and the foundation rock mass is set as granite with a uniaxial compressive strength of 120 MPa and an internal friction angle of 35 degrees. Based on the three-dimensional geomechanical finite element model, coupled fluid dynamics control equations, a dynamic simulation module for dam failure process is established. The initial failure triggering condition is set as continuous heavy rain causing the reservoir water level to exceed the design flood level by three meters and the duration is greater than twelve hours, or the peak ground acceleration is greater than 0.3 grams and the continuous vibration period is greater than five seconds. Run the dynamic simulation module for the dam break process and output the temporal and spatial distribution map of the dam break flood wavefront arrival, the distribution map of the maximum flood velocity field, the distribution map of the maximum flood depth, and the spatiotemporal evolution sequence of sediment transport concentration. Based on the spatial distribution map of the arrival time of the flood wavefront, the time threshold for the flood wavefront to reach the nearest populated area downstream is calculated. If the time is less than 30 minutes, the candidate dam site is marked as a high-risk site. Based on the distribution maps of maximum flood velocity and maximum flood depth, and by overlaying spatial coordinate data of important infrastructure, the total value loss of submerged infrastructure is calculated and quantified using an integral function of infrastructure replacement cost multiplied by the ratio of submerged water depth to structural flood resistance height. Based on the spatiotemporal evolution sequence of sediment transport concentration and combined with boundary vector data of ecologically sensitive areas, the damage index of sediment deposition to wetland ecosystem service functions is calculated. The index is defined as the ratio of sediment coverage area to total area of ​​sensitive areas multiplied by the ratio of average sediment deposition thickness to ecological threshold thickness. A multi-objective decision function is constructed, whose input variables include time threshold, total value loss, damage index, estimated engineering construction cost of candidate dam sites, and expected annual power generation revenue of candidate dam sites. Its output is a comprehensive risk-return score, which is obtained by linearly weighting the sum of variables after determining the weights of each variable using the entropy weight method. All candidate dam sites in the initial candidate dam site set are sorted in descending order according to their comprehensive risk-return scores, and the top three candidate dam sites with the highest scores are selected as the final recommended dam site schemes.

[0007] According to another aspect of the present invention, a dam site selection system integrating dam failure scenario simulation is provided, comprising: The multi-source spatial data acquisition module is used to acquire digital elevation model data of the entire target watershed, hydrogeological map data, historical rainfall intensity distribution data, spatial distribution data of densely populated areas, spatial coordinate data of important infrastructure, and boundary vector data of ecologically sensitive areas. The candidate dam site screening module is used to extract the topology of the watershed network based on digital elevation model data, identify potential confluence paths and topographic convergence areas, and combine the distribution of rock permeability coefficient and fault structure orientation information in hydrogeological map data to screen out an initial set of candidate dam sites that meet the basic geological stability requirements. The 3D geomechanical modeling module is used to construct a 3D geomechanical finite element model for each candidate dam site in the initial candidate dam site set. The structural material parameters of the dam body are set as follows: concrete compressive strength grade 50, elastic modulus 30,000 MPa, Poisson's ratio 0.2, and the foundation rock mass is set as granite with a uniaxial compressive strength of 120 MPa and an internal friction angle of 35 degrees. The dam break dynamic simulation module is used to establish a dynamic simulation module of the dam break process by coupling fluid dynamics control equations on the basis of a three-dimensional geomechanical finite element model. The initial dam break triggering condition is set as continuous heavy rain causing the reservoir water level to exceed the design flood level by three meters and the duration is greater than twelve hours, or the peak ground acceleration is greater than 0.3 grams and the continuous vibration period is greater than five seconds. The disaster impact field output module is used to run the dynamic simulation module of the dam break process and output the temporal and spatial distribution map of the dam break flood wavefront arrival, the distribution map of the maximum flood velocity field, the distribution map of the maximum flood depth, and the spatiotemporal evolution sequence of sediment transport concentration. The population risk quantification module is used to calculate the time threshold for the flood wavefront to reach the nearest downstream population settlement based on the spatial distribution map of the arrival time of the flood wavefront. If the time is less than 30 minutes, the candidate dam site is marked as a high-risk site. The economic damage assessment module is used to calculate the total value loss of submerged infrastructure based on the distribution map of the maximum flow velocity field and the distribution map of the maximum water depth of the flood, overlaid with the spatial coordinate data of important infrastructure. The total value loss is quantified by the integral function of the infrastructure replacement cost multiplied by the ratio of the submerged water depth to the structure's flood resistance height. The ecological damage calculation module is used to calculate the damage index of sediment deposition to wetland ecosystem service functions based on the spatiotemporal evolution sequence of sediment transport concentration and combined with the boundary vector data of ecologically sensitive areas. The index is defined as the ratio of sediment coverage area to the total area of ​​sensitive areas multiplied by the ratio of average sediment deposition thickness to ecological threshold thickness. The comprehensive scoring and decision module is used to construct a multi-objective decision function. Its input variables include time threshold, total value loss, damage index, estimated engineering construction cost of candidate dam sites, and expected annual power generation revenue of candidate dam sites. Its output is a comprehensive risk-return score, which is obtained by linearly weighting the sum of the weights of each variable using the entropy weight method. The dam site selection module is used to sort all candidate dam sites in the initial candidate dam site set in descending order according to the comprehensive risk-return score, and select the top three candidate dam sites as the final recommended dam site schemes.

