Pumped storage power plant upper reservoir siting method and system based on digital elevation model

Abstract

The application belongs to the field of power systems, and relates to a pumped storage power station upper reservoir site selection method and system based on a digital elevation model; digital elevation model data is obtained and preprocessed, a priority flood filling algorithm is used to detect closed depressions, a water level reservoir capacity curve is used to solve the water level that meets the required reservoir capacity through binary search, an extended disc geometric constraint is applied to limit the reservoir inundation range, a closed reservoir boundary polygon is generated and dam body volume is estimated, and output is ranked by the number of engineering conditions met. The system includes a data preprocessing module, a depression detection module, a water level solving module, a geometric constraint module, a boundary generation module, a dam body estimation module, a hierarchical ranking module and a sensitivity analysis module. The application realizes end-to-end automated site selection from digital elevation model data to construction-level candidate sites, does not require manual topographic interpretation, can retain boundary candidates in a hierarchical ranking manner, and supports real-time sensitivity analysis.

Description

Technical Field

[0001] This application belongs to the field of power systems, specifically relating to a method and system for selecting the site of the upper reservoir of a pumped storage power station based on a digital elevation model. Background Technology

[0002] With the accelerated global energy structure transformation and the continuous increase in the penetration rate of renewable energy sources such as wind power and photovoltaics, the demand for pumped storage power station installed capacity is showing a rapid growth trend. The site selection of the upper reservoir of a pumped storage power station is a key part of the project feasibility study. The upper reservoir must occupy a natural topographic depression to store the required storage capacity with minimal dam construction, because dam investment often dominates the total investment of the entire project.

[0003] Among them, automated reservoir site selection based on digital elevation models (DEMs) is an important research direction in the current engineering technology field. DEMs can accurately express surface elevation information in digital form, providing a data foundation for quantitatively assessing the engineering suitability of candidate basins. By performing topographic analysis on DEMs, closed depressions can be automatically identified, reservoir capacity parameters can be calculated, and boundary polygons can be generated, thus replacing the traditional method of manually interpreting topographic maps.

[0004] However, existing technologies have the following shortcomings: Early site selection methods based on Geographic Information Systems (GIS) employed terrain feature extraction and spatial query techniques, but these typically required extensive manual parameter tuning and could only produce qualitative suitability maps, failing to provide quantitatively ranked candidate solutions with boundary polygons, making them difficult to directly apply to engineering design. While recently developed GIS multi-criteria decision analysis frameworks have expanded to topology-driven site selection models, they still lack automated depression identification and closed boundary generation functions, failing to achieve end-to-end automation from terrain data to engineering candidates. The dam axis model perpendicular to valley lines proposed in Chinese patent literature exhibits instability in rugged terrain and multi-saddle areas; local valley line tracing may result in dam sections extending several kilometers around the reservoir perimeter instead of crossing true closed boundaries, leading to severe distortion in dam volume estimation. Furthermore, while karst depression identification technology has been applied to pumped-storage reservoir construction in karst areas, it is limited to specific geological environments and cannot be extended to general terrain conditions. A deeper limitation lies in the fact that existing methods treat the site selection problem as a binary classification problem, only determining suitability or unsuitability, discarding potentially feasible boundary candidate solutions under different design parameters. Since the required reservoir capacity is a project-level variable, a basin deemed too shallow for a project with a large required reservoir capacity might actually be an ideal site for a project with a smaller required reservoir capacity. Binary screening methods cannot identify this parameter sensitivity. Therefore, there is an urgent need for an end-to-end site selection method that can automatically identify all closed depressions from digital elevation model data, quantitatively assess the engineering suitability of each depression, generate construction-level boundary polygons, and retain boundary candidates in a hierarchical ranking manner. Summary of the Invention

[0005] To achieve the above objectives, this application provides the following technical solution: The purpose of this invention is to provide a method and system for selecting the upper reservoir of a pumped storage power station based on a digital elevation model, so as to realize end-to-end automated site selection from digital elevation model data to construction-level candidate sites, without the need for manual terrain interpretation, and can retain boundary candidates in a hierarchical ranking manner and support real-time sensitivity analysis. It can effectively solve the problems of long site selection process, strong subjectivity, inability to automate end-to-end, and inability to identify parameter sensitivity in the existing technology.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is as follows: A method for selecting the site of the upper reservoir of a pumped storage power station based on a digital elevation model includes the following steps: Step 1: Obtain digital elevation model data for the target area and preprocess it. Generate a regular grid digital elevation model from the scattered elevation data using an interpolation algorithm and perform smoothing processing. Step 2: Use the priority flood filling algorithm to detect depressions in the preprocessed digital elevation model, identify closed depressions and extract feature parameters; Step 3: For each depression, use a binary search based on the water level-capacity curve to find the water level that satisfies the required reservoir capacity; Step 4: Apply extended disk geometric constraints to each depression to limit the reservoir inundation range; Step 5: Extract water level contour lines and perform Douglas-Peucker simplification and minimum side length constraint to generate a closed boundary polygon; Step 6: Estimate the dam volume and dam length based on the low elevation zone of the perimeter; Step 7: Sort and output the candidate sites according to the number of sites that meet the engineering conditions; Step 8: Support real-time parameter adjustment and ranking updates based on pre-calculated water level curves.

[0007] Furthermore, the interpolation algorithm in step 1 is a bilinear interpolation method or a nearest neighbor interpolation method, and the smoothing process is performed using a Gaussian kernel filter.

[0008] Furthermore, in step 2, the min-heap priority queue is initialized with all boundary cells. Boundary cells are cells on the outer edge of the grid or cells adjacent to invalid value cells. Invalid value cells are regarded as walls, i.e., infinite elevation. Pop the lowest elevation cell from the priority queue and propagate inwards, raising the inner cells to their overflow predecessor elevation; Connectivity component labeling is performed on the difference between the filled digital elevation model and the original digital elevation model; When using the priority flood filling algorithm for depression detection, it also includes: Applying a high-altitude constraint, and only taking depressions whose seed elevation is not less than a preset threshold as candidates; Initializing a min-heap priority queue with all boundary grid cells, where the boundary grid cells are grid cells on the outer edge of the grid or grid cells adjacent to no-data NaN grid cells, and the NaN grid cells are regarded as walls, that is, infinite elevation; Popping the grid cell with the lowest elevation from the priority queue, propagating inward, and raising the internal grid cell to the elevation of its overflow predecessor, which is equivalent to filling all depressions to the overflow elevation; Performing connected component labeling on the difference between the filled DEM and the original DEM, and extracting each closed depression as well as its overflow elevation, bottom elevation and natural storage capacity.

