River flood capacity intelligent analysis method and system
By constructing an intelligent response model and a grid search strategy, the flood discharge capacity of river channels is analyzed, which solves the problems of high computational complexity and insufficient prediction ability in existing technologies. It realizes accurate analysis of flood discharge capacity of river channels and scientific prediction of influencing factors, supporting engineering planning and scheduling decisions.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- BUREAU OF HYDROLOGY CHANGJIANG WATER RESOURCES COMMISSION
- Filing Date
- 2026-03-10
- Publication Date
- 2026-06-12
Smart Images
Figure CN122196475A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of hydrology and water resources, and more specifically, to a method and system for intelligent analysis of river flood discharge capacity. Background Technology
[0002] River flood carrying capacity is a crucial indicator for ensuring flood control safety in a river basin, often quantified as the maximum safe flow rate under a guaranteed water level. This definition implicitly assumes a one-to-one correspondence between water level and flow rate. However, influenced by flood fluctuations, the water level-flow rate relationship at a river cross-section exhibits a counter-clockwise loop curve, with significant differences in flow rate at the same water level during the rising and falling flood phases. Therefore, this invention proposes the following definition of river flood carrying capacity: the critical flood process through which a river can safely pass under guaranteed water level constraints. The critical flood process refers to a flood during which the highest water level at the control section precisely reaches the guaranteed water level. This flood process is characterized by features such as flood magnitude, flow temporal distribution, and flood peak morphology.
[0003] In addition, it is constantly evolving due to factors such as backwater (such as changes in water levels in downstream reservoirs, lakes, and oceans), riverbed scouring and sedimentation (manifested as the evolution of hydraulic elements such as cross-sectional area, roughness, and water surface gradient), and human activities (such as the construction of water-blocking structures like wharves and bridges, as well as control hubs like pumping stations and reservoirs, and river dredging and other management measures).
[0004] Therefore, accurately analyzing the flood discharge capacity of rivers and its evolution, and quantitatively identifying the mechanisms and contributions of various influencing factors, are core issues and technical prerequisites for the precise implementation of flood control scheduling and river management measures. Existing research largely relies on hydrodynamic principles to construct models, using numerical simulations to calculate the flood discharge capacity of rivers under given boundary conditions, and drawing flood discharge threshold maps based on multi-scenario simulation results to provide a basis for flood control scheduling decisions. However, this method has high computational complexity, requires significant computing resources, and is limited by static boundary conditions; once these conditions change, the model needs to be recalibrated. Furthermore, this method lacks a systematic exploration of the historical evolution of river flood discharge capacity and does not possess the ability to scientifically predict changes in flood discharge capacity. Against this backdrop, there is an urgent need for an intelligent method for analyzing river flood discharge capacity that can integrate historical evolution data, quantitatively analyze multiple influencing factors, and balance accuracy and efficiency, in order to address the shortcomings of existing technologies. Summary of the Invention
[0005] To address the aforementioned problems, the present invention aims to provide an intelligent analysis technology for river flood discharge capacity, which is intended to achieve intelligent analysis of river flood discharge capacity, systematic identification of influencing factors, and scientific prediction of changes, thereby providing technical support for engineering planning and scheduling decisions.
[0006] To achieve the above technical objectives, this application provides an intelligent method for analyzing the flood discharge capacity of river channels, comprising the following steps: The flood discharge capacity of river channels and its influencing factors are quantitatively analyzed and the indicators are quantified. By introducing the design flood, a water level and discharge data sample based on the hydrodynamic model is constructed. A smart response model of "flood process - highest water level" is constructed, and expanded water level and flow data samples are added during model training; and the flood discharge capacity of the river is analyzed by solving the critical flood characteristic combination in which the highest water level just reaches the guaranteed water level.
[0007] Preferably, when quantifying the indicators, a long series of basic data of the target river channel is collected, including water level and flow data, measured cross-sectional data, and human activity data, and a quantifiable indicator system is constructed; the data is preprocessed, missing data is imputed, and abnormal samples are removed to form a basic database.
