Reservoir capacity calculation and reservoir capacity loss interval identification method for large and medium-sized reservoirs

Abstract

The invention provides a storage capacity calculation and storage capacity loss interval identification method for a large and medium-sized reservoir. The method comprises the following steps: extracting a storage capacity characteristic curve within an elevation range below a dead water level by using an underwater terrain actual measurement method; performing water body identification on the remote sensing image based on an AI technology to extract a reservoir capacity characteristic curve in a dead water level-normal storage water level elevation range; based on a digital elevation model, extracting a reservoir capacity characteristic curve within a normal storage water level-check flood level elevation range; performing cross precision verification after extending the three sections of storage capacity characteristic curves; and a complete water level-reservoir capacity relation curve covering the whole operation water level range is constructed based on the three sections of reservoir capacity characteristic curves, the reservoir capacity increment of each water level section is calculated section by section, and a reservoir capacity abnormal loss interval is judged through a set threshold value of the reservoir capacity increment. According to the method, the reservoir capacity values of the large and medium-sized reservoirs under different elevations can be quickly and accurately calculated, the elevation interval with serious reservoir capacity loss is effectively identified, and the intelligent level and timeliness of reservoir capacity calculation are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the field of storage area management of large and medium-sized reservoirs, and in particular to a method for calculating storage capacity and identifying storage capacity loss intervals of large and medium-sized reservoirs. Background Art

[0002] Reservoir capacity, a core parameter in the construction and operation management of water conservancy projects, serves as a crucial basis for formulating flood control and scheduling plans, monitoring dam safety, ensuring downstream watershed security, and evaluating the effectiveness of water conservancy projects. However, some reservoirs currently face numerous challenges, such as blurred management and protection boundaries, inadequate supervision, and illegal occupation of reservoir land. This, coupled with increasingly severe sedimentation, has resulted in the encroachment of some reservoirs' flood control capacity, resulting in varying degrees of capacity loss. This has reduced the reservoirs' flood control capabilities and impacted their ability to function properly.

[0003] Factors affecting the accuracy of reservoir capacity calculations primarily include the calculation method and the topographic data of the reservoir area. For operational reservoirs, the traditional method for obtaining and calibrating reservoir capacity curves utilizes ultrasonic principles to set up sections at regular intervals throughout the reservoir area and measure the water depth at each section. This depth measurement is primarily performed using echo sounders. An underwater topographic map is then constructed based on the cross-sectional measurements. Planimeter, grid, and grid point methods are then used to calculate the reservoir's surface area at different water levels. Volume calculations are performed manually, with lines drawn and the corresponding area read at specific water level intervals to estimate the reservoir capacity. This method suffers from long measurement cycles, high labor costs, and limited accuracy, making it difficult to meet the real-time and accurate storage capacity data requirements for dynamic reservoir operation management and intelligent decision-making. With the development of information-based reservoir management and digital twin water conservancy, the shortcomings of traditional methods in practicality and timeliness have become a major bottleneck restricting the advancement of modern reservoir management.

[0004] In recent years, the rapid development of artificial intelligence (AI) and spatial information technology has provided a new path to addressing these issues. AI possesses powerful capabilities in image recognition, information reasoning, and data fusion. Combined with spatial information technology tools such as remote sensing imagery, unmanned underwater surveying platforms, and DEM terrain modeling, it can efficiently extract reservoir morphology, analyze storage capacity changes, and identify areas of significant storage capacity loss. This technology demonstrates promising technical feasibility and application prospects.

[0005] Therefore, with the continuous development of remote sensing technology, geographic information systems (GIS), and artificial intelligence (AI), reservoir capacity calculation methods based on multi-source spatial information and intelligent algorithms have gradually become a hot topic in research and application. Especially in the management of large and medium-sized reservoirs, how to achieve efficient calculation and dynamic update of reservoir capacity has become a technical difficulty in the current construction of water conservancy informatization. Summary of the Invention

[0006] This invention aims to address these issues by proposing a method for calculating the storage capacity of large and medium-sized reservoirs and identifying storage capacity loss intervals. This method utilizes spatial data resources such as remote sensing imagery, GIS data, and digital elevation models (DEMs), combined with AI image recognition and analysis technology, to quickly and accurately calculate the storage capacity of large and medium-sized reservoirs at different elevations and effectively identify elevation intervals with significant storage capacity loss. This significantly improves the intelligence and timeliness of reservoir capacity calculations, providing technical support for scientific scheduling, safety monitoring, and refined management of reservoirs.

