Reservoir capacity curve checking method based on multi-source data fusion

Abstract

The application discloses a reservoir capacity curve checking method based on multi-source data fusion, which comprises the following steps: S1, DEM data preprocessing: at least two sets of DEM data of different periods or different sources of a reservoir are subjected to data quality enhancement respectively; then, a plurality of sets of DEM data after data quality enhancement are fused to obtain a set of accurate new DEM data; S2, reservoir capacity curve acquisition: a preliminary reservoir capacity curve is generated based on the new DEM data; a DEM region corresponding to a to-be-corrected region in the preliminary reservoir capacity curve is optimized to generate a theoretical reservoir capacity curve; S3, reservoir capacity curve correction: a water storage capacity and water level relationship of the theoretical reservoir capacity curve is constructed, a correction equation is established for a water level subinterval, and a correction parameter in the correction equation is fitted by using measured scattered point data, so that an accurate reservoir capacity curve is obtained. The method solves the problems of single data, static model error and insufficient checking in the traditional method, and realizes high-precision and generalizable reservoir capacity curve generation.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of water conservancy and geographic information technology, and in particular to a reservoir capacity curve checking method based on multi-source data fusion. BACKGROUND

[0002] The reservoir capacity curve is the core basis of reservoir operation and management, and directly determines the reliability of flood control scheduling, water resource optimization, and engineering safety evaluation. In flood control scheduling, the reservoir capacity curve accurately calculates the flood storage capacity through the water level-storage capacity relationship, providing a basis for flood discharge decision-making; in water resource management, its accuracy affects the scheduling and distribution of comprehensive benefits such as irrigation, power generation, and water supply; in safety evaluation, the calculation of siltation volume and dam stability analysis all need to take the reservoir capacity curve as a reference. Studies have shown that if the error of the reservoir capacity curve exceeds 5%, it may lead to misjudgment of flood storage capacity, failure of irrigation plans, and even cause dam breach risks. Therefore, a high-precision reservoir capacity curve is the "lifeline" of reservoir safety and efficient operation.

[0003] Despite the rapid development of hydrological monitoring technology (such as satellite remote sensing, unmanned aerial survey, and the popularization of automatic water level stations), the dynamic updating of the reservoir capacity curve still faces severe challenges. On the one hand, traditional methods rely on historical surveying and mapping data or a single DEM to generate static curves, although they are combined with periodic cross-section measurements for verification, the update cycle is long (usually 3-5 years) and the cost is high, making it difficult to reflect the dynamic changes in the reservoir area (such as siltation, erosion, and landslide entry into the reservoir) in a timely manner. On the other hand, the existing monitoring system has accumulated a large amount of real-time data (such as water level-flow time series and multi-period DEM), but it forms a "data island" with historical achievements: real-time hydrological data is only used for short-term early warning, and is not used to drive the correction of the reservoir capacity curve; due to differences in resolution and collection time, multi-source DEM data lack a fusion verification mechanism, resulting in insufficient spatiotemporal continuity in terrain modeling. Therefore, how to connect the fusion channel of real-time monitoring data and historical multi-source DEM, and build a closed-loop checking system of "data-driven correction-terrain dynamic modeling", has become a key challenge to improve the reliability of the reservoir capacity curve. SUMMARY

[0004] The present application relates to the field of water conservancy and geographic information technology, and in particular to a reservoir capacity curve checking method based on multi-source data fusion.