[0008] Preferably, when acquiring expanded water level and flow rate data samples, a hydrodynamic model is constructed based on the river channel conditions of the year in which the flood occurred. The model parameters are calibrated using the measured flood process of that year. The calculation results are compared with the measured data to verify the model and determine the model parameters for that year. The generated design flood process is input into the hydrodynamic model to simulate the water level process at the control section. Based on this, flood characteristic indicators corresponding to each design flood process are extracted to expand the water level and flow rate data sample for each process.
[0009] Preferably, when constructing the intelligent response model, flood characteristics, river channel conditions, and human activities are used as model inputs, and the highest water level of the control section is used as the model output. An interpretable machine learning method is used to calculate the characteristic importance of each input factor to the change of the highest water level of the control section, and key influencing factors are screened out. The model input is then changed to the key influencing factors, and the model is trained.
[0010] Preferably, when obtaining key influencing factors, the key influencing factors include the cross-sectional area of the water passage, the peak flow rate, the time taken for the flow rate to reach the peak flow rate, the maximum 7-day flood volume, the initial water level, the initial flow rate, and the hydraulic radius.
[0011] Preferably, when training the model, the intelligent response model is trained using the multiple linear regression method, and the model parameters are optimized using leave-one-out cross-validation.
[0012] Preferably, when analyzing the flood discharge capacity of a river channel, a grid search strategy is adopted to solve for the critical flood characteristic combination where the highest water level just reaches the guaranteed water level, thereby analyzing the flood discharge capacity of the river channel.
[0013] Based on the same inventive concept, this invention discloses an intelligent analysis system for river flood discharge capacity, comprising: The data sample construction module is used to quantitatively analyze the flood discharge capacity of the river channel and its influencing factors, and to quantify the indicators. By introducing the design flood, it constructs a water level and flow rate data sample based on the hydrodynamic model. The analysis module is used to analyze the flood discharge capacity of the river by solving the critical flood characteristic combination that just reaches the guaranteed water level, based on the constructed "flood process-highest water level" intelligent response model. During model training, expanded water level and flow rate data samples are added.
[0014] The present invention discloses the following technical effects: This invention fully considers the dynamic evolution of river flood discharge capacity under the influence of flood processes, cross-sectional changes, and human activities. Through in-depth mining of multi-source data, it achieves intelligent analysis of river flood discharge capacity, systematic identification of influencing factors, and scientific prediction of changes, providing technical support for engineering planning and scheduling decisions. Attached Figure Description
[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0016] Figure 1 This is a flowchart of the intelligent analysis method for river flood discharge capacity described in this invention; Figure 2 This is a schematic diagram of the hydrodynamic model simulation described in this invention; Figure 3 This is a schematic diagram illustrating the screening of key influencing factors as described in this invention; Figure 4 This is a schematic diagram illustrating the prediction effect of the intelligent response model described in this invention; Figure 5 This is a schematic diagram illustrating the flood discharge capacity described in this invention; Figure 6 This is a schematic diagram of the critical flood process line described in this invention. Detailed Implementation
[0017] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. The components of the embodiments of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely represents selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.
[0018] like Figures 1-6 As shown, this invention provides an intelligent method for analyzing the flood discharge capacity of river channels, comprising the following steps: Step S1: Data Collection and Indicator Quantification. To quantitatively analyze the flood control capacity of a river and its influencing factors, it is necessary to collect a long series of basic data on the target river, including water level and flow data, measured cross-sectional data, and human activity data, and to construct a quantifiable indicator system; the data is preprocessed, missing data is imputed, outlier samples are removed, and a basic database is formed.
[0019] Step S2: Introduce the design flood and expand the water level and flow rate data sample based on the hydrodynamic model.
[0020] Step S3: Construct a smart response model for the "flood process - highest water level". The model inputs include flood characteristics, river channel conditions, and human activities, and the model output is the highest water level at the control section. Interpretable machine learning methods are used to calculate the characteristic importance of each input factor to the change in the highest water level at the control section, and key influencing factors are screened out. The key influencing factors are used as model inputs, and methods such as multiple linear regression are used to train the model.
