Method for forecasting flood in future session based on peak rainstorm distribution characteristics

Abstract

The invention belongs to the technical field of hydrological forecasting, and particularly provides a method for forecasting future session flood based on peak rainstorm distribution characteristics, and the method comprises the steps: collecting and reorganizing data, and processing the data into a time sequence with equal time intervals; obtaining space-time distribution of the flood peak formation rainstorm amount of each session of the research basin; calculating the rainfall process of each sub-basin of the flood to be forecasted, and analyzing the spatial and temporal distribution characteristics of peak rainstorm; performing similarity analysis on the spatial and temporal distribution characteristics of the peak storm of the to-be-forecasted flood and the historical session flood to obtain a session flood which is highly similar to the spatial and temporal distribution characteristics of the peak storm of the to-be-forecasted flood and is close to the peak storm amount, and taking the session flood as a to-be-calibrated session flood; establishing a multi-objective function; optimizing the objective function to obtain an optimal parameter of the hydrological model; and performing hydrological forecasting on the to-be-forecasted flood. According to the method, the parameters of the hydrological model for forecasting the flood in the future session with high accuracy can be obtained, so that a high-accuracy flood forecasting result is obtained.

Description

Technical Field

[0001] The present invention belongs to the technical field of hydrological forecasting, and in particular, relates to a method for forecasting future floods based on peak rainstorm distribution characteristics. Background Art

[0002] At present, in order to study issues such as reservoir scheduling and flood disaster prevention, the calibration of the parameters of the basin hydrological model is one of the indispensable steps. When calibrating the parameters of the hydrological model, the first step is to select the flood events to be calibrated. The widely used method for selecting flood events to be calibrated is to select representative floods, that is, the selected floods to be calibrated must include medium, small and large floods, and then analyze the advantages and disadvantages of the selected hydrological model parameters by the qualified rate of simulated floods when the hydrological model parameters are driven by the hydrological model parameters. The hydrological model parameters selected by this method can reflect the average runoff law of the study basin, and have certain advantages in the number of qualified forecasts of historical flood events. The optimal parameters selected by this method for hydrological forecasting of future floods are based on the fact that the runoff law of future floods conforms to the average runoff law of the study basin.

[0003] However, due to the difference in peak rainstorms, the size of future floods will also be different, and the corresponding runoff patterns will also be different, and may not necessarily conform to the average runoff patterns of the basin. At this time, the accuracy of using the hydrological model parameters to predict future flood processes is low, and it is necessary to select historical floods with high similarity to the peak rainstorm distribution of future floods as the floods to be rated. Summary of the invention

[0004] The technical problem to be solved by the present invention is to provide a method for predicting future floods based on the distribution characteristics of peak rainstorms, so as to improve the accuracy of the forecast results of future floods.

[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is: a method for predicting future floods based on the distribution characteristics of peak rainstorms, comprising the following steps: S1. Data collection and compilation: Collect the rainfall excerpts from the rainfall stations and hydrological stations in the study basin and the flood element excerpts from the hydrological stations, and process them into time series with equal time intervals; S2. Obtaining the spatiotemporal distribution characteristics of historical flood peak rainstorms: Calculate the rainfall time series of each sub-basin divided in step S1 to obtain the spatiotemporal distribution of the peak rainstorms of each flood in the study basin; S3. Analyze the temporal and spatial distribution characteristics of peak rainstorms in future floods: calculate the rainfall process of each sub-basin of the flood to be predicted, and analyze the temporal and spatial distribution characteristics of peak rainstorms; S4. Screening floods to be rated: Perform similarity analysis on the spatiotemporal distribution characteristics of the peak rainstorms of the predicted floods and historical floods, and obtain floods with high similarity to the spatiotemporal distribution characteristics of the peak rainstorms of the predicted floods and similar peak rainstorm amounts as the floods to be rated; S5. Establishing multi-objective functions: Establishing objective functions for the runoff processes of the historical floods selected in step S5 , ,… , the set of objective functions is the constructed multi-objective function ,in, The number of floods to be rated; S6. Optimize the objective function to obtain the optimal parameters of the hydrological model; S7. Perform hydrological forecast on the flood to be forecasted: substitute the optimal parameters of the hydrological model obtained in step S6 into the hydrological model, input rainfall data of a certain flood into the hydrological model, and forecast the flood process of the certain flood.

