A method for predicting water head increment under river channel expansion condition

Abstract

The present application relates to the technical field of water head increment prediction, and particularly relates to a water head increment prediction method under river channel expansion conditions, comprising the following steps: establishing a benchmark hydraulic state set based on multi-source observation data and historical working condition data; converting the cross section and topographic changes formed by the expansion project into an expansion form description set, and generating a form-hydraulic increment table under the reference of the benchmark hydraulic state set, wherein the form-hydraulic increment table comprises a local energy loss increment map, a channel connectivity adjustment map and a key structure influence area; taking the benchmark hydraulic state set and the form-hydraulic increment table as inputs, combining a target boundary scenario, generating a water head increment prediction grid and a key cross section water head increment curve, and outputting a prediction reliability map and a sensitive element ranking. Through the three-step construction of the benchmark hydraulic state set, the expansion form-hydraulic increment table and the coupling prediction model, the present application breaks through the process from multi-source data fusion, model parameter calibration and connected topology identification.

Description

Technical Field

[0001] This invention relates to the field of water head increment prediction technology, and in particular to a method for predicting water head increment under the condition of river channel widening. Background Technology

[0002] River widening, as a common water conservancy and urban infrastructure engineering measure, is applied to various engineering purposes such as improving flood control capacity, improving navigation conditions, and expanding aquatic ecological space. Widening activities typically involve large-scale topographic reshaping, such as cross-section widening, bottom elevation adjustment, and slope reconstruction. These changes will disturb the original hydrodynamic structure and affect key hydraulic indicators such as mainstream river velocity, water surface morphology, and the connectivity of retention systems, thereby affecting scheduling operations, levee safety, and the stability of structures.

[0003] However, existing technologies for predicting head response under river widening conditions still have the following shortcomings: First, they lack a unified data-driven analysis framework. Most current prediction methods rely on single hydrological simulations or two-dimensional hydrodynamic models, lacking effective integration of multi-source heterogeneous data from historical hydrological conditions, remote sensing topography, and widening design drawings. This makes it impossible to comprehensively construct a representative baseline hydraulic state system, resulting in weak spatial adaptability and temporal extensibility of the prediction results. Second, the hydraulic response to widening disturbances lacks structured expression. The resulting river morphology changes are complex, with phenomena such as sudden changes in local flow velocity, increased energy loss, and obstructed connectivity of storage units often occurring simultaneously. Existing methods struggle to independently quantify these local disturbances, making it difficult to accurately capture the impact characteristics on key structures or sensitive areas. Third, the model results lack the ability to express uncertainty and interpret risk. Most existing models use deterministic input parameters for single simulations, ignoring the uncertainties in key elements such as river parameters, boundary conditions, and widening morphology. This makes it impossible to assess the confidence level of the prediction results or identify the most sensitive controlling factors, hindering risk management in engineering design and control strategies. Summary of the Invention

[0004] This invention provides a method for predicting head increments under river widening conditions. This method integrates multi-source data, possesses the ability to express widening response mechanisms, and can perform uncertainty assessment and sensitivity ranking to support more scientific and refined widening scheme comparison and subsequent scheduling and operation design.

[0005] A method for predicting head increment under river widening conditions includes the following steps: S1: Based on multi-source observation data and historical operating condition data, establish a benchmark hydraulic state set, which includes benchmark head distribution, roughness partitioning and boundary condition time series, channel-storage unit connectivity relationship and key section water conveyance-storage response curves. S2: The cross-section and topographic changes formed by the widening project are transformed into a widening morphology description set, and a morphology-hydraulic increment table is generated under the reference of the baseline hydraulic state set. The morphology-hydraulic increment table includes a local energy loss increment map, a channel connectivity adjustment map, and a key structure influence zoning. The morphology-hydraulic increment table is used to drive head response inference. S3: Using the reference hydraulic state set and the morphological-hydraulic increment table as input, and combining the target boundary scenario, generate a head increment prediction grid and a head increment curve for key sections, and output a prediction confidence map and a ranking of sensitive elements; the head increment prediction grid and the head increment curve for key sections serve as the direct basis for engineering comparison and verification.

[0006] Optionally, the multi-source observation data includes river topographic point clouds, historical hydrological sequences, remote sensing images, and field survey records; the historical operating condition data includes historical flood events, typical scheduling processes, and corresponding monitoring head data.

[0007] Optionally, S1 specifically includes: S11: Establish a baseline head distribution by assimilating multi-source observation data with historical operating condition data; S12: Driven by the reference head distribution and the historical hydrological sequence, the Manning roughness of the river channel and storage unit is calibrated through hydrodynamic model inversion and system identification methods, and roughness partitioning is performed according to the underlying surface properties and calibration results. At the same time, the time series of boundary conditions required by the model is determined. S13: Based on the river topography and remote sensing images, identify potential water storage units, and use GIS spatial analysis technology to determine the connection position, elevation threshold and initial hydraulic width between the river and each water storage unit, and establish the river-water storage unit connectivity relationship. S14: Based on the historical hydrological sequence and the corresponding monitoring head data, select key control sections, and derive the cross-sectional water conveyance-storage response curves of the key control sections through statistical analysis or hydrodynamic model simulation.

[0008] Optionally, S1 further includes constructing the benchmark hydraulic state set by combining the benchmark head distribution, roughness partitioning and boundary condition time series, channel-storage unit connectivity and key section water conveyance-storage response curves, and performing uncertainty analysis and verification on it using historical operating condition data that were not involved in the construction.

[0009] Optionally, in S2, based on the design drawings, BIM model or measured topographic data of the excavation project, the geometric parameters of the river cross section, changes in riverbed elevation, slope gradient and revetment type after the project is implemented are extracted to form a digital excavation morphology description set.