[0021] Step S4: Based on the intelligent response model, a grid search strategy is adopted to solve the critical flood characteristic combination where the highest water level just reaches the guaranteed water level, and to analyze the flood discharge capacity of the river channel.
[0022] Step S1 includes: 1) Hydrological conditions: This type of data is used to characterize the hydrological conditions and morphological evolution of river control sections, mainly including water level and discharge data from flood events and measured cross-sectional data, among which: Water level and flow rate data: Flood data from hydrological stations are collected and processed to obtain flow rate and water level hydrographs for different years and flood events at the control sections. If the backwater effect is significant, water level hydrographs for the corresponding time periods at downstream control stations also need to be collected. Based on this, characteristic indicators for each flood event are extracted, including but not limited to the initial water level (…). ), initial flow ( ), peak flow ( The time it takes for the flow rate to reach the peak flow rate ( ), highest water level ( Maximum 1-day, 3-day, and 7-day flood volumes ( , , )wait.
[0023] Measured cross-sectional data: Historical measured cross-sectional data from hydrological stations were collected to obtain measured cross-sectional data (distance from the starting point and riverbed elevation sequence) for different measurement dates. Based on this, and combined with the guaranteed water level of the control station, the cross-sectional area (A) and hydraulic radius (R) at that water level were extracted.
[0024] 2) Human activities: This type of data is used to characterize the alteration of river hydraulic conditions by human activities, and mainly includes three categories: water-blocking structures, control hubs, and river management measures. Water-blocking structures: Bridges, wharves, and other river-crossing projects compress the flow cross-section, creating a localized water-blocking effect and causing water levels to rise. Therefore, it is necessary to collect and compile the project's construction time and its design parameters, such as the water-blocking ratio (…). ).
[0025] Regulating and storage hubs: Reservoirs, dams, pumping stations, and other engineering projects alter flood processes by regulating or pumping water, thus adjusting the temporal distribution of water levels and flow along the river. Their regulatory capacity can be quantified as the total width of the gate opening (…). ), Design discharge capacity ( ), and the design flow rate of the pumping station ( )wait.
[0026] River management measures: Dredging and silt removal measures can improve the flow pattern of the river, reduce the roughness (n) of the river, and enhance the flood discharge capacity of the river.
[0027] After completing the above data collection and indicator quantification, all data are preprocessed, missing data are imputed, and outlier samples are removed to form a standardized basic database.
[0028] Step S2 includes: Considering the sparsity of flood data, for years with fewer measured flood events, design floods can be introduced to expand the data sample. The specific method is as follows: Based on the design flood results of the hydrological station, design flood process lines at different frequencies are generated using the same frequency or same ratio amplification method; a hydrodynamic model is constructed according to the river channel conditions of that year, and the roughness and other parameters of the model are calibrated using the measured flood process of that year; the generated design flood process is input into the hydrodynamic model to simulate the water level process of the control section, thereby expanding the water level and flow data.
[0029] Step S3 includes: A smart response model is constructed, taking flood characteristics (water level and flow data), river channel conditions (measured cross-sectional data), and human activities as inputs, and the highest water level at the control section as output, establishing the following mapping relationship: In the formula, i represents the year in which the flood occurred; j represents the flood event. , , The characteristics of the flood, the state of the river channel, and human activities are determined respectively, and the highest water level of the j-th flood in year i is determined by mapping F together. .
[0030] Based on the above mapping relationship, interpretable machine learning methods such as gradient boosting trees (XGBoost / LightGBM) are used to evaluate the feature importance of each input factor to the change of the highest water level at the control section by calculating the average information gain of the features. After normalization, the factors are ranked, and factors with importance below a preset threshold are removed to select key influencing factors to simplify the intelligent response model. ; ; In the formula, Let K be the total gain of feature j across all decision trees, and K be the total number of decision trees. Let be the sum of the gains from splitting feature j in the k-th tree. Let j be the set of nodes used for splitting in the k-th tree. The information gain gained when splitting node n in the k-th tree using feature j. for After normalization, p represents the total number of features.