[0006] In the preferred scheme, in the step S1, the digital elevation map and land use map of the study basin are collected, and the sub-basins of the study basin are divided through GIS. If there is a hydrological station or reservoir in the study basin, the hydrological station or reservoir is used as the basin exit point of the sub-basin, and Thiessen polygons are divided to determine the influencing rainfall stations and corresponding weights of each sub-basin; the land use map is loaded into GIS, and the impermeability of each sub-basin is calculated year by year; the rainfall excerpt tables of the rainfall stations and hydrological stations in the study basin and the flood element excerpt tables of the hydrological stations are collected, and they are processed into time series with equal time intervals.

[0007] In a preferred solution, in step S4, the correlation coefficient and the certainty coefficient are selected as indicators for similarity analysis of peak rainstorm distribution between historical floods and floods to be forecasted, wherein the correlation coefficient is calculated as follows: (1); The calculation method of the coefficient of certainty is: (2); In the formula, r represents the correlation coefficient, DC represents the coefficient of certainty, It represents the peak rainstorm series of historical floods in each sub-basin; It represents the peak rainstorm series of the flood to be forecasted in each sub-basin; It represents the mean value of the peak rainstorm series of historical floods in each sub-basin; It represents the mean of the peak rainstorm series of the flood to be forecast in each sub-basin.

[0008] In the preferred scheme, in step S5, the objective function of each flood is determined according to the fitting effect of a single flood, and the same type or different types of objective functions are set. The objective function adopts a peak weighted root mean square error function or a residual sum of squares function or a comprehensive objective function that comprehensively considers the impact of flood peaks and flood volumes.

[0009] In a preferred solution, the peak weighted root mean square error function is: (3); In formula (3), Indicates i Objective function of flooding; For the i The number of ordinates of the flow process line of the flood; The measured flood was j Traffic flow at the moment; The output calculated with the selected model parameters is given by i The flood in j Traffic flow at the moment; It is the average value of the measured flood flow.

[0010] In a preferred solution, the residual square sum function is: (4).

[0011] In a preferred solution, the comprehensive function that comprehensively considers the impact of flood peak and flood volume is: (5); In the formula, The first j The weight corresponding to each traffic sample point; in Calculated by formula (6): (6); In the formula, For flood j The runoff volume in each period; is the total runoff of the flood; in, Calculated by formula (7): (7); In the formula, The time interval for the runoff data with equal time intervals; Calculated by formula (8): (8).

[0012] In the preferred scheme, in step S6, if the number of multi-objective functions established in step S5 is one, the objective function is optimized by a single-objective optimization algorithm, and the optimization result is directly used as the optimal solution for the hydrological model parameters; if the number of multi-objective functions established in step S5 is two or three, the objective function is optimized by a multi-objective optimization algorithm; if the number of multi-objective functions established in step S5 is more than three, the objective function is optimized by a super-multi-objective large-scale optimization algorithm.

[0013] In the preferred solution, the single-objective optimization algorithm uses a pattern search method, the multi-objective algorithm uses a DGEA algorithm, and the super-multi-objective large-scale optimization algorithm uses a LMOCSO algorithm.

[0014] In the preferred solution, when the multi-objective optimization algorithm or the super-multi-objective large-scale optimization algorithm is used to optimize the objective function, each group of parameters in the Pareto solution obtained after optimization should be brought into the hydrological model, the flood to be rated should be simulated, and the hydrological forecast accuracy should be evaluated respectively. The final optimal solution of the hydrological model parameters is determined by combining the qualified rate with the Euclidean distance of the Pareto solution. The calculation method of the Euclidean distance is as follows: = (9); In the formula, Indicates i The Euclidean distance of the Pareto solutions; Indicates i The vector of points of the Pareto solutions; A vector representing a reference point; represents the dimension of the Pareto solution vector; Indicates i The first Pareto solution j The value of the dimension; Indicates the reference point j The value of a dimension.