[0010] Optionally, the generation of the morphology-hydraulic increment table specifically includes: S21: Under the reference of the above-mentioned hydraulic state set, compare the degree of abrupt change and geometric shape of the river channel cross section before and after the widening, and quantify the additional head loss caused by the cross section expansion, contraction and flow direction change based on hydraulic empirical formulas or local three-dimensional flow field simulation, and generate a spatially distributed local energy loss increment map. S22: Based on the channel-storage unit connectivity relationship in the above-mentioned widening morphology description set and the benchmark hydraulic state set, recalculate the connection elevation and hydraulic width between the widened channel and the storage unit, quantify the change in connectivity through a hydraulic coupling model or empirical function, and generate a channel connectivity adjustment map. S23: Identify existing key structures within the scope of the excavation project's influence, and analyze their foundation elevation, flow regime, and relative spatial relationship with the excavation project; through hydrodynamic simulation or engineering experience, delineate areas where hydraulic conditions have significantly changed due to riverbed incision and flow velocity variations, and form a key structure influence zoning.

[0011] Optionally, the generation of the morphological-hydraulic increment table may also include integrating the local energy loss increment map, the channel connectivity adjustment map, and the key structure influence zoning to form the morphological-hydraulic increment table.

[0012] Optionally, S3 specifically includes: S31: Couple the hydrodynamic model of the reference hydraulic state set with the local energy loss increment map and channel connectivity adjustment map in the morphology-hydraulic increment table to construct a coupled model for head increment prediction. S32: Define one or more target boundary scenarios representing future hydrological conditions and engineering scheduling rules; input each scenario into the coupled model, calculate the hydraulic state before and after the excavation, and obtain the difference to get the head increment prediction grid. S33: Based on the head increment prediction grid, for the key sections defined in the reference hydraulic state, extract the head increment under different target boundary scenarios, including different flow rates, and generate the key section head increment curve with flow rate or time as the horizontal axis and head increment as the vertical axis.

[0013] Optionally, S3 may also include uncertainty analysis and sensitivity ranking.

[0014] Optionally, the uncertainty analysis and sensitive element ranking specifically include: Identify and quantify the uncertainty range of key input parameters in the baseline hydraulic state set and the morphological-hydraulic increment table; Based on the Monte Carlo simulation sampling method, a large number of random samples and model calculations are performed within the above uncertainty range to generate the probability distribution of the water head increment, and a prediction confidence map representing the spatial confidence of the prediction results is drawn accordingly. The sensitivity analysis method of analysis of variance is used to calculate the degree of influence of each input parameter on the water head increment result and output the ranking of sensitive elements.

[0015] The beneficial effects of this invention are: This invention establishes a complete process from multi-source data fusion, model parameter calibration, and connected topology identification to coupled simulation and uncertainty analysis through a three-step construction of a benchmark hydraulic state set, an excavation morphology-hydraulic increment table, and a coupled prediction model. This avoids the problems of relying on experience-based judgment and fragmented results in traditional methods.

[0016] This invention proposes a method for constructing a morphological-hydraulic increment table, which quantitatively expresses cross-sectional expansion, changes in channel connectivity, and the impact of key structures in a structured manner, and incorporates them as intermediate inputs into the coupled simulation. This structure enables the systematic capture and quantitative modeling of microscale hydraulic disturbances such as the impact of excavation on local energy loss, changes in the connectivity of storage units, and the reconfiguration of flow paths.

[0017] This invention incorporates Monte Carlo simulation and sensitivity analysis methods such as Sobol / Morris into the prediction stage to systematically evaluate the impact of input parameter fluctuations on the prediction results, outputting a prediction credibility map and a ranking of sensitive elements. This enhances the interpretability and credibility of the model in practical engineering decision-making and strengthens the engineering guidance value of predicting the impact of excavation widening. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only for this invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a head prediction diagram according to an embodiment of the present invention; Figure 2 This is a flowchart of a method according to an embodiment of the present invention; Detailed Implementation

[0020] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. For some well-known technologies, those skilled in the art may also use other alternative methods to implement the invention. Moreover, the accompanying drawings are only for more specific description of the embodiments and are not intended to specifically limit the present invention.

[0021] like Figure 1-2As shown, a method for predicting head increment under river widening conditions includes the following steps: S1: Based on multi-source observation data and historical operating data, a baseline hydraulic state set is established. The baseline hydraulic state set includes the baseline head distribution, roughness partitioning and boundary condition time series, channel-storage unit connectivity relationship, and key section water conveyance-storage response curves. S1 specifically includes: S11: Data fusion and assimilation integrates multi-source observation data and historical operational data. Multi-source observation data includes river topographic point clouds, historical hydrological sequences, remote sensing imagery, and field survey records. Historical operational data includes historical flood events, typical scheduling processes, and corresponding monitoring head data. Through data assimilation technology, the observation field and the model calculation field are dynamically fused to obtain the baseline head distribution at time t. This fusion process allows for the maintenance of a spatiotemporally continuous baseline head distribution even under conditions of sparse or anomalous observation data. The update form is represented as: ; This expression represents a weighted fusion of model results and measured data to obtain a baseline head that more closely approximates reality. (Model head) Reflecting calculations and predictions, observing water head The difference between the actual measurement and the coefficient reflects the model error; The adjustment ratio of the model to the control observations. Among them, This represents the baseline head distribution, and the merged head value at position x and time t. This indicates that the model calculates the head field; Indicates the observed head field; The weighted assimilation coefficient represents the weight balance between the model and the observations. The value range is 0 to 1. When using the weighted assimilation coefficient... In this scheme, the weight balance between the model and observations is achieved through error control to realize the weighted assimilation coefficient. This determines the relative influence of model calculation results and observed data in the fusion process. When observation accuracy is high, error is small, and data resolution is high, the larger value should be chosen. This allows observations to dominate the results; when the model has been fully calibrated, has good stability, or when observations are incomplete, a smaller value should be taken. The model determines the outcome. This is based on the principle that smaller errors have larger weights and larger errors have smaller weights, generally determined by the model error variance. With observation error variance The relative size is determined. When When the fusion result is closer to the model, the result is more accurate. When the results are close to the observations, the fusion results are closer; when the two are comparable, a trade-off is made.