[0007] In order to achieve the above object, the present invention adopts the following technical solutions:

[0008] A method for calculating the storage capacity of large and medium-sized reservoirs and identifying storage capacity loss intervals includes the following steps:

[0009] Step S1: using the underwater terrain measurement method to extract the reservoir capacity characteristic curve within the elevation range below the dead water level;

[0010] Step S2: Perform water body recognition on the remote sensing image based on AI technology to extract the reservoir capacity characteristic curve within the range of dead water level to normal water level elevation;

[0011] Step S3: extracting the reservoir capacity characteristic curve within the range of normal water storage level to verified flood level elevation based on the digital elevation model;

[0012] Step S4: Extend the three sections of reservoir capacity characteristic curves obtained in steps S1-S3 upstream and downstream respectively to form extended curve sections. Within the overlapping water level interval of the extended curve sections, calculate the unit water level section reservoir capacity difference values ​​of the three sections of reservoir capacity characteristic curves in the overlapping part at a water level interval of 1 meter. By comparing the unit water level section reservoir capacity difference values, the cross-precision verification of the three sections of reservoir capacity characteristic curves is completed.

[0013] Step S5: Based on the three-segment reservoir capacity characteristic curves obtained in steps S1-S3, a complete water level-reservoir capacity relationship curve covering the entire operating water level range is constructed, the reservoir capacity increment of each water level segment is calculated segment by segment, and the abnormal reservoir capacity loss interval is determined by the set reservoir capacity increment threshold.

[0014] Furthermore, the step S1 includes:

[0015] Step S1.1: Use an unmanned vessel equipped with a multi-beam bathymetry system to measure underwater topography in the reservoir area. The multi-beam bathymetry system transmits sound waves and receives reflected signals to obtain raw multi-beam data of the underwater topography.

[0016] Step S1.2: Preprocess the collected raw multi-beam data, including acoustic signal denoising, signal strength correction, data georeferencing, and generate three-dimensional data of the reservoir underwater topography based on GIS software;

[0017] Step S1.3: Based on the underwater topography 3D data and the reservoir geometry, calculate the reservoir capacity at different water levels using triangulation, grid method, or volume method;

[0018] Step S1.4: Based on the storage capacity of the reservoir at different water levels calculated in step S1.3, draw a storage capacity characteristic curve within the elevation range below the dead water level.

[0019] Furthermore, the step S2 includes:

[0020] Step S2.1: Obtain a long-term remote sensing image of the study area including the maximum flooded area of ​​the reservoir;

[0021] Step S2.2: Preprocess the acquired remote sensing image, including radiation correction, atmospheric correction, and cloud removal;

[0022] Step S2.3: Train the deep learning model to identify water bodies and non-water bodies. The deep learning model automatically adjusts its parameters to maximize recognition accuracy by learning water body and non-water body samples in the training dataset.

[0023] Step S2.4: Calculate the area of ​​the water body identified in step S2.3, find the water level value corresponding to the imaging time of each image, fit the water body area and the corresponding water level on the day, select the function with the highest accuracy as the fitting function, and obtain the water level-area curve of the reservoir in this water level range;

[0024] Based on the reservoir capacity calculation model of remote sensing data and the water level-area curve of the reservoir in the water level range, the interval storage capacity of the reservoir in the water level range is calculated, and then accumulated to different water levels to obtain the water level-capacity data;

[0025] Step S2.5: Based on the calculation results of step S2.4, draw the reservoir capacity characteristic curve within the range of dead water level to normal water level elevation.

[0026] Furthermore, the reservoir capacity calculation model in step S2.4 adopts the trapezoidal or pyramidal volume method. The model first determines whether the reservoir is a trapezoidal or pyramidal body based on its type, and then divides the water body into n layers according to different water levels. The reservoir capacity is calculated by accumulating the volume of the n layers of trapezoidal or pyramidal bodies.

[0027] The trapezoidal formula is:

[0028]

[0029] The pyramid formula is:

[0030]

[0031] The formula for cumulative storage capacity is:

[0032]

[0033] In the above formula, is the reservoir capacity difference between two adjacent water levels; is the water level difference between two adjacent water levels; 、 are the water surface areas corresponding to two adjacent water levels; i is the ordinal number; n is the cumulative number; is the initial reservoir capacity; V is the cumulative reservoir volume.

[0034] Furthermore, step S3 includes:

[0035] Step S3.1: Obtain DEM data of the reservoir area by photogrammetry, ground surveying, or digitizing existing topographic maps, wherein the DEM data provides topographic information of the reservoir, including water surface extent, topography, and elevation data;

[0036] Step S3.2: Preprocess the DEM data, including coordinate system correction, filling depressions, interpolation, and clipping;

[0037] Step S3.3: Using the pre-processed DEM data of the reservoir area as the background, use the surface volume tool in the GIS software system toolbox to calculate the volume of the area between the reservoir bottom terrain surface and the reference plane, that is, the reservoir capacity;

[0038] Step S3.4: Based on the calculation results, draw the reservoir capacity characteristic curve within the range of normal water storage level to verification flood level elevation.

[0039] Furthermore, step S4 includes:

[0040] Step S4.1: Based on the reservoir capacity curve fitting equations of methods S1, S2, and S3, the three obtained reservoir capacity curves are extended 3 to 5 meters toward both ends;

[0041] Step S4.2: Calculate the storage capacity difference of each section of the three storage capacity curves at the overlapping part of the water level of 1 meter;

[0042] Step S4.3: Check whether the 1-meter tolerance of each section of the three reservoir capacity characteristic curves in the overlapping part is within the tolerance range, and further verify the reliability of the reservoir capacity calculation methods S1, S2 and S3.