[0015] The method provided by the present invention for predicting future floods based on peak rainstorm distribution characteristics has the following beneficial effects: according to the similarity between the peak rainstorm distribution characteristics of future floods and the peak rainstorm distribution characteristics of historical floods, the flood events similar to the future flood runoff law are determined as the flood events to be calibrated, multiple floods are calibrated simultaneously through a multi-objective large-scale optimization algorithm, and the optimal hydrological model parameters are determined from the Pareto solution of the hydrological model parameters by using the qualified rate combined with the Euclidean distance method. The present invention can still obtain high-precision future flood forecast results even when the technicians' flood forecasting experience and computer computing performance are not very good. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The present invention will be further described below in conjunction with the accompanying drawings and embodiments: Figure 1 This is the model map of the Xiagushan watershed; Figure 2 Thiessen polygons for the Lower Lone Mountain watershed; Figure 3 is the Pareto solution of the hydrological model parameters; Figure 4 is the optimal parameter of the hydrological model; Figure 5 This is a comparison chart of the simulated flow and measured flow of the 19940703 flood; Figure 6 This is a comparison chart of the simulated flow and measured flow of the 19950821 flood; Figure 7 This is a comparison chart of the simulated flow and measured flow of the 19960804 flood; Figure 8 This is a comparison chart of the simulated flow and measured flow of the 20000716 flood; Fig. 9 This is a comparison chart of the simulated flow and measured flow of the 20020627 flood; Fig.10 This is the predicted flow process line for flood No. 20130526. DETAILED DESCRIPTION

[0017] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0018] A method for predicting future floods based on peak rainstorm distribution characteristics, comprising the following steps: S1. Data collection and compilation: Collect rainfall excerpts from rain gauges and hydrological stations in the study basin, as well as flood element excerpts from hydrological stations, and process them into time series with equal time intervals.

[0019] Collect digital elevation maps and land use maps of the study basin, divide the study basin into sub-basins through GIS, and use the hydrological station or reservoir as the basin exit point of the sub-basin if there is a hydrological station or reservoir in the study basin, divide the Thiessen polygon, determine the influencing rainfall stations and corresponding weights of each sub-basin; load the land use map in GIS, calculate the imperviousness of each sub-basin year by year; collect the rainfall excerpts of the rainfall stations and hydrological stations in the study basin and the flood element excerpts of the hydrological stations, and process them into time series with equal time intervals. Determine the influencing rainfall stations and corresponding weights of each sub-basin, and then calculate the rainfall series of each sub-basin through the Thiessen polygon method; calculate the imperviousness of each sub-basin year by year as the input parameter of the hydrological model.

[0020] S2. Obtain the spatiotemporal distribution characteristics of historical flood peak rainstorms: Calculate the rainfall time series of each sub-basin divided in step S1 to obtain the spatiotemporal distribution of each flood peak rainstorm in the study basin. Specifically, the rainfall time series of each sub-basin is calculated by the Thiessen polygon method.

[0021] S3. Analyze the temporal and spatial distribution characteristics of peak rainstorms in future floods: Calculate the rainfall process of each sub-basin of the flood to be forecasted, and analyze the temporal and spatial distribution characteristics of peak rainstorms. Specifically, the rainfall process of each sub-basin of the flood to be forecast is calculated by the Thiessen polygon method. Analyze the total amount of rainstorms in each sub-basin, the shape of the rainstorm process line, and the time and location of the rain peak.

[0022] S4. Screening floods to be calibrated: Conduct similarity analysis on the spatiotemporal distribution characteristics of the peak rainfall of the forecasted floods and historical floods, and obtain floods with high similarity in spatiotemporal distribution characteristics with the peak rainfall of the forecasted floods and similar peak rainfall as the floods to be calibrated.

[0023] The correlation coefficient and the certainty coefficient are selected as the indicators for the similarity analysis of the peak rainstorm distribution of historical floods and floods to be forecasted. The calculation method of the correlation coefficient is: (1); The calculation method of the coefficient of certainty is: (2); In the formula, r represents the correlation coefficient, DC represents the coefficient of certainty, It represents the peak rainstorm series of historical floods in each sub-basin; It represents the peak rainstorm series of the flood to be forecasted in each sub-basin; It represents the mean value of the peak rainstorm series of historical floods in each sub-basin; It represents the mean of the peak rainstorm series of the flood to be forecast in each sub-basin.

[0024] S5. Establishing multi-objective functions: Establishing objective functions for the runoff processes of the historical floods selected in step S5 , ,… , the set of objective functions is the constructed multi-objective function ,in, The number of floods to be rated.

[0025] The objective function of each flood is determined according to the fitting effect of a single flood. The same or different types of objective functions are set. The objective function adopts the peak weighted root mean square error function or the residual sum of squares function or a comprehensive objective function that comprehensively considers the impact of flood peak and flood volume.

[0026] The peak weighted root mean square error function is: (3); In formula (3), Indicates i Objective function of flooding; For the i The number of ordinates of the flow process line of the flood; The measured flood was j Traffic flow at the moment; The output calculated with the selected model parameters is given by i The flood in j Traffic flow at the moment; It is the average value of the measured flood flow.