[0031] The input factors are changed to the selected key influencing factors, and an intelligent response model is constructed based on the mapping relationship. Multiple linear regression or ensemble learning algorithms with strong fitting ability to nonlinear relationships, such as random forest, are preferred. The model loss function uses the root mean square error (RMSE) of the predicted and actual water levels, and the coefficient of determination is used. Evaluate the accuracy of the regression model.
[0032] Step S4 includes: Based on a trained intelligent response model, and given boundary conditions such as guaranteed water level, initial water level, initial flow rate, river channel conditions, and human activities, the problem aims to solve for the critical flood characteristic combination where the highest water level just reaches the guaranteed water level. The problem can be described as follows: ; ; ; In the formula, The guaranteed water level at the control section; X is the critical flood characteristic combination to be solved; C is the given boundary condition; The feasible region is characterized by flood features.
[0033] To solve the above problems, this invention adopts a grid search strategy to discretize flood characteristics within the feasible region to form a multi-dimensional parameter network; inputs the feature combination corresponding to all grid points into the intelligent response model to calculate the corresponding maximum water level, and solves the flood feature combination that satisfies the maximum water level as the guaranteed water level.
[0034] Based on this, and according to the typical flood hydrograph of the cross section, a selection is made from the critical flood characteristic combination. The combination of flood peak and flood volume is consistent with typical processes. The typical process line is scaled up using the same frequency amplification method to deduce the critical flood process that allows the river channel to pass safely under water level constraints, thereby achieving an intuitive representation of the river channel's flood discharge capacity and providing technical support for engineering planning and scheduling decisions.
[0035] Example: This invention discloses an intelligent analysis method for river flood discharge capacity, the process of which is as follows: Figure 1 As shown, it includes the following steps: Step S1: Data Collection and Indicator Quantification. To quantitatively analyze the flood control capacity of a river and its influencing factors, it is necessary to collect a long series of basic data on the target river, including water level and flow data, measured cross-sectional data, and human activity data, and to construct a quantifiable indicator system; the data is preprocessed, missing data is imputed, outlier samples are removed, and a basic database is formed.
[0036] Step S2: Introduce the design flood and expand the water level and flow rate data sample based on the hydrodynamic model.
[0037] Step S3: Construct a "flood process - highest water level" intelligent response model. The input of the model is flood characteristics, river state, and human activities, and the output of the model is the highest water level of the control section. Use interpretable machine learning methods to calculate the characteristic importance of each input factor to the change of the highest water level of the control section, and screen out key influencing factors. Change the model input to the key influencing factors and train the model.
[0038] Step S4: Based on the intelligent response model, a grid search strategy is adopted to solve the critical flood characteristic combination where the highest water level just reaches the guaranteed water level, and to analyze the flood discharge capacity of the river channel.
[0039] In step S1, historical flood data from hydrological stations from 1980 to 2023 are collected, and peak flow rates are screened. The data from eight flood events were compiled, yielding flow and water level profiles. For each flood event, flood characteristics were extracted, including the initial water level (…). ), initial flow ( ), peak flow ( The time it takes for the flow rate to reach the peak flow rate ( ), highest water level ( ), maximum 1d / 3d / 7d flood volume ( , , This forms a sample of water level and flow rate data for each event.
[0040] Measured cross-sectional data for each year of flood event were collected, and the starting distance of the control sections and the measured sequence of riverbed elevations for each year were compiled. Using a guaranteed water level of 48m as a benchmark, the cross-sectional area (A) and hydraulic radius (R) below this guaranteed water level were calculated, forming a river channel state characteristic dataset that corresponds one-to-one with each flood event. Table 1 shows the cross-sectional area and hydraulic radius of the control sections at the guaranteed water level of 48m.
[0041] Table 1. Cross-sectional area and hydraulic radius of the control section to ensure a water level of 48m. In this embodiment, the control section is minimally affected by human activities, and there are no water-blocking structures or water storage hubs nearby.
