CN118095562A - Lake flood end water storage strategy intelligent optimization method based on hydrologic hydrodynamic model - Google Patents

Abstract

The invention discloses an intelligent optimization method for a lake water storage strategy at the tail of a flood, which is based on a hydrological hydrodynamic model and comprises the following steps: acquiring a water storage strategy sample library containing water storage time and water storage scheduling rules, inputting the water storage strategy sample library into a hydrological hydrodynamic model which is jointly applied to a pre-constructed lake and a water storage diapause area thereof, and acquiring a three-dimensional mapping relation of the water storage time, the water storage scheduling rules and evaluation indexes of each water storage strategy sample; extracting characteristic value data of an evaluation index in the three-dimensional mapping relation by adopting a KNN nearest neighbor algorithm according to the three-dimensional mapping relation; and taking the extracted evaluation index characteristic value data as a calculation basis, and adopting an NSGA-II intelligent optimization algorithm to perform optimal solution search iteration on the water storage time and the water storage scheduling rule so as to obtain an optimal water storage strategy. The method can realize the refined simulation of the flood evolution process and the flooding dynamics, can effectively avoid the local optimal solution, and can further improve the water storage efficiency while reducing the lake flood control risk.

Description

Lake flood end water storage strategy intelligent optimization method based on hydrologic hydrodynamic model

Technical Field

The application belongs to the technical field of flood reclamation, relates to water resource management, flood prediction, prevention and control and optimal scheduling of water resources, and in particular relates to an intelligent optimization method for a lake flood end water storage strategy based on a hydrologic hydrodynamic model.

Background

The water resource time course distribution in the semi-arid and semi-humid areas of China is extremely uneven under the influence of the monsoon climate. In recent years, with the rapid development of economy and the continuous increase of social water demand, the contradiction between flood control and disaster reduction difficulties in flood season and water resource shortage in non-flood season in a plurality of watersheds is more and more prominent. The lake is used as a multifunctional regulation reservoir for 'accumulation and drainage and raising', is a key node for causing upstream waterlogging caused by flood, and also is an important task for carrying water resource supply. Therefore, on the premise of ensuring the safety of lake embankment and downstream flood control, how to scientifically and reasonably excavate flood resources of the lake basin and realize the transition from 'disaster water' to 'resource water' is a hot spot problem of the current research.

At present, the water level stage application of flood limit and the advanced water storage scheduling of flood end become the common means of "flood recycling", compared with the traditional reservoir water storage scheduling, the response degree of the hydrologic elements of the lake to the runoff components of the lake entering different areas is more sensitive, and meanwhile, the combined application of different discharging channels of the lake is also not negligible to the flood prevention risks possibly brought by the reservoir flood areas around the lake area, so that the hydrologic hydrodynamic model numerical simulation is generally adopted to replace the reservoir flood control calculation process in the production practice process.

However, the hydrographic hydrodynamic model is generally only used for high-precision numerical simulation calculation, and when coordination of complex river network topological relations of lakes and watercourses and different levels of flood control and drainage engineering scheduling is faced, a target-oriented optimization decision model needs to be further constructed if intelligent comprehensive management and scheme decision of water resources are to be realized.

Disclosure of Invention

The invention aims to overcome the defects in the prior art, and provides an intelligent optimization method for a lake flood end water storage strategy based on a hydrologic hydrodynamic model, which is used for organically combining a flood evolution model with engineering optimization scheduling by coupling the hydrologic hydrodynamic model with a multi-objective optimization algorithm, so as to realize the fine simulation of water flow in a local complex river network area and the comprehensive treatment of water resources.

In order to solve the technical problems, the invention is realized by adopting the following technical scheme:

in a first aspect, the invention provides an intelligent optimization method for a lake water storage strategy at the tail of a flood, based on a hydrological hydrodynamic model, comprising the following steps:

acquiring a water storage strategy sample library containing water storage time and water storage scheduling rules, inputting the water storage strategy sample library into a hydrological hydrodynamic model which is jointly applied to a pre-constructed lake and a water storage diapause area thereof, and acquiring a three-dimensional mapping relation of the water storage time, the water storage scheduling rules and evaluation indexes of each water storage strategy sample;

extracting characteristic value data of an evaluation index in the three-dimensional mapping relation by adopting a KNN nearest neighbor algorithm according to the three-dimensional mapping relation;

And taking the extracted evaluation index characteristic value data as a calculation basis, and adopting an NSGA-II intelligent optimization algorithm to perform optimal solution search iteration on the water storage time and the water storage scheduling rule so as to obtain an optimal water storage strategy.