[0009] Further, step 2 further includes applying a depression area threshold filter to exclude depressions with an area smaller than a preset threshold.

[0010] Further, the search interval of the binary search in step 3 is determined according to the relationship between the natural storage capacity and the required storage capacity; When the required storage capacity is less than or equal to the natural storage capacity, the search interval is from the bottom elevation to the overflow elevation; When the required storage capacity is greater than the natural storage capacity, the search interval is from the overflow elevation to the sum of the overflow elevation and a preset maximum dam height, wherein the dam height is equal to the water level minus the overflow elevation.

[0011] Further, in step 4, taking the depression seed as the center of a circle to determine the radius of the expanded disk geometric constraint; Limiting the reservoir inundation extent to the intersection of the connected region below the water level and the disk; The radius Rdisk of the expanded disk geometric constraint satisfies Rdisk=α·sqrt(Vreq / (dmin·π)), where α is an expansion factor with a value of 1.0-2.0, dmin is the minimum acceptable average water depth, and Vreq is the required storage capacity; The expanded disk geometric constraint limits the reservoir inundation extent to the intersection of the connected region {z<W} and the disk. When the reservoir cannot store Vreq under the disk constraint, the depression is marked as insufficient storage capacity and a penalty is imposed in the ranking.

[0012] Further, the process of generating a closed boundary polygon in step 5 includes: Extracting the contour line of the water level on the inundation extent mask, and selecting the polygon containing the seed point; Performing polygon simplification with a set tolerance; Applying a minimum side length constraint, and iteratively deleting vertices whose adjacent side lengths are less than a set value; Repairing self-intersecting polygons, and if the area loss of the corrected version exceeds a preset ratio, rolling back to the unsimplified polygon; The process of estimating the dam volume in step 5 includes: Identify the low-elevation zone outside the inundation area, and estimate the dam volume using a triangular cross section for each dam cell.

[0013] Furthermore, the engineering conditions described in step 7 include: The average water depth is within the set window, the dam height is not greater than the set maximum value, the dam length is not greater than the set maximum value, and the inundation range does not touch the data boundary of the digital elevation model. Candidate sites are sorted by the number of sites that meet the criteria. Within each number level, they are arranged in ascending order of dam volume and descending order of reservoir area. Sites that are close to the boundary or have insufficient reservoir capacity are penalized by volume and then ranked last.

[0014] Furthermore, step 8 includes: For each depression, the water level-storage capacity curve, water surface area curve, and dam volume curve are pre-calculated at multiple water levels at equal intervals. Since the water level-storage capacity curve is monotonically increasing, any required reservoir capacity can be solved by interpolation in a constant time. Users can adjust parameters in real time to obtain updated rankings.

[0015] According to a second aspect of the present invention, the present invention claims protection for a pumped storage power station upper reservoir site selection system based on a digital elevation model, for executing the aforementioned pumped storage power station upper reservoir site selection method based on a digital elevation model, comprising: The data preprocessing module acquires and preprocesses the digital elevation model data of the target area, generates a regular grid digital elevation model from the scattered elevation data through an interpolation algorithm, and performs smoothing processing. The depression detection module uses a priority flood filling algorithm to detect depressions in the preprocessed digital elevation model, identify closed depressions and extract feature parameters. The water level calculation module uses a binary search to find the water level that meets the required reservoir capacity for each depression based on the water level-reservoir capacity curve. The geometric constraint module applies extended disk geometric constraints to each depression to limit the reservoir's inundation range; The boundary generation module extracts water level contour lines and performs Douglas-Peucker simplification and minimum side length constraints to generate closed boundary polygons. The dam body estimation module estimates the dam volume and dam length based on the low elevation zone of the perimeter. The hierarchical ranking module sorts and outputs candidate sites according to the number of sites that meet the engineering conditions. The sensitivity analysis module supports real-time parameter adjustment and updates based on pre-calculated water level curves, ranking 7th.

[0016] Compared with the prior art, the present invention has the following beneficial effects: 1. End-to-end automated site selection capability: This invention realizes end-to-end automated site selection from digital elevation model data to construction-level candidate sites, eliminating the need for manual terrain interpretation and significantly improving site selection efficiency. Traditional manual site selection takes a long time, while this fully automated process greatly shortens the site selection cycle.

[0017] 2. Quantitative sorting and engineering-level output: This invention can not only identify closed depressions, but also generate closed boundary polygons that can be directly used for engineering design, and output quantitative sorting candidate schemes, which solves the problem that the existing technology only produces qualitative suitability maps and cannot provide construction-level boundary polygons.

[0018] 3. Parameter sensitivity and real-time analysis capability: This invention retains boundary candidates by using a hierarchical ranking method, which can identify parameter sensitivity. Through pre-calculated water level curves and interpolation technology, users can quickly obtain updated rankings after adjusting parameters, breaking through the limitation of existing binary screening methods that cannot achieve real-time sensitivity analysis.