[0005] To achieve the above-mentioned purpose, the technical solution adopted by the present application is as follows:

[0006] A reservoir capacity curve checking method based on multi-source data fusion, comprising the following steps,

[0007] S1, DEM data preprocessing: at least two sets of DEM data of different periods or different sources of the reservoir are respectively subjected to coordinate correction, resolution matching and noise removal processing to realize data quality enhancement; the multi-set DEM data after data quality enhancement is fused by using conflict area refinement processing combined with an adaptive weighted fusion algorithm to obtain a new accurate DEM data;

[0008] S2, reservoir capacity curve acquisition: the new DEM data are used to calculate the reservoir capacity to generate a preliminary reservoir capacity curve; the to-be-corrected area of the preliminary reservoir capacity curve is determined based on the comparison result of the preliminary reservoir capacity curve and the historical reservoir capacity curve, and the theoretical reservoir capacity curve is generated by optimizing the DEM area corresponding to the to-be-corrected area;

[0009] S3, reservoir capacity curve correction: the relationship between the storage capacity and the water level of the theoretical reservoir capacity curve is constructed, the correction equation is established for each water level sub-interval, and the correction parameters in the correction equation are fitted by using the measured scattered point data to obtain an accurate reservoir capacity curve.

[0010] Preferably, step S1 specifically includes the following contents,

[0011] S11, each set of DEM data is unified to the CGCS2000 coordinate system to ensure the consistency of the plane datum; then, the resolution of each set of DEM data is matched, bilinear interpolation is performed on the DEM data with low resolution, and resampling is avoided on the DEM data with high resolution to introduce high-frequency noise; finally, each set of DEM data is subjected to noise removal processing, the local elevation standard deviation is calculated, the abnormal pixels with the local elevation standard deviation exceeding the preset standard deviation are removed, and the reservoir boundary is subjected to mask processing to exclude the interference of the surrounding terrain;

[0012] S12, for the conflict area in each set of DEM data, if the difference between each set of DEM data is too large, CAD drawings or remote sensing satellite images are used for dynamic matching;

[0013] S13, the noise part in each set of DEM data is removed, the accurate part in each set of DEM data is retained, and an adaptive weighted fusion algorithm is used to fuse the DEM data, so as to cut out a new accurate DEM data.

[0014] Preferably, the adaptive weighted fusion algorithm dynamically allocates weights to each set of DEM data according to the vertical accuracy, acquisition time and terrain complexity of each set of DEM data to realize the fusion of multiple sets of DEM data.

[0015] Preferably, step S2 specifically includes the following contents,

[0016] S21, calculate the storage capacity by using the new DEM data, and generate a preliminary storage capacity curve by theoretically analyzing the storage capacity curve according to the historical storage capacity curve and the reservoir siltation condition;

[0017] S22, compare the preliminary storage capacity curve with the historical storage capacity curve, determine the to-be-corrected area in the preliminary storage capacity curve based on the relative error between the water levels of the two curves, reconstruct the terrain surface of the DEM area corresponding to the to-be-corrected area, minimize the storage capacity deviation between the preliminary storage capacity curve and the historical storage capacity curve, and generate a theoretical storage capacity curve.

[0018] Preferably, step S21 specifically comprises the following steps:

[0019] Preferably, step S22 specifically comprises the following steps:

[0020] S221, calculate the relative error by aligning the preliminary storage capacity curve with the historical storage capacity curve according to the water level, and mark the points in the preliminary storage capacity curve with a relative error greater than a first preset error as the to-be-corrected area;

[0021] S222, after completing the marking of the to-be-corrected area, optimize the DEM area corresponding to the to-be-corrected area, reconstruct the terrain surface of the DEM area corresponding to the to-be-corrected area by using the moving least square method, and minimize the storage capacity deviation between the preliminary storage capacity curve and the historical storage capacity curve;

[0022] S223, recalculate the storage capacity by using the locally optimized DEM data until the error between the preliminary storage capacity curve and the historical storage capacity curve is less than or equal to a second preset error and the trend of the preliminary storage capacity curve is consistent with the trend of the historical storage capacity curve, if not, repeat the above operation until the preliminary storage capacity curve meets the above two indexes, and generate a theoretical storage capacity curve.

[0023] Preferably, the second preset error is smaller than the first preset error.