[0008] In the dynamic simulation module of the dam break process, the fluid dynamics governing equations adopt the two-dimensional shallow water wave equations, and its continuity equation is: , Its momentum equation is ; ; Where h is the water depth, u is the flow velocity in the x direction, v is the flow velocity in the y direction, g is the gravitational acceleration, zb is the riverbed elevation, and Sfx and Sfy are the friction slope terms, calculated using the Manning formula. The Manning roughness coefficient is assigned a value according to the land cover type: 0.035 for forest land, 0.025 for cultivated land, 0.04 for bare rock, and 0.015 for hardened urban ground.

[0009] The calculation of the spatiotemporal evolution sequence of sediment transport concentration uses the sediment continuity equation: ; Where C is the sediment concentration, Dx and Dy are the sediment diffusion coefficients in the x and y directions, respectively, and are taken as one-tenth of the flow velocity multiplied by the characteristic length. w is the sediment settling velocity, which is calculated using the Stokes formula based on the sediment particle size. E is the riverbed scour source term, which is activated when the bottom shear stress is greater than the critical starting stress. The critical starting stress is taken as 0.5 N per square meter.

[0010] The specific steps for determining the weights of each variable using the entropy weight method are as follows: Construct the original data matrix of n candidate dam sites on m evaluation indicators; positively process the benefit-type indicators and negatively process the cost-type indicators; calculate the proportion pij of the i-th scheme under the j-th indicator; and calculate the entropy value of the j-th indicator. , Where k = 1 / lnn, the difference coefficient of the j-th indicator is calculated. Final weight .

[0011] The estimated construction cost is calculated based on the dam axis length, maximum dam height, foundation excavation volume, and total concrete pouring volume at the candidate dam site. ; The formula is: multiply the comprehensive unit price of dam body per unit length by the length of dam axis, add the incremental cost of dam body per unit height by the maximum dam height, add the unit price of earthwork excavation per unit volume by the foundation excavation volume, and add the unit price of concrete materials and construction per unit volume by the total amount of concrete poured.

[0012] The expected annual power generation revenue is calculated based on the multi-year average runoff at the candidate dam site, the design head, the overall turbine efficiency, and the annual utilization hours. The calculation formula is as follows: Annual power generation revenue = annual power generation × grid connection price, where the comprehensive efficiency of the turbine is taken as 85%, the annual utilization hours are taken as 4,000 hours, and the grid connection price is taken as 0.4 yuan per kilowatt-hour.

[0013] In the spatial distribution data of population settlements, each settlement is assigned a population density attribute. When calculating the time threshold for the arrival of the flood front in the nearest downstream population settlement, the total number of affected people is also output. The calculation method is to sum the product of the population density and the area of ​​the settlement for all settlements whose arrival time is less than the threshold.

[0014] Spatial coordinate data for critical infrastructure includes the spatial location and structural attributes of bridges, substations, railway lines, highway hubs, hospitals, and schools. The flood resistance height of the structure is set according to different facility types: five meters for bridge piers, three meters for substation equipment floors, two meters for railway subgrades, one meter for highway pavements, and 1.5 meters for the ground floor of hospitals and schools.

[0015] The boundary vector data of ecologically sensitive areas includes the spatial boundaries of the core area of ​​nature reserves, wetland parks, habitats of rare species, and water conservation areas. The ecological threshold thickness is set according to different ecological types: 0.3 meters for wetland parks, 0.1 meters for habitats of rare species, and 0.5 meters for water conservation areas.

[0016] In the preliminary screening module for candidate dam sites, the threshold for rock permeability coefficient is set to less than 0.1 meters per day, areas where the angle between the fault structure and the dam axis is less than 30 degrees are excluded, and the threshold for the convergence angle of the terrain convergence area is set to greater than 120 degrees.