[0019] 4. Precise coding of engineering judgments: This invention extends the geometric constraints of the disk and utilizes the natural indicator of average water depth to fully encode engineering judgments within a single interval. The depth window accurately captures engineering intuition, making the site selection results more in line with actual engineering needs. Attached Figure Description

[0020] Figure 1 This is a flowchart illustrating the process of a method for selecting the site of an upper reservoir for a pumped storage power station based on a digital elevation model, according to the present invention. Figure 2 This is a structural module diagram of a pumped storage power station upper reservoir site selection system based on a digital elevation model according to the present invention. Figure 3 This is a three-dimensional topographic map of the research area for the proposed method of selecting the upper reservoir of a pumped storage power station based on a digital elevation model. Figure 4 This is a three-dimensional panoramic view of the closed boundary polygon of the candidate reservoir site for the pumped storage power station site selection method based on digital elevation model according to the present invention. Figure 5 This is a two-dimensional overlay of the closed boundary polygon of the candidate reservoir site in the pumped storage power station site selection method based on digital elevation model according to the present invention. Figure 6 This is a close-up image of the Top-8 candidate sites for the pumped storage power station upper reservoir site selection method based on digital elevation model according to the present invention. Figure 7 This is a three-dimensional partial view of the optimal site #1 of the pumped storage power station upper reservoir site selection method based on digital elevation model according to the present invention. Figure 8This is a schematic diagram comparing the key engineering indicators of the first 8 candidate sites for the pumped storage power station upper reservoir site selection method based on digital elevation model according to the present invention. Figure 9 This is a sensitivity analysis diagram showing the optimal site index of a pumped storage power station reservoir site selection method based on digital elevation model according to the present invention, as a function of the required reservoir capacity Vreq. Detailed Implementation

[0021] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0022] The terms "first," "second," and "third" in this application are for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first," "second," or "third" may explicitly or implicitly include at least one of those features. In the description of this application, "multiple" means at least two, such as two, three, etc., unless otherwise explicitly specified. All directional indications in the embodiments of this application, such as up, down, left, right, front, back, etc., are only used to explain the relative positional relationships and movements between components in a specific orientation as shown in the accompanying drawings. If the specific orientation changes, the directional indications will change accordingly. Furthermore, the terms "including" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or devices.

[0023] References to embodiments herein mean that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a mutually exclusive, independent, or alternative embodiment. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0024] According to the first embodiment of the present invention, referring to Figure 1 This invention claims protection for a method for selecting the site of the upper reservoir of a pumped storage power station based on a digital elevation model, comprising the following steps: Step 1: Obtain digital elevation model data for the target area and preprocess it. Generate a regular grid digital elevation model from the scattered elevation data using an interpolation algorithm and perform smoothing processing. Step 2: Use the priority flood filling algorithm to detect depressions in the preprocessed digital elevation model, identify closed depressions and extract feature parameters; Step 3: For each depression, use a binary search based on the water level-capacity curve to find the water level that satisfies the required reservoir capacity; Step 4: Apply extended disk geometric constraints to each depression to limit the reservoir inundation range; Step 5: Extract water level contour lines and perform Douglas-Peucker simplification and minimum side length constraint to generate a closed boundary polygon; Step 6: Estimate the dam volume and dam length based on the low elevation zone of the perimeter; Step 7: Sort and output the candidate sites according to the number of sites that meet the engineering conditions; Step 8: Support real-time parameter adjustment and ranking updates based on pre-calculated water level curves.

[0025] Furthermore, the interpolation algorithm in step 1 is a bilinear interpolation method or a nearest neighbor interpolation method, and the smoothing process is performed using a Gaussian kernel filter.

[0026] Furthermore, in step 2, the min-heap priority queue is initialized with all boundary cells. Boundary cells are cells on the outer edge of the grid or cells adjacent to invalid value cells. Invalid value cells are regarded as walls, i.e., infinite elevation. Pop the lowest elevation cell from the priority queue and propagate inwards, raising the inner cells to their overflow predecessor elevation; Connectivity component labeling is performed on the difference between the filled digital elevation model and the original digital elevation model; When using the priority flood filling algorithm for depression detection, it also includes: Apply a high altitude constraint, and only depressions with a seed elevation not less than a preset threshold are considered as candidates; Initialize the min-heap priority queue with all boundary cells, where the boundary cells are the outer edge cells of the grid or cells adjacent to cells with no data (NaN), and the NaN cells are considered as walls, i.e., infinite elevation. Pop the lowest elevation cell from the priority queue and propagate inwards, raising the internal cells to their overflow precursor elevation, which is equivalent to filling all depressions to the overflow elevation. Connectivity components were marked for the difference between the filled DEM and the original DEM, and each closed depression and its overflow elevation, bottom elevation and natural reservoir capacity were extracted.

[0027] Furthermore, step 2 also includes applying a depression area threshold filter to exclude depressions with an area smaller than a preset threshold.

[0028] Further, the search interval of the binary search described in step 3 is determined according to the relationship between the natural storage capacity and the required storage capacity; When the required storage capacity is less than or equal to the natural storage capacity, the search interval ranges from the bottom elevation to the overflow elevation; When the required storage capacity is greater than the natural storage capacity, the search interval ranges from the overflow elevation to the sum of the overflow elevation and a preset maximum dam height, where the dam height is equal to the water level minus the overflow elevation.

[0029] Further, in step 4, the radius of the geometric constraint of the expanding disk is determined with the depression seed as the center of a circle; The inundated area of the reservoir is limited to the intersection of the connected area below the water level and the disk; The radius of the geometric constraint of the expanding disk Rdisk=α·sqrt(Vreq / (dmin·π)), where α is an expansion factor with a value of 1.0-2.0, dmin is the minimum acceptable average water depth, and Vreq is the required storage capacity; The geometric constraint of the expanding disk limits the inundated area of the reservoir to the intersection of the connected region {z<W} and the disk. When the reservoir cannot store Vreq under the disk constraint, the depression is marked as insufficient storage capacity and a penalty is imposed in the ranking.

[0030] Further, the process of generating the closed boundary polygon in step 5 includes: Extracting the contour line of the water level on the inundation range mask, and selecting the polygon containing the seed point; Performing polygon simplification with a set tolerance; Imposing a minimum side length constraint, and iteratively deleting vertices whose adjacent side lengths are less than a set value; Repairing the self-intersecting polygon, and rolling back to the unsimplified polygon if the area loss of the corrected version exceeds a preset proportion; The process of estimating the dam volume in step 5 includes: Identifying the peripheral low-elevation zone outside the inundated area, and estimating the dam volume for each dam grid element using a triangular cross-section.

[0031] Further, the engineering conditions described in step 7 include: The average water depth is within a set window, the dam height is not greater than a set maximum value, the dam length is not greater than a set maximum value, and the inundated area does not touch the boundary of the digital elevation model data; Candidate sites are sorted by the number of satisfied conditions, within each number level, they are sorted in ascending order of dam volume and descending order of reservoir area, and sites that are adjacent to the boundary and have insufficient storage capacity are placed at the end after a volume penalty is imposed.