[0027] The residual sum of squares function is: (4).

[0028] The comprehensive function that comprehensively considers the impact of flood peak and flood volume is: (5); In the formula, The first j The weight corresponding to each traffic sample point; in Calculated by formula (6): (6); In the formula, For flood j The runoff volume in each period; is the total runoff of the flood; in, Calculated by formula (7): (7); In the formula, The time interval for the runoff data with equal time intervals; Calculated by formula (8): (8).

[0029] S6. Optimize the objective function to obtain the optimal parameters of the hydrological model.

[0030] If the number of multi-objective functions established in step S5 is one, the objective function is optimized by a single-objective optimization algorithm, and the optimization result is directly used as the optimal solution for the hydrological model parameters; if the number of multi-objective functions established in step S5 is two or three, the objective function is optimized by a multi-objective optimization algorithm; if the number of multi-objective functions established in step S5 is more than three, the objective function is optimized by a super-multi-objective large-scale optimization algorithm.

[0031] The single-objective optimization algorithm adopts, for example, the pattern search method; the multi-objective algorithm adopts, for example, the DGEA algorithm (Directionguided evolutionary algorithm); the super-multi-objective large-scale optimization algorithm adopts, for example, the LMOCSO algorithm (Large-scale multi-objective competitive swarm optimization algorithm).

[0032] When the multi-objective optimization algorithm or the super-multi-objective large-scale optimization algorithm is used to optimize the objective function, each group of parameters in the Pareto solution obtained after optimization should be brought into the hydrological model, the flood to be rated should be simulated, the hydrological forecast accuracy should be evaluated respectively, and the final optimal solution of the hydrological model parameters should be determined by the qualified rate combined with the Euclidean distance of the Pareto solution. Specifically, when the multi-objective optimization algorithm or the super-multi-objective large-scale optimization algorithm is used to optimize the objective function, the calculation result of the optimization algorithm is the Pareto solution of the hydrological model parameters; the accuracy of the hydrological forecast is evaluated from three aspects: the peak flow error, the flood volume error and the peak time; the hydrological model parameters that make the hydrological model simulate the highest qualified rate of floods are selected as the optimal solution of the hydrological model. When the qualified rates of the simulated floods of each group of parameters in the Pareto solution are the same, the Euclidean distance method is used to determine the optimal solution of the hydrological model parameters, that is, the solution with the smallest Euclidean distance is selected as the final optimal solution of the hydrological model parameters.

[0033] The calculation method of Euclidean distance is as follows: = (9); In the formula, Indicates i The Euclidean distance of the Pareto solutions; Indicates i The vector of points of the Pareto solutions; A vector representing a reference point; represents the dimension of the Pareto solution vector; Indicates i The first Pareto solution j The value of the dimension; Indicates the reference point j The value of a dimension.

[0034] S7. Perform hydrological forecast on the flood to be forecasted: substitute the optimal parameters of the hydrological model obtained in step S6 into the hydrological model, input rainfall data of a certain flood into the hydrological model, and forecast the flood process of the certain flood.

[0035] Implementation Cases like Figures 1 to 10 As shown, this embodiment takes the future flood forecast in the basin above the Xiagushan hydrological station in the Huaihe River basin as an example to demonstrate the effect achieved by the present invention.

[0036] Xiagushan Hydrological Station is a hydrological station in the upper reaches of the Huaihe River, located on the Dangze River, a tributary of the Shahe River, with a controlled basin area of ​​358.6 . The basin has abundant groundwater, a wide distribution of deciduous broad-leaved forests, and a vegetation coverage rate of about 90%. The basin is located in the continental monsoon climate zone, and local or large-scale rainstorms often occur. The average annual precipitation is 800 mm, of which 2 / 3 are concentrated in the flood season; the flood forecast period in the basin is short, the suddenness is strong, and the flood peaks rise and fall sharply. There are 5 rain gauges, including Beizi, Jinghe, Tumen, Wawu, and Yeping, and 6 measuring stations, including Xiagushan Hydrological Station, in the Xiagushan Basin. Among them, Xiagushan Station has both water volume observation and flow observation tasks. In this embodiment, the floods that occurred from 8:00 on October 8, 1980 to 8:00 on August 15, 2012 are used as historical floods to calibrate the parameters of the hydrological model, and the 20130526 flood that occurred in 2013 is used as a future flood to predict the flood process. The steps of the method for predicting future floods based on the distribution characteristics of peak rainstorms in this embodiment are as follows: S1: Collect digital elevation maps and land use maps of the Xiagushan Basin. Use GIS to divide the sub-basins of the basin. The sub-basins of the basin are shown in the following figure. Figure 1 As shown. Divide the Thiessen polygons to determine the influencing rainfall stations and corresponding weights of each sub-basin. The divided Thiessen polygons are as follows Figure 2As shown, the weights of each influencing rainfall station in each sub-basin are shown in Table 1; the land use map in S1 is loaded into GIS, and the imperviousness of each sub-basin is calculated year by year. The imperviousness of each sub-basin each year is shown in Table 2; the rainfall excerpts of the rainfall stations and hydrological stations in the study basin and the flood element excerpts of the hydrological stations are collected and processed into hourly time series.