[0017] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. This invention constructs a dynamic simulation engine for the physical process of dam breakage, transforming the failure behavior of the dam body under extreme conditions into a quantifiable spatiotemporal disaster field, breaking through the limitations of traditional dam site selection methods that rely solely on static terrain parameters and empirical coefficients; by introducing a triple assessment dimension of population risk time threshold, economic damage integral function, and ecological damage composite index, it achieves a systematic, spatial, and quantitative characterization of the secondary disaster impact of dam breakage.

[0018] 2. By establishing a comprehensive scoring and decision-making model that integrates engineering costs, power generation revenue, and multidimensional risk factors, the dam site selection results simultaneously meet the requirements of economic feasibility, safety redundancy, and ecological sustainability. By embedding two-dimensional shallow water wave equations and sediment transport equations into the simulation module, and adopting Manning coefficient zoning assignment and Stokes settlement model, the simulation accuracy of flood evolution and sediment diffusion range is significantly improved. By automatically determining the weights of each evaluation index using the entropy weight method, the subjective bias of human weighting is avoided, enhancing the objectivity and comparability of the decision results. By parameterizing and solidifying all dam material parameters, geomechanical conditions, trigger thresholds, economic parameters, and ecological thresholds into the system module, the repeatability and engineering applicability of the method are ensured. Attached Figure Description

[0019] Figure 1 This is a schematic diagram of the structure of the unmanned aerial vehicle platform in an embodiment of the present invention; Figure 2 This is a flowchart of the dynamic path planning in an embodiment of the present invention; Figure 3 This is a diagram of the multi-source heterogeneous data fusion and analysis architecture in an embodiment of the present invention. Detailed Implementation

[0020] Example 1: This invention provides a dam site selection method that integrates dam-break scenario simulation. Its core lies in constructing a candidate dam site preliminary screening model driven by multi-source geospatial data, combining it with a dynamic simulation engine for dam-break physical processes and a multi-dimensional disaster impact assessment system. This enables the quantification and comprehensive ranking of the systematic risks of candidate dam sites under extreme conditions, thereby embedding the concept of full life-cycle safety redundancy design into the engineering planning stage, enhancing the scientific rigor, foresight, and disaster resilience of dam site decisions. This method transforms the dam-break physical process into a calculable and spatially applicable disaster impact field, and couples it with multi-dimensional indicators such as population safety, economic losses, ecological damage, engineering costs, and power generation revenue for analysis. Ultimately, it outputs recommended dam site schemes with clear ranking criteria, completely overcoming the limitations of traditional methods that rely solely on static terrain parameters and empirical coefficients.

[0021] The method includes the following steps: acquiring digital elevation model data of the entire target watershed, hydrogeological map data, historical rainfall intensity distribution data, spatial distribution data of densely populated areas, spatial coordinate data of important infrastructure, and boundary vector data of ecologically sensitive areas; extracting the topology of the watershed's river system network based on the digital elevation model data, identifying potential confluence paths and topographic convergence areas, and combining the distribution of rock permeability coefficients and fault orientation information in the hydrogeological map data to select an initial set of candidate dam sites that meet the basic geological stability requirements; for each candidate dam site in the initial set of candidate dam sites, constructing a three-dimensional geomechanical finite element model and defining the dam body. The structural material parameters are: concrete compressive strength grade 50, elastic modulus 30,000 MPa, Poisson's ratio 0.2; the dam foundation rock mass is set as granite with a uniaxial compressive strength of 120 MPa and an internal friction angle of 35 degrees. Based on a three-dimensional geomechanical finite element model, coupled with fluid dynamics control equations, a dynamic simulation module for the dam break process is established. The initial failure triggering conditions are set as either continuous heavy rainfall causing the reservoir water level to exceed the design flood level by 3 meters for a duration greater than 12 hours, or a peak ground acceleration greater than 0.3 g for a duration greater than 5 seconds. The dynamic simulation module for the dam break process is run, outputting the temporal and spatial distribution map of the dam break flood wavefront arrival time, the maximum flood flow, and other parameters. The study analyzed velocity field distribution maps, maximum flood depth distribution maps, and the spatiotemporal evolution sequence of sediment transport concentration. Based on the spatiotemporal distribution map of the flood wavefront arrival time, it calculated the time threshold for the flood wavefront to reach the nearest downstream population settlement; if this time is less than 30 minutes, the candidate dam site is marked as a high-risk site. Using the maximum flood velocity field distribution map and the maximum flood depth distribution map, overlaid with spatial coordinate data of important infrastructure, the study calculated the total value loss of submerged infrastructure, quantifying it using an integral function of the infrastructure replacement cost multiplied by the ratio of submerged water depth to the structure's flood-resistant height. Finally, based on the spatiotemporal evolution sequence of sediment transport concentration and combined with boundary vector data of ecologically sensitive areas, it calculated the sediment transport concentration. The damage index of sedimentation to wetland ecosystem services is defined as the ratio of sediment-covered area to the total area of ​​the sensitive area multiplied by the ratio of average sediment deposition thickness to ecological threshold thickness. A multi-objective decision function is constructed, with input variables including time threshold, total value loss, damage index, estimated engineering construction cost of candidate dam sites, and expected annual power generation revenue of candidate dam sites. The output is a comprehensive risk-return score, which is obtained by linearly weighting the variables after determining their weights using the entropy weight method. All candidate dam sites in the initial candidate dam site set are sorted in descending order according to their comprehensive risk-return scores, and the top three candidate dam sites are selected as the final recommended dam site schemes.