[0024] Preferably, step S3 specifically comprises the following steps:

[0025] S31, based on the water balance, inverse the change of the storage capacity in a certain range, select at least 30 days of continuous water level time series data and corresponding inflow and outflow, and integrate the water balance equation according to the time step;

[0026] S32, adopt a piecewise correction method, construct a relationship between the storage capacity and the water level, divide the water level range into subintervals with a preset interval distance, and establish a correction equation for each subinterval;

[0027] S33, the correction parameter in the correction equation is fitted by using the measured scattered point data through a weighted least square method, fine dynamic correction of the reservoir capacity in the range is realized, and an accurate reservoir capacity curve is obtained.

[0028] Preferably, step S3 further comprises,

[0029] S34, 10% of the measured scattered point data is selected as a verification set, the root mean square error of the accurate reservoir capacity curve is calculated to check the accuracy of the correction result; if the root mean square error is greater than a preset root mean square error, the correction parameter needs to be adjusted according to the correction equation until the root mean square error is less than or equal to the preset root mean square error.

[0030] Preferably, the DEM data must include the reservoir capacity and the surrounding catchment area.

[0031] The beneficial effects of the present application are: 1. The traditional method relies on single DEM data, which is easily affected by data source errors and time lag, especially in complex terrain areas such as vegetation coverage and river channels, the reservoir capacity calculation error is generally 8%-15%. The present application method effectively eliminates data blind area and local contradiction through multi-source DEM adaptive weighted fusion and fine processing of conflict areas (such as CAD drawings, historical image verification), the terrain modeling accuracy is improved by more than 50%, and the reservoir capacity calculation error after fusion can be controlled within 3%. 2. The traditional method relies on manual section measurement, which takes months to update and costs more than 100,000 yuan, and cannot respond to sudden topographic changes. The present application method uses real-time water level-flow data to complete dynamic updating of the reservoir capacity curve within 30 days through water balance inversion and piecewise linear correction model, and the operation and maintenance cost is reduced by 90%. 3. The traditional method is limited by high cost and data demand, and is only suitable for large reservoirs, while the present application method can adapt to small and medium-sized reservoirs (coverage rate increased by 40%) through multi-source data fusion and local correction mechanism, and is compatible with ArcGIS tool chain in the whole process without custom development. In practical application, the reservoir capacity error ≤3% ensures the safety of flood control, reduces the frequency of manual inspection, and has safety and economy, which has significant popularization value. 4. The present application method can make full use of historical data and accurately correct the reservoir capacity curve according to the monitoring information, which solves the problems of single data, static model error and insufficient verification in the traditional method, and realizes the generation of high-precision and popular reservoir capacity curve. BRIEF DESCRIPTION OF DRAWINGS

[0032] Figure 1 is a flow chart of the method in the embodiment of the present application;

[0033] Figure 2 is a schematic diagram of underwater DEM in the embodiment of the present application;

[0034] Figure 3 is a schematic diagram of noise in the DEM in the embodiment of the present application;

[0035] Figure 4 is a water DEM schematic diagram in the embodiment of the present application;

[0036] Figure 5 is a DEM schematic diagram after data fusion in the embodiment of the present application;

[0037] Figure 6 is a comparison diagram of the fused DEM and the underwater DEM in the embodiment of the present application;

[0038] Figure 7 is a comparison diagram of the fused DEM and the water DEM in the embodiment of the present application;

[0039] Figure 8 is a comparison schematic diagram of the 1996 reservoir capacity curve, the underwater DEM, the checked reservoir capacity curve and the water DEM in the embodiment of the present application;

[0040] Figure 9 is a comparison schematic diagram of the checked curve and the 1996 reservoir capacity curve in the embodiment of the present application. DETAILED DESCRIPTION

[0041] In order to make the purpose, technical scheme and advantages of the present application more clear, the present application is further described in detail below with reference to the drawings. It should be understood that the specific embodiments described herein are only used to explain the present application, and are not used to limit the present application.