[0042] After completing the above data collection and indicator quantification, all data are preprocessed, missing data are imputed, and outlier samples are removed to form a standardized basic database.
[0043] In step S2, a hydrodynamic model is constructed based on the river channel conditions of the year in which the flood occurred. The model's roughness and other parameters are calibrated using the measured flood process of that year. The calculated results are compared with the measured data to verify the model and determine the model parameters for that year.
[0044] The generated design flood process is input into the hydrodynamic model to simulate the water level process at the control section, such as... Figure 2 As shown, flood characteristic indicators corresponding to each design flood process are extracted to expand the data sample of water level and flow rate for each event.
[0045] In step S3, an intelligent response model is constructed, taking flood characteristics (water level and discharge data) and river channel conditions (measured cross-sectional data) as inputs and the highest water level at the control section as output, and establishing the following mapping relationship: ; In the formula, i represents the year in which the flood occurred; j represents the flood event. , The characteristics of the flood and the state of the river channel are determined respectively, and the highest water level of the j-th flood in year i is determined by mapping F. .
[0046] Based on the basic database formed in step S1, feature importance is calculated using LightGBM: ; ; In the formula, Let K be the total gain of feature j across all decision trees, and K be the total number of decision trees. Let be the sum of the gains from splitting feature j in the k-th tree. Let j be the set of nodes used for splitting in the k-th tree. The information gain gained when splitting node n in the k-th tree using feature j. for After normalization, p represents the total number of features.
[0047] After normalization, the feature importance is obtained as follows: Figure 3 As shown. An importance threshold of 0.05 was set, and factors with importance below the threshold were removed, ultimately resulting in seven key influencing factors, ranked from highest to lowest importance as follows: cross-sectional area (A), peak flow rate (A), and peak discharge rate (B). The time it takes for the flow rate to reach the peak flow rate ( ), maximum 7-day flood volume ( ), initial water level ( ), initial flow ( ), hydraulic radius (R).
[0048] Using the selected key influencing factors as input and the highest water level at the cross-section as output, a smart response model was trained using a multiple linear regression method that performs well in small sample scenarios. The model parameters were then optimized using leave-one-out cross-validation, resulting in the following response relationship: Model prediction results are as follows Figure 4 As shown, the root mean square error (RMSE) is 0.543, and the coefficient of determination is... =0.773, indicating that the constructed intelligent response model can accurately fit the response relationship of "flood process - highest water level", meeting the accuracy requirements of subsequent intelligent analysis of flood discharge capacity.
[0049] In step S4, based on the trained intelligent response model, the river's flood discharge capacity (ensuring a water level of 48m) is analyzed. The river's condition is determined based on the latest cross-sectional measurement data, including the cross-sectional area of the water passage. Hydraulic radius R = 5.31m; initial flow rate set. initial water level After setting the above boundary conditions, the problem of solving for the critical flood characteristic combination where the highest water level just reaches the guaranteed water level can be described as follows: ; ; ; In the formula, The guaranteed water level at the control section; X is the critical flood characteristic combination to be solved; C is the given boundary condition; The feasible region is characterized by flood features.
[0050] To solve the above problem, this invention employs a grid search strategy to discretize flood characteristics within the feasible region, forming a multi-dimensional parameter network. The feature combinations corresponding to all grid points are input into the intelligent response model to calculate the corresponding maximum water level. The invention then solves for the flood feature combinations that satisfy the condition that the maximum water level is the guaranteed water level (error less than 0.01m), such as... Figure 5 As shown, this enables the scientific prediction of the flood discharge capacity of a river channel after changes in boundary conditions.
[0051] Based on this, and according to the typical flood hydrograph of the cross section, a selection is made from the critical flood characteristic combination. The peak and volume combinations are consistent with typical processes. The typical process curve is scaled using the same frequency amplification method to deduce the critical flood process that ensures safe passage of the river under water level constraints. The critical flood process obtained by scaling one of the peak and volume combinations is as follows: Figure 6 As shown in the figure. This allows for a direct representation of the flood discharge capacity of a river channel, providing technical support for engineering planning and scheduling decisions.
