A large and medium-sized reservoir capacity calculation and capacity loss interval identification method

Abstract

The present application provides a large and medium-sized reservoir storage capacity calculation and storage capacity loss interval identification method, comprising: using underwater topography measurement method to extract the storage capacity characteristic curve in the elevation range below the dead water level; based on AI technology, water body is identified from remote sensing images to extract the storage capacity characteristic curve in the elevation range from the dead water level to the normal storage water level; based on the digital elevation model, the storage capacity characteristic curve in the elevation range from the normal storage water level to the check flood water level is extracted; the three-section storage capacity characteristic curves are extended and then cross-precision verification is performed; based on the three-section storage capacity characteristic curves, a complete water level-storage capacity relationship curve covering the entire operating water level range is constructed, the storage capacity increment of each water level section is calculated, and the abnormal storage capacity loss interval is determined through the threshold value of the set storage capacity increment. The present application can quickly and accurately calculate the storage capacity value of large and medium-sized reservoirs at different elevations, effectively identify the elevation interval with serious storage capacity loss, and significantly improve the intelligent level and timeliness of reservoir storage capacity calculation.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of large and medium-sized reservoir area management, and particularly relates to a large and medium-sized reservoir storage capacity calculation and reservoir storage capacity loss interval identification method. BACKGROUND

[0002] Reservoir storage capacity is one of the core parameters in water conservancy construction and operation management, and is an important basis for formulating flood control scheduling schemes, dam safety monitoring, ensuring the safety of downstream river basins, and evaluating the effectiveness of water conservancy projects. However, current part of the reservoirs are facing many problems, such as fuzzy management and protection boundaries, inadequate supervision, illegal occupation of reservoir land, and increasingly serious sediment accumulation, which leads to the occupation of part of the flood control storage capacity of the reservoir and the occurrence of reservoir storage capacity loss to varying degrees, thereby reducing the flood control capacity of the reservoir and affecting the normal function of the reservoir.

[0003] The factors affecting the accuracy of reservoir storage capacity calculation mainly include the reservoir storage capacity calculation method and the topographic data of the reservoir area. For the reservoirs that have been put into operation, the traditional reservoir capacity curve acquisition and checking method is to set a cross section at a certain interval in the reservoir area range by using the ultrasonic principle, measure the water depth of the cross section, mainly rely on the echo sounder for depth measurement, then draw the underwater topographic map according to the cross section measurement results, and then calculate the water surface area of the reservoir at different water levels by using methods such as the planimeter method, the grid method, and the grid point method. The volume calculation is carried out by manually setting lines and reading the corresponding area at certain water level intervals to calculate the reservoir storage capacity. This kind of method has long measurement period, high labor cost, and limited accuracy of the results, and it is difficult to meet the real-time and accuracy requirements of reservoir dynamic operation management and intelligent decision-making for reservoir capacity data. With the development of reservoir information management and digital twin water conservancy, the deficiencies of traditional methods in practicality and timeliness have become an important bottleneck restricting the improvement of the level of modern management of reservoirs.

[0004] In recent years, the rapid development of artificial intelligence (AI) and spatial information technology has provided a new path to solve the above problems. AI has strong image recognition, information reasoning and data fusion capabilities, combined with remote sensing images, unmanned platform underwater measurement, DEM terrain modeling and other spatial information technology means, can efficiently extract the shape of the reservoir area, analyze the change of reservoir capacity, and identify the significant area of reservoir capacity loss, and has good technical feasibility and application prospect.

[0005] Therefore, with the continuous development of remote sensing technology, geographic information system (GIS) technology and artificial intelligence (AI) technology, the reservoir capacity calculation method based on multi-source spatial information and intelligent algorithm has gradually become a research and application hotspot. Especially in the management of large and medium-sized reservoirs, how to realize the efficient calculation and dynamic update of reservoir capacity has become a technical difficulty in the current water conservancy information construction. SUMMARY

[0006] The present application aims to solve the above problems, and proposes a large and medium-sized reservoir capacity calculation and capacity loss interval identification method. The method comprehensively utilizes spatial data resources such as remote sensing images, GIS data, digital elevation model (DEM), and combines AI image recognition and analysis technology, to quickly and accurately calculate the reservoir capacity value of large and medium-sized reservoirs at different elevations, and effectively identify the elevation interval with serious reservoir capacity loss, thereby significantly improving the intelligent level and timeliness of reservoir capacity calculation, and providing technical support for scientific scheduling, safety monitoring and fine management of reservoirs.