[0032] Further, step 8 includes: For each depression, the water level-storage capacity curve, water surface area curve, and dam volume curve are pre-calculated at multiple water levels at equal intervals. Since the water level-storage capacity curve is monotonically increasing, any required reservoir capacity can be solved by interpolation in a constant time. Users can adjust parameters in real time to obtain updated rankings.

[0033] According to a second embodiment of the present invention, referring to Figure 2 This invention claims protection for a system for selecting the upper reservoir of a pumped-storage power station based on a digital elevation model, used to execute the aforementioned method for selecting the upper reservoir of a pumped-storage power station based on a digital elevation model, comprising: The data preprocessing module acquires and preprocesses the digital elevation model data of the target area, generates a regular grid digital elevation model from the scattered elevation data through an interpolation algorithm, and performs smoothing processing. The depression detection module uses a priority flood filling algorithm to detect depressions in the preprocessed digital elevation model, identify closed depressions and extract feature parameters. The water level calculation module uses a binary search to find the water level that meets the required reservoir capacity for each depression based on the water level-reservoir capacity curve. The geometric constraint module applies extended disk geometric constraints to each depression to limit the reservoir's inundation range; The boundary generation module extracts water level contour lines and performs Douglas-Peucker simplification and minimum side length constraints to generate closed boundary polygons. The dam body estimation module estimates the dam volume and dam length based on the low elevation zone of the perimeter. The hierarchical ranking module sorts and outputs candidate sites according to the number of sites that meet the engineering conditions. The sensitivity analysis module supports real-time parameter adjustment and ranking updates based on pre-calculated water level curves.

[0034] According to an embodiment of the present invention, taking the site selection of the upper reservoir of a pumped storage power station in a high-altitude mountainous area as an application scenario, the specific implementation process of the site selection method based on the digital elevation model is described in detail. The technical solution of this embodiment covers the complete process of data acquisition and preprocessing, automatic detection of depressions, iterative solution of water level, extended disk geometric constraints, closed boundary polygon generation, dam volume estimation, and hierarchical sorting output.

[0035] The pumped storage power station reservoir site selection system based on digital elevation model described in this invention includes a data preprocessing module, a depression detection module, a water level solution module, a geometric constraint module, a boundary generation module, a dam body estimation module, a hierarchical ranking module, and a sensitivity analysis module. The modules are connected through standardized data interfaces to form an end-to-end data processing pipeline.

[0036] The data preprocessing module receives raw topographic scatter data and outputs a regular grid digital elevation model; the depression detection module automatically analyzes the regular grid digital elevation model and outputs all detected closed depressions and their characteristic parameters; the water level calculation module calculates the water level value that meets the design reservoir capacity requirements based on the topographic characteristics of each depression; the geometric constraint module applies extended disk constraints to each candidate reservoir and selects schemes that meet the engineering layout requirements; the boundary generation module extracts closed boundary polygons within the inundation range that meets the constraints; the dam body estimation module calculates the required dam volume for each candidate scheme; the hierarchical ranking module performs a comprehensive evaluation and ranking of all candidate sites; and the sensitivity analysis module pre-calculates key parameter curves, supporting real-time parameter adjustment and scheme comparison.

[0037] The specific functions and data interaction relationships of each module are as follows: the output of the data preprocessing module serves as the input of the depression detection module; the depression feature parameters output by the depression detection module are simultaneously supplied to the water level solution module and the geometric constraint module; the calculation results of the water level solution module and the constraint conditions of the geometric constraint module are jointly input into the boundary generation module; the polygon data generated by the boundary generation module are simultaneously sent to the dam estimation module and the classification and ranking module; the sensitivity analysis module can run independently and update the classification and ranking results in real time according to the changes in the parameters input by the user.

[0038] The data preprocessing module performs the acquisition and preprocessing operations of the digital elevation model data for the target area. Raw terrain data is typically stored in scattered format, containing the three-dimensional coordinates of each measurement point. In this embodiment, the terrain dataset for the target area contains 752,276 scattered elevation data points. The measurement points cover the entire study area, with a sampling density of approximately 15 to 20 points per square kilometer, which effectively reflects the micro-undulations of the land surface.

[0039] The preprocessing process includes two main steps: gridded interpolation and smoothing / denoising. Gridded interpolation transforms discrete measurement point data into a regularly arranged grid, forming a regular grid digital elevation model. The interpolation algorithm employs bilinear interpolation, which calculates the elevation value at any location within a rectangular cell formed by four adjacent known points through linear interpolation. Compared to nearest-neighbor interpolation, bilinear interpolation offers higher accuracy and preserves the continuity of terrain features; compared to more advanced interpolation methods, it is computationally efficient and suitable for processing large-scale terrain data.

[0040] In this embodiment, the interpolated regular grid digital elevation model has a resolution of 20m, meaning each grid cell represents a 20m × 20m area on the ground; the grid size is 420 × 280 cells, covering an area of ​​approximately 84km². 2After generating the regular grid, the elevation values ​​range from 1320m to 2509m, reflecting the significant elevation difference from valley to mountaintop in this area. The DEM is smoothed and denoised using a Gaussian kernel with σ=1 grid. Figure 3 The image shown is a three-dimensional topographic map of the study area.

[0041] Smoothing and denoising are performed using a Gaussian kernel filter on the regular grid digital elevation model. The standard deviation parameter of the Gaussian kernel function controls the smoothing intensity; in this embodiment, the kernel standard deviation is set to 1 cell. The mathematical essence of Gaussian filtering is to perform a weighted average of each cell and its neighborhood. The neighborhood weights follow a two-dimensional Gaussian distribution, with cells closer to the center having larger weights. Smoothing eliminates measurement noise and minor fluctuations in the terrain data, making the subsequent depression detection algorithm more stable. For a Gaussian filter with a kernel standard deviation of 1 cell, its effective filtering window radius is approximately 3 times the standard deviation, covering a 7×7 neighborhood around the central cell.

[0042] The output of the data preprocessing module includes: regular grid digital elevation model data, grid geometric parameter resolution, number of rows and columns, geographic coordinate range, and metadata information; these data are stored in the form of two-dimensional arrays, with elevation values ​​represented as floating-point numbers, and areas without data marked with a specific no-value symbol.