[0022] In this process, acquiring digital elevation model data (DEM) of the entire target watershed, hydrogeological map data, historical rainfall intensity distribution data, spatial distribution data of densely populated areas, spatial coordinate data of critical infrastructure, and boundary vector data of ecologically sensitive areas are the prerequisites and foundations for the entire method implementation. The DEM data must cover the entire target watershed with a spatial resolution of no less than 30 meters to accurately depict topographic relief and confluence paths; the hydrogeological map data must include rock strata type, permeability coefficient distribution map, and fault strike and dip information to assess geological stability; the historical rainfall intensity distribution data must include the spatial distribution and intensity levels of extreme rainfall events over the past 30 years to set dam-break trigger thresholds; the spatial distribution data of densely populated areas must be stored in a planar vector format, with each area assigned a population density attribute, in units of people per square kilometer; and the spatial coordinate data of critical infrastructure... The data should include the spatial location and structural attributes of bridges, substations, railway lines, highway hubs, hospitals, and schools. Structural attributes include the flood resistance height: 5 meters for bridge piers and abutments, 3 meters for substation equipment floors, 2 meters for railway subgrades, 1 meter for highway surfaces, and 1.5 meters for the ground floor of hospitals and schools. The boundary vector data for ecologically sensitive areas should include the spatial boundaries of the core areas of nature reserves, wetland parks, rare species habitats, and water conservation areas. Each ecological type should be assigned an ecological threshold thickness: 0.3 meters for wetland parks, 0.1 meters for rare species habitats, and 0.5 meters for water conservation areas. All data must be projected in a unified coordinate system, using the Gauss-Krüger projection, with the central meridian aligned with the longitude of the target watershed center to ensure geometric consistency in the spatial overlay analysis.

[0023] In this step, the topology of the watershed's river network is extracted based on digital elevation model (DEM) data. Potential confluence points and topographic convergence areas are identified. Combined with information on rock permeability distribution and fault orientation from hydrogeological maps, an initial set of candidate dam sites that meet the basic geological stability requirements is selected. This step first fills in depressions in the DEM data to eliminate false confluence points caused by data noise. Then, the flow direction matrix and cumulative confluence are calculated, setting a threshold of 5% of the watershed area to extract the main river network. Based on the river network, confluence points are identified—where two or more tributaries meet, as these points naturally possess water storage potential. Simultaneously, topographic convergence areas are identified—river valley sections where contour lines curve inwards on both sides with a convergence angle greater than 120 degrees. These areas are conducive to dam axis layout and reservoir formation. After obtaining potential dam site candidates, hydrogeological data were overlaid to exclude areas with a rock permeability coefficient greater than 0.1 meters per day, as these areas are prone to seepage leading to dam foundation instability. Areas where the angle between the fault strike and the proposed dam axis is less than 30 degrees were also excluded, as these areas are susceptible to fault displacement under seismic loads, potentially damaging the dam's structural integrity. The final selected candidate points constitute the initial set of candidate dam sites. Each candidate point is recorded with its geographic coordinates, cumulative runoff volume, convergence angle, rock permeability coefficient, distance from the nearest fault, and the angle between the convergence and convergence points.

