A positive and negative calculation method for joint flood control optimization scheduling of reservoir and flood area

Abstract

This invention discloses a forward and inverse calculation method for joint flood control optimization scheduling of reservoirs and floodplains. The method includes: constructing and calibrating a two-dimensional hydrodynamic model based on basic data; constructing and calibrating a piecewise Muskingen model based on the river simulation results of the calibrated two-dimensional hydrodynamic model; fitting the floodplain outlet flood diversion function based on the river flow and floodplain diversion process simulated by the two-dimensional hydrodynamic model to construct a floodplain outlet flood diversion model; setting the objective function and constraints for scheduling; performing inverse optimization of the scheduling scheme based on the piecewise Muskingen model and the floodplain outlet flood diversion model to obtain the optimal scheduling scheme; and inputting the inverse optimization results into the two-dimensional hydrodynamic model for forward calculation to obtain the scheduling result of the optimal scheduling scheme. The advantages are: it considers the floodplain diversion process under the influence of the river flow loop curve, enabling rapid and accurate simulation of flood evolution and floodplain diversion processes, and improving the timeliness and feasibility of considering floodplain diversion in the optimized scheduling model.

Description

Technical Field

[0001] This invention relates to the field of reservoir scheduling technology, and in particular to a forward and inverse calculation method for joint flood control optimization scheduling of reservoirs and floodplains. Background Technology

[0002] Flood control optimization scheduling calculations are one of the most important fundamental tasks in reservoir operation. With climate change and the increasing frequency of extreme weather events, the threat posed by floods to the socio-economic and ecological environment is intensifying. Floodplains, as the last line of defense against floods, store floodwaters, reduce peak flow, and mitigate the impact of floods on downstream large urban clusters, buying time for downstream flood control and emergency response measures. However, due to urbanization, many farms, factories, and small towns have emerged within floodplains, and their activation inevitably endangers the lives and property of the people living there. Because of the heavy computational load on hydraulics, it is difficult to couple hydraulic models to achieve optimized scheduling of floodplain activation. Currently, decision-makers mostly activate floodplains by rules, which is not conducive to advance evacuation of residents and reducing flood inundation losses. Summary of the Invention

[0003] The purpose of this invention is to provide a forward and inverse calculation method for the joint flood control optimization scheduling of reservoirs and floodplains, thereby solving the aforementioned problems existing in the prior art.

[0004] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0005] A forward and inverse calculation method for joint flood control optimization scheduling of reservoirs and floodplains includes the following steps:

[0006] S1. Construction of a two-dimensional hydrodynamic model: Based on the collected basic data of reservoirs and floodplains, a two-dimensional hydrodynamic model is constructed and calibrated.

[0007] S2. Flood diversion data generation in the floodplain: Based on the calibrated two-dimensional hydrodynamic model, simulate the flood diversion flow at the floodplain outlet, the flow in the river channel outside the floodplain outlet, and the flow in the river channel at the key protection section.

[0008] S3. Construction of Muskingen Model: A segmented Muskingen model was constructed and calibrated based on the river channel simulation results of the two-dimensional hydrodynamic model.

[0009] S4. Construction of flood diversion model for floodplain: Based on the river flow and flood diversion process simulated by the two-dimensional hydrodynamic model, the flood diversion function of the floodplain is fitted to construct the flood diversion model of the floodplain.

[0010] S5. Forward and Reverse Calculation: Set the objective function and constraints for scheduling, perform reverse optimization of the scheduling scheme based on the calibrated segmented Muskingen model and the flood diversion model of the floodplain, and obtain the optimal scheduling scheme; then input the reverse optimization result into the calibrated two-dimensional hydrodynamic model for forward calculation to obtain the scheduling result of the optimal scheduling scheme.

[0011] Preferably, step S1 specifically involves determining the piecewise roughness, simulation time, and initial lake water volume parameters of the two-dimensional hydrodynamic model based on the measured river flow and flood inundation depth of the floodplain.