[0043] Furthermore, step S5 includes:

[0044] Step S5.1: Integrate the reservoir capacity curves calculated by methods S1, S2, and S3 according to their applicable elevation ranges, eliminate overlaps and breakpoints, and construct a complete water level-reservoir capacity relationship curve covering the entire operating water level range;

[0045] Step S5.2: Calculate the difference between the complete water level-storage capacity relationship curve and the original design storage capacity curve segment by segment, taking 1 meter as the water level interval, and obtain the storage capacity increment for each water level segment;

[0046] Step S5.3: Through statistical analysis of storage capacity increments, based on historical data or engineering experience, a storage capacity increment threshold is set to identify abnormal storage capacity loss intervals.

[0047] Furthermore, the threshold value of the fixed storage capacity increment is 10% or 20%.

[0048] Furthermore, step S5 further includes:

[0049] Step S5.4: Use field measurements or other independent data sources to verify the reservoir storage capacity curve calculated according to S1, S2, and S3, as well as the identification of storage capacity loss intervals and related calculated values, and propose corresponding maintenance or repair measures based on the storage capacity loss situation.

[0050] Compared with the prior art, the present invention has the following beneficial effects:

[0051] (1) The present invention integrates multi-source spatial information technologies such as underwater multi-beam bathymetry, remote sensing image recognition, and DEM terrain modeling, significantly improving the automation and calculation accuracy of reservoir capacity data acquisition, and can meet the management needs of rapid storage capacity updates for large and medium-sized reservoirs.

[0052] (2) The proposed segmented reservoir capacity curve fitting and cross-validation mechanism improves the adaptability and overall continuity of each measurement method in the transition water level section, and enhances the calculation consistency after data fusion.

[0053] (3) The differential analysis method based on 1-meter reservoir capacity difference proposed in this paper can accurately identify the elevation intervals with significant reservoir capacity loss, providing data support for reservoir siltation detection, morphological change identification and safety management.

[0054] (4) The results of this invention can be directly used for the dynamic verification of basic data on reservoir operation, which is helpful for scientific decision-making on key functions such as flood control scheduling, water resources allocation, and hydropower generation. It has good engineering applicability and promotion value. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 The present invention is a flowchart of a method for calculating storage capacity of large and medium-sized reservoirs and identifying storage capacity loss intervals.

[0056] Figure 2 This is a flow chart of the reservoir capacity characteristic curve method within the dead water level to normal water level elevation range based on AI remote sensing image water body recognition technology.

[0057] Figure 3This is a schematic diagram of the reservoir capacity curve calculation process based on DEM data.

[0058] Figure 4 This is a relationship diagram of reservoir water level and area within the range of dead water level and normal water storage level of a reservoir in South China.

[0059] Figure 5 This is a schematic diagram of the storage capacity curve of a reservoir in South China reconstructed based on the method of the present invention.

[0060] Figure 6 This is a comparison chart of the storage capacity per meter of the original storage capacity curve of a reservoir in South China and the storage capacity per meter of the storage capacity curve reconstructed by the present invention.

[0061] Figure 7 This is a graph showing the difference between the 2024 storage capacity increase and the original storage capacity increase at a 1-meter section of a reservoir in South China. DETAILED DESCRIPTION

[0062] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0063] like Figure 1 As shown, this embodiment provides a method for calculating the storage capacity of large and medium-sized reservoirs and identifying storage capacity loss intervals. It is based on a large reservoir in South China. The dead water level of the reservoir is 175 m, the normal water level is 185 m, and the verified flood level is 195 m. The specific implementation method is carried out according to the following steps:

[0064] Step S1: Use the underwater terrain measurement method to extract the reservoir capacity characteristic curve within the elevation range below the dead water level.

[0065] Using an underwater topography measurement method, an unmanned vessel equipped with a multi-beam bathymetry system is used to map the area below the dead water level of the reservoir, obtaining raw topographic data within this elevation range. The measured data is preprocessed, fitted, and modeled, converted into a water level-reservoir capacity relationship, and a reservoir capacity characteristic curve below the dead water level is formed, reflecting the volume changes in the area at different water levels. Step S1 specifically includes:

[0066] Step S1.1: Underwater topography measurement is performed using a multibeam bathymetry system aboard an unmanned vessel. Post-processing software is used to generate three-dimensional underwater topography data from the collected underwater measurement data. Before measurement, ensure that the unmanned vessel and multibeam bathymetry system are fully installed and debugged, and that the required measurement software is available. Plan bathymetry lines are laid out as required, with spacing of 3 to 4 cm on a 1:10,000 topographic map. The bathymetry lines should be parallel to the axis of the reservoir's main dam. A bathymetry check line is laid out as required, intended to be perpendicular to the bathymetry plan line. Before and after each operation, the rover should be moved to a known point (or checkpoint), and the coordinates of that known point (or checkpoint) recorded and compared with the known coordinates of the point. Operations can only be carried out when the difference meets regulatory requirements to ensure the accuracy of bathymetry point positioning.