[0024] In this step, a three-dimensional geomechanical finite element model is constructed for each candidate dam site in the initial set of candidate dam sites. The structural material parameters of the dam body are set as follows: concrete compressive strength grade 50, elastic modulus of 30,000 MPa, and Poisson's ratio of 0.2. The foundation rock mass is set as granite with a uniaxial compressive strength of 120 MPa and an internal friction angle of 35 degrees. The modeling process first extracts cross-sectional elevation data along the dam axis based on the topographic profile at the candidate dam site to construct the geometric outline of the dam body. The dam crest width is uniformly set to 10 meters, and the upstream and downstream slope ratio is set to 1:0.8. The dam foundation extends downward to 20 meters below the bedrock surface and laterally to 50 meters outside the dam abutment to ensure that the stress diffusion zone is completely included in the model. Regarding material properties, the dam concrete was modeled using a linear elastic constitutive model with a density of 2,400 kg / m³ and a coefficient of thermal expansion of 10⁻⁵ degrees Celsius. The dam foundation granite was modeled using a Mohr-Coulomb elastoplastic constitutive model with a density of 2,700 kg / m³, an elastic modulus of 50,000 MPa, a Poisson's ratio of 0.25, and a cohesion of 10 MPa. The model boundary conditions were set as follows: fixed constraint at the bottom, horizontal constraint applied laterally but allowing free vertical deformation, and a free boundary at the top. Tetrahedral elements were used for mesh generation, with element size controlled within two meters in the dam body region, densified to one meter in the stress concentration zone of the dam foundation, and widened to five meters in the far-field region. The total number of elements was controlled between 500,000 and 800,000 to ensure a balance between computational accuracy and efficiency.

[0025] In this step, based on the three-dimensional geomechanical finite element model, fluid dynamics control equations are coupled to establish a dynamic simulation module for the dam-break process. The initial dam-break triggering condition is set as either continuous heavy rainfall causing the reservoir water level to exceed the design flood level by three meters for a duration greater than twelve hours, or a peak ground acceleration greater than 0.3 grams and a sustained vibration period greater than five seconds. The dam-break simulation module uses a two-dimensional shallow water wave equation set as the control equation, and its continuity equation is: , Its momentum equation is: ; ; Where h is the water depth, u is the flow velocity in the x-direction, v is the flow velocity in the y-direction, g is the gravitational acceleration, zb is the riverbed elevation, and Sfx and Sfy are the friction slope terms, calculated using the Manning formula. The Manning roughness coefficient is assigned a value according to the surface cover type: 0.035 for forest, 0.025 for cultivated land, 0.04 for bare rock, and 0.015 for hardened urban surfaces. The simulation area covers a range from 5 km upstream to 20 km downstream of the dam site, with a grid resolution of 10 meters, a time step of 0.5 seconds, and a total simulation duration set as the time required for the flood wave to propagate to the farthest affected point after dam failure plus a two-hour margin. The initial failure morphology is set as a gap of 50 meters wide and 10 meters deep appearing in the middle of the dam body. The gap expansion rate is updated in real time based on fluid-structure interaction calculations until the dam body completely fails. The trigger condition monitoring module reads rainfall data from meteorological stations or seismic monitoring network data in real time. When any trigger condition is met, the dam failure simulation process is automatically started.

[0026] In this step, the dynamic simulation module for the dam-break process is run, outputting the spatiotemporal distribution map of the flood wavefront arrival time, the maximum flood velocity field distribution map, the maximum flood depth distribution map, and the spatiotemporal evolution sequence of sediment transport concentration. The flood wavefront arrival time-space distribution map records the time point at which the flood front first arrives at each spatial grid, with a time resolution of one second and a spatial resolution of ten meters; the maximum flood velocity field distribution map records the maximum instantaneous velocity value at each grid point during the simulation, in meters per second; the maximum flood depth distribution map records the maximum instantaneous water depth value at each grid point during the simulation, in meters; the spatiotemporal evolution sequence of sediment transport concentration uses the sediment continuity equation: ; Where C represents the sediment concentration, Dx and Dy are the sediment diffusion coefficients in the x and y directions, respectively, and are taken as one-tenth of the flow velocity multiplied by the characteristic length, with the characteristic length being a grid size of ten meters. w is the sediment settling velocity, calculated using the Stokes formula based on the sediment particle size, with a particle size range of 0.01 mm to 0.1 mm, corresponding to a settling velocity of 0.001 m / s to 0.01 m / s. E is the riverbed scour source term, activated when the bottom shear stress exceeds the critical initiation stress by 0.5 N / m², and the scour rate is proportional to the overstress value, with a proportionality coefficient of 0.001 m / s. The sediment concentration is output as three-dimensional raster data with a time step of ten minutes, covering the entire simulation area.

[0027] In this step, based on the spatial distribution map of the flood wavefront arrival time, the time threshold for the flood wavefront to reach the nearest downstream populated area is calculated. If this time is less than 30 minutes, the candidate dam site is marked as a high-risk site. The calculation process first extracts the boundary polygons of all populated areas and calculates the coordinates of the geometric center point of each area. Then, in the spatial distribution map of the flood wavefront arrival time, the arrival time value corresponding to the center point of each area is queried. The minimum arrival time value among all areas is taken as the time threshold for the candidate dam site. Simultaneously, the total affected population is calculated by iterating through all areas. If the arrival time of the center point of an area is less than 30 minutes, the population density of that area is multiplied by its area and added to the total population. If the time threshold is less than 30 minutes, a penalty coefficient of 0.5 is applied to the candidate dam site in the subsequent decision function, forcibly reducing its overall score and ensuring that high-risk sites are effectively excluded.