[0012] The two-dimensional hydrodynamic model employs a water storage unit method, neglecting non-dominant physical processes and simplifying the solution of the hydrodynamic equations. The change in unit water volume is derived from the cumulative water exchange between units. The flow rates of adjacent units are calculated using an explicit shallow-water inertial scheme, as shown in the formula.

[0013]

[0014] Where, q t S represents the unit width flow rate at the cell boundary at time t; surf Where g is the slope of the water surface, n is the acceleration due to gravity, and h is the roughness coefficient. flow The critical flow depth is given; the time step Δt is constrained by the Courant-Freidrichs-Levy condition, as shown in the formula.

[0015]

[0016] Wherein, △t max For the maximum time step, h max The maximum water depth in the study area, k is the constraint coefficient; Δx is the edge length of the cell grid.

[0017] The formula for calculating the change in unit water volume after unit flow exchange is as follows:

[0018] △V (r,c) =Q x,(r-1,c) -Q x,(r,c) +Q y,(r,c-1) -Q y,(r,c) (10)

[0019] Among them, Q x Q y ΔV represents the flow rate in the x and y directions; r and c are the cell row and column indices; ΔV is the cell water volume.

[0020] Preferably, step S2 specifically involves simulating the discharge process of key protected sections, river sections where the outlets are located, and the flood diversion process of the outlets based on a calibrated two-dimensional hydrodynamic model, and recording the corresponding discharge process and simulated flow process.

[0021] Preferably, step S3 specifically involves constructing and calibrating a segmented Muskingen model of the river channel according to the key protection sections and the river channel sections where the estuaries are located;

[0022] The segmented Muskingen model estimates flow rate by simulating the retention and movement of water in a river channel. The core of this model is to represent the river channel as a water storage unit, whose outflow is related to the inflow and retention status. Each segment of the channel has two parameters representing retention time and scaling degree; the formula is...

[0023] Q t+1 =C0I t+1 +C1I t +C2Q t (11)

[0024]

[0025] Where C0, C1, and C2 are the weighting coefficients of the corresponding terms; K is the retention time coefficient of the river channel; X is the degree of flattening of the inflow; Δt is the time step; I t and I t+1 The river inflow at time t and time t+1 are respectively; Q t and Q t+1 The outflow from the river at time t and time t+1 are respectively; by calibrating K and X, a rapid simulation of river floods is achieved.

[0026] Preferably, the parameters of the segmented Muskingen model are optimized and calibrated using the differential evolution algorithm with the goal of maximizing Nash efficiency.

[0027] Preferably, step S4 specifically involves using a well-defined two-dimensional hydrodynamic model to simulate the river flow process and the flood diversion process in the floodplain. The river flow is considered as a loop curve with the current and previous river flow as independent variables, and the flood diversion function of the floodplain is fitted based on multiple linear or neural network to obtain the flood diversion model of the floodplain.

[0028] Preferably, step S5 specifically includes the following:

[0029] S51. Reverse optimization solution: Based on actual scheduling needs, set the objective function and constraints of scheduling, and perform reverse optimization on the segmented Muskingen model and flood diversion model of flood flood area with fixed utilization of multiple scheduling schemes to obtain the optimal reservoir discharge process and gate opening time under given objective function and constraints.

[0030] S52. Forward simulation calculation: Using the optimal reservoir discharge process and gate opening time obtained from the reverse optimization as inputs to the calibrated two-dimensional hydrodynamic model, the flood flooding process and flood flooding channel flow process under the optimal scheduling scheme are calculated in the forward calculation.

[0031] S53. Result Judgment: Determine whether the forward calculation result meets the expected requirements. If not, adjust the constraints and re-perform the reverse optimization solution and forward simulation calculation until the forward calculation result meets the expected requirements.

[0032] Preferably, in step S5, the objective function includes minimizing the peak discharge from the reservoir, minimizing the water storage in the floodplain, and minimizing the peak discharge in the floodplain outlet channel.