[0067] Step S1.2: ① Verification and processing of raw data: First, transfer the collected water depth data into the data processing software to verify the water depth measurement results. Distorted water depth points are eliminated, unqualified parts are repaired, and some parts are smoothed. ② Water depth correction and data conversion: The data file containing the plane coordinates of the sounding points, water surface elevation, water depth values, and other information is automatically processed by the sounding software (where the underwater point elevation H is calculated by subtracting the water depth value from its water surface elevation). It is then converted into a data format compatible with digital mapping software, containing only the point number, code, and X, Y, and H values. ③ Draw underwater contour lines: The sounding points are plotted on the computer screen at the mapping scale. By building a digital terrain model (DTM), generating a triangulated network, and calculating and drawing underwater contour lines, the underwater contour lines are smoothed and represented as solid lines.

[0068] Step S1.3: First, obtain the pour point of the reservoir outlet, then use the watershed tool to calculate the catchment area of ​​the reservoir, and use the raster to polygon tool to extract the vector surface data of the reservoir catchment area; then use the raster to polygon tool to extract the vector surface data of the inundation area, and use the conditional function to extract the raster data below the reservoir level; crop the DEM of the reservoir area, use the conditional function to extract the DEM of the inundation area, and finally set up a loop, lowering each 1 meter, and use the surface volume tool to calculate the area and volume of the reservoir under different water level conditions.

[0069] Step S1.4: Based on the calculated area and volume of the reservoir under different water level conditions, the reservoir capacity curve for the elevation range below the dead water level of the reservoir is finally obtained.

[0070] Step S2: AI-based water body identification is performed on remote sensing images to extract reservoir capacity characteristic curves within the elevation range from dead water level to normal storage level. Since the water level of most reservoirs fluctuates year-round between the dead water level and normal storage level, remote sensing monitoring offers wide coverage, convenient data acquisition, and high timeliness, providing a reliable basis for analyzing reservoir capacity trends. AI algorithms are used to extract water body characteristics from remote sensing images, including spectral, texture, and shape features. A deep learning model, such as a CNN, is then trained to identify water bodies and non-water bodies. Based on the identified water body areas and corresponding elevation data, reservoir capacity is calculated at different water levels. Finally, based on the calculated results, a reservoir capacity characteristic curve is plotted within the elevation range from dead water level to normal storage level.

[0071] Step S2 specifically includes:

[0072] Step S2.1: Download optical images covering the reservoir with cloud cover less than 10% from 2002 to 2023. The requirements for remote sensing images are as follows: the image should be as large as possible and the study area should be as cloud-free as possible. If one image cannot cover the study area, multiple images can be stitched together.

[0073] Step S2.2: Preprocess the remote sensing image, including radiation correction, atmospheric correction, and cloud removal, to improve data quality.

[0074] Step S2.3: ① Based on the preprocessed remote sensing image, a total of 1,000 sample points were collected using the visual interpretation method, including 500 water body and non-water body sample points; ② With the assistance of the image analysis software segmentation tool, the optimal segmentation scale was determined and the sample points were used as constraints to achieve multi-scale segmentation and classification of the original image; ③ The center of gravity of the shape after multi-scale segmentation was calculated, and with the center of gravity as the center, a training sample of homogeneous pixel image blocks suitable for the CNN input port was generated; ④ A CNN model was constructed and the training samples were input into the model for training; ⑤ The trained model was used to classify the long-term images of the study area to generate water body extraction results.

[0075] Step S2.4: Calculate the water area based on the extracted water body information, and search the water level value corresponding to the imaging time of each image on the hydrological telemetry system, and fit and establish the water level-area relationship curve, such as Figure 4 Table 1 shows the water surface area and corresponding water levels extracted from different remote sensing images of the reservoir. In addition to ensuring accuracy, corresponding reservoir area data must be available for each meter-level water level. Reservoir capacity is calculated using area data and corresponding water level data to determine the reservoir shape. An appropriate mathematical model for capacity calculation (generally a trapezoidal formula or a pyramidal formula) is selected to calculate the cumulative capacity error. If the minimum capacity data cannot be obtained, the capacity curve is reconstructed and combined with the original low-water-level interval capacity curve for smoothing to obtain the reconstructed reservoir capacity value.