[0043] The depression detection module uses a priority flood filling algorithm to automatically analyze the preprocessed digital elevation model and identify all closed depressions that meet the criteria. The core idea of ​​the priority flood filling algorithm is to simulate the process of floodwater overflowing from the edge of the basin inwards. When the floodwater overflows the lowest point at the edge of the basin, the interior of the basin will be completely filled.

[0044] During the algorithm initialization phase, all boundary cells are first identified and added to the min-heap priority queue. Boundary cells are divided into two categories: the first category is the outermost cells of the digital elevation model grid, namely all cells in the first row, last row, first column and last column of the grid; the second category is cells adjacent to cells with no value. Cells with no value are regarded as insurmountable walls in the algorithm, and their elevation is set to infinity to ensure that the algorithm will not cross the data missing area.

[0045] The core operation of the priority queue is to pop the cell with the lowest current elevation from the queue for processing. Each pop operation has a logarithmic time complexity, ensuring the overall efficiency of the algorithm. The algorithm maintains a marker array to record whether each cell has been processed. After popping a cell, the algorithm checks all its unprocessed adjacent cells. If an adjacent cell's elevation is lower than the current cell's, its elevation is raised to the current cell's elevation, and it is added to the priority queue. This process simulates the characteristics of flood overflow: water always accumulates first in low-lying areas and only overflows when the low-lying areas are full.

[0046] After filling, the new elevation of each cell reflects the water level elevation at that location after the flood has completely overflowed. By comparing the difference between the filled elevation and the original elevation, all areas that form closed depressions can be identified; cells with a difference greater than zero belong to the interior of the depression, while cells with a difference equal to zero belong to the edge or outer area of ​​the depression.

[0047] Connectivity component labeling is performed on the difference between the filled digital elevation model and the original digital elevation model to extract the geometric features of each closed depression; in this embodiment, the area threshold is set to 2000m. 2 That is, only areas with an area of ​​not less than 2000 m² are retained. 2 Closed depressions were identified; this threshold can be adjusted according to project needs, as depressions with too small an area are generally not economically viable. Through connected component labeling, 25 closed depressions meeting the criteria were identified. For each depression, the overflow elevation *s*, bottom elevation *zmin*, depression area *Adep*, and natural reservoir capacity *Vfree* were extracted. After applying a high-altitude constraint *zseed ≥ 2000m*, some depressions were retained as qualified candidates. Figure 4 , Figure 5 The image shown is a 3D panoramic view and a 2D overlay of the closed boundary polygon of the candidate site.

[0048] The following feature parameters are extracted for each closed depression: overflow elevation, which is the lowest elevation point at the edge of the depression, determining the maximum water level at which the reservoir can store water without building a dam; bottom elevation, which is the lowest elevation point inside the depression; depression area, which is the number of cells inside the depression multiplied by the cell area; and natural storage capacity, which is the total storage volume below the overflow elevation and above the original topography.

[0049] The algorithm also needs to impose a high-altitude constraint to filter out candidate depressions with excessively low altitudes. In this embodiment, the seed elevation threshold is set to 2000m, meaning that only depressions with a maximum elevation of at least 2000m are considered qualified candidates. This constraint reflects the design requirement that pumped-storage power stations typically require a certain elevation to achieve the required head difference for their upper reservoirs. The high-altitude constraint can be adjusted according to the terrain characteristics and engineering requirements of the project area.

[0050] After high-altitude constraint screening, some depressions were retained as qualified candidates and entered into the subsequent processing flow. The data output by the depression detection module includes: the location range of each qualified candidate depression, seed point coordinates, overflow elevation, bottom elevation, depression area, and natural reservoir capacity.

[0051] The water level calculation module calculates the water level value that meets the design requirements for each qualified depression; the required reservoir capacity is a key parameter in the design of pumped storage power stations, determining the reservoir's energy storage capacity and installed capacity; in this embodiment, the required reservoir capacity is set to 3 million m³. 3 .

[0052] The functional relationship between water level and reservoir capacity is constructed as follows: for each possible water level value, all cells within the depression are traversed, the volume of cells below that water level is calculated and summed; specifically, for a given water level W, the reservoir capacity V(W) is equal to the sum of the volumes of all cells with elevations below W; the volume of each cell is equal to the product of the elevation difference of that cell and the cell area, and the cell area is equal to the square of the resolution of the digital elevation model.

[0053] The water level is determined using a binary search algorithm, which iteratively searches for a water level value that satisfies the required reservoir capacity within the effective interval. The boundary of the binary search interval is determined based on the relationship between the natural reservoir capacity and the required reservoir capacity.

[0054] When the required reservoir capacity is less than or equal to the natural reservoir capacity, the water level is below the overflow elevation, and the reservoir can reach the required capacity without the need for a dam. In this case, the lower bound of the binary search is the elevation of the depression bottom, and the upper bound is the overflow elevation. The search continues within this interval until a water level is found that exactly meets the required reservoir capacity.

[0055] When the required reservoir capacity exceeds the natural reservoir capacity, the water level must be higher than the overflow elevation, necessitating the construction of a dam to block the water flow. In this case, the lower bound of the binary search is the overflow elevation, and the upper bound is the overflow elevation plus the maximum allowable dam height. In this embodiment, the maximum allowable dam height is set to 300m, a value that can be adjusted based on engineering experience and terrain conditions. The required dam height is obtained by subtracting the overflow elevation from the searched water level.

[0056] The convergence accuracy of the binary search is set to the sub-centimeter level, that is, the water level error is less than 0.01m. Each iteration calculates the reservoir capacity value corresponding to the current intermediate water level, and adjusts the upper or lower bound of the search interval according to the relationship between the reservoir capacity and the required reservoir capacity. In this embodiment, the binary search converges to meet the accuracy requirements within 55 iterations.

[0057] The data output by the water level solution module includes: the solved water level for each candidate depression, the dam height (if a dam needs to be built), and the corresponding reservoir capacity curve data. This data provides input for subsequent geometric constraints and boundary generation.