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

[0008] A large and medium-sized reservoir capacity calculation and capacity loss interval identification method, comprising the following steps:

[0009] Step S1: using underwater topographic survey method to extract the reservoir capacity characteristic curve in the elevation range below the dead water level;

[0010] Step S2: based on AI technology, water body recognition is performed on the remote sensing image to extract the reservoir capacity characteristic curve in the elevation range from the dead water level to the normal storage water level;

[0011] Step S3: based on the digital elevation model, the reservoir capacity characteristic curve in the elevation range from the normal storage water level to the check flood water level is extracted;

[0012] Step S4: the three reservoir capacity characteristic curves obtained in steps S1-S3 are respectively extended upstream and downstream to form extended curve segments, and in the overlapping water level interval of the extended curve segments, the unit water level segment reservoir capacity difference of the three reservoir capacity characteristic curves in the overlapping part is calculated at an interval of 1 meter water level, and the intersection precision verification of the three reservoir capacity characteristic curves is completed by comparing the unit water level segment reservoir capacity difference;

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

[0014] Further, the step S1 comprises:

[0015] Step S1.1: using an unmanned ship to carry a multi-beam sounding system to measure the underwater topography in the reservoir area, the multi-beam sounding system acquires original multi-beam data of underwater topography by emitting sound waves and receiving reflected signals;

[0016] Step S1.2: pre-processing the collected original multi-beam data, including denoising of sound wave signals, correction of signal intensity, geographical reference of data, and generating three-dimensional data of reservoir underwater topography based on GIS software;

[0017] Step S1.3: Based on the three-dimensional data of underwater topography, combined with the geometric shape of the reservoir, the reservoir capacity at different water levels is calculated by using triangulation, grid method or volume method;

[0018] Step S1.4: According to the reservoir capacity at different water levels calculated in step S1.3, the reservoir capacity characteristic curve in the elevation range below the dead water level is drawn.

[0019] Further, the step S2 comprises:

[0020] Step S2.1: Obtain remote sensing images containing the maximum inundation range of the reservoir in the long-term study area;

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

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

[0023] Step S2.4: Calculate the area of the water body region identified in step S2.3, and find the water level value corresponding to the imaging time of each image. Fit the water body area and the corresponding daily water level data, select the function with the highest accuracy as the fitting function, and obtain the water level-area curve of the reservoir in the water level interval;

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

[0025] Step S2.5: According to the calculation result of step S2.4, draw the reservoir capacity characteristic curve in the elevation range between the dead water level and the normal storage water level.

[0026] Further, the reservoir capacity calculation model in step S2.4 adopts the trapezoidal or prism volume method. The model first determines whether the reservoir is a trapezoidal or prism body according to the reservoir type, then divides the water body into n layers according to different water levels, and calculates the reservoir capacity by accumulating the volumes of the n layers of trapezoidal or prism bodies;

[0027] The trapezoidal formula is:

[0028]

[0029] The prism formula is:

[0030]

[0031] The cumulative reservoir capacity formula is:

[0032]

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

[0034] Further, the step S3 comprises:

[0035] Step S3.1: Obtain the DEM data of the reservoir area by photogrammetry, ground measurement or digitization of existing topographic maps, which provides topographic information of the reservoir, including water surface range, terrain and elevation data;

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

[0037] Step S3.3: Based on the preprocessed DEM data of the reservoir area, use the surface volume tool in the GIS software system toolbox to calculate the volume between the reservoir bottom terrain surface and the reference plane, i.e. the reservoir storage;

[0038] Step S3.4: According to the calculation result, draw the reservoir storage characteristic curve within the normal storage level~check flood level elevation range.

[0039] Further, step S4 comprises:

[0040] Step S4.1: Extend the three-section reservoir storage curves obtained by the reservoir storage curve fitting relationship of S1, S2 and S3 to both ends by 3~5 meters;

[0041] Step S4.2: Calculate the 1-meter-section reservoir storage difference of the overlapping part of the three-section reservoir storage curves;

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

[0043] Further, step S5 comprises:

[0044] Step S5.1: Integrate the reservoir storage curves calculated by S1, S2 and S3 respectively according to their applicable elevation range, eliminate overlaps and breakpoints, and construct a complete water level~reservoir storage 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 in 1-meter water level intervals, and obtain the storage capacity increment of each water level section;

[0046] Step S5.3: Set a threshold value for the storage capacity increment based on historical data or engineering experience through statistical analysis of the storage capacity increment, and identify the storage capacity abnormal loss interval.