[0076] Table 1 Water surface areas and corresponding water levels extracted from different remote sensing images

[0077] Remote sensing image period (yyyyMMddHHmm) <![CDATA[Area (km 2 )]]> Water level (m) 200202061054 91.3 186.68 200205291054 88.7 183.06 200206141053 57.3 182.57 200210201053 96.2 185.93 200311081054 81.8 181.15 200402121054 68.8 179.17 200405021054 66.2 175.68 200410091053 58.1 172.10 200410251054 55.3 171.16 200411101054 57.6 171.27 200412121054 65.8 171.41 200505211054 57.1 165.11 200511131055 83.1 182.04 200602011055 79.6 180.41 200604061055 72.1 177.69 200604221055 71.4 176.97 200610151054 73.7 176.3 200801061056 73.9 178.03 200803101055 66.9 176.12 200906011055 57.4 176.75 200908201055 70.4 177.14 200909211055 71.4 178.27 201303081101 97.4 185.42 201401061102 107.2 188.73 201401221101 106.1 188.56 201501251104 103.6 186.79 201504151104 96.2 184.2 201508051105 77.2 179.49 201509061105 73.7 177.92 201602291107 57.2 175.87 201608071108 55.1 171.26 201610261108 93.7 184.62 201705221108 96.2 184.21 201709111107 77.1 181.85 201709271108 84.7 182.31 201801171107 91.1 184.53 201811011102 98.6 187.9 201904101057 94.6 183.97 201310261107 102.9 186.7 201312291106 108.9 188.66 201401301106 94.8 188.37 201403191106 103.9 186.87 201501011105 103.1 186.75 201501171104 103.3 186.78 201506261104 85.7 181.03 201509301106 61.2 177.19 201510161105 68.0 177.5 201511171105 71.7 177.36 201606121104 65.2 173.09 201612051105 102.2 187.06 201709191106 82.7 182.2 201802101104 93.3 183.88 201905201104 92.0 182.59 201906051105 90.2 182.54 201910111105 90.4 183.91 202001151105 92.1 183.37 202005061104 79.1 178.51 202007091105 70.8 175.42 202007251106 63.9 174.53 202008101105 63.4 173.94 202009111105 61.3 173.17 202101011105 81.3 180.83 202105091104 77.1 178.37 202112031106 93.3 183.89 201901231122 102.0 186.83 201902071122 98.5 186.36 201902271122 99.0 185.65 201903091122 99.6 185.26 201903191122 98.2 184.87 201904081122 94.4 184.08 201904231122 92.1 183.48 201905181122 93.6 182.65 201906071122 87.6 182.47 201909301122 94.4 184.26 201910051122 92.1 184.1 201911241122 90.3 183.92 201912091122 89.6 183.92 202002021122 88.9 182.91 202002121122 87.4 182.55 202003081122 85.3 181.43 202003131122 83.8 181.17 202003281122 82.0 180.43 202005071122 77.1 178.46 202006211122 70.9 176.28 202007061122 69.1 175.58 202007161122 68.1 175.02 202007261122 65.8 174.47 202008101122 61.8 173.94 202011131122 78.5 179.67 202101021122 80.3 180.83 202101121122 81.6 180.8 202102161122 79.3 180.2 202102211122 79.0 180.07 202102261122 79.4 179.95 202105121122 77.8 178.26 202105171122 78.0 178.07 202105221122 73.7 177.85 202109291122 59.3 173.69 202112031122 91.5 183.89 202203081122 99.2 182.75 202204271122 93.0 181.45 202303231122 95.0 181.82 202305171122 96.9 179.62 202305221122 87.5 179.39 202307111122 86.3 177.64 202311181122 93.4 181.26 202312081122 111.9 181.55

[0078] The fitting relationship between reservoir water level and area obtained based on different fitting functions is as follows:

[0079] (1) Quadratic polynomial function: y = 0.00664x 2 +0.5996x-241.9(R 2 =0.95)

[0080] (2) Cubic polynomial function: y = -0.003453x 3 +1.872x 2 -335x+19880(R 2 =0.96)

[0081] In the above formula, y represents the water surface area (in km²) and x represents the water level (in meters).

[0082] Based on the above comparison, R 2 The larger cubic polynomial fitting function is used as the reservoir water level~area fitting function.

[0083] The reservoir capacity calculation model based on remote sensing data adopts the prism volume method and the prism formula to calculate the reservoir interval capacity in the range from dead water level to normal water level (175 m~185 m). The capacity is then accumulated to different water levels to obtain the water level-capacity data.

[0084] Step S2.5: Based on the calculation results of step S2.4, draw the reservoir capacity characteristic curve within the range of dead water level to normal water level elevation.

[0085] Step S3: Extract the reservoir capacity characteristic curve for the range from the normal water level to the verified flood level using the digital elevation model (DEM). The digital elevation model (DEM) method generates a terrain model from elevation data and combines this with water level data to estimate reservoir capacity. DEM data for the range from the normal water level to the verified flood level is collected. The "Surface Volume" tool in the GIS software toolbox is used to calculate the volume between the reservoir bottom terrain surface and the reference plane, i.e., the reservoir capacity. Based on this calculation, the reservoir capacity characteristic curve for the range from the normal water level to the verified flood level is plotted. Software for calculating reservoir capacity includes, but is not limited to, QGIS.