[0022] The core idea of ​​S11 is to leverage the complementarity of model and observational data to establish a spatiotemporally continuous, physically consistent, and realistic benchmark head distribution through weighted assimilation techniques. Considering that river monitoring data often suffers from spatial heterogeneity, temporal discontinuities, and uncertainties in error, relying solely on models or observations is insufficient to accurately reflect the true hydraulic state. Therefore, the model-calculated head is considered the carrier of continuity and physical constraints, while the measured head is considered the basis for realism and local correction. Weighted assimilation coefficients are used. The design achieves a balance between the two in the fusion results. When the quality of the observation data is high, the system tends to use the observation information for correction; when there are missing measurements or large errors in the observations, the model is relied upon to maintain overall smoothness and energy conservation. Based on the optimal estimation principle in data assimilation theory, i.e., the smaller the error, the greater the weight, the more adaptively multi-source information can be fused in different hydrological stages and spatial regions by dynamically adjusting the weights, thereby maximizing the use of effective observations while maintaining physical rationality. The final benchmark head distribution can reflect the continuous evolution of the model and correct deviations in real time, providing a reliable initial state basis for subsequent widening morphology-hydraulic coupling and incremental prediction.

[0023] S12: Based on the reference head distribution Driven by historical hydrological sequences Q(t), a hydrodynamic model inversion and system identification method is used to calibrate the Manning roughness of the river channel and storage units. When the objective function converges, the calibrated roughness partitioning map and boundary condition time series function are obtained. Based on the underlying surface properties, including sand, clay, vegetation, or revetment, roughness zones are classified and spatially mapped. The objective function is expressed as: ; This expression represents the objective function used in roughness calibration to measure the difference between the model's calculated results and actual observed values. It involves continuously adjusting the Manning roughness of each partition. The water head value calculated by the model With the corresponding observed head The roughness distribution that best reflects the true hydraulic characteristics is determined by minimizing the sum of squared differences between the model and the observations. This is based on the least squares principle, which states that the deviation between the model and the observations can be measured by the sum of squared errors. By minimizing this sum of squared errors, the statistically optimal roughness calibration result can be obtained, balancing local errors with the overall fitting effect, so that the roughness partition in the reference hydraulic state has the minimum systematic deviation and maintains physical consistency.

[0024] in, This represents the objective function for roughness calibration; Indicates the first Partition roughness The simulated head value below; This indicates the corresponding observed head value; Indicates the first The roughness of the roughness zone; This indicates the total number of partitions.

[0025] The roughness calibration objective function optimizes the roughness parameters by minimizing the deviation between the model-predicted head and the observed head, thereby improving the fitting accuracy of the hydrodynamic model to the river flow. This mainly relies on comparing the simulation of local flow with actual observations. The process of optimizing the roughness parameters using the above objective function includes the following steps: Based on the riverbed sediment type, flow regime, historical data, or literature values, the Manning roughness ratio for each region is initially set. These initial values ​​provide a starting point for parameter optimization.

[0026] Error calculation: Run the hydrodynamic model using the initial roughness value. Calculate the simulated head value and compare it with the actual observed head. By comparison, the sum of squared errors is obtained. .

[0027] Optimization algorithms: Optimization algorithms are employed, including gradient descent, genetic algorithms, and Newton's method to adjust the roughness of each partition. To minimize the objective function During the optimization process, the algorithm adjusts the roughness based on the error gradient or global search method, gradually approaching the optimal solution.

[0028] Dynamic adjustment and convergence: As iterations proceed, the optimization algorithm dynamically adjusts the value of each roughness partition until the objective function is achieved. Converging to a minimum value indicates that the difference between the simulated head and the observed head has been minimized, meaning that the optimal roughness distribution has been found.

[0029] Manning roughness The roughness parameter reflects the resistance of the riverbed and sidewalls to water flow and is closely related to topographic type (gravel bed, clay, rock), flow regime (laminar or turbulent), and bottom surface roughness. During widening, changes in the riverbed cross-section will lead to dynamic changes in the roughness parameter, thus affecting the energy loss and head distribution of the water flow. By optimizing the roughness, the model can simulate changes in water flow resistance caused by widening, helping to identify which areas experience changes in head increment.

[0030] Boundary condition timing function This is to simulate and describe the changes in upstream and downstream boundary conditions (water level, flow rate, or operational status of control facilities) over time at different time steps, and to use these changes as input to a hydraulic model to drive head increment prediction. River boundary conditions (such as upstream flow rate, downstream water level control, or gate scheduling) are represented as a continuous time series function in the time dimension. The boundary conditions are then combined with other parameters, namely roughness partitioning, in the hydrodynamic model to predict head increments. Boundary conditions typically come from several data sources: hydrological data (flow and precipitation data from historical hydrological sequences); scheduling rules (reservoir scheduling and river control facilities, i.e., gate and pump operation plans); measured monitoring (downstream water level, river depth, etc., obtained through real-time monitoring systems); and remote sensing data (water level or flow information provided by remote sensing or satellite imagery). Data acquisition and processing involve collecting historical hydrological data, engineering scheduling data, etc., and preprocessing them to form a time series. Future boundary scenarios can be hypothesized based on historical patterns or simulation results. The time series function uses smoothing techniques, including moving averages and interpolation, to smooth the collected boundary condition data, ensuring the continuity and stability of the time series. Boundary conditions can be divided into several different stages, including dry season, wet season, and flood season. The boundary conditions for each stage can be described by different functions, ultimately forming a time series function that conforms to actual operating conditions. In practice, boundary conditions may change, so it is necessary to update the time series function B(t) regularly to ensure that the model is always consistent with the latest hydrological and engineering scheduling conditions.

[0031] S12 uses model inversion and system identification methods to quantitatively calibrate the Manning roughness of river channels and retention units, thereby establishing a roughness zoning map that conforms to actual hydraulic characteristics. During the design phase, it was considered that river channel roughness is significantly affected by substrate type, bank slope morphology, vegetation distribution, and structures, exhibiting significant spatial differences. Using a uniform roughness value would lead to accumulated head calculation errors, failing to accurately reproduce actual water surface changes. In the specific implementation, firstly, the baseline head distribution and historical hydrological sequences are used as model-driven inputs to calculate the deviation between the model's output head value and the measured head. Then, the objective function is constructed using the least squares principle, with the sum of squared deviations across all time and spatial points as the optimization index. By iteratively adjusting the Manning roughness of each zone, the error function is minimized, ultimately obtaining the optimal roughness combination.