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

[0048] Further, step S5 further comprises:

[0049] Step S5.4: Verify the reservoir storage capacity curve calculated according to S1, S2, S3, and the identification and related calculation values of the storage capacity loss interval using field measurement or other independent data sources, and propose corresponding maintenance or repair measures according to the storage capacity loss.

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

[0051] (1) The present application integrates underwater multi-beam sounding, remote sensing image recognition, DEM terrain modeling and other multi-source spatial information technologies, significantly improves the automation and calculation accuracy of reservoir capacity data acquisition, and can meet the management needs of rapid updating of large and medium-sized reservoir capacity.

[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 improves the calculation consistency after data fusion.

[0053] (3) The difference analysis method based on 1-meter section reservoir capacity difference proposed by the present application can accurately identify the elevation interval with significant storage capacity loss, and provide data support for reservoir sedimentation detection, morphological change identification and safety management.

[0054] (4) The present application can be directly used for dynamic checking of reservoir operation basic data, which is helpful for scientific decision-making of key functions such as flood control scheduling, water resource allocation, and hydropower generation, and has good engineering applicability and popularization value. BRIEF DESCRIPTION OF DRAWINGS

[0055] Figure 1 is a flowchart of a large and medium-sized reservoir capacity calculation and reservoir capacity loss interval identification method according to an embodiment of the present application.

[0056] Figure 2 is a flowchart of a reservoir capacity feature curve method in the dead water level-normal impoundment elevation range based on AI remote sensing image water body recognition technology.

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

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

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

[0060] Figure 6 is a comparison diagram of per meter capacity of the original reservoir capacity curve of a reservoir in South China and per meter capacity of the reconstructed reservoir capacity curve.

[0061] Figure 7 is a difference change diagram of 1-meter section 2024 reservoir capacity increment and original reservoir capacity increment of a reservoir in South China. DETAILED DESCRIPTION

[0062] To make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0063] As shown in Figure 1 , the present embodiment provides a large and medium-sized reservoir capacity calculation and reservoir capacity loss interval identification method, which takes a large reservoir in South China as the object, the dead water level of the reservoir is 175 m, the normal storage water level is 185 m, and the check flood level is 195 m, and the specific implementation method is performed according to the following steps:

[0064] Step S1: using underwater topographic survey method to extract capacity characteristic curve in the elevation range below the dead water level.

[0065] The underwater topographic survey method is adopted, the multi-beam sounding system is carried on the unmanned ship, the underwater topography of the area below the dead water level of the reservoir is surveyed, and the original topographic data in the elevation range is obtained. The measurement data is preprocessed, fitted and modeled, converted into water level-capacity corresponding relationship, and the capacity characteristic curve below the dead water level is formed, which reflects the volume change under different water levels in the area. Step S1 specifically includes:

[0066] Step S1.1: Underwater topographic survey adopts multi-beam sounding system carried by unmanned ship. After processing the collected underwater survey data, the underwater topographic three-dimensional data is generated. Before measurement, ensure that the unmanned ship and multi-beam sounding system have been installed and debugged, and the required measurement software is ready. The sounding plan line is laid out according to the requirements, and the laying interval of the sounding plan line is 3 to 4 cm on the 1:10000 topographic map. The sounding line is preferably parallel to the main dam axis of the reservoir. The sounding check line is laid out according to the requirements, and is perpendicular to the sounding plan line. Before and after each operation, the flow station should be moved to the known point (or check point), the coordinates of the known point (or check point) are recorded, and compared with the known coordinates of the point. The difference meets the specification requirements before operation to ensure the accuracy of the sounding point positioning.