[0086] Step S3 includes:

[0087] Step S3.1: Collect the real-scene three-dimensional 2-meter grid DEM data of the reservoir.

[0088] Step S3.2: Preprocess the DEM data, including coordinate system correction, filling depressions, interpolation, clipping, etc., to ensure the accuracy and reliability of the data.

[0089] Step S3.3: Using the preprocessed DEM data of the reservoir area as the background, extract the contour lines. Then, based on the contour lines, existing measured elevation points, and contour lines, regenerate a digital elevation model of regular or irregular shapes. Use the surface volume tool in the GIS software system toolbox to calculate the volume of the area between the reservoir bottom terrain surface and the reference plane, i.e., the reservoir capacity.

[0090] Step S3.4: Based on the calculation results, draw the reservoir capacity characteristic curve within the range of normal water storage level to verification flood level elevation.

[0091] Step S4: Based on the reservoir capacity curve fitting equations from methods S1, S2, and S3, the three reservoir capacity curves are extended 3 to 5 meters in both directions and the reservoir tolerance difference of the three reservoir capacity curves at the water level in the overlapping part is calculated. The reservoir tolerance difference of the three reservoir capacity characteristic curves at the water level in the overlapping part is checked to see if it is within the tolerance range. This further verifies the reliability of the reservoir capacity calculation method and finally obtains the new, verified reservoir capacity curve.

[0092] Specifically, the reservoir capacity characteristic curve within the elevation range below the dead water level, the reservoir capacity characteristic curve within the elevation range below the dead water level, and the reservoir capacity characteristic curve within the elevation range from the normal water level to the verified flood level are extended 3 to 5 meters upstream and downstream to form extended curve segments, so as to enhance the continuity and comparability of different methods in the overlapping water level interval. In the extended overlapping water level interval, the unit water level section reservoir capacity difference value corresponding to each method is calculated at a water level interval of 1 meter to reflect the difference in reservoir capacity calculation of different methods at the same water level. By comparing the unit water level section reservoir capacity difference value, the cross-precision verification of the reservoir capacity curves of different methods is completed. If the difference is small, it means that the calculation results of these methods in this interval are consistent and the accuracy is high; on the contrary, if the difference is large, the calculation model needs to be further optimized or adjusted. Based on the results of cross-validation, the reservoir capacity curve with the smallest error and the highest accuracy is selected to obtain a new verified reservoir capacity curve.

[0093] Table 2 Calculation table of reservoir capacity for each meter of water level using three verification methods

[0094] Water level (m) <![CDATA[Method S1 (hundred million m 3 > <![CDATA[Method S2 (billion m 3 > <![CDATA[Method S3 (hundred million m 3 )]]> 173 0.6044 0.5943 174 0.6134 0.6196 175 0.6182 0.6465 176 0.6204 0.6746 177 0.6223 0.7038 178 0.6274 0.7338 179 0.7747 0.7716 180 0.9130 0.8778 181 0.9708 0.9277 182 1.0022 0.9591 183 1.0111 0.9747 184 1.0089 185 1.0602 186 1.0755 1.1191 187 1.0909 1.1655 188 1.1062 1.2092

[0095] The calculation table of reservoir capacity per meter of water level for the three methods is shown in Table 2. It can be found that in the overlapping part of method S1 and method S2 (173 m~183 m interval), the average absolute value of the per-meter reservoir capacity difference between the two methods is 0.0407 million m3 (accounting for 0.12% of the original total reservoir capacity), and the maximum value is 0.1064 million m3 when the water level is 178 m. 3(accounting for 0.32% of the original total reservoir capacity), with a minimum value of 179 m3 at 0.0031 billion m3 3 (accounting for 0.01% of the original total storage capacity).

[0096] In the overlapping part of methods S2 and S3, the absolute value of the per-meter reservoir tolerance of the two methods is 0.0436 billion m3 when the water level is 186 m. 3 (accounting for 0.13% of the original total reservoir capacity), and 7.46 million m3 at 187 m 3 (accounting for 0.22% of the original total reservoir capacity), and 10.30 million m3 at 188 m 3 (accounting for 0.31% of the original total storage capacity).

[0097] Through cross-validation, it can be found that the reservoir capacity difference of each 1-meter section of water level in the overlapping part of the three curves is less than 0.5% of the original reservoir capacity, and the error is small, which further verifies the reliability of the segmented fitting method proposed in this method.

[0098] The storage capacity curve of the reservoir based on the method of the present invention is obtained by fitting three sections respectively: (1) 120 m to 175 m: underwater terrain measurement method; (2) 175 m to 185 m: AI-based remote sensing image water body recognition method; (3) 185 m to 195 m: land surface digital elevation model method. Figure 5 This is the new storage capacity curve of the reservoir obtained based on the method of the present invention.

[0099] Step S5: Based on the three-segment reservoir capacity characteristic curves obtained in steps S1-S3, a complete water level-reservoir capacity relationship curve covering the entire operating water level range is constructed. The reservoir capacity increment of each water level segment is calculated segment by segment, and the abnormal reservoir capacity loss interval is determined by the set reservoir capacity increment threshold (for example, 10% or 20%).