[0033] The constraints include the maximum flow rate at key river sections downstream of the reservoir, reservoir water level constraints, and the maximum flood diversion volume in the floodplain.

[0034] Preferably, the procedure before step S1 includes:

[0035] S0. Basic Data Collection: Collect and organize basic data on reservoirs and floodplains, including reservoir information, measured data, two-dimensional topographic data, and floodgate information for flood diversion areas.

[0036] Preferably, in step S0,

[0037] Reservoir information includes reservoir capacity curves, as well as highest water level, lowest water level, and water level fluctuation constraints;

[0038] The measured data include measured river flow, outlet opening time, outlet size, and flooded area of ​​the floodplain. These data need to come from the same flood event.

[0039] The two-dimensional topographic data is underwater topography, covering the floodplain, the floodplain, and the connecting waterways before the reservoir;

[0040] Information on floodgates includes location, depth, and length.

[0041] The beneficial effects of this invention are as follows: The method fully considers the relationship between water level and flow rate during flood evolution, particularly during rising and receding water levels. It fits the outlet flow process using the current and previous river flow rates, eliminating the need for hydraulic calculations of river level. Therefore, the Muskingan model can be used to simulate the river flow process, improving calculation speed. Combined with optimization algorithms, the optimal scheduling result under given constraints is calculated in reverse. Finally, a forward two-dimensional hydraulic model is used to calculate the floodplain inundation process and the floodplain outlet channel flow process. This method saves significant time, is easy to implement, and provides technical support for watershed flood management. Attached Figure Description

[0042] Figure 1This is a flowchart of the method in an embodiment of the present invention;

[0043] Figure 2 This is a schematic diagram of the discharge process of Guanting Reservoir in an embodiment of the present invention;

[0044] Figure 3 This is a schematic diagram showing the hydraulic modeling scope and simulation results in an embodiment of the present invention;

[0045] Figure 4 This is a schematic diagram of a typical flood scenario in an embodiment of the present invention;

[0046] Figure 5 This is a schematic diagram of the segmented Muskingu topology in an embodiment of the present invention;

[0047] Figure 6 This is a schematic diagram of the fitting results of Nanshikou Gate in an embodiment of the present invention;

[0048] Figure 7 This is a schematic diagram of the preferred scheduling result in an embodiment of the present invention. Detailed Implementation

[0049] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0050] Example 1

[0051] This embodiment addresses the problem that traditional floodplain activation planning relies heavily on hydraulics and weir flow formulas, making it difficult to quickly determine the optimal scheduling scheme. It provides a forward and inverse calculation method for joint flood control optimization scheduling of reservoirs and floodplains. This method fits the outlet flow process using the current and previous river flow rates, eliminating the need for hydraulic calculations of river levels. Therefore, the Muskingan model can be used to simulate the river flow process, improving calculation speed. Combined with optimization algorithms, the optimal scheduling result under given constraints is calculated in reverse. Finally, a forward two-dimensional hydraulic model is used to calculate the floodplain inundation process and the floodplain outlet river flow process. This method saves significant time, is easy to implement, provides technical support for watershed flood management, and offers a new approach to flood control scheduling considering floodplain operations. Figure 1 As shown, the method specifically includes the following parts:

[0052] I. Basic Data Collection

[0053] Collect and organize basic data on reservoirs and floodplains, including reservoir information, measured data, two-dimensional topographic data, and flood diversion gate information.

[0054] (1) Reservoir information includes reservoir capacity curves, maximum water level, minimum water level, water level fluctuation and other constraints.

[0055] (2) The measured data include river flow, opening time and size of the outlet, and flooded area, and must come from the same flood event.

[0056] (3) Two-dimensional topographic data is underwater topography, covering floodplains, the connecting river channels between floodplains and reservoirs, and the smaller the resolution of the topographic data, the better.

[0057] (4) Information on the floodgates of the floodplain refers to their location, depth and length.