[0052] This invention discloses an intelligent method for analyzing river flood discharge capacity. It constructs a quantifiable index system by collecting flood data from various events and river cross-sections; introduces design floods and expands the water level-discharge data sample based on a hydrodynamic model; builds an intelligent response model for the flood process and maximum water level, screens key influencing factors, and optimizes the training model; and analyzes river flood discharge capacity using a grid search strategy based on the intelligent response model. This invention fully considers the dynamic evolution of river flood discharge capacity under the influence of flood processes, cross-sectional changes, and human activities. Through in-depth mining of multi-source data, it achieves intelligent analysis of river flood discharge capacity, systematic identification of influencing factors, and scientific prediction of changes, providing technical support for engineering planning and scheduling decisions.
[0053] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.
[0054] In the description of this invention, it should be understood that the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.
[0055] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.
Claims
1. A method for intelligent analysis of river flood discharge capacity, characterized in that, Includes the following steps: The flood discharge capacity of river channels and its influencing factors are quantitatively analyzed and the indicators are quantified. By introducing the design flood, a water level and discharge data sample based on the hydrodynamic model is constructed. A smart response model for "flood process - highest water level" is constructed, and expanded water level and flow rate data samples are added during model training. The flood discharge capacity of the river is analyzed by solving the critical flood characteristic combination where the highest water level just reaches the guaranteed water level.
2. The intelligent analysis method for river flood discharge capacity according to claim 1, characterized in that: When quantifying indicators, a long series of basic data on the target river channel are collected, including water level and flow data, measured cross-sectional data, and human activity data, and a quantifiable indicator system is constructed. The data is then preprocessed, missing data is imputed, and outlier samples are removed to form a basic database.
3. The intelligent analysis method for river flood discharge capacity according to claim 2, characterized in that: When acquiring expanded water level and flow rate data samples, a hydrodynamic model is constructed based on the river channel conditions in the year of the flood. The model parameters are calibrated using the measured flood process of that year. The calculation results are compared with the measured data to verify the model and determine the model parameters for that year. The generated design flood process is input into the hydrodynamic model to simulate the water level process at the control section. Based on this, flood characteristic indicators corresponding to each design flood process are extracted to expand the water level and flow rate data sample for each process.
4. The intelligent analysis method for river flood discharge capacity according to claim 3, characterized in that: When constructing the intelligent response model, flood characteristics, river conditions, and human activities are used as model inputs, and the highest water level at the control section is used as the model output. Interpretable machine learning methods are used to calculate the characteristic importance of each input factor to the change of the highest water level at the control section, and key influencing factors are screened out. The model input is then changed to the key influencing factors, and the model is trained.
5. The intelligent analysis method for river flood discharge capacity according to claim 4, characterized in that: When obtaining key influencing factors, the key influencing factors include cross-sectional area of water passage, peak flow, time taken for flow to reach peak flow, maximum 7-day flood volume, initial water level, initial flow, and hydraulic radius.
6. The intelligent analysis method for river flood discharge capacity according to claim 5, characterized in that: When training the model, the intelligent response model is trained using the multiple linear regression method, and the model parameters are optimized using leave-one-out cross-validation.
7. The intelligent analysis method for river flood discharge capacity according to claim 6, characterized in that: When analyzing the flood discharge capacity of a river channel, a grid search strategy is adopted to solve for the critical flood characteristic combination where the highest water level just reaches the guaranteed water level, thereby analyzing the flood discharge capacity of the river channel.
8. A smart analysis system for river flood discharge capacity, used to implement the smart analysis method for river flood discharge capacity as described in claim 1, characterized in that, include: The data sample construction module is used to quantitatively analyze the flood discharge capacity of the river channel and its influencing factors, and to quantify the indicators. By introducing the design flood, it constructs a water level and flow rate data sample based on the hydrodynamic model. The analysis module is used to analyze the flood discharge capacity of the river by solving the critical flood characteristic combination that just reaches the guaranteed water level based on the constructed "flood process-highest water level" intelligent response model. During model training, expanded water level and flow rate data samples are added.