[0028] In this step, based on the maximum flood velocity distribution map and the maximum flood depth distribution map, and overlaying the spatial coordinate data of important infrastructure, the total value loss of the submerged infrastructure is calculated. This loss is quantified using an integral function that multiplies the infrastructure replacement cost by the ratio of the submerged water depth to the structural flood resistance height. Specifically, the calculation process is as follows: For each type of infrastructure, its spatial location and structural flood resistance height are extracted; in the maximum flood depth distribution map, the maximum water depth value of the grid where the facility is located is queried; if the maximum water depth value is greater than the structural flood resistance height, the facility is determined to be submerged; the submerged depth is calculated as the maximum water depth value minus the structural flood resistance height; the loss ratio function is defined as the integral of the ratio of the submerged depth to the structural flood resistance height, with an upper limit of the integral equal to the actual submerged depth and a lower limit of zero. The integrand is a piecewise linear function; when the ratio is less than one, the function value equals the ratio; when the ratio is greater than one, the function value is always one; the final loss value is the infrastructure replacement cost multiplied by the loss ratio. The replacement cost of bridges is estimated at 500,000 yuan per linear meter, substations at 500 million yuan per station, railway lines at 100 million yuan per kilometer, highway hubs at 300 million yuan per location, hospitals at 200 million yuan per institution, and schools at 100 million yuan per school. The total value loss is obtained by summing the losses of all submerged facilities.

[0029] In this step, based on the spatiotemporal evolution sequence of sediment transport concentration and combined with boundary vector data of ecologically sensitive areas, a damage index of sediment deposition on wetland ecosystem services is calculated. This index is defined as the ratio of sediment-covered area to the total area of ​​the sensitive area multiplied by the ratio of average sediment deposition thickness to the ecological threshold thickness. The calculation process first sets a concentration threshold of 10 kg per cubic meter for the sediment concentration raster data at each time step; areas exceeding this threshold are identified as sediment-covered areas. The covered areas from all time steps are spatially superimposed, and the union is taken to obtain the final total sediment-covered area. The ratio of the covered area to the total area of ​​the ecologically sensitive area is calculated. Simultaneously, for each grid within the covered area, the sediment deposition is integrated. Deposition equals concentration multiplied by deposition velocity multiplied by the time step, and the sum is used to obtain the total deposition thickness. The spatial average value is taken as the average deposition thickness. The ratio of the average deposition thickness to the ecological threshold thickness is calculated. The two ratios are multiplied to obtain the damage index. This index is dimensionless, ranging from zero to positive infinity; a larger value indicates more severe ecological damage.

[0030] In this step, a multi-objective decision function is constructed. Its input variables include a time threshold, total value loss, damage index, estimated construction cost of the candidate dam site, and expected annual power generation revenue of the candidate dam site. The output is a comprehensive risk-return score, which is obtained by linearly weighting the variables after determining their weights using the entropy weight method. The estimated construction cost is calculated based on the dam axis length, maximum dam height, foundation excavation volume, and total concrete pouring volume at the candidate dam site. This is achieved by multiplying the comprehensive unit price per unit length of the dam body by the dam axis length, adding the incremental cost per unit height of the dam body multiplied by the maximum dam height, adding the unit price per unit volume of earthwork excavation multiplied by the foundation excavation volume, and adding the unit price per unit volume of concrete materials and construction multiplied by the total concrete pouring volume. Specifically, the comprehensive unit price per unit length of the dam body is set at 80,000 yuan per meter, the incremental cost per unit height of the dam body is set at 50,000 yuan per meter of height increase, the unit price per unit volume of earthwork excavation is set at 25 yuan per cubic meter, and the unit price per unit volume of concrete materials and construction is set at 800 yuan per cubic meter.