[0032] S13: Identify potential waterlogging units based on river channel topographic point clouds and remote sensing imagery. Furthermore, the connection elevation threshold between the main channel of the river and the retention unit was determined through GIS spatial analysis. With initial hydraulic width This topology is used for subsequent coupled energy analysis and head response propagation calculations. The channel-retention unit connectivity is represented as follows: ; in, This represents the set of connectivity relationships between river channels and storage units; Representing the main channel unit; Indicates the first Storage unit; Indicates the total number of storage units; This represents the initial hydraulic width, used to characterize connectivity. This represents the elevation threshold connecting the retention area to the main channel; the value range is determined based on the difference between the bottom elevation of the main channel and the bottom elevation of the retention area, and is taken between 0.5 and 3.0 above the bottom elevation of the main channel; when the retention area is adjacent to the main channel and the terrain is flat, A lower value, 0.5–1.0, indicates that a rise in water level is sufficient to form connectivity; when the retention area is slightly higher than the main channel and is separated by banks or protective dikes, Versions 1.0 to 2.0 are connected only during floods or high water levels; when the retention area is higher than the main channel and overflow connection is only formed during extremely high water levels, Values ​​range from 2.0 to 3.0.

[0033] The core of S13 is to clarify the connectivity between the main channel and surrounding retention units through spatial topology analysis, establishing a realistic water exchange structure for subsequent hydraulic calculations. Considering the presence of depressions, flood detention areas, old channels, or riverbanks along the river, these areas will periodically connect with or isolate from the main channel under different water level conditions. Their connectivity characteristics affect the overall head changes and energy distribution of the river. The location and extent of each potential retention unit are identified using river topographic point clouds and remote sensing imagery. Then, GIS spatial analysis is used to determine the elevation threshold and initial hydraulic width between each retention unit and the main channel. The elevation threshold is used to determine the initiation condition of connectivity; that is, when the river water level exceeds this elevation, water flow begins to enter the retention area. The initial hydraulic width is used to measure the scale of the connectivity channel and its hydraulic transmission capacity.

[0034] S14: Based on historical hydrological sequences With monitoring head data Select key control sections Through statistical analysis or hydrodynamic model simulation, the water conveyance-storage response curves of key sections are derived. Stable water conveyance-storage response relationships can be obtained through multi-condition verification, which can be used to characterize the hydraulic-morphological coupling features. The functional relationship of the water conveyance-storage response curves of key sections can be expressed as: ; in, Indicates the head at the critical section; Indicates the flow rate at the critical section; This indicates the water storage capacity of adjacent storage units; This represents the nonlinear response function obtained by fitting historical data or extrapolating from a model. In the hydraulic model after widening, the nonlinear response function establishes a nonlinear relationship between water conveyance and storage by analyzing head changes at key river sections under different flow rates and storage conditions. This relationship describes the complex interaction between water flow and storage units, especially during flood season or high water levels, where the relationship between water conveyance capacity and water volume in the storage area may exhibit significant nonlinear characteristics. The water conveyance-storage response curve represents the head change. With water flow or water storage volume The relationship between them. In the model after widening, the increase in water head at the key cross-section. It is the result of the combined effect of changes in water volume caused by both flow rate and water storage units. Water volume in the water storage area. Closely related to changes in water head, the response curve will involve the following variables: Q: Water flow rate, which indicates the rate at which water flows through this cross-section; : Retention capacity, representing the amount of water stored in the retention unit; : Water head increment, indicating the difference in water head change before and after the excavation.

[0035] The flow and storage responses under different working conditions were simulated using a baseline hydraulic model (model before excavation) and a coupled hydraulic model (model after excavation). Hydrodynamic simulations were used to calculate different flow conditions. Increase in water head and changes in water volume within the storage unit. During the simulation, the trend of head increment can be obtained by changing the flow rate and the amount of water stored, and then the nonlinear function of the water conveyance-storage response can be derived. After obtaining the head increment... and traffic or water storage volume After obtaining a series of data, nonlinear regression analysis or empirical fitting methods can be used to establish the nonlinear response relationship between water flow and storage. The nonlinear function is expressed as: ; in, The coefficients to be fitted represent the nonlinear relationship between head increment, flow rate, and retained water volume. During the fitting process, the coefficients are adjusted based on actual data to minimize the error of the fitted curve. The accuracy of the nonlinear response function is verified by comparing it with measured data. If the fitting result has a large error, the form of the nonlinear function can be adjusted according to the actual situation, or more adjustment parameters can be added for optimization. This process can also further identify the influence of key parameters on the response function through sensitivity analysis, thereby optimizing the model accuracy.

[0036] S14 analyzes the head variation patterns at key cross-sections under different flow rates and storage conditions to establish a water conveyance-storage relationship that reflects the actual hydraulic response characteristics. Considering that river head is affected not only by flow rate but also by the combined effects of the storage and discharge processes of storage units, local roughness changes, and backwater conditions, exhibiting significant nonlinear and hysteretic characteristics, the study first selects key control cross-sections using historical hydrological sequences and measured head data, extracting multi-period flow-head-storage combination samples. Subsequently, through statistical analysis or hydrodynamic model simulation, a head model is established. With traffic , water storage capacity nonlinear functional relationship This function employs polynomial fitting, piecewise functions, or empirical regression to characterize the different rates and magnitudes of head response to storage during flood rise and recession. In actual river channels, energy loss and water level changes are not linearly related, especially during high flow, backflow, or storage phases, where head rise and changes in water storage exhibit lag and amplification effects. By establishing a nonlinear response function, the dynamic characteristics of head changes with flow and storage can be realistically reproduced in the model.

[0037] S15 benchmark set integration and verification: S151: Distribute the reference head Roughness partitioning results Timing with boundary conditions River-storage unit connectivity Key section water conveyance-storage response curve Integrate them to form a complete reference hydraulic state set. , is represented as: ; S152: Subsequently, historical operating data not involved in the construction were selected for uncertainty analysis and verification, and error indices were calculated. ; This expression represents the average error used to evaluate the difference between model predictions and actual observations during the validation of the baseline hydraulic state set. It is calculated by taking the predicted head values ​​at each observation point at different times. Compared with the measured water head value The absolute error between samples is calculated, and the average of all samples is taken to obtain the overall error index. The smaller the value, the closer the model prediction is to the actual observation, and the higher the reliability of the baseline hydraulic state set; among which, This indicates the number of samples involved in the error calculation.