[0037]

[0038]

[0039] S2: Calculate the rainfall time series of each sub-basin divided in step S1, so as to obtain the temporal and spatial distribution of the peak rainstorms of each flood in the study basin. Taking the 19950821 flood as an example, the time series of the peak rainstorms of this flood in each sub-basin is shown in Table 3:

[0040] S3: The temporal and spatial distribution of future floods 20130526 is shown in Table 4:

[0041] S4: Through the analysis of the peak rainfall, it is concluded that the peak size of the flood to be calibrated should be between 1100-2530m 3 / s; The similarity analysis of the peak rainstorm of each historical flood and the 20130526 flood is shown in Table 5. The floods whose peaks meet the requirements and whose peak rainstorm process correlation is greater than 0 are selected from the table for calibration. It can be concluded that the five floods of 19940703, 19950821, 19960804, 20000716, and 20020627 meet the requirements, so these five floods are taken as the floods to be calibrated.

[0042]

[0043] S5: The objective function selects the peak weighted root mean square error function.

[0044] S6: The multi-objective large-scale optimization algorithm LMOCSO is used to optimize the multi-objective function, and the Pareto solution of the parameters is as follows: Figure 3 The optimal parameters of the hydrological model are shown in Figure 4 The simulated flow rate and measured flow rate curves are shown in Figure 5~Figure 9 shown.

[0045] S7: The flow process line of the predicted flood of 20130526 is as follows Fig.10 shown.

[0046] In the practical application of the present invention, the specific flow process of the predicted flood is unknown at the time of forecasting. However, if you want to illustrate the effect, you can compare the measured flow process of the flood. Compared with the measured flow process of the 20130526 flood, the relative error of the predicted flood peak flow is 13.52%, and the flood peak forecast is qualified; the peak time error is 1 hour, and the peak time forecast is qualified. It will be easily understood by those skilled in the art that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for predicting future floods based on peak rainstorm distribution characteristics, characterized in that: The following steps are involved: S1. Data collection and compilation: Collect the rainfall excerpts from the rainfall stations and hydrological stations in the study basin and the flood element excerpts from the hydrological stations, and process them into time series with equal time intervals; S2. Obtaining the spatiotemporal distribution characteristics of historical flood peak rainstorms: Calculate the rainfall time series of each sub-basin divided in step S1 to obtain the spatiotemporal distribution of the peak rainstorms of each flood in the study basin; S3. Analyze the temporal and spatial distribution characteristics of peak rainstorms in future floods: calculate the rainfall process of each sub-basin of the flood to be predicted, and analyze the temporal and spatial distribution characteristics of peak rainstorms; S4. Screening floods to be rated: Perform similarity analysis on the spatiotemporal distribution characteristics of the peak rainstorms of the predicted floods and historical floods, and obtain floods with high similarity to the spatiotemporal distribution characteristics of the peak rainstorms of the predicted floods and similar peak rainstorm amounts as the floods to be rated; S5. Establishing multi-objective functions: Establishing objective functions for the runoff processes of the historical floods selected in step S5 , ,… , the set of objective functions is the constructed multi-objective function ,in, The number of floods to be rated; S6. Optimize the objective function to obtain the optimal parameters of the hydrological model; S7. Perform hydrological forecast on the flood to be forecasted: substitute the optimal parameters of the hydrological model obtained in step S6 into the hydrological model, input rainfall data of a certain flood into the hydrological model, and forecast the flood process of the certain flood.