[0042] Embodiment one

[0043] As shown in the figure, in the embodiment, a reservoir capacity curve checking method based on multi-source data fusion is provided, which includes the following parts, Figure 1

[0044] I. DEM data preprocessing

[0045] At least two sets of DEM data of different periods or different sources of the reservoir (the DEM data must contain the reservoir capacity and the surrounding catchment area) are respectively subjected to coordinate correction, resolution matching and noise removal processing to realize data quality enhancement; the multi-set DEM data after data quality enhancement is fused by using conflict area refinement processing combined with an adaptive weighted fusion algorithm to obtain a set of accurate new DEM data.

[0046] The coordinate system, resolution, etc. of each set of DEM data may be different, so these different DEM data need to be preprocessed. The preprocessing includes data quality enhancement and data fusion.

[0047] ​First, the coordinates need to be corrected. All DEM data need to be unified into the CGCS2000 coordinate system using the "Project Raster" tool in ArcGIS to ensure consistent plane datums. Subsequently, the two sets of DEM data need to be matched in resolution. For the low-resolution DEM, a bilinear difference is performed and it is converted to a high-resolution grid to avoid high-frequency noise introduced by resampling. Finally, the two sets of DEM data need to be denoised. By using the "Focal Statistics" tool to calculate the local elevation standard deviation, abnormal pixels that exceed the preset elevation standard deviation (such as ±3σ) are eliminated. At the same time, the reservoir boundary is masked to eliminate interference from the surrounding terrain.

[0048] In order to crop more accurate DEM data, it is necessary to fuse the two sets of DEM data. After the above operation, the two sets of DEM data are completely consistent in scale and coordinates, laying the foundation for fusion. Remove the noise part of the two sets of DEM data and retain the accurate part of the two sets of DEM data, so as to crop a set of accurate DEM data. In order to achieve accurate cropping, it is necessary to fuse the two sets of DEM data and use an adaptive weighted fusion algorithm to dynamically assign weights based on the vertical accuracy (RMSE) of the DEM, the acquisition time and the complexity of the terrain. The weight formula is,

[0049]

[0050] Among them, W i is the weight of the i-th set of DEM data; RMSE i is the vertical accuracy of the i-th set of DEM data; T i is the sampling time of the i-th set of DEM data; T current is the current time; λ is the time attenuation coefficient (0.1-0.3 is recommended).

[0051] For conflicting areas between two sets of DEMs, if the two sets of DEM data themselves are too different, dynamic matching is required using CAD drawings or remote sensing satellite images to ensure that the fused data is more reasonable and accurate than the data before fusion.

[0052] 2. Storage Capacity Curve Acquisition

[0053] The reservoir capacity is calculated based on the new DEM data to generate a preliminary reservoir capacity curve; the area to be corrected of the preliminary reservoir capacity curve is determined based on the comparison results with the historical reservoir capacity curve, and the theoretical reservoir capacity curve is generated by optimizing the area to be corrected.

[0054] After completing data quality enhancement and fusion, it is equivalent to generating a new DEM data. First, it is necessary to use the new DEM data to calculate the reservoir capacity and conduct a theoretical analysis of the reservoir capacity curve based on the historical reservoir capacity curve and the reservoir siltation situation, and generate a preliminary reservoir capacity curve.

[0055] Specifically: first, the new DEM data elevation band needs to be divided, according to the amplitude of the reservoir water level, taking the dead water level and the check flood level as the two extreme values, and the spatial discrete difference is carried out according to the preset distance (such as 0.1m), and the "Surface Volume" tool in ArcGIS is used to calculate the corresponding reservoir capacity of each elevation band layer by layer, and the reservoir capacity calculation formula is as follows:

[0056]

[0057] Among them, V k is the reservoir capacity corresponding to the kth elevation band; H k is the water level of the kth elevation band; H j is the elevation of the jth pixel in the new DEM data; A is the pixel area; N is the total number of pixels in the new DEM data.