Further, the water storage strategy sample library containing water storage time and water storage scheduling rules comprises:

The scheme set of water storage opportunity in the later flood season: t= { T ₀+t,T₀+2t,…,T_e };

The water storage time is the starting time of water storage scheduling, T is the flood duration corresponding to 95% frequency of the annual flood duration frequency discharge fitting curve, T ₀ is the first water storage time, T _e is the last water storage time, and T _e is smaller than the end date of the later flood period;

A water storage scheduling rule scheme set in the later flood period:

The water storage scheduling rule refers to a permutation and combination form of a plurality of down-leakage gates for combining the down-leakage flow under different water level conditions;

j is a down-flow gate, j=1, 2, …, M; i is a water storage scheme, i=1, 2, …, N;

For the downward leakage flow rate of the ith scheme of the jth gate when the lake water level is at the flood limit water level,/> The method is characterized in that the method is the discharging flow of the jth gate in the ith scheme when the lake water level is at the normal water storage level; /(I)Q _u、Q_d is the upper and lower limits of the total amount of leakage respectively;

And (3) arranging and combining the water storage opportunity scheme set and the water storage scheduling rule scheme set to obtain a water storage strategy sample library.

Furthermore, the hydrologic hydrodynamic model for combined use of the lake and the flood storage area thereof is constructed by the following steps:

Based on MIKE11, a one-dimensional river network hydrodynamic model is established, and the basic equation is as follows:

Wherein: x and t are space coordinates and time coordinates, respectively; z is the average water level of the section; q is the section flow; a is the cross-sectional area; c is the thank you coefficient; q is the lateral inflow; g is gravity acceleration; r is the hydraulic radius; alpha is a momentum correction coefficient; |q| is the absolute value of Q;

based on MIKE, a two-dimensional flood area hydrodynamic model is established, and the basic equation is as follows:

Wherein: x and y are Cartesian coordinate systems; t is time; h is the water depth; u and v are average flow velocity in x and y directions respectively; s is a source sink item; g is gravity acceleration; η is the elevation of the river bottom; ρ is the density of water; ρ ₀ is the relative density of water; p _a is atmospheric pressure; τ _ax、τ_ay is wind load force ,τ_ax＝ρ_aC_Ds|ω_ax|ω_ax,τ_ay＝ρ_aC_Ds|ω_ay|ω_ay;ρ_a is air density; omega _ax、ω_ay is the wind speed at 10m above the water surface; c _Ds is the drag coefficient; τ _bx、τ_by is the resistance of the river bed, N is the river bed roughness; f is the Coriolis force coefficient,/>Omega is the angular velocity of the earth rotation 0.729 x 10 ^-4,Is latitude; s _xx、s_xy、s_yx、s_yy is the radiation stress component; t _xx、T_xy、T_yx、T_yy is the horizontal viscous stress component; u _s、v_s is the source sink stream flow rate.

Establishing a one-dimensional and two-dimensional coupled flood evolution model based on MIKEFLOOD, including:

Establishing standard connection, mapping connecting lines to boundaries of one or more grid units to form coupled lines, providing boundary flow for a two-dimensional model by a one-dimensional model, and returning an average water level value on the coupled lines to the one-dimensional model by the two-dimensional model;

Establishing lateral connection, connecting grid cells of the two-dimensional model to a river reach of the one-dimensional model from the side, determining coupling lines by coordinates, wherein the grid cells mapped to the coupling lines need to participate in coupling, and calculating the water flow through the lateral connection by adopting a flow formula of the hydraulic building.

Furthermore, the hydrological hydrodynamic model for the combined application of the lakes and the flood storage areas of the lakes further comprises the steps of carrying out parameter calibration and verification on the river course roughness of the one-dimensional river network hydrodynamic model and the mat surface roughness of the two-dimensional flood storage areas under the hydrodynamic model:

and selecting actual measurement rainfall runoff and water level data in the live year, comparing the average water level difference absolute value and the maximum water level difference absolute value of the actual measurement water level and the simulated water level of the hydrologic station in the research area, and observing whether the water level change process and the peak time are basically consistent.