[0100] By identifying the intervals with large storage capacity changes, reservoir managers can better understand the dynamic characteristics of the reservoir and formulate more scientific and effective management strategies to cope with various hydrological conditions and needs. Therefore, the original storage capacity curve per meter of the water level range from 120 m to 195 m is compared with the reconstructed storage capacity curve per meter of this method to identify the storage capacity loss interval. The comparison diagram is shown in Figure 2. Figure 6 shown. Figure 7 The difference between the reservoir capacity increment reconstructed by this method and the original reservoir capacity increment is shown in the figure for each meter interval. Combining the two figures, it can be found that when the water level changes from 174 m to 181 m, the difference between the two is the largest (greater than 100 million m 3 ).

[0101] In the range of 174 m to 181 m, the total storage capacity of the reservoir is reduced by 93 million m3 compared with the original storage capacity curve. 3, accounting for the reduction in the entire reservoir capacity (120m to 195m interval) (176 million m 3 In other words, compared with the original reservoir capacity curve, the reservoir capacity loss is most obvious in the water level range of 174 m to 181 m, accounting for 53% of the total reduction.

[0102] This phenomenon is closely related to reservoir siltation caused by soil erosion in the reservoir area. Soil erosion leads to reservoir siltation and river clogging. Due to the fertile land and readily available water resources surrounding the reservoir and along the river, much of the water conservation forest has been cleared for land reclamation and deforestation, causing severe soil erosion. Large amounts of sediment from the slopes have been eroded, transported, and deposited in downstream reservoirs and river channels. This leads to siltation, reducing the effective storage capacity of the reservoirs, impacting their flood control and water supply functions, shortening their service life, and clogging the river channel, raising the riverbed and affecting flood flow.

[0103] In summary, compared to the original storage capacity curve, the reservoir's storage capacity shows a decreasing trend at the same water level. The loss is most pronounced in the 174-181 m water level range. This phenomenon is likely related to increased sedimentation in the reservoir caused by soil erosion in the reservoir area. Reservoir management should strengthen soil and water conservation efforts. Furthermore, field research indicates that the areas of significant storage capacity loss identified by this method are consistent with the actual conditions of the reservoir.

[0104] This invention not only significantly improves the intelligence of reservoir capacity calculations, enabling timely detection of flood control capacity encroachment, but also provides strong support for the safe operation, scientific scheduling, and resource management of reservoirs. Its achievements will help promote the refinement and dynamism of reservoir capacity representation, which is of great significance for the rational allocation of water resources and the scientific formulation of related policies.

[0105] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present invention should be included in the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

Claims

1. A method for calculating storage capacity of large and medium-sized reservoirs and identifying storage capacity loss intervals, characterized in that: The steps include: Step S1: using the underwater terrain measurement method to extract the reservoir capacity characteristic curve within the elevation range below the dead water level; Step S2: Perform water body recognition on the remote sensing image based on AI technology to extract the reservoir capacity characteristic curve within the range of dead water level to normal water level elevation; Step S3: extracting the reservoir capacity characteristic curve within the range of normal water storage level to verified flood level elevation based on the digital elevation model; Step S4: Extend the three sections of reservoir capacity characteristic curves obtained in steps S1-S3 upstream and downstream respectively to form extended curve sections. Within the overlapping water level interval of the extended curve sections, calculate the unit water level section reservoir capacity difference values ​​of the three sections of reservoir capacity characteristic curves in the overlapping part at a water level interval of 1 meter. By comparing the unit water level section reservoir capacity difference values, the cross-precision verification of the three sections of reservoir capacity characteristic curves is completed. Step S5: Based on the three-segment reservoir capacity characteristic curves obtained in steps S1-S3, a complete water level-reservoir capacity relationship curve covering the entire operating water level range is constructed, the reservoir capacity increment of each water level segment is calculated segment by segment, and the abnormal reservoir capacity loss interval is determined by the set reservoir capacity increment threshold.

2. The method for calculating storage capacity of large and medium-sized reservoirs and identifying storage capacity loss intervals according to claim 1, wherein: The step S1 comprises: Step S1.1: Use an unmanned vessel equipped with a multi-beam bathymetry system to measure underwater topography in the reservoir area. The multi-beam bathymetry system transmits sound waves and receives reflected signals to obtain raw multi-beam data of the underwater topography. Step S1.2: Preprocess the collected raw multi-beam data, including acoustic signal denoising, signal strength correction, data georeferencing, and generate three-dimensional data of the reservoir underwater topography based on GIS software; Step S1.3: Based on the underwater topography 3D data and the reservoir geometry, calculate the reservoir capacity at different water levels using triangulation, grid method, or volume method; Step S1.4: Based on the storage capacity of the reservoir at different water levels calculated in step S1.3, draw a storage capacity characteristic curve within the elevation range below the dead water level.