[0058] According to the above formula, the reservoir capacity value of the new DEM data is calculated, and the preliminary reservoir capacity curve discrete point graph is generated.

[0059] After generating the new reservoir capacity curve discrete point graph, it needs to be compared with the historical curve, and the terrain is modified again to generate the theoretical reservoir capacity curve.

[0060] Specifically: first, the preliminary curve and the historical curve are aligned according to the water level, the relative error is calculated, and the points with an error greater than the first preset error (such as 5%) are marked as the area to be modified.

[0061]

[0062] Among them, E(H) is the relative error of the preliminary reservoir capacity curve and the historical reservoir capacity curve; V n (H) is the preliminary reservoir capacity; V0(H) is the historical reservoir capacity.

[0063] After marking the area to be modified, the local DEM data needs to be optimized, and the DEM area corresponding to the difference interval is reconstructed using the moving least squares method (MLS) to minimize the deviation of the preliminary reservoir capacity curve and the historical curve capacity.

[0064]

[0065] Among them, Ω is the difference interval; V c (H) is the calculated reservoir capacity; V0(H) is the historical reservoir capacity.

[0066] The optimized DEM is recalculated for reservoir capacity until the error E(H)≤second preset error (such as 3%), and the trend of the new reservoir capacity curve is consistent with that of the historical reservoir capacity curve, if not, the above operation needs to be repeated until the reservoir capacity curve meets the above two indicators, and then the theoretical check curve is generated.

[0067] III. Correction of storage curve

[0068] After the above two steps, the theoretical storage curve needs to be dynamically corrected using the measured hydrological data to generate an accurate storage curve. By constructing the relationship between storage and water level of the theoretical storage curve, a correction equation is established for each water level sub-interval, and the correction parameters in the correction equation are fitted using the measured scattered data to obtain an accurate storage curve.

[0069] Based on the principle of water balance, the change of storage is inverted. At least 30 days of continuous water level time series data (sampling interval ≤1 hour) and corresponding inflow and outflow are selected. According to the time step Δt (such as 1 day), the water balance equation is integrated:

[0070] ΔV 实际 =∫(Q in -Q out -Q e -Q s )dt

[0071] Where, ΔV 实际 is the actual change of storage; Q in is the inflow of the reservoir; Q out is the outflow of the reservoir; Q e is the evaporation flow of the reservoir; Q s is the leakage flow of the reservoir; t is the time step. Q in and Q out are measured values; Q e and Q s are empirical values generated based on historical data.

[0072] Based on the method of water balance, when the inflow, outflow, water level, and storage change process are known according to the monitoring information, and compared with the previously generated theoretical storage curve, the storage difference at each point can be known, and then the correction is completed. However, in order to improve the correction efficiency, a segmented correction method is adopted, a relationship between storage and water level is constructed, the water level range is divided into sub-intervals with a preset interval distance (such as 0.1 meters), and a correction equation is established for each sub-interval:

[0073] V x =a·V c +b·H+c

[0074] Where, V x is the corrected storage; V c is the theoretical storage; H is the water level; a, b, c are correction parameters.

[0075] The measured scattered point data is used to fit the parameters a, b and c by the weighted least square method, and the weight is determined by the density of the measured data in the subinterval. Through the above correction, the storage capacity in the range can be finely and dynamically corrected as long as the measured data includes the interval. In order to test the accuracy of the correction, 10% of the collected data can be selected as the verification set, and the root mean square error (RMSE) of the corrected curve is calculated:

[0076]

[0077] wherein V x (H j ) is the storage capacity value of the corrected curve; V s (H j ) is the measured storage capacity value.

[0078] If the RMSE is greater than a preset root mean square error (such as 3%), the correction model parameters need to be adjusted according to the correction equation until the root mean square error is less than 3%.