[0038] S153: If the error index If the value is below the preset threshold, the baseline hydraulic state set is verified and can be used for subsequent excavation morphology-hydraulic coupling analysis.

[0039] The core of S15 is to unify and integrate the results obtained from the preceding steps into a reusable and verifiable reference hydraulic state set. Considering the strong coupling relationships between the reference head distribution, roughness partitioning, boundary condition time series, channel-storage connectivity, and key section response curves, only by unifying this information into a single data structure can the consistency and traceability of input conditions in subsequent widening analyses be guaranteed. The assimilated head field, the calibrated roughness partitioning, the constructed boundary time series functions, connectivity relationships, and nonlinear response functions are integrated according to spatial and temporal dimensions to form the reference hydraulic state set. Subsequently, independent verification was performed using historical operating data not involved in the construction, by calculating the average error between the predicted head and the observed head. This is used to evaluate the reliability of the integrated results. If the error is less than a preset threshold, it indicates that the reference set matches the actual situation in terms of overall hydraulic characteristics.

[0040] S2: The cross-section and topographic changes formed by the widening project are transformed into a widening morphology description set, and a morphology-hydraulic increment table is generated under the reference of the baseline hydraulic state set. The morphology-hydraulic increment table includes a local energy loss increment map, a channel connectivity adjustment map, and a key structure impact zoning. The morphology-hydraulic increment table is used to drive the head response inference. S2 specifically includes: S21: Based on the design drawings, BIM model, or measured topographic data of the widening project, extract the geometric parameters of the river channel cross-section after widening, specifically including: cross-section width. Riverbed elevation Slope gradient Types of riverbank protection Spatial variables. Constructing a set of excavation morphology descriptions: ; in, Indicates the coordinates of the location along the river course; Indicates the width of the cross-section after widening; Indicates the bottom elevation of the cross-section after widening; Indicates the slope gradient after widening and excavation; Indicates the type or material code of the revetment after the excavation; This represents a set of excavation morphology descriptions used to characterize the cross-sectional geometric changes brought about by excavation.

[0041] S21 extracts key geometric information from the widened river channel to construct a structured description set of widening morphology, used to comprehensively characterize the impact of widening projects on the river channel cross-sectional morphology. Widening directly alters the river channel width, bottom elevation, slope structure, and revetment type. These changes affect local velocity distribution, energy loss, reservoir connectivity, and the risk of hydraulic scour to surrounding structures. Therefore, these parameters are organized as functions along the river channel mileage to form a unified dataset, serving as the basis for subsequent hydraulic response analysis. From the cross-sectional width Riverbed elevation Slope gradient and bank protection type The composition clearly expresses the geometric morphological characteristics of the river channel at different locations after widening and dredging. These elements are interconnected, including the influence of changes in riverbed elevation on the hydraulic radius of the cross-section, the influence of slope gradient and revetment type on boundary resistance and stability, and the determination of cross-sectional width on flow capacity and flow pattern.

[0042] S22: Under the reference hydraulic state set, compare the changes in cross-sectional morphology before and after widening, and quantify the local additional energy loss through hydraulic empirical formulas or three-dimensional flow field simulation. Loss coefficient The specific values ​​can be determined by parameters such as the expansion ratio, the narrowing ratio, and the curvature of the bend, and can be based on three-dimensional simulation results or empirical formulas from standards. The additional energy loss increment can be expressed as: ; This expression describes how, when a river channel undergoes abrupt changes in cross-section due to dredging, including expansion, contraction, and bends, the water flow experiences turbulence or separation at these locations, resulting in additional energy loss. Indicates the increase in local energy loss; The local loss coefficient, representing the morphological abrupt change, is determined by factors such as cross-sectional changes and flow direction inversions. This indicates the flow velocity under baseline conditions; Represents gravitational acceleration; The spatial distribution forms a map of local energy loss increments. This represents the local hydraulic loss coefficient.

[0043] S22 quantitatively analyzes the geometrical abrupt changes in the river cross-section caused by the widening project, under the reference of a baseline hydraulic state set, assesses the resulting increase in local energy loss, and expresses it in the form of a spatial distribution map. This scheme was designed because after widening, significant cross-sectional expansion, contraction, or flow direction reversal may occur in local areas of the river channel. These changes can cause turbulence, backflow, eddies, and other phenomena, leading to additional energy losses and affecting the local head distribution, particularly at bends, downstream of bridges, or at points where reservoirs connect.

[0044] S23: Based on the channel-retention unit connectivity relationship in the widening morphology description set and the benchmark hydraulic state set, compare the changes in connection elevation and hydraulic width before and after widening. Through comparison... Compared with the benchmark value , and This allows us to obtain changes in connectivity and generate a channel connectivity adjustment map.

[0045] The new connectivity elevation is represented as: ; The new hydraulic width is denoted as: ; in, Indicates the first The new connectivity elevation threshold of the storage unit determines the starting water level at which the storage unit participates in the flow. Indicates the relationship between the expanded excavation and the first The hydraulic width of the storage unit; This represents the effective hydraulic transport index function at the connection section. It varies depending on different topographic features and excavation methods. The range of values ​​is: Low-connectivity retention units include depressions, farmland, and low-lying beaches. This indicates that a connection will occur as soon as the water level rises slightly, and it is suitable for areas with frequent water storage and regulation.

[0046] The central interconnected type includes high beaches, the inner area of ​​the embankment, and the riverbank terraces. This indicates that a higher water level is required for connectivity, and it typically only functions during flood season.

[0047] High connectivity thresholds include embankment-enclosed areas and old riverbed water retention zones. It is activated only during extremely high water levels and is typically used for passive water storage or flood buffer zones. Effective hydraulic transport index function It is a discriminant function that describes whether a certain location has effective water flow capacity after widening and is used to measure the spatial connection strength between the main channel and the storage unit. It can be designed as a piecewise assignment function or a weighting function, using a binary logic function (discriminative type): ; in The connectivity tolerance is set to 0.2m, meaning only areas close to the connectivity elevation are counted. This scheme is suitable for quickly determining connectivity and generating clear connectivity widths.