[0058] Topographic data must be underwater topographic data and should be collected from the local river management agency. Information on the activation status of floodplains and the inundated area can be obtained through satellite imagery, news reports, or the local hydrological bureau; this is crucial for the calibration of the two-dimensional hydrodynamic model.

[0059] II. Construction of Two-Dimensional Hydrodynamic Model

[0060] Based on the collected basic data of reservoirs and floodplains, a two-dimensional hydrodynamic model was constructed and calibrated.

[0061] Two-dimensional hydrodynamic models can be implemented using existing commercial software or self-written programs. This embodiment uses a self-written two-dimensional hydrodynamic model. Based on the measured river flow and flood inundation depth, parameters such as the piecewise roughness, simulation time step, and initial lake water volume of the two-dimensional hydrodynamic model are determined.

[0062] This two-dimensional hydrodynamic model employs the storage cell method, neglecting non-dominant physical processes to simplify the solution of the hydrodynamic equations. Changes in the storage capacity of a cell are derived from the cumulative exchange of water between cells, while the flow rates between adjacent cells are calculated using an explicit shallow-water inertial scheme.

[0063]

[0064] In the formula, q t Let S be the unit width flow rate at the cell boundary at time t (m² / s). surf Water surface slope, g, gravitational acceleration (m / s²), n, roughness (s / m¹ / ³), h flow Critical flow depth (m).

[0065] The time step is constrained by the Courant-Freidrichs-Levy (CFL) condition.

[0066]

[0067] In the formula, △t max For the maximum time step, h maxThe maximum water depth within the study area, k is the constraint coefficient, typically ranging from 0.2 to 0.7; Δx is the element grid side length. After obtaining the element flow exchange, the element water volume change can be calculated.

[0068] △V (r,c) =Q x,(r-1,c) -Q x,(r,c) +Q y,(r,c-1) -Q y,(r,c) (17)

[0069] Among them, Q x Q y ΔV represents the flow rate in the x and y directions; r and c are the row and column indices of the cell; ΔV is the water volume per cell.

[0070] Compared to the complete solution of the shallow water equation, the computation process of this scheme is significantly simplified and can be accelerated by GPU parallel processing for fast simulation. Previous studies have fully demonstrated that this method has good reliability in both experimental models and practical flood forecasting applications.

[0071] Constructing a two-dimensional hydrodynamic model requires accurately modeling the location and size of the floodgates to invert the actual flood diversion process. The roughness of the two-dimensional hydrodynamic model needs to be adjusted in segments based on measured river flood peak processes and floodplain inundation conditions. Simultaneously, it's necessary to observe the initial conditions, such as the initial conditions of lakes and some reservoirs, and the initial water level of the river. Preheating with historical data over a longer period is also acceptable.

[0072] III. Flood Diversion Data Generation

[0073] Based on a calibrated two-dimensional hydrodynamic model, the flood diversion flow at the flood flood inlet, the flow in the river channel outside the flood flood inlet, and the flow in the river channel at the key protection section are simulated.

[0074] Specifically: Given various flood discharge processes (actual or simulated), based on a calibrated two-dimensional hydrodynamic model, simulate the key protection section, the river section where the outlet is located, and the flood diversion process of the outlet, and record the corresponding discharge process and simulated flow process.

[0075] IV. Construction of the Muskingan Model

[0076] A segmented Muskingen model was constructed and calibrated based on the river channel simulation results from a two-dimensional hydrodynamic model.

[0077] The Muskingum model is a linear hydrological model used for flood runoff calculations, widely applied in the prediction and management of watershed floods in rivers and reservoirs. It estimates flow rate by simulating the retention and movement of water in a river channel. The core of the model is representing the river channel as a water storage unit, whose outflow is related to the inflow and retention status. Each segment of the channel has two parameters representing retention time and scavenging degree. In applications, this model typically requires discretization to facilitate numerical calculations.