2. A method for predicting future floods based on peak rainstorm distribution characteristics according to claim 1, characterized in that: In the step S1, a digital elevation map and a land use map of the study basin are collected, and the sub-basins of the study basin are divided through GIS. If there is a hydrological station or a reservoir in the study basin, the hydrological station or the reservoir is used as the basin exit point of the sub-basin, and Thiessen polygons are divided to determine the influencing rainfall stations and corresponding weights of each sub-basin; the land use map is loaded into GIS, and the impermeability of each sub-basin is calculated year by year; the rainfall excerpt tables of the rainfall stations and hydrological stations in the study basin and the flood element excerpt tables of the hydrological stations are collected, and they are processed into time series with equal time intervals.

3. The method for predicting future floods based on peak rainstorm distribution characteristics according to claim 1, characterized in that: In step S4, the correlation coefficient and the certainty coefficient are selected as indicators for similarity analysis of peak rainstorm distribution between historical floods and floods to be forecasted, wherein the correlation coefficient is calculated as follows: (1)； The calculation method of the coefficient of certainty is: (2)； In the formula, r represents the correlation coefficient, DC represents the coefficient of certainty, It represents the peak rainstorm series of historical floods in each sub-basin; It represents the peak rainstorm series of the flood to be forecasted in each sub-basin; It represents the mean value of the peak rainstorm series of historical floods in each sub-basin; It represents the mean of the peak rainstorm series of the flood to be forecast in each sub-basin.

4. A method for predicting future floods based on peak rainstorm distribution characteristics according to claim 1, characterized in that: In step S5, the objective function of each flood is determined according to the fitting effect of a single flood, and the same type or different types of objective functions are set. The objective function adopts a peak weighted root mean square error function or a residual sum of squares function or a comprehensive objective function that comprehensively considers the impact of flood peaks and flood volumes.

5. A method for predicting future floods based on peak rainstorm distribution characteristics according to claim 4, characterized in that: The peak weighted root mean square error function is: (3)； In formula (3), Indicates i Objective function of flooding; For the i The number of ordinates of the flow process line of the flood; The measured flood was j Traffic flow at the moment; The output calculated with the selected model parameters is given by i The flood in j Traffic flow at the moment; It is the average value of the measured flood flow.

6. A method for predicting future floods based on peak rainstorm distribution characteristics according to claim 4, characterized in that: The residual sum of squares function is: (4)。 7. A method for predicting future floods based on peak rainstorm distribution characteristics according to claim 4, characterized in that: The comprehensive function that comprehensively considers the impact of flood peak and flood volume is: (5)； In the formula, The first j The weight corresponding to each traffic sample point; in Calculated by formula (6): (6)； In the formula, For flood j The runoff volume in each period; is the total runoff of the flood; in, Calculated by formula (7): (7)； In the formula, The time interval for the runoff data with equal time intervals; Calculated by formula (8): (8)。 8. The method for predicting future floods based on peak rainstorm distribution characteristics according to claim 1, characterized in that: In step S6, if the number of multi-objective functions established in step S5 is one, the objective function is optimized by a single-objective optimization algorithm, and the optimization result is directly used as the optimal solution of the hydrological model parameters; If the number of multi-objective functions established in step S5 is two or three, a multi-objective optimization algorithm is used to optimize the objective function; if the number of multi-objective functions established in step S5 is more than three, a super-multi-objective large-scale optimization algorithm is used to optimize the objective function.

9. A method for predicting future floods based on peak rainstorm distribution characteristics according to claim 8, characterized in that: The single-objective optimization algorithm uses the pattern search method, the multi-objective algorithm uses the DGEA algorithm, and the super-multi-objective large-scale optimization algorithm uses the LMOCSO algorithm.

10. The method for predicting future floods based on peak rainstorm distribution characteristics according to claim 8, characterized in that: When the multi-objective optimization algorithm or the super-multi-objective large-scale optimization algorithm is used to optimize the objective function, each group of parameters in the Pareto solution obtained after optimization should be brought into the hydrological model, the flood to be rated should be simulated, and the hydrological forecast accuracy should be evaluated respectively. The optimal solution of the hydrological model parameters can be determined by combining the qualified rate with the Euclidean distance of the Pareto solution. The calculation method of the Euclidean distance is as follows: = (9)； In the formula, Indicates i The Euclidean distance of the Pareto solutions; Indicates i A vector of Pareto solution points; A vector representing a reference point; represents the dimension of the Pareto solution vector; Indicates i The first Pareto solution j The value of the dimension; Indicates the reference point j The value of a dimension.

Images