Furthermore, the hydrologic hydrodynamic model for combined use of the lake and the flood storage areas thereof further comprises input conditions:

The input conditions of the one-dimensional river network hydrodynamic model comprise river network files, section files, boundary files and parameter files;

The input conditions of the two-dimensional flooding area hydrodynamic model comprise a mesh irregular triangular mesh file, a roughness file and an initial water depth file;

2. The input conditions of the two-dimensional coupled flood model comprise a link mode.

Further, the method includes the steps of obtaining a water storage strategy sample library containing water storage time and water storage scheduling rules, inputting the water storage strategy sample library into a hydrologic hydrodynamic model for combined application of a pre-constructed lake and a water storage diapause area thereof, and obtaining a three-dimensional mapping relation of the water storage time, the water storage scheduling rules and evaluation indexes of each water storage strategy sample, wherein the three-dimensional mapping relation comprises the following steps:

inputting a water storage strategy sample library into a constructed hydrologic hydrodynamic model, and simulating a flood process in a later flood season:

Z_t＝z_i,i＝t_end (9)

Wherein: r _e is the water level overrun of the lake representative hydrologic station in the water storage process of the later flood season, Z _max is the highest water storage level of the lake representative hydrologic station in the water storage process of the later flood season, W _d is the total amount of abandoned water of the lake representative hydrologic station in the water storage process of the later flood season, Z _t is the water storage level of the lake representative hydrologic station at the flood end in the water storage process of the later flood season, wherein two items R _e and Z _max reflect flood control indexes, and two items W _d and Z _t reflect water storage indexes;

t _i is the time period number that the lake water level exceeds the flood limit water level in the later flood period; z _i is the mean lake level of the ith period; t ₁ is the time period number of the later flood period; t ₂ is the time period number of the main flood season; q _ij is the average down-flow rate of the jth gate in the ith period; t _end is the end date of the later flood season; z _l is a flood limit water level process line;

Forming a time-quantity-scale three-dimensional mapping relation and an image of a water storage time (time), a water storage scheduling rule (quantity) and an evaluation index (scale) by adopting a uniform distribution method

Further, according to the three-dimensional mapping relation, extracting characteristic value data of an evaluation index in the three-dimensional mapping relation by adopting a KNN nearest neighbor algorithm, wherein the method comprises the following steps of:

Performing data interpolation on the acquired index data based on a KNN nearest neighbor algorithm to acquire data samples of flood control indexes and water storage indexes under any water storage time and water storage scheduling rules; the Euclidean distance formula between two water storage strategy samples is as follows:

Wherein: d is the Euclidean distance between two samples; t ₁、t₂ is the start-up time; q ₁、q₂ is the total amount of leakage.

Furthermore, the method uses the extracted characteristic value data of the evaluation index as a calculation basis, adopts an NSGA-II intelligent optimization algorithm to perform optimal solution search iteration on the water storage time and the water storage scheduling rule to obtain an optimal water storage strategy, and comprises the following steps:

setting an objective function as a flood control objective and a water storage objective;

Setting decision variables as water storage time and total drainage;

setting constraint conditions;

performing optimal solution search iteration by adopting an NSGA-II intelligent optimization algorithm to obtain a non-inferior leading edge curve between two targets of a flood end water storage strategy problem;

And deriving the Pareto front fitting function to obtain a displacement relation curve between the flood control target and the water storage target, thereby obtaining an optimal water storage strategy.

Further, the objective function is:

Wherein: f ₁ is a flood control target; for the t day, discharging the highest water storage level when the total amount is q; z _max、Z_minZ_min is the upper and lower limits Z ^max-gap＝Z_max-Z_min of the highest water storage level under all possible water storage strategies; /(I) For the t day, the water level overrun when the total amount of the leakage is q; /(I)The upper limit and the lower limit of the water level overrun rate are used for all the feasible water storage schemes; alpha ₁、β₁ is a weight factor of the flood control index;

f ₂ is a water storage target; For the t day, discharging the flood end water storage level when the total amount is q; z' _tZ_target is the target water level of the water storage; z _t,max,/> Upper and lower limits of last flood water level for all feasible water storage strategies For the t day, discharging the water discarding amount when the total amount is q; w _d,max,/>Upper and lower limits for water reject volume for all viable water storage strategiesAlpha ₂、β₂ is a weight factor of the water storage index;

The constraint conditions are as follows:

T_min≤t≤T_max (13)

Q_min≤q≤Q_max (14)

Z_max,tq≤Z_max≤Z_w (15)

Z_r≤Z_t,tq≤Z′_r (16)

R_min≤R_e,tq≤R_max (17)

wherein: equation 13 is a water storage time constraint, equation 14 is a control flow constraint, equation 15 is a highest water level constraint, equation 16 is a target water level constraint, and equation 17 is a water level overrun constraint;

t _min、T_max is the start and stop time of the later flood period respectively; q _max、Q_min is the maximum value and the minimum value of the total leakage amount of the gate when the lake water level reaches the normal water storage level respectively; z _w is the flood-stagnation warning water level of the broken leves; z _r、Z'_r is the flood limit water level before and after the flood period stage control adjustment respectively; r _max、R_min is the upper limit and the lower limit of the water level overrun constraint respectively.