[0048] The purpose of S23 is to reassess the changes in connectivity between the river channel and the retention units after the widening project, and to express this in the form of a channel connectivity adjustment map. Widening directly alters the riverbed elevation, slope structure, and cross-sectional shape, thus affecting the effective flow into the retention units, as well as the scale and hydraulic efficiency of the connecting channels. First, based on the topographic data after widening, the connection area between each retention unit and the main channel is identified, and the lowest riverbed elevation within this area is extracted as the new connectivity elevation threshold. First, it is used to determine when the water level can enter the storage area; second, it is used to construct an effective hydraulic transmission index function. The system uses a 0-1 scale to determine whether each location has effective water flow capacity, and integrates over the entire connected area to calculate the hydraulic width of the connection after widening. Finally, these new values ​​are compared with the baseline state before the excavation to determine whether the excavation improved or weakened connectivity.

[0049] S24: Identify key structures such as existing bridges, dams, and diversion piers within the impact area of ​​the excavation project, and extract their foundation elevations. Location coordinates In conjunction with the velocity distribution after widening, With hydraulic gradient The process involves determining whether the structure triggers hydraulic changes in the vicinity. If one of the following conditions is met, the area near the structure is included in the critical structure influence zone, which will be treated as a special boundary region in subsequent coupled simulations. The conditions are expressed as follows: Basic relative erosion: ; Sudden change in flow rate: ; in, This indicates the set threshold for changes in basic elevation; This represents the set threshold for flow velocity change, with a value of 0.3; the foundation elevation change threshold is used to determine whether the excavation causes a change in the relative scour of the structure's foundation. Value range: If the structures are shallow foundation piers, revetment walls, and guide dikes, If the structure is a deep foundation gate pier or pile foundation structure, The value range is 0.6 1.0; Elevation changes exceeding 1 typically require a dedicated scour stability review. The velocity change threshold is used to identify whether abrupt changes in flow conditions around a structure may trigger scour or hydrodynamic disturbances. The value range is... For exposed foundations and non-reinforced bank slopes, a smaller value should be used. For areas with good protection and solid structure, If the change value exceeds 0.6, it is usually considered a major change in hydraulic conditions and should be included in the impact zoning.

[0050] The purpose of S24 is to identify potential changes in hydraulic conditions caused by dredging projects to existing hydraulic structures, including bridges, dams, and guide piers, thereby delineating the critical structure impact zone and providing a basis for subsequent model boundary setting and engineering risk analysis. Dredging projects may affect the stability of structure foundations, the degree of water scouring, or the stress state by altering riverbed elevation or flow velocity distribution; therefore, the impact range needs to be identified and marked in advance. First, the foundation elevation and spatial location of the structure are extracted, and then compared with the riverbed elevation and flow velocity distribution after dredging. If dredging causes downcutting of the riverbed at the structure's location (i.e., increased foundation exposure) or a sudden increase in flow velocity (i.e., enhanced local hydraulic scouring capacity), then the structure is considered potentially affected. This judgment uses two engineering experience thresholds as criteria: a foundation elevation change threshold of 0.5 and a flow velocity change threshold of 0.3. If either condition is met, the structure and its surrounding area are included in the critical structure impact zone.

[0051] S25: Map of local energy loss increments Channel connectivity adjustment diagram Key Structure Impact Zone Set A unified, integrated morphological-hydraulic increment table is constructed. This table serves as the direct input for subsequent head increment prediction steps, driving the response analysis of excavation work to changes in hydraulic state. The constructed morphological-hydraulic increment table is represented as follows: ; in, Representation of form - hydraulic increment table; This indicates the set of river channel spaces covered by the impact zone of key structures. The core of S25 is to unify and integrate the results of the dredging impact obtained from the previous sub-steps into a structured dataset called the morphological-hydraulic increment table, which serves as the direct input for subsequent head increment prediction. This increment table includes three aspects: the local energy loss increment caused by dredging, information on changes in channel connectivity, and the impact range of key structures. It comprehensively characterizes the impact of dredging projects on the river hydraulic system from the perspectives of hydraulic loss, channel connectivity, and structural safety. River dredging alters hydraulic conditions in various ways, and individual analyses often fail to reflect the overall impact. However, these impacts are needed as input conditions to collaboratively predict the actual head response. Therefore, the local energy loss increment map is used to characterize the flow disturbance and additional head loss caused by dredging; the channel connectivity adjustment map is used to reflect changes in storage capacity and lateral replenishment capacity; and the key structure impact zoning is used to identify potential risk areas and set boundary constraints. Integrating these three into a unified table helps to standardize model input, automate spatial analysis, and ensure the traceability of the subsequent response process.

[0052] S3: Using the baseline hydraulic state set and the morphological-hydraulic increment table as input, and combining the target boundary scenario, generate a head increment prediction grid and a head increment curve for key sections, and output a prediction confidence map and a ranking of sensitive elements; the head increment prediction grid and the head increment curve for key sections serve as the direct basis for engineering comparison and verification.

[0053] S3 specifically includes: S31: Hydrodynamic model that integrates the reference hydraulic state Local energy loss increment map in the morphological-hydraulic increment table Channel connectivity adjustment diagram By coupling, a prediction model for the case after expanded excavation is obtained. , is represented as: ; in, This represents a hydrodynamic model under the baseline hydraulic state. This represents a coupled prediction model under expanded excavation conditions; This indicates the model superposition coupling operation; A map showing the increase in local energy loss; , This indicates the channel connectivity adjustment parameter.