[0058] The extended disk geometric constraint module applies spatial layout constraints to each candidate depression and selects candidate schemes that meet engineering design preferences. The core of the extended disk constraint is to set a circular area with a limited radius, with the depression seed point as the center. The flooding range of the reservoir must not exceed the intersection of this circular area and the original depression flooding range.

[0059] The expansion factor is a coefficient greater than 1, used to appropriately enlarge the disk based on the theoretical minimum size, leaving room for engineering margins; in this embodiment, the expansion factor is set to 1.5, and the minimum acceptable average water depth is set to 40m.

[0060] Extending the circular constraint achieves three engineering objectives: The first objective is compactness: limiting the reservoir's inundation area to a circular region ensures a compact, approximately circular reservoir shape; a compact reservoir shape means a shorter dam line length, thus reducing the amount of dam work and the length of hydraulic structures; The second objective is depth assurance: when the inundation area is limited, the water level must rise to a higher position to achieve the required reservoir capacity, thereby increasing the average depth of the reservoir; the average depth equals the required capacity divided by the inundation area, and when the area is limited by the circular constraint, the average depth increases, tending towards a preset depth window; The third objective is engineering judgment coding: the average water depth is a key indicator for evaluating the economic efficiency of reservoir engineering; a reservoir that is too shallow has low capacity efficiency, while a reservoir that is too deep may involve adverse geological conditions; the preset depth window accurately captures engineering intuition, and in this embodiment, the depth window is set to 40m to 60m.

[0061] The reservoir inundation area is limited to the intersection of the connected region below the water level and the disk, resulting in the actual inundation area after applying the extended disk constraint. Under this constraint, the water level is usually higher than the unconstrained water level to ensure that the required reservoir capacity is achieved within the restricted area.

[0062] The data output by the extended disk geometry constraint module includes: the disk radius of each candidate depression, the constrained inundation range mask, and the corresponding average water depth; this data is used by the subsequent boundary generation module. The boundary generation module extracts closed boundary polygons on the inundation range mask after applying extended disk constraints and estimates the dam body engineering volume.

[0063] The extraction process of the closed boundary polygon is as follows: First, extract contour lines with the same elevation as the solved water level value on the inundation area mask; contour lines are boundary lines formed by connecting grid cells with specific elevation values ​​in the digital elevation model, and may consist of multiple independent closed curves; then, select the contour line containing the seed point as the closed boundary of the reservoir; the seed point is the grid cell containing the highest elevation point in the depression, located at the center of the final inundation area of ​​the reservoir.

[0064] The extracted initial contour lines may contain a large number of vertices, which would lead to excessive data volume if used directly in engineering design. The module uses the Douglas-Peucker algorithm to simplify the polygon. This algorithm removes redundant vertices recursively, which greatly reduces the number of vertices while maintaining shape features. In this embodiment, the simplification tolerance is set to 15m, that is, the maximum deviation between the simplified polygon and the original polygon does not exceed 15m.

[0065] After the algorithm is simplified, there may be vertices with excessively short side lengths that cause numerical problems in subsequent processing. The module applies a minimum side length constraint and iteratively deletes vertices with adjacent side lengths less than 30m. After each deletion, the adjacent side lengths are recalculated, and this process is repeated until all adjacent side lengths meet the minimum side length requirement.

[0066] The original terrain may contain local depressions or saddles, causing self-intersections in the extracted contour lines. The module repairs these self-intersecting polygons using a geometric repair algorithm to reconnect the self-intersecting parts into valid simple polygons. If the area loss of the repaired polygon exceeds 50% of the original area, the repair is abandoned, and the module reverts to the unsimplified original polygon. Figure 6 The image shows close-ups of the top 8 candidate sites. Each sub-image displays local topography, cyan water surface, blue line of closed boundary, and red outline of dam. Figure 7 The image shows a 3D partial view of the optimal site #1, where cyan represents the water surface and blue lines represent the closed boundary.

[0067] The dam volume estimation is performed outside the inundation zone. The module first identifies a low-elevation zone within a certain distance outside the boundary of the inundation zone, which corresponds to the possible location of the dam. For each cell within the low-elevation zone, a triangular cross-section is used to estimate the required dam volume at that location. The height of the triangular cross-section equals the water level plus the freeboard minus the ground elevation. The freeboard is a safety margin to ensure that floodwaters will not overflow the dam crest; in this embodiment, the freeboard is set to 2m. The dam volume for each cell equals the square of the cross-sectional height multiplied by the cell width and then multiplied by a shape factor. The shape factor reflects the shape characteristics of the dam cross-section; in this embodiment, the shape factor is set to 0.5, corresponding to the triangular cross-section. The total dam volume is obtained by summing the volumes of all dam cells. The total dam length equals the number of dam cells multiplied by the cell resolution.

[0068] The boundary generation module outputs data including: the vertex sequence of the closed boundary polygon for each candidate site, the dam volume, dam height, and dam length. This data is used by the ranking module.

[0069] The ranking module performs a comprehensive evaluation and outputs a ranking of all candidate sites. The evaluation uses four engineering conditions to measure the engineering feasibility of the candidate sites from different perspectives.

[0070] The first condition C1 requires that the average water depth be within a preset depth window. In this embodiment, the depth window is set to 40m to 60m. An average water depth of less than 40m means that the reservoir is too shallow and the area occupied per unit capacity is too large; an average water depth of more than 60m may mean that the terrain is too steep or there are adverse geological conditions.

[0071] The second condition, C2, requires that the dam height not exceed the maximum allowable value. In this embodiment, the maximum allowable dam height is set at 300m. A dam exceeding this height will significantly increase the engineering difficulty and investment.

[0072] The third condition, C3, requires that the dam length not exceed the maximum allowable value. In this embodiment, the maximum allowable dam length is set at 5000m. An excessively long dam line will reduce the economic efficiency of the project and increase construction risks.

[0073] The fourth condition, C4, requires that the inundation area does not touch the data boundaries of the digital elevation model. Candidate sites that touch the boundaries mean that part of the reservoir area is outside the data coverage, making accurate engineering assessment impossible, and should be excluded.

[0074] Each candidate site receives a pass count (npass) based on the number of sites that meet the above conditions, with a value ranging from 0 to 4. Sites with a higher pass count have better project feasibility. The module first sorts sites by pass count from high to low, and within the same pass count, it sorts them by dam volume from low to high. A lower dam volume means lower project investment. For sites with the same dam volume, they are sorted by reservoir area from large to small, with larger sites having stronger energy storage capacity.