[0079] Example Two

[0080] As an important water source in Beijing, the storage capacity curve of the Miyun Reservoir (established in 1996) has not been updated for a long time. Due to the continuous siltation and topographic changes in the reservoir area, there is a certain deviation between the actual storage capacity and the curve, which seriously affects the flood control scheduling and water resource utilization efficiency. In order to improve the accuracy, in this embodiment, the method of the present application is used to fuse multi-source data, combined with the real-time hydrological monitoring data in 2024, to dynamically correct the historical storage capacity curve.

[0081] I. DEM data preprocessing

[0082] Two sets of DEM data of the Miyun Reservoir are collected, which are the underwater DEM data set of the Miyun Reservoir prepared by the Natural Resources Righting and Registration Cadastral Survey Project of the City Surveying and Mapping Institute in 23 (hereinafter referred to as underwater DEM) and the water DEM data set of the Miyun Reservoir provided by the Digital Intelligence Institute of the Miyun Reservoir Management Department in 22 (hereinafter referred to as water DEM). At the same time, the 1:2000 CAD drawings of the surrounding area of the Miyun Reservoir prepared by the Hebei Second Institute of Surveying and Mapping in 2005 and the water regime monitoring data of the Miyun Reservoir in 2024 are also collected, including the changes of inflow, water level, water volume and other information in the reservoir area.

[0083] Through the analysis of the two sets of DEM data, it is found that although the coordinate system and resolution of the two sets of DEM data are consistent, there are still some problems. The DEM data of the underwater DEM can be found through the grid histogram distribution that the grid is generally distributed below 151m, and there is a sharp drop in the interval greater than 151m, which does not conform to the actual situation, so the area above 151m in the underwater DEM is marked, see Figure 2Secondly, the image of DEM can also clearly find that there are strip patterns of noise, which affects the accuracy of DEM, see Figure 3 . By analyzing the DEM data on the water with ArcGIS, it is found that it is close to a plane below 131.2m, which obviously does not conform to the actual situation, so the area below 131.2m of the DEM data on the water is marked as an inaccurate area and needs to be discarded, see Figure 4 . Therefore, the approximate accurate interval of DEM data is constructed, considering the underwater DEM below 131.2m as accurate data, the interval of 131.2m-151m is controversial, and the DEM on the water or underwater may be accurate, and the data above 151m is accurate. According to this interval, the CAD drawings provided by Hebei No. 2 Institute are introduced to verify the conflicting area of 131.2m-151m, and finally it is found that the underwater DEM data is more accurate. At this point, the preliminary DEM data screening work is completed, and two sets of DEM data need to be fused. By using the adaptive weighted fusion algorithm and the screened interval for segmented fusion, the fused DEM and the comparison with the underwater and water DEM are shown in Figures 5-7 .

[0084] II. Reservoir capacity curve acquisition

[0085] After the fusion of DEM data, the reservoir capacity is calculated with ArcGIS and compared with the 1996 reservoir capacity curve to analyze the trend and error. The comparison effect diagram is shown in Figure 8 . In this way, the preliminary reservoir capacity curve based on DEM data is completed, and the theoretical reservoir capacity curve is obtained by correcting the to-be-corrected area in the preliminary reservoir capacity curve.

[0086] III. Reservoir capacity curve correction

[0087] Through the image, it can be found that the fused DEM data above 151m has obvious fluctuations, so the next step is to use the water regime monitoring information of Miyun Reservoir in 2024 to accurately correct the theoretical reservoir capacity curve generated by the fused DEM. Because the current reservoir capacity problems mainly exist above 151m, the water level of the reservoir rises from 150.99m to 155.35m in August-October 2024, covering the water level interval that needs to be corrected, so August, September and October 2024 are selected as the water regime data for correction. According to the water balance method, the water level range is divided into 92 sub-intervals according to 0.1m, and each sub-interval is fitted and corrected. The fitting coefficient and formula are as follows:

[0088] V x = 0.98·V c + 0.15·H-12.5

[0089] The revised curve is compared with the storage curve in 1996 as shown in the figure Figure 9 The fitted values are compared with the storage values in 1996 as shown in Table 1. It can be found from the figure and the table that the newly fitted curve is slightly lower than that in 1996, which is related to the siltation of the Miyun Reservoir and conforms to the actual situation, so it is considered that the revised DEM data conforms to the actual change trend.