[0054] S31 constructs a head increment prediction model that can accurately reflect the impact of widening, called the coupled prediction model. The design concept is to introduce new influencing factors brought about by widening into the existing baseline hydrodynamic model to simulate and predict the hydraulic state after widening. The hydrodynamic model of the baseline hydraulic state set constructed in S1 serves as the basic calculation framework, possessing a complete watershed hydraulic structure and physical parameters. The morphological-hydraulic increment table obtained in S2 includes key influencing factors such as the increase in local energy loss and changes in channel connectivity caused by widening. These influencing quantities are added to the baseline model in a model coupling manner to form a prediction model under widening conditions. Although widening of the river channel does not change the basic structure of the hydrodynamic model, it will cause a series of changes in local parameters, including increased local resistance and enhanced stagnant connectivity. If these changes are not included in the model calculation, it will be difficult to accurately predict the head change trend. Therefore, by keeping the model structure unchanged and superimposing incremental information, a balance is achieved between model reuse and local adjustment, maintaining overall computational stability while effectively responding to local disturbances caused by widening.

[0055] S32 sets the target boundary scenarios and executes simulations to define the target boundary scenario set. Represented as: ; in, This represents the number of target boundary scenarios, i.e. Total number Indicates the first Each target boundary scenario represents a time-varying boundary condition input, including inflow, control water level, and scheduling rules; each set of boundary scenarios This represents a set of time-series inputs that may indicate future hydrological processes, engineering scheduling rules, or downstream control conditions. Under each scenario, the baseline model and the coupled model are run separately to calculate the head distribution before and after the widening. and The head increment prediction grid is defined as follows: ; S32 simulates the head response of the river channel before and after widening by setting multiple representative target boundary scenarios, thereby obtaining spatially distributed head increment prediction results. The purpose is to assess the potential head changes caused by the widening project under different future hydrological conditions, downstream control methods, or scheduling rules, ensuring that the prediction results have scenario adaptability and engineering reference value.

[0056] S33: Extracting the Head Increment Curve of Key Sections. Let the set of key sections defined in the reference hydraulic state set be... For each cross section and each target boundary scenario Extract the corresponding head increment as a function of flow rate. or time The relationship of change: ; This expression represents the first flow rate under different flow conditions. The increase in water head at a key cross-section due to dredging. For any given flow rate. Water head value of the cross section after widening Compared with the water head value before excavation The difference between the flow rate and the head change is the change in head at that cross-section under that flow rate, reflecting the hydraulic response brought about by the expanded excavation. In the comparative analysis of expanded excavation, the head change at key cross-sections is a direct indicator for judging the project's effectiveness, including changes in water control, flood discharge, and storage capacity. The flow-head increment curve reveals the cross-section's sensitivity to different time periods and boundary conditions, providing a quantitative expression of head prediction results at typical locations, facilitating project comparison and risk assessment.

[0057] The core of S33 is based on the head increment prediction grid in the simulation results. It extracts the head change difference of a set of preset key sections under different flow or time conditions to form a head increment curve for the key sections. The purpose is to focus the spatially distributed prediction results on the locations most sensitive to engineering decisions and display the head response in the form of a curve, which helps to carry out comparability analysis and verification of typical sections. First, based on the predefined key section locations of the reference hydraulic state set, including control sections, water level observation points, and upstream sections of structures, the head values ​​of these locations before and after widening are extracted from the model results. Then, the difference between the two is calculated to form the relationship curve of head increment with flow or time.

[0058] S34 performs uncertainty analysis and ranks sensitive elements: S341: Let the set of input parameters be... The corresponding range of uncertainties is Construct the parameter space: ; in, Indicates the first The minimum possible value of each input parameter; Indicates the first The most likely values ​​that each input parameter can take; The parameter space represents the complete set of all combinations of input parameters within their uncertainty range; S342: In N samplings were performed using Monte Carlo sampling, and a coupled model was run to obtain the predicted head increment results for the nth simulation. Based on all simulation results, statistical spatial location was determined. Predicted mean and confidence interval at ... Draw a prediction confidence map: ; in, Indicates position The lower confidence limit for the prediction includes the 5th percentile; Indicates position The upper confidence limit for the prediction is included at the 95th percentile.

[0059] S343: Sensitive element ranking employs Sobol index based on variance decomposition, absolute value ranking of regression coefficients, or Morris local perturbation gradient analysis to calculate each input parameter. For the output results The degree of influence Si, and ranked: ; Output a sorted list of sensitive elements: ; in, Indicates the parameter to be performed The functional expression for sensitivity analysis, This represents a list of input parameters sorted from highest to lowest sensitivity; the specific approach in functional form involves establishing the input parameters through sensitivity analysis. With output results The functional mapping relationship between (i.e., the increase in head) is used to measure the degree of influence of parameter changes on the results. The functional form varies depending on the analysis method and may include: 1. Using a regression function, construct a linear or nonlinear regression model, taking the water head increment as the basis. For dependent variable, parameter Using as the independent variable, regression coefficients are obtained through fitting, and the sensitivity index is the absolute value or standardized value of the coefficient.

[0060] 2. Variance decomposition function form (such as the Sobol method): This decomposes the total variance of the head increment. Decompose the sensitivity into components caused by each input parameter individually or interactively, and define the sensitivity function as follows: ; in, Indicates by Variance contribution caused by a single factor.

[0061] 3. Locally perturbative function form (such as the Morris method): In the input space... Perform a small perturbation and observe the output. The change in sensitivity can be expressed as the local increment slope or the average disturbance response.

[0062] S34 mainly comprises two parts: first, generating a prediction confidence map through uncertainty simulation; and second, outputting a ranking of sensitive factors through sensitivity analysis. Together, these constitute the confidence assessment of the model results and the identification of key influencing factors. Regarding the prediction confidence map, a Monte Carlo sampling method is used to randomly sample a large number of samples from within the uncertainty range of the input parameters. Each sample group represents a possible combination of model inputs. By running the coupled model on each sample group, the corresponding head increment prediction results are obtained. Then, the distribution characteristics of all simulation results are statistically analyzed at each spatial location, and the mean and upper and lower confidence limits are extracted to draw the prediction confidence map. This map visually reflects the degree of fluctuation in results under various uncertain inputs in different regions, identifying high-risk or stable areas of head change. Regarding the ranking of sensitive factors, the scheme calculates the influence of each parameter on the change in head increment results by performing variance decomposition, regression coefficient evaluation, or local perturbation analysis (such as the Sobol index or Morris method) on all input parameters, and ranks them according to the strength of their influence.