[0078] Q t+1 =C0I t+1 +C1I t +C2Q t (18)

[0079]

[0080] In the formula, C0, C1, and C2 are the weighting coefficients of the corresponding terms; K is the retention time coefficient of the river channel; X represents the degree of flattening of the inflow; I is the inflow; and Q is the outflow. By calibrating K and X, rapid simulation of river floods can be achieved.

[0081] The parameters of the Muskingan model can be calibrated using the differential evolution algorithm, with the goal of maximizing Nash efficiency during the calibration process.

[0082] V. Construction of Flood Diversion Model at the Floodplain

[0083] Based on the simulation of river flow and flood diversion process in the floodplain using a two-dimensional hydrodynamic model, the flood diversion function of the floodplain outlet is fitted to construct the flood diversion model of the floodplain outlet.

[0084] Specifically: A well-defined two-dimensional hydrodynamic model is used to simulate the river flow process and the flood diversion process in the floodplain. The river flow is considered as a loop curve with the current and previous flow as independent variables and the current outlet flow as the dependent variable. The flood diversion function of the floodplain outlet is fitted based on multiple linear or neural network to obtain the flood diversion model of the floodplain outlet.

[0085] In this embodiment, the current and previous flow rates of the river channel outside the outlet are used to fit the outlet flow process. Using flow rate fitting at two different times can effectively simulate the problem of the loop curve during river flooding and receding (i.e., the same flow rate corresponds to different river levels during flooding and receding, thus the outlet flood diversion flow is also different), improving the estimation accuracy of the outlet flood diversion flow. The fitting method can use multiple linear regression or neural network methods.

[0086] In this invention, by inputting historical reservoir discharge data, the flow rates at key river sections and the flood diversion discharge at the outlets, simulated by a two-dimensional hydrodynamic model, can be obtained. Based on the river flow simulated by the two-dimensional hydrodynamic model, a Muskingen model is fitted. Simultaneously, the current and previous river flow rates outside the outlets are used to fit the outlet flow process. This yields the Muskingen model and the flood diversion model for the floodplain. These two models replace the preceding two-dimensional hydrodynamic model for rapid simulation. In subsequent scheduling, the two-dimensional hydrodynamic model is not used; instead, the Muskingen model is used for river simulation, and the simulated river flow is input into the flood diversion model to obtain the flood diversion discharge. In application, the upstream discharge is input into the Muskingen model, a section of the Muskingen model is run to obtain the river flow before the outlet, and this river flow is input into the flood diversion model to obtain the flood diversion discharge. Then, the input flow for the next section of the Muskingen model is obtained for simulation, and so on until the next flood diversion outlet is reached, at which point the flow rate of the second flood diversion outlet is obtained, and then the next section of the Muskingen model is simulated.

[0087] In the past, the flow rate at river mouths was calculated using the weir flow formula. However, the weir flow formula requires river level to calculate the flow capacity, which typically necessitates a two-dimensional hydrodynamic model. The time-consuming and unstable nature of two-dimensional hydrodynamic model calculations poses significant challenges to joint optimization scheduling. This invention utilizes the simulation results from the two-dimensional hydrodynamic model to construct a segmented Muskingen model and a flood diversion model for the floodplain. Subsequently, the Muskingen model and the flood diversion model are used to replace the two-dimensional hydrodynamic model for simulation.

[0088] VI. Forward and Reverse Calculations

[0089] The objective function and constraints of the scheduling are set, and the scheduling scheme is back-optimized based on the calibrated segmented Muskingen model and the flood diversion model of the floodplain to obtain the optimal scheduling scheme. The back-optimization result is then input into the calibrated two-dimensional hydrodynamic model to perform forward calculation to obtain the scheduling result of the optimal scheduling scheme.

[0090] 6.1 Reverse optimization solution: Based on actual scheduling requirements, set the objective function and constraints for scheduling, and perform reverse optimization on the segmented Muskingen model and flood diversion model of the flood flood area with the utilization rate of multiple scheduling schemes to obtain the optimal reservoir discharge process and gate opening time under the given objective function and constraints.