[0075] For sites that are close to the boundary or have insufficient storage capacity, the module applies a volume penalty, lowering their ranking to the bottom of the same level. Sites close to the boundary have uncertainty due to incomplete data, and sites with insufficient storage capacity require a higher dam to reach the required storage capacity.

[0076] In this embodiment, after evaluation and sorting, the top 12 candidate stations and their detailed information are output; stations #1 to #9 passed 3 times, and the reason for failure was that the average water depth did not meet the depth window requirement, with an average water depth of 22m to 37m, which is lower than the lower limit of the depth window of 40m; station #10 met the depth window requirement but failed to pass the boundary touch condition; stations #11 to #12 passed 2 times, and failed to meet both conditions; Figure 8 The chart shown is a comparison of key engineering indicators for the top 8 candidate sites.

[0077] The sensitivity analysis module pre-calculates key parameter curves for each depression, allowing users to quickly obtain updated rankings after adjusting parameters in real time.

[0078] Pre-calculations were performed at multiple equally spaced water levels to calculate water level-storage capacity curves, water surface area curves, and dam volume curves. In this embodiment, 48 equally spaced water levels were pre-calculated, covering the entire range from the bottom elevation to the maximum possible water level. The water level-storage capacity curve reflects the monotonically increasing relationship between reservoir capacity and water level, the water surface area curve reflects the trend of submerged area changing with water level, and the dam volume curve reflects the required dam volume at different water levels. For each depression, water level-storage capacity curves, water surface area curves, and dam volume curves were pre-calculated at 48 equally spaced water levels. Sensitivity analysis showed that increasing Vreq from 3×10... 6 m³ decreased to approximately 1-1.5 × 10 6 When the depth is m³, the average water depth d of the optimal station #1 falls within the [40, 60] m depth window, achieving full compliance with npass=4. This diagnostic capability is a direct result of the hierarchical ranking, which cannot be achieved through binary filtering. For example... Figure 9 The figure shown is a sensitivity analysis chart.

[0079] Because the water level-reservoir capacity curve has a monotonically increasing characteristic, any required reservoir capacity can be solved by interpolation in constant time without repeated binary search. When the user adjusts parameters such as the required reservoir capacity, minimum average water depth, and maximum dam height, the system can complete the re-evaluation and sorting update of all candidate sites within seconds.

[0080] Sensitivity analysis shows that when the required storage capacity increases from 3 million m³, 3 Reduced to 1 million to 1.5 million m³ 3 At that time, the average water depth of the optimal site #1 changed from less than 40m to a depth window of 40m to 60m, achieving full compliance with the number of passes of 4; this diagnostic capability helps engineers understand the impact of parameter changes on site selection results and provides a scientific basis for project decision-making.

[0081] To further illustrate the applicability of the method of the present invention under different terrain conditions, the specific implementation of the alternative technical solution is described using the site selection of pumped storage power stations in low-altitude hilly areas as a scenario.

[0082] In low-altitude hilly areas, the elevation difference is relatively small, and the elevation of depressions is usually lower than that of high-altitude mountainous areas. In this embodiment, the topographic scatter data of the target area obtained by the data preprocessing module is used to generate a regular grid digital elevation model with a resolution of 10m through bilinear interpolation, and the grid coverage area is approximately 120km². 2 The elevation range is from 580m to 890m.

[0083] The depression detection module uses the same priority flood filling algorithm to identify a greater number of small depressions under low-altitude terrain conditions; in this embodiment, a total of 47 depressions with an area of ​​not less than 1000 m² were identified. 2For closed depressions, a seed elevation threshold of 600m is set to screen out candidates with too low an altitude.

[0084] The water level calculation module performs a binary search on each of the 47 qualified depressions, with the required reservoir capacity set at 5 million m³. 3 Since some depressions have large natural reservoir capacity, the search range needs to be determined based on the specific conditions of each depression.

[0085] The parameter settings of the extended disk geometric constraint module are different from those in Example 1: the extension factor is 1.8, the minimum acceptable average water depth is set to 30m, and the depth window is set to 30m to 50m; the head difference in low-altitude areas is relatively small, so the depth window needs to be adjusted appropriately to obtain a better engineering solution.

[0086] The process of generating closed boundary polygons and estimating dam body is the same as in Example 1. The final output is a list of candidate sites ranked according to the evaluation of four engineering conditions. The sensitivity analysis function allows users to adjust parameters such as reservoir capacity and depth window and observe the trend of ranking changes in real time.

[0087] The foregoing has shown and described the basic principles, main features, and advantages of the present invention; those skilled in the art should understand that the present invention is not limited to the above embodiments. In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces, or indirect coupling or communication connection between apparatuses or units, and may be electrical, mechanical, or other forms.

[0088] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated units described above can be implemented in hardware or as software functional units. The above are merely embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made based on the description and drawings of this application, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

[0089] The specific embodiments of the invention have been described in detail above, but they are only examples. This application is not limited to the specific embodiments described above. For those skilled in the art, any equivalent modifications or substitutions to the invention are also within the scope of this application. Therefore, all equivalent changes, modifications, and improvements made without departing from the spirit and principles of this application should be covered within the scope of this application.

Claims

1. A method for selecting the site of the upper reservoir of a pumped storage power station based on a digital elevation model, characterized in that, Includes the following steps: Step 1: Obtain digital elevation model data for the target area and preprocess it. Generate a regular grid digital elevation model from the scattered elevation data using an interpolation algorithm and perform smoothing processing. Step 2: Use the priority flood filling algorithm to detect depressions in the preprocessed digital elevation model, identify closed depressions and extract feature parameters; Step 3: For each depression, use a binary search based on the water level-capacity curve to find the water level that satisfies the required reservoir capacity; Step 4: Apply extended disk geometric constraints to each depression to limit the reservoir inundation range; Step 5: Extract water level contour lines and perform Douglas-Peucker simplification and minimum side length constraint to generate a closed boundary polygon; Step 6: Estimate the dam volume and dam length based on the low elevation zone of the perimeter; Step 7: Sort and output the candidate sites according to the number of sites that meet the engineering conditions; Step 8: Support real-time parameter adjustment and ranking updates based on pre-calculated water level curves.