[0063] This invention encompasses any substitutions, modifications, equivalent methods, and solutions made within the spirit and scope of this invention. To provide the public with a thorough understanding of this invention, specific details are described in detail in the following preferred embodiments; however, those skilled in the art will fully understand the invention even without these details. Furthermore, to avoid unnecessary misunderstanding of the essence of this invention, well-known methods, processes, procedures, components, and circuits are not described in detail.

[0064] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for predicting head increase under river widening conditions, characterized in that, Includes the following steps: S1: Based on multi-source observation data and historical operating condition data, establish a benchmark hydraulic state set, which includes benchmark head distribution, roughness partitioning and boundary condition time series, channel-storage unit connectivity relationship and key section water conveyance-storage response curves. S2: The cross-section and topographic changes formed by the widening project are transformed into a widening morphology description set, and a morphology-hydraulic increment table is generated under the reference of the baseline hydraulic state set. The morphology-hydraulic increment table includes a local energy loss increment map, a channel connectivity adjustment map, and a key structure influence zoning. The morphology-hydraulic increment table is used to drive head response inference. S3: Using the reference hydraulic state set and the morphological-hydraulic increment table as input, and combining the target boundary scenario, generate a head increment prediction grid and a head increment curve for key sections, and output a prediction confidence map and a ranking of sensitive elements; the head increment prediction grid and the head increment curve for key sections serve as the direct basis for engineering comparison and verification.

2. The method for predicting head increase under river widening conditions according to claim 1, characterized in that, The multi-source observation data includes river topographic point clouds, historical hydrological sequences, remote sensing images, and field survey records; the historical operating condition data includes historical flood events, typical scheduling processes, and corresponding monitoring head data.

3. The method for predicting head increase under river widening conditions according to claim 2, characterized in that, S1 specifically includes: S11: Establish a baseline head distribution by assimilating multi-source observation data with historical operating condition data; S12: Driven by the reference head distribution and the historical hydrological sequence, the Manning roughness of the river channel and storage unit is calibrated through hydrodynamic model inversion and system identification methods, and roughness partitioning is performed according to the underlying surface properties and calibration results. At the same time, the time series of boundary conditions required by the model is determined. S13: Based on the river topography and remote sensing images, identify potential water storage units, and use GIS spatial analysis technology to determine the connection position, elevation threshold and initial hydraulic width between the river and each water storage unit, and establish the river-water storage unit connectivity relationship. S14: Based on the historical hydrological sequence and the corresponding monitoring head data, select key control sections, and derive the cross-sectional water conveyance-storage response curves of the key control sections through statistical analysis or hydrodynamic model simulation.

4. The method for predicting water head increment under river widening conditions according to claim 3, characterized in that, S1 further includes constructing the benchmark hydraulic state set by combining the benchmark head distribution, roughness partitioning and boundary condition time series, channel-storage unit connectivity relationship and key section water conveyance-storage response curve, and performing uncertainty analysis and verification on it using historical operating condition data that were not involved in the construction.

5. The method for predicting water head increment under river widening conditions according to claim 1, characterized in that, In S2, based on the design drawings, BIM model or measured topographic data of the excavation project, the geometric parameters of the river cross section, changes in riverbed elevation, slope gradient and bank protection type after the project is implemented are extracted to form a digital excavation morphology description set.

6. The method for predicting the increase in water head under the condition of river widening as described in claim 5, characterized in that, The generation of the morphology-hydraulic increment table specifically includes: S21: Under the reference of the above-mentioned hydraulic state set, compare the degree of abrupt change and geometric shape of the river channel cross section before and after the widening, and quantify the additional head loss caused by the cross section expansion, contraction and flow direction change based on hydraulic empirical formulas or local three-dimensional flow field simulation, and generate a spatially distributed local energy loss increment map. S22: Based on the channel-storage unit connectivity relationship in the above-mentioned widening morphology description set and the benchmark hydraulic state set, recalculate the connection elevation and hydraulic width between the widened channel and the storage unit, quantify the change in connectivity through a hydraulic coupling model or empirical function, and generate a channel connectivity adjustment map. S23: Identify existing key structures within the scope of the excavation project's influence, and analyze their foundation elevation, flow regime, and relative spatial relationship with the excavation project; through hydrodynamic simulation or engineering experience, delineate areas where hydraulic conditions have significantly changed due to riverbed incision and flow velocity variations, and form a key structure influence zoning.

7. The method for predicting water head increment under river widening conditions according to claim 6, characterized in that, The generation of the morphological-hydraulic increment table also includes integrating the local energy loss increment map, the channel connectivity adjustment map, and the key structure influence zoning to form the morphological-hydraulic increment table.

8. The method for predicting water head increment under river widening conditions according to claim 1, characterized in that, S3 specifically includes: S31: Couple the hydrodynamic model of the reference hydraulic state set with the local energy loss increment map and channel connectivity adjustment map in the morphology-hydraulic increment table to construct a coupled model for head increment prediction. S32: Define one or more target boundary scenarios representing future hydrological conditions and engineering scheduling rules; input each scenario into the coupled model, calculate the hydraulic state before and after the excavation, and obtain the difference to get the head increment prediction grid. S33: Based on the head increment prediction grid, for the key sections defined in the reference hydraulic state, extract the head increment under different target boundary scenarios, including different flow rates, and generate the key section head increment curve with flow rate or time as the horizontal axis and head increment as the vertical axis.

9. The method for predicting head increase under river widening conditions according to claim 8, characterized in that, S3 also includes uncertainty analysis and sensitivity ranking.

10. The method for predicting head increase under river widening conditions according to claim 9, characterized in that, The uncertainty analysis and sensitive element ranking specifically include: Identify and quantify the uncertainty range of key input parameters in the baseline hydraulic state set and the morphological-hydraulic increment table; Based on the Monte Carlo simulation sampling method, a large number of random samples and model calculations are performed within the above uncertainty range to generate the probability distribution of the water head increment, and a prediction confidence map representing the spatial confidence of the prediction results is drawn accordingly. The sensitivity analysis method of analysis of variance is used to calculate the degree of influence of each input parameter on the water head increment result and output the ranking of sensitive elements.

Images