[0091] The objective functions include minimizing the peak flood discharge from the reservoir, minimizing the water storage in the floodplain, and minimizing the peak flood discharge at the floodplain outlet. Constraints include the maximum flow rate at key downstream river sections, reservoir water level constraints, and the maximum flood diversion capacity in the floodplain.

[0092] 6.2 Forward simulation calculation: The optimal reservoir discharge process and gate opening time obtained from the reverse optimization are input into the calibrated two-dimensional hydrodynamic model to calculate the flood flood inundation process and flood flood channel flow process under the optimal scheduling scheme (such as the flow process of each section of the river, the flood diversion process of the gate, the inundation area, the peak flow of the downstream outlet of the flood flood area, etc.).

[0093] 6.3 Result Judgment: Determine whether the forward calculation result meets the expected requirements. If not, adjust the constraints and repeat the reverse optimization solution and forward simulation calculation until the forward calculation result meets the expected requirements.

[0094] Example 2

[0095] In this embodiment, the Yongding River Basin is taken as an example. The water conservancy relationship in the Yongding River Basin is complex, starting from the Guanting Reservoir (such as...). Figure 2 As shown, there are two reservoirs, one Lugouqiao floodgate, one Xiaoqinghe flood diversion gate, and five floodplains leading to Qujiadian. During flood control scheduling, the safety of Beijing, Tianjin, floodplains (flood storage areas), and reservoirs needs to be comprehensively considered. This invention provides a forward and inverse calculation method for joint flood control optimization scheduling of reservoirs and floodplains, quickly yielding the Lugouqiao flood discharge process, the flood diversion gate process, and the flood diversion process of the floodplains. Specific execution processes and results are shown below:

[0096] I. Basic Data Collection

[0097] Based on research needs, collect two-dimensional topographic data of river channels and floodplains, reservoir water level and capacity curves, scheduling targets, inflow and outflow, and monitoring water level and flow at key river sections.

[0098] II. Construction of Two-Dimensional Hydrodynamic Model

[0099] Based on two-dimensional topographic data, the entire area is divided into regular square grids (20m*20m), and floodgates are set. A two-dimensional hydrodynamic model is constructed based on the ITF-Flood model of the China Institute of Water Resources and Hydropower Research. This model employs a two-dimensional shallow water equation explicit difference scheme and CUDA framework programming, enabling GPU parallel acceleration. It also couples modules for hydrological runoff generation, infiltration, gates, and floodgates, allowing for the simulation of the entire hydrological and hydrodynamic process. Figure 3 As shown. Input the measured discharge flow from the reservoir, and calibrate the model parameters using the measured water level and flow data from the river channel.

[0100] III. Flood Diversion Data Generation

[0101] Input the three typical floods of the Yongding River Basin in 1956 and 1963 (e.g., 237). Figure 4 As shown in the figure, multiple sets of input data are generated using the same magnification method, input into a two-dimensional hydrodynamic model for simulation, and multiple sets of datasets are recorded.

[0102] IV. Construction of the Muskingan Model

[0103] A segmented Muskingan model was established and calibrated based on the key protection sections and the river channel sections at the floodgate locations, such as... Figure 5 As shown.

[0104] V. Construction of Flood Diversion Model at the Floodplain

[0105] Multiple linear fitting was used, with the current and previous flow rates of the river channels at each floodgate location as independent variables, to fit the flood diversion process at the corresponding floodgate at the current time, such as... Figure 6 As shown.

[0106] VI. Forward and Reverse Calculations

[0107] The objective function is to minimize the peak discharge and the flood diversion volume in the floodplain. Constraints include the highest and lowest reservoir water levels, water level fluctuations, maximum reservoir discharge, and maximum flow capacity of key river sections. The optimization model iteratively optimizes multiple proposed solutions using the piecewise Muskingan model and the flood diversion model at the floodplain outlet.