2. The method for selecting the upper reservoir of a pumped storage power station based on a digital elevation model according to claim 1, characterized in that, The interpolation algorithm described in step 1 is either a bilinear interpolation method or a nearest neighbor interpolation method, and the smoothing process is performed using a Gaussian kernel filter.

3. The method for selecting the upper reservoir of a pumped storage power station based on a digital elevation model according to claim 1, characterized in that, In step 2, the min-heap priority queue is initialized with all boundary cells. Boundary cells are cells on the outer edge of the grid or cells adjacent to invalid value cells. Invalid value cells are regarded as walls, i.e., infinite elevation. Pop the lowest elevation cell from the priority queue and propagate inwards, raising the inner cells to their overflow predecessor elevation; Connectivity component labeling is performed on the difference between the filled digital elevation model and the original digital elevation model; When using the priority flood filling algorithm for depression detection, it also includes: Apply a high altitude constraint, and only depressions with a seed elevation not less than a preset threshold are considered as candidates; Initialize the min-heap priority queue with all boundary cells, where the boundary cells are the outer edge cells of the grid or cells adjacent to cells with no data (NaN), and the NaN cells are considered as walls, i.e., infinite elevation. Pop the lowest elevation cell from the priority queue and propagate inwards, raising the internal cells to their overflow precursor elevation, which is equivalent to filling all depressions to the overflow elevation. Connectivity components were marked for the difference between the filled DEM and the original DEM, and each closed depression and its overflow elevation, bottom elevation and natural reservoir capacity were extracted.

4. The method for selecting the upper reservoir of a pumped storage power station based on a digital elevation model according to claim 3, characterized in that, Step 2 also includes applying a depression area threshold filter to exclude depressions with an area smaller than a preset threshold.

5. The method for selecting the upper reservoir of a pumped storage power station based on a digital elevation model according to claim 1, characterized in that, The search interval for the binary search in step 3 is determined based on the relationship between the natural reservoir capacity and the required reservoir capacity; When the required storage capacity is less than or equal to the natural storage capacity, the search range is from the bottom elevation to the overflow elevation; When the required reservoir capacity is greater than the natural reservoir capacity, the search range is from the overflow elevation to the overflow elevation plus the preset maximum dam height, where the dam height is equal to the water level minus the overflow elevation.

6. The method for selecting the upper reservoir of a pumped storage power station based on a digital elevation model according to claim 1, characterized in that, In step 4, the radius of the geometric constraint of the extended disk is determined with the depression seed as the center; The inundation area of ​​the reservoir is limited to the intersection of the connected area below the water level and the disk; The radius of the extended disk geometric constraint is Rdisk = α·sqrt(Vreq / (dmin·π)), where α is the extension factor with a value of 1.0-2.0, dmin is the minimum acceptable average water depth, and Vreq is the required reservoir capacity. The extended disk geometric constraint limits the inundated area of the reservoir to the intersection of the connected region {z<W} and the disk. When the reservoir cannot store Vreq under the disk constraint, the depression is marked as having insufficient storage capacity and a penalty is imposed in the ranking.

7. The method for selecting the upper reservoir of a pumped storage power station based on a digital elevation model according to claim 1, characterized in that, The process of generating the closed boundary polygon in Step 5 includes: Extracting the contour line of the water level on the inundation range mask, and selecting the polygon containing the seed point; Performing polygon simplification with a set tolerance; Applying the minimum side length constraint, and iteratively deleting vertices whose adjacent side lengths are smaller than the set value; Repairing the self-intersecting polygon, and rolling back to the unsimplified polygon if the area loss of the corrected version exceeds the preset ratio; The process of estimating the dam volume in Step 5 includes: Identifying the peripheral low-elevation zone outside the inundation range, and estimating the dam volume for each dam grid cell using a triangular cross-section.

8. The method for selecting the upper reservoir of a pumped storage power station based on a digital elevation model according to claim 1, characterized in that, The engineering conditions described in Step 7 include: The average water depth is within the set window, the dam height is not greater than the set maximum value, the dam length is not greater than the set maximum value, and the inundation range does not touch the boundary of the digital elevation model data; Candidate sites are sorted by the number of satisfied conditions, and within each number level, sorted by ascending dam volume and descending reservoir area. Sites adjacent to the boundary and sites with insufficient storage capacity are placed at the end after volume penalty is imposed.

9. The method for selecting the upper reservoir of a pumped storage power station based on a digital elevation model according to claim 1, characterized in that, Step 8 includes: Pre-calculating the water level-storage capacity curve, water surface area curve and dam volume curve at multiple equally spaced water levels for each depression. Since the water level-storage capacity curve is monotonically increasing, any required storage capacity can be solved in constant time through interpolation, which supports obtaining an updated ranking after the user adjusts parameters in real time.

10. A system for selecting the upper reservoir of a pumped-storage power station based on a digital elevation model, used to execute the method for selecting the upper reservoir of a pumped-storage power station based on a digital elevation model as described in any one of claims 1-9, characterized in that, Includes: A data preprocessing module, which acquires and preprocesses the digital elevation model data of the target area, generates a regular grid digital elevation model from scattered point elevation data through an interpolation algorithm, and performs smoothing processing; A depression detection module, which uses a priority flood filling algorithm to detect depressions on the preprocessed digital elevation model, identifies closed depressions and extracts characteristic parameters; A water level solving module, which solves the water level meeting the required storage capacity for each depression through binary search based on the water level-storage capacity curve; A geometric constraint module, which imposes an extended disk geometric constraint on each depression to limit the inundated area of the reservoir; A boundary generation module, which extracts water level contour lines, performs Douglas-Peucker simplification and minimum side length constraint to generate a closed boundary polygon; A dam estimation module, which estimates the dam volume and dam length based on the peripheral low-elevation zone; A hierarchical ranking module, which performs hierarchical sorting and output of candidate sites according to the number of satisfied engineering conditions; A sensitivity analysis module, which supports real-time parameter adjustment and updated ranking based on the pre-calculated water level curves.

Images