[0090] Table 1 Comparison of fitted storage values with storage values in 1996

[0091]

[0092]

[0093]

[0094] By adopting the technical scheme disclosed in the present application, the following beneficial effects are obtained:

[0095] The present application provides a reservoir storage curve checking method based on multi-source data fusion. The method effectively eliminates data blind spots and local contradictions through multi-source DEM adaptive weighted fusion and conflict area refinement (such as CAD drawings and historical image verification), and improves the terrain modeling accuracy by more than 50%. The error of the reservoir storage calculation after fusion can be controlled within 3%. The method uses real-time water level-flow data to complete dynamic updating of the reservoir storage curve within 30 days through water balance inversion and piecewise linear correction model, and reduces the operation and maintenance cost by 90%. The method can adapt to small and medium-sized reservoirs (coverage rate increased by 40%) through multi-source data fusion and local correction mechanism, and is compatible with the ArcGIS tool chain without the need for custom development. In practical application, the reservoir storage error of ≤3% ensures the safety of flood control, reduces the frequency of manual inspection, and has safety and economy, and has significant promotion value. The method can make full use of historical data and accurately correct the reservoir storage curve according to monitoring information, solve the problems of single data, static model error and insufficient verification in the traditional method, and realize high-precision and generalizable reservoir storage curve generation.

[0096] The above only describes the preferred embodiments of the present application. It should be noted that for those skilled in the art, without departing from the principles of the present application, several improvements and refinements can be made, and these improvements and refinements should be considered as the protection scope of the present application.

Claims

1. A reservoir capacity curve calibration method based on multi-source data fusion, characterized by: The following steps are included: S1. DEM data preprocessing: coordinate correction, resolution matching and noise reduction are performed on at least two sets of DEM data of different periods or sources of the reservoir to enhance data quality; multiple sets of DEM data with enhanced data quality are fused using conflict area refinement processing combined with adaptive weighted fusion algorithm to obtain a set of accurate new DEM data; S2. Acquisition of storage capacity curve: Calculate the storage capacity based on the new DEM data to generate a preliminary storage capacity curve; determine the area to be corrected for the preliminary storage capacity curve based on the comparison results with the historical storage capacity curve, and generate a theoretical storage capacity curve by optimizing the DEM area corresponding to the area to be corrected; Step S2 specifically includes the following contents: S21. Calculate the reservoir capacity using the new DEM data and conduct a theoretical analysis of the reservoir capacity curve based on historical reservoir capacity curves and reservoir sedimentation conditions to generate a preliminary reservoir capacity curve. S22. Compare the preliminary storage capacity curve with the historical storage capacity curve, determine the area to be corrected in the preliminary storage capacity curve based on the relative error between the water levels of the two, reconstruct the terrain surface of the DEM area corresponding to the area to be corrected, so as to minimize the storage capacity deviation between the preliminary storage capacity curve and the historical storage capacity curve, and generate a theoretical storage capacity curve; S3. Reservoir capacity curve correction: Construct the storage capacity and water level relationship of the theoretical reservoir capacity curve, establish correction equations for each water level sub-interval, and use measured scattered data to fit the correction parameters in the correction equation to obtain an accurate reservoir capacity curve.