[0031] The expected annual power generation revenue is calculated based on the multi-year average runoff at the candidate dam site, the design head, the overall turbine efficiency, and the annual utilization hours. The calculation formula is as follows: ; Annual power generation = multi-year average runoff × design head × turbine comprehensive efficiency × 9.81 / 36 million kilowatts; annual power generation revenue = annual power generation × grid connection price, where the turbine comprehensive efficiency is taken as 85%, the annual utilization hours are taken as 4,000 hours, and the grid connection price is taken as 0.4 yuan per kilowatt-hour. The specific steps for determining the weights of each variable using the entropy weight method are as follows: construct the original data matrix of n candidate dam sites on m evaluation indicators. ; Benefit-oriented indicators are positively evaluated, while cost-oriented indicators are negatively evaluated. ; Calculate the proportion pij of the ith option under the j-th indicator, and calculate the entropy value of the j-th indicator. ,in Calculate the coefficient of difference for the j-th indicator. Final weight The formula for linear weighted summation is: , Standardization is achieved using the range method.

[0032] In this step, all candidate dam sites in the initial candidate dam site set are sorted in descending order of their comprehensive risk-reward score, and the top three candidate dam sites with the highest scores are selected as the final recommended dam site schemes. A stable sorting algorithm is used in the sorting process to ensure that candidate dam sites with the same score are sorted in ascending order by their dam site numbers. The output includes the comprehensive score value for each recommended dam site, the original values ​​of each sub-indicator, a weight allocation table, and an index of key results from the dam break simulation. The final recommended scheme is accompanied by a decision-making explanation document, which details the comparison of the advantages and disadvantages of each dam site and the reasons for the recommendation, for the reference of engineering decision-makers.

[0033] Through the aforementioned steps, this invention achieves full automation and quantification of the entire process from data acquisition, candidate screening, physical simulation, risk quantification to comprehensive decision-making, significantly improving the scientific rigor and risk resistance of dam site selection. By introducing dynamic simulation of dam breaks and multi-dimensional impact assessment, site selection decisions are no longer limited to topography and cost, but comprehensively incorporate the triple constraints of safety, economy, and ecology, providing solid technical support for the resilient planning of modern water conservancy projects.

Claims

1. A dam site selection method integrating dam break scenario simulation, characterized in that, Includes the following steps: S1. Obtain multi-source spatial data of the target watershed, including digital elevation model data, hydrogeological map data, historical rainfall intensity distribution data, spatial distribution data of densely populated areas, spatial coordinate data of important infrastructure, and boundary vector data of ecologically sensitive areas; S2. Based on the digital elevation model data, extract the topology of the watershed network, identify potential confluence paths and topographic convergence areas, and combine the distribution of rock permeability coefficients and fault orientation information in the hydrogeological map data to select an initial set of candidate dam sites that meet the basic geological stability requirements. S3. For each candidate dam site in the initial candidate dam site set, construct a three-dimensional geomechanical finite element model; S4. Based on the three-dimensional geomechanical finite element model, couple the fluid dynamics control equations to establish a dynamic simulation module for the dam failure process, and set the initial failure triggering conditions. S5. Run the dynamic simulation module of the dam break process and output the temporal and spatial distribution map of the arrival time of the dam break flood wavefront, the distribution map of the maximum flow velocity field of the flood, the distribution map of the maximum water depth of the flood, and the temporal and spatial evolution sequence of sediment transport concentration. S6. Based on the spatial distribution map of the arrival time of the dam-break flood wavefront, the distribution map of the maximum flow velocity field of the flood, the distribution map of the maximum water depth of the flood, and the spatiotemporal evolution sequence of the sediment transport concentration, calculate the population risk quantification index, the economic damage assessment index, and the ecological damage calculation index. S7. Construct a multi-objective decision function, whose input variables include the population risk quantification index, the economic damage assessment index, the ecological damage calculation index, the estimated engineering construction cost of the candidate dam site, and the expected annual power generation revenue of the candidate dam site. Its output is a comprehensive risk-return score, which is obtained by linearly weighting the sum of the weights of each variable using the entropy weight method. S8. Sort all candidate dam sites in the initial candidate dam site set in descending order according to the comprehensive risk-return score, and select the candidate dam sites with the highest scores as the final recommended dam site scheme.

2. The method according to claim 1, characterized in that, The steps for selecting the initial set of candidate dam sites that meet the basic geological stability requirements specifically include: Identify potential confluence path intersections and terrain convergence regions in the digital elevation model data; Based on the aforementioned hydrogeological map data, areas with a rock stratum permeability coefficient greater than 0.1 meters per day or an angle of less than 30 degrees between the fault structure and the dam axis were excluded. Regions with convergence angles greater than 120 degrees in the topographic convergence area were selected as the initial set of candidate dam sites that meet the basic geological stability requirements.

3. The method according to claim 1, characterized in that, The specific steps for constructing the three-dimensional geomechanical finite element model include: The structural material parameters of the dam are set as follows: concrete compressive strength grade 50, elastic modulus of 30,000 MPa, and Poisson's ratio of 0.