[0067] Step S1.2: ①Original data checking and processing: first, the collected water depth data is input into the data processing software, and the water depth measurement results are checked. Some distorted water depth points are removed, and unqualified parts are repaired and measured. Local smoothing processing can be performed. ②Water depth correction and data conversion: the data file containing the plane coordinates of the sounding point, water surface elevation, water depth value and other information is automatically processed by the sounding software (wherein the underwater point elevation H is calculated by subtracting the water depth value from the water surface elevation), and then converted into a data format file containing only point number, code and X, Y, H, which conforms to the digital mapping software. ③Drawing underwater contour lines: the sounding points are drawn on the computer screen according to the mapping scale, and the underwater contour lines are calculated and drawn by establishing a digital terrain model (DTM), generating a triangular network, and calculating and drawing underwater contour lines. The underwater contour lines are represented by solid lines.

[0068] Step S1.3: First, the pour point of the reservoir outlet position is obtained, and then the catchment area of the reservoir is calculated using the watershed tool, and the vector surface data of the catchment area of the reservoir is extracted using the raster-to-surface tool. Then, the vector surface data of the submerged area is extracted using the raster-to-surface tool, and the raster data below the reservoir level is extracted using the conditional function. The DEM of the reservoir area is cut out, and the DEM of the submerged area is extracted using the conditional function. Finally, a loop is set, and each 1 meter is sequentially lowered, and the area and volume of the reservoir under different water level conditions are calculated using the surface volume tool.

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

[0070] Step S2: Water body recognition based on AI technology to extract the reservoir capacity characteristic curve in the dead water level~normal storage level elevation range. Since most reservoir water levels fluctuate between the dead water level and the normal storage level year-round, and remote sensing monitoring has the characteristics of wide coverage, convenient data acquisition, and strong timeliness, it provides a reliable basis for reservoir capacity trend analysis. AI algorithm is used to extract water body features from remote sensing images, including spectral features, texture features, and shape features, etc. Then a deep learning model such as CNN is trained to identify water bodies and non-water bodies. Based on the identified water body area and corresponding elevation data, the reservoir capacity at different water levels is calculated. Finally, according to the calculation results, the reservoir capacity characteristic curve in the dead water level~normal storage level elevation range is drawn.

[0071] Step S2 specifically includes:

[0072] Step S2.1: Download optical images covering the reservoir from 2002 to 2023 with cloud cover less than 10%. The requirements for remote sensing images are: as many maps as possible, and as little cloud coverage as possible in the study area; if one image cannot cover the study area, multiple images can be spliced together.

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

[0074] Step S2.3: ① Based on the preprocessed remote sensing images, 1000 sample points are collected using visual interpretation method, with 500 water body and non-water body sample points each; ② The optimal segmentation scale is determined with the aid of image analysis software segmentation tool, and the original image is segmented and classified at multiple scales with sample points as constraints; ③ The center of gravity of the multi-scale segmented shape is calculated, and a homogeneous pixel image block training sample suitable for the input port of CNN is generated with the center of gravity as the center; ④ A CNN model is constructed, and the training sample is input into the model for training; ⑤ The trained model is used to classify the long-term image of the study area, and the water body extraction result is generated.

[0075] Step S2.4: Calculate the water area based on the extracted water body information, and find the water level value corresponding to the imaging time of each image on the hydrological telemetry system, and fit to establish the water level~area relationship curve, as shown in Figure 4 Table 1 shows the water surface area extracted from different remote sensing images of the reservoir and the corresponding water level. In addition to ensuring accuracy, there must be corresponding reservoir area data for each meter-level water level. The area data and corresponding water level data are used to calculate the reservoir capacity, determine the reservoir shape, and select the appropriate reservoir capacity calculation mathematical model (generally trapezoidal formula or prism formula) to calculate the cumulative reservoir capacity difference. If the lowest reservoir capacity data cannot be obtained, the reservoir capacity curve is reconstructed, and the original low water level interval reservoir capacity curve is combined for smoothing processing to obtain the reconstructed reservoir capacity value.

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

[0077] Remote sensing image date (yyyyMMddHHmm) 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 the water level and the area of the reservoir 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 (unit: km²), and x represents the water level (unit: m).

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

[0083] Based on the reservoir storage calculation model of remote sensing data, the prismatic volume method is adopted, and the prismatic formula is used to calculate the interval storage of the reservoir in the dead water level~normal storage level (175 m~185 m) range, which is then accumulated to different water levels to obtain the water level~storage data.