3. The method for calculating storage capacity of large and medium-sized reservoirs and identifying storage capacity loss intervals according to claim 1, wherein: The step S2 comprises: Step S2.1: Obtain a long-term remote sensing image of the study area including the maximum flooded area of ​​the reservoir; Step S2.2: Preprocess the acquired remote sensing image, including radiation correction, atmospheric correction, and cloud removal; Step S2.3: Train the deep learning model to identify water bodies and non-water bodies. The deep learning model automatically adjusts its parameters to maximize recognition accuracy by learning water body and non-water body samples in the training dataset. Step S2.4: Calculate the area of ​​the water body identified in step S2.3, find the water level value corresponding to the imaging time of each image, fit the water body area and the corresponding water level on the day, select the function with the highest accuracy as the fitting function, and obtain the water level-area curve of the reservoir in this water level range; Based on the reservoir capacity calculation model of remote sensing data and the water level-area curve of the reservoir in the water level range, the interval storage capacity of the reservoir in the water level range is calculated, and then accumulated to different water levels to obtain the water level-capacity data; Step S2.5: Based on the calculation results of step S2.4, draw the reservoir capacity characteristic curve within the range of dead water level to normal water level elevation.

4. The method for calculating storage capacity of large and medium-sized reservoirs and identifying storage capacity loss intervals according to claim 3, characterized in that: The reservoir capacity calculation model in step S2.4 adopts the trapezoidal or pyramidal volume method. The model first determines whether the reservoir is a trapezoidal or pyramidal body based on its type. Then, the water body is divided into n layers according to different water levels. The reservoir capacity is calculated by accumulating the volume of the n layers of trapezoidal or pyramidal bodies. The trapezoidal formula is: ； The pyramid formula is: ； The formula for cumulative storage capacity is: ； In the above formula, is the reservoir capacity difference between two adjacent water levels; is the water level difference between two adjacent water levels; 、 are the water surface areas corresponding to two adjacent water levels; i is the ordinal number; n is the cumulative number; is the initial storage capacity; V is the cumulative reservoir volume.

5. The method for calculating storage capacity of large and medium-sized reservoirs and identifying storage capacity loss intervals according to claim 1, characterized in that: The step S3 comprises: Step S3.1: Obtaining DEM data of the reservoir area by photogrammetry, ground surveying, or digitizing existing topographic maps, wherein the DEM data provides topographic information of the reservoir, including water surface extent, topography, and elevation data; Step S3.2: Preprocess the DEM data, including coordinate system correction, filling depressions, interpolation, and clipping; Step S3.3: Using the pre-processed DEM data of the reservoir area as the background, use the surface volume tool in the GIS software system toolbox to calculate the volume of the area between the reservoir bottom terrain surface and the reference plane, that is, the reservoir capacity; Step S3.4: Based on the calculation results, draw the reservoir capacity characteristic curve within the range of normal water storage level to verification flood level elevation.

6. The method for calculating storage capacity of large and medium-sized reservoirs and identifying storage capacity loss intervals according to claim 1, characterized in that: Step S4 includes: Step S4.1: Based on the reservoir capacity curve fitting equations of methods S1, S2, and S3, the three obtained reservoir capacity curves are extended 3 to 5 meters toward both ends; Step S4.2: Calculate the storage capacity difference of each section of the three storage capacity curves at the overlapping part of the water level of 1 meter; Step S4.3: Check whether the 1-meter tolerance of each section of the three reservoir capacity characteristic curves in the overlapping part is within the tolerance range, and further verify the reliability of the reservoir capacity calculation methods S1, S2 and S3.

7. The method for calculating storage capacity of large and medium-sized reservoirs and identifying storage capacity loss intervals according to claim 1, characterized in that: Step S5 includes: Step S5.1: Integrate the reservoir capacity curves calculated by methods S1, S2, and S3 according to their applicable elevation ranges, eliminate overlaps and breakpoints, and construct a complete water level-reservoir capacity relationship curve covering the entire operating water level range; Step S5.2: Calculate the difference between the complete water level-storage capacity relationship curve and the original design storage capacity curve segment by segment, taking 1 meter as the water level interval, and obtain the storage capacity increment for each water level segment; Step S5.3: Through statistical analysis of storage capacity increments, based on historical data or engineering experience, a storage capacity increment threshold is set to identify abnormal storage capacity loss intervals.

8. The method for calculating storage capacity of large and medium-sized reservoirs and identifying storage capacity loss intervals according to claim 7, characterized in that: The threshold value of the fixed storage capacity increment is 10% or 20%.

9. The method for calculating storage capacity of large and medium-sized reservoirs and identifying storage capacity loss intervals according to claim 8, characterized in that: Step S5 further includes: Step S5.4: Use field measurements or other independent data sources to verify the reservoir storage capacity curve calculated according to S1, S2, and S3, as well as the identification of storage capacity loss intervals and related calculated values, and propose corresponding maintenance or repair measures based on the storage capacity loss situation.

Images