[0108] The optimal joint scheduling scheme is input into a two-dimensional hydrodynamic model for forward calculation, providing more accurate results for scheduling decisions, such as... Figure 7 As shown. If the expected result is not met, repeat the forward and reverse calculations and adjust the constraint values.

[0109] In summary, the forward and inverse calculation method for joint flood control optimization scheduling of reservoirs and floodplains provided by this invention fully considers the impact of dynamic reservoir capacity caused by uneven water surface during flood evolution, thus improving the accuracy of inflow calculation. This invention uses minute-level measured water level data for data assimilation, making full use of measured data information and effectively reconstructing the flood peak process. Furthermore, the derived inflow process exhibits small fluctuations, good continuity, and a fine time scale, providing better guidance for reservoir scheduling.

[0110] By adopting the above-disclosed technical solution of this invention, the following beneficial effects are obtained:

[0111] This invention provides a forward and inverse calculation method for the joint flood control optimization scheduling of reservoirs and floodplains. It fully considers the relationship between water level and flow rate during flood evolution, fitting the outlet flow process using the current and previous river flow rates. This eliminates the need for hydraulic calculations of river level, allowing the use of the Muskingan model to simulate the river flow process and improve calculation speed. Combined with optimization algorithms, the optimal scheduling result under given constraints is calculated in reverse. Finally, a forward two-dimensional hydraulic model is used to calculate the floodplain inundation process and the floodplain outlet river flow process. This method saves significant time, is easy to implement, and provides technical support for watershed flood management.

[0112] 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 forward and inverse calculation method for joint flood control optimization scheduling of reservoirs and floodplains, characterized in that: Includes the following steps, S1. Construction of a two-dimensional hydrodynamic model: Based on the collected basic data of reservoirs and floodplains, a two-dimensional hydrodynamic model is constructed and calibrated. S2. Flood diversion data generation in the floodplain: Based on the calibrated two-dimensional hydrodynamic model, simulate the flood diversion flow at the floodplain outlet, the flow in the river channel outside the floodplain outlet, and the flow in the river channel at the key protection section. S3. Construction of Muskingen Model: A segmented Muskingen model was constructed and calibrated based on the river channel simulation results of the two-dimensional hydrodynamic model. S4. Construction of flood diversion model for floodplain: Based on the river flow and flood diversion process simulated by the two-dimensional hydrodynamic model, the flood diversion function of the floodplain is fitted to construct the flood diversion model of the floodplain. S5. Forward and Reverse Calculation: Set the objective function and constraints for scheduling, perform reverse optimization of the scheduling scheme based on the calibrated segmented Muskingen model and the flood diversion model of the floodplain, and obtain the optimal scheduling scheme; then input the reverse optimization result into the calibrated two-dimensional hydrodynamic model to perform forward calculation to obtain the scheduling result of the optimal scheduling scheme; Step S4 specifically involves using a well-defined two-dimensional hydrodynamic model to simulate the river flow process and the flood diversion process in the floodplain. The river flow is considered as a loop curve with the current and previous flow as independent variables and the current outlet flow as the dependent variable. The flood diversion function of the floodplain outlet is fitted based on multiple linear or neural network to obtain the flood diversion model of the floodplain outlet. Step S5 specifically includes the following: S51. Reverse optimization solution: Based on actual scheduling needs, set the objective function and constraints of scheduling, and perform reverse optimization on the segmented Muskingen model and flood diversion model of flood flood area with fixed utilization of multiple scheduling schemes to obtain the optimal reservoir discharge process and gate opening time under given objective function and constraints. S52. Forward simulation calculation: Using the optimal reservoir discharge process and gate opening time obtained from the reverse optimization as inputs to the calibrated two-dimensional hydrodynamic model, the flood flooding process and flood flooding channel flow process under the optimal scheduling scheme are calculated in the forward calculation. S53. Result Judgment: Determine whether the forward calculation result meets the expected requirements. If not, adjust the constraints and re-perform the reverse optimization solution and forward simulation calculation until the forward calculation result meets the expected requirements.