2. The foundation rock mass of the dam is set to be granite, with a uniaxial compressive strength of 120 MPa and an internal friction angle of 35 degrees.

4. The method according to claim 1, characterized in that, The steps for setting the initial failure trigger conditions specifically include: The initial failure trigger condition is set as continuous heavy rainfall causing the reservoir water level to exceed the design flood level by three meters and lasting for more than twelve hours; or The initial collapse trigger condition is set as a peak ground acceleration greater than 0.3 g and a continuous vibration period greater than 5 seconds.

5. The method according to claim 1, characterized in that, In the dynamic simulation module for the dam break process, the fluid dynamics governing equations adopt the two-dimensional shallow water wave equations, and its continuity equation is: ； Its momentum equation is: ； ； Where h is the water depth, u is the flow velocity in the x direction, v is the flow velocity in the y direction, g is the gravitational acceleration, zb is the riverbed elevation, and Sfx and Sfy are the friction slope terms. The friction slope terms are calculated using the Manning formula. The Manning roughness coefficient is assigned a value according to the land cover type: 0.035 for forest land, 0.025 for cultivated land, 0.04 for bare rock, and 0.015 for hardened urban ground.

6. The method according to claim 1, characterized in that, The calculation of the spatiotemporal evolution sequence of sediment transport concentration adopts the sediment continuity equation: ； Wherein, C is the sediment concentration, Dx and Dy are the sediment diffusion coefficients in the x and y directions, respectively, and the diffusion coefficient is taken as one-tenth of the flow velocity multiplied by the characteristic length, w is the sediment settling velocity, which is calculated using the Stokes formula based on the sediment particle size, and E is the riverbed scour source term, which is activated when the bottom shear stress is greater than the critical starting stress, and the critical starting stress is taken as 0.5 N per square meter.

7. The method according to claim 1, characterized in that, The specific steps for calculating the population risk quantification index, economic damage assessment index, and ecological damage calculation index include: Based on the spatial distribution map of the arrival time of the flood wavefront, the time threshold for the flood wavefront to reach the nearest populated area downstream is calculated. If the time is less than 30 minutes, the candidate dam site is marked as a high-risk site, and the total number of affected people is output at the same time. The total number of affected people is calculated by summing the product of the population density and the area of ​​the populated area for all populated areas where the flood wavefront arrival time is less than the time threshold. Based on the maximum flow velocity field distribution map and the maximum water depth distribution map of the flood, and superimposed with the spatial coordinate data of the important infrastructure, the total value loss of the submerged infrastructure is calculated, and quantified by the integral function of the infrastructure replacement cost multiplied by the ratio of the submerged water depth to the structure's flood resistance height. The spatial coordinate data of the critical infrastructure includes the spatial location and structural attributes of bridges, substations, railway lines, highway hubs, hospitals, and schools. The flood resistance height of the structure is set according to different facility types.

8. The method according to claim 1, characterized in that, The steps for calculating the damage index of sedimentation to wetland ecosystem services specifically include: The damage index is defined as the ratio of the sediment-covered area to the total area of ​​the ecologically sensitive zone multiplied by the ratio of the average sediment deposition thickness to the ecological threshold thickness. The boundary vector data of the ecologically sensitive areas includes the spatial boundaries of the core area of ​​nature reserves, wetland parks, habitats of rare species, and water conservation areas; The ecological threshold thickness is set according to different ecological types.

9. The method according to claim 1, characterized in that, The estimated construction cost of the project is calculated based on the length of the dam axis, the maximum dam height, the volume of foundation excavation, and the total amount of concrete poured at the candidate dam site. It is calculated by multiplying the comprehensive unit price of the dam body per unit length by the length of the dam axis, adding the incremental cost of the dam body per unit height by the maximum dam height, adding the unit price of earthwork excavation per unit volume by the volume of foundation excavation, and adding the unit price of concrete materials and construction per unit volume by the total amount of concrete poured. The calculation of the expected annual power generation revenue is based on the multi-year average runoff at the candidate dam site, the design head height, the overall efficiency of the turbine, and the annual utilization hours. The annual power generation revenue is equal to the annual power generation multiplied by the online electricity price.

10. The method according to claim 1, characterized in that, The steps for determining the weights of each variable using the entropy weight method specifically include: Construct the original data matrix of n candidate dam sites on m evaluation indicators; Benefit-oriented indicators are positively evaluated, while cost-oriented indicators are negatively evaluated. Calculate the proportion pij of the i-th option under the j-th indicator; Calculate the entropy value of the j-th index. , where k = 1 / lnn; Calculate the difference coefficient of the j-th indicator: dj = 1 − ej; Final weight .

Images