2. The reservoir capacity curve verification method based on multi-source data fusion according to claim 1 is characterized in that: Step S1 specifically includes the following contents: S11. Unify all DEM data sets into the CGCS2000 coordinate system to ensure consistent planar datums. Then, match the resolutions of all DEM data sets, perform bilinear interpolation on low-resolution DEM data, and avoid resampling high-frequency noise for high-resolution DEM data. Finally, perform denoising on all DEM data sets, calculate the local elevation standard deviation, and remove abnormal pixels whose local elevation standard deviation exceeds the preset standard deviation. At the same time, mask the reservoir boundary to eliminate interference from the surrounding terrain. S12. For conflicting areas in each set of DEM data, if the differences between the DEM data sets are too large, dynamic matching is required using CAD drawings or remote sensing satellite images. S13. Remove the noise part in each set of DEM data, retain the accurate part in each set of DEM data, and use the adaptive weighted fusion algorithm to fuse these DEM data, so as to cut out a set of accurate new DEM data.

3. The reservoir capacity curve verification method based on multi-source data fusion according to claim 2 is characterized in that: The adaptive weighted fusion algorithm dynamically assigns weights to each set of DEM data according to its vertical accuracy, acquisition time and terrain complexity, so as to achieve the fusion of multiple sets of DEM data.

4. The reservoir capacity curve calibration method based on multi-source data fusion according to claim 1 is characterized in that: Specifically, step S21 divides the new DEM data into elevation bands, takes the dead water level and the verified flood level as the two extreme values ​​according to the variation of the reservoir water level, performs spatial discrete interpolation according to the preset interpolation distance, and calculates the corresponding reservoir capacity of each elevation band layer by layer to generate a preliminary discrete point diagram of the reservoir capacity curve.

5. The reservoir capacity curve calibration method based on multi-source data fusion according to claim 1 is characterized in that: Step S22 specifically includes the following contents: S221, aligning the preliminary storage capacity curve with the historical storage capacity curve according to the water level, calculating the relative error, and marking the points in the preliminary storage capacity curve with relative errors greater than a first preset error as areas to be corrected; S222. After the area to be corrected is marked, the DEM area corresponding to the area to be corrected is optimized, and the topographic surface of the DEM area corresponding to the area to be corrected is reconstructed using the moving least squares method to minimize the storage capacity deviation between the preliminary storage capacity curve and the historical storage capacity curve; S223. Recalculate the storage capacity using the locally optimized DEM data until the error between the preliminary storage capacity curve and the historical storage capacity curve is less than or equal to the second preset error and the trend of the preliminary storage capacity curve is consistent with the trend of the historical storage capacity curve. If they are inconsistent, repeat steps S221-S223 until the preliminary storage capacity curve meets the requirements and generates a theoretical storage capacity curve.

6. The reservoir capacity curve calibration method based on multi-source data fusion according to claim 5 is characterized in that: The second preset error is smaller than the first preset error.

7. The reservoir capacity curve calibration method based on multi-source data fusion according to claim 1 is characterized in that: Step S3 specifically includes the following contents: S31. Based on the water balance inversion, calculate the change in reservoir capacity within a certain range, select at least 30 days of continuous water level time series data and corresponding inflow and outflow, and integrate the water balance equation according to the time step; S32. Using a segmented correction method, construct a relationship between storage capacity and water level, divide the water level range into sub-intervals with preset intervals, and establish a correction equation for each sub-interval; S33. Using the measured scattered data, the correction parameters in the correction equation are fitted by weighted least squares method to achieve refined dynamic correction of the storage capacity within the range and obtain an accurate storage capacity curve.

8. The reservoir capacity curve calibration method based on multi-source data fusion according to claim 7 is characterized in that: Step S3 also includes, S34. Select 10% of the measured scattered data as a validation set and calculate the root mean square error of the precise storage capacity curve to check the accuracy of the correction result; If the RMS error is greater than the preset RMS error, the correction parameters need to be readjusted according to the correction equation until the RMS error is less than or equal to the preset RMS error.

9. The reservoir capacity curve calibration method based on multi-source data fusion according to claim 7, characterized in that: The DEM data must include the reservoir capacity and the surrounding catchment area.

Images