2. The forward and inverse calculation method for joint flood control optimization scheduling of reservoirs and floodplains according to claim 1, characterized in that: Step S1 specifically involves determining the piecewise roughness, simulation time, and initial lake water volume parameters of the two-dimensional hydrodynamic model based on the measured river flow and flood inundation depth of the floodplain. The two-dimensional hydrodynamic model employs a water storage unit method, neglecting non-dominant physical processes and simplifying the solution of the hydrodynamic equations. The change in unit water volume is derived from the cumulative water exchange between units. The flow rates of adjacent units are calculated using an explicit shallow-water inertial scheme, as shown in the formula. ； in, Let t be the unit width flow rate at the cell boundary; Where is the water surface slope, g is the acceleration due to gravity, and n is the roughness coefficient. Critical flow depth; time step Subject to the Courant-Freidrichs-Levy conditions, the formula is as follows: ； in, For the maximum time step, The maximum water depth in the study area, k is the constraint coefficient; The side length of the cell grid; The formula for calculating the change in unit water volume after unit flow exchange is as follows: ； in, , r and c represent the flow rates in the x and y directions, respectively; r and c are the cell row and column indices. Unit water volume; 3. The forward and inverse calculation method for joint flood control optimization scheduling of reservoirs and floodplains according to claim 1, characterized in that: Step S2 specifically involves simulating the discharge process of key protected sections, river sections where the outlets are located, and the flood diversion process of the outlets based on a calibrated two-dimensional hydrodynamic model, and recording the corresponding discharge process and simulated flow process.

4. The forward and inverse calculation method for joint flood control optimization scheduling of reservoirs and floodplains according to claim 1, characterized in that: Step S3 specifically involves constructing and calibrating a segmented Muskingen model of the river channel according to the key protection sections and the river channel sections where the estuaries are located. The segmented Muskingen model estimates flow rate by simulating the retention and movement of water in a river channel. The core of this model is to represent the river channel as a water storage unit, whose outflow is related to the inflow and retention status. Each segment of the channel has two parameters representing retention time and scaling degree; the formula is... ； ； ； ； in, , , These are the weighting coefficients for the corresponding items; This represents the retention time coefficient of the river channel; The degree of flatness of the incoming flow; For time step; and The river inflows at time t and time t+1 are respectively. and River outflow at time t and time t+1, respectively; through-rate and This enables rapid simulation of river floods.

5. The forward and inverse calculation method for joint flood control optimization scheduling of reservoirs and floodplains according to claim 4, characterized in that: The parameters of the segmented Muskingen model are optimized and calibrated using the differential evolution algorithm with the goal of maximizing Nash efficiency.

6. The forward and inverse calculation method for joint flood control optimization scheduling of reservoirs and floodplains according to claim 5, characterized in that: In step S5, the objective function includes minimizing the peak discharge from the reservoir, minimizing the water storage in the floodplain, and minimizing the peak discharge from the floodplain outlet channel. The constraints include the maximum flow rate at key river sections downstream of the reservoir, reservoir water level constraints, and the maximum flood diversion volume in the floodplain.

7. The forward and inverse calculation method for joint flood control optimization scheduling of reservoirs and floodplains according to any one of claims 1 to 6, characterized in that: Step S1 is preceded by, S0. Basic Data Collection: Collect and organize basic data on reservoirs and floodplains, including reservoir information, measured data, two-dimensional topographic data, and floodgate information for flood diversion areas.

8. The forward and inverse calculation method for joint flood control optimization scheduling of reservoirs and floodplains according to claim 7, characterized in that: In step S0, Reservoir information includes reservoir capacity curves, as well as highest water level, lowest water level, and water level fluctuation constraints; The measured data include measured river flow, outlet opening time, outlet size, and flooded area of ​​the floodplain. These data need to come from the same flood event. The two-dimensional topographic data is underwater topography, covering the floodplain, the floodplain, and the connecting waterways before the reservoir; Information on floodgates includes location, depth, and length.

Images