[0084] Step S2.5: According to the calculation results of step S2.4, draw the reservoir capacity characteristic curve in the dead water level~normal storage level elevation range.

[0085] Step S3: Based on the digital elevation model, extract the reservoir capacity characteristic curve in the normal storage level~check flood level elevation range; the digital elevation model (DEM) method generates a terrain model through elevation data, and estimates the reservoir capacity combined with water level data. Collect the DEM data in the normal storage level~check flood level elevation range, 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, according to the calculation results, draw the reservoir capacity characteristic curve in the normal storage level~check flood level elevation range. The reservoir capacity calculation software includes but is not limited to QGIS, etc.

[0086] Step S3 includes:

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

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

[0089] Step S3.3: Based on the preprocessed DEM data of the reservoir area, contour lines are extracted, and then a regular or irregular digital elevation model is regenerated according to the contour lines, existing measured elevation points, and contour lines. The volume of the area between the reservoir bottom terrain surface and the reference plane is calculated using the surface volume tool in the GIS software system toolbox, which is the reservoir capacity.

[0090] Step S3.4: Based on the calculation results, the reservoir capacity characteristic curve within the normal water level ~ check flood water level elevation range is drawn.

[0091] Step S4: Based on the fitting relationship of the reservoir capacity curves of the three methods S1, S2, and S3, the three-section reservoir capacity curves are obtained, and the two ends are extended by 3-5 meters. The water level 1-meter-section reservoir capacity difference in the overlapping part is calculated. The water level 1-meter-section reservoir capacity difference in the overlapping part of the three-section reservoir capacity characteristic curves is checked to further verify the reliability of the reservoir capacity calculation method, and then the new reservoir capacity curve after checking is obtained.

[0092] Specifically, the reservoir capacity characteristic curve below the dead water level, the reservoir capacity characteristic curve below the dead water level, and the reservoir capacity characteristic curve within the normal water level ~ check flood water level elevation range are respectively extended upstream and downstream by 3-5 meters to form extended curve sections, 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 1-meter water level interval to reflect the reservoir capacity calculation difference 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 precision is high; otherwise, 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 precision is selected to obtain the new reservoir capacity curve after checking.

[0093] Table 2 Water level 1-meter-section reservoir capacity calculation table of three review methods

[0094] Water level (m) Method S1 (billion m 3 ])]] Method S2 (billion m 3 ])]] Method S3 (billion 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 water level 1-meter-section reservoir capacity calculation table of 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 absolute value of the per-meter reservoir capacity difference of the two methods is 0.0407 million m3 (0.12% of the original total reservoir capacity), and the maximum value is 0.1064 million 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 53% of the total capacity reduction (176 million m 3 In other words, compared with the original reservoir capacity curve, the 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 the reservoir siltation caused by soil erosion in the reservoir area. Soil erosion leads to siltation of the reservoir and blockage of the river channel. Due to the fertile land and convenient water use, most of the water conservation forests around the reservoir and on both sides of the river channel have been destroyed and deforested, causing serious soil erosion. A large amount of slope sediment is eroded and transported and then deposited in the downstream reservoir and river channel. This leads to siltation of the reservoir, reduction of effective reservoir capacity, and affects the play of its flood control, water supply and other functions, also reduces its service life, and also blocks the river channel, raises the river bed, and affects the river channel flood discharge.

[0103] In summary, compared with the original reservoir capacity curve, the reservoir capacity shows a decreasing trend at the same water level, especially in the water level range of 174 m to 181 m, the capacity loss is most obvious. This phenomenon may be related to the increase of reservoir siltation caused by soil erosion in the reservoir area, and the subsequent reservoir management department should strengthen soil and water conservation work. At the same time, through field research, the larger area of reservoir capacity loss identified by the method is consistent with the actual situation of the reservoir.

[0104] The present application not only can significantly improve the intelligent level of reservoir capacity calculation and timely find the occupation of flood control capacity, but also can provide strong support for the safe operation, scientific scheduling and resource management of the reservoir. The results are helpful to promote the refinement and dynamic of reservoir capacity expression, and have important significance for the rational allocation of water resources and the scientific formulation of related policies.

[0105] The above is only a specific embodiment of the present application, but the protection scope of the present application is not limited thereto. Any changes or replacements within the technical scope disclosed by the present application can be easily thought of by any person skilled in the art, and should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope 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 is 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