Taking the water storage time and the total drainage amount as decision variables, taking the data extracted by the time-quantity-standard three-dimensional mapping relation as the calculation basis of the targets, carrying out optimal solution search iteration by adopting an NSGA-II intelligent optimization algorithm to obtain a non-inferior leading edge curve between the two targets of the flood end water storage strategy problem, deriving a Pareto leading edge fitting function to obtain a displacement relation curve between the flood control target and the water storage target, and further comparing and selecting the water storage strategy to obtain the optimal water storage strategy.

In a second aspect, the present invention provides a computer storage medium, on which a computer program is stored, which when executed by a processor, implements the above-mentioned intelligent optimization method for a lake flood and end water storage strategy based on a hydrographic model.

Compared with the prior art, the invention has the beneficial effects that:

(1) The plain lake hydrologic model is coupled with the multi-target genetic algorithm, so that the flood evolution process and the submerged dynamic can be finely simulated, and the accurate assessment of the flood resource utilization risk and benefit is facilitated;

(2) The multi-objective optimizing strategy based on the intelligent algorithm can effectively avoid the local optimal solution, further improve the water storage efficiency while reducing the flood control risk of the lake, further improve the reliability of optimizing and regulating, and has good effect on solving the problem of maximizing the flood resources.

Drawings

FIG. 1 is a flow chart of an intelligent optimization method for a lake water storage strategy at the tail of a flood, based on a hydrological hydrodynamic model;

FIG. 2 is a schematic view of a basin according to an embodiment of the present invention;

3-1 to 3-3 are schematic diagrams of MIKE one-dimensional river network hydrodynamics models;

FIGS. 4-1 to 4-3 are schematic diagrams of a MIKE two-dimensional flood area hydrodynamic model;

FIG. 5 is a schematic diagram of a MIKE FLOOD one-dimensional coupled flood model;

FIGS. 6-1 to 6-10 are schematic diagrams of simulated water level errors for hydrokinetic models;

7-1 to 7-2 are three-dimensional mapping relation diagrams of water storage time, water storage scheduling rules and flood control risk indexes;

8-1 to 8-2 are three-dimensional mapping relation diagrams of water storage time, water storage scheduling rules and water storage benefit indexes;

9-1 to 9-4 are schematic diagrams of error points between a target sample extracted by the KNN nearest neighbor algorithm and a predicted value thereof;

Fig. 10 is a schematic diagram of a Pareto front curve of a flood end water storage strategy;

FIG. 11 is a schematic diagram of a flood control-reservoir displacement relationship;

Fig. 12 is a schematic view of a Pareto non-inferior break-up point of a flood end water storage strategy;

fig. 13-1 to 13-2 are schematic diagrams showing comparison of water level and discharge processes of an optimal water storage strategy under a multi-objective optimization scheme.

Detailed Description

The technical conception of the invention is as follows: the hydrologic hydrodynamic model and the NSGA-II multi-objective intelligent optimization algorithm are coupled, and the flood evolution simulation and the engineering optimization scheduling are organically combined, so that the reliability and the optimality of an optimization regulation model are improved, and references and bases are provided for flood resource utilization of the water-type lakes while the water flow fine simulation in the complex river network area is realized.

The following description of the embodiments of the present application will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present application, but not all embodiments. The following description of at least one exemplary embodiment is merely exemplary in nature and is in no way intended to limit the application, its application, or uses.

Example 1

Referring to fig. 1, the embodiment discloses an intelligent optimization method for a lake flood end water storage strategy based on a hydrological hydrodynamic model, which comprises the following steps:

acquiring a water storage strategy sample library containing water storage time and water storage scheduling rules, inputting the water storage strategy sample library into a hydrological hydrodynamic model which is jointly applied to a pre-constructed lake and a water storage diapause area thereof, and acquiring a three-dimensional mapping relation of the water storage time, the water storage scheduling rules and evaluation indexes of each water storage strategy sample;

extracting characteristic value data of an evaluation index in the three-dimensional mapping relation by adopting a KNN nearest neighbor algorithm according to the three-dimensional mapping relation;

And taking the extracted evaluation index characteristic value data as a calculation basis, and adopting an NSGA-II intelligent optimization algorithm to perform optimal solution search iteration on the water storage time and the water storage scheduling rule so as to obtain an optimal water storage strategy.

According to the method, an intelligent optimization model of the lake water storage strategy at the tail of the flood can be constructed, and the specific steps are as follows:

S10, constructing a hydrologic hydrodynamic model for combined use of lakes and flood storage areas;

s20, constructing a water storage strategy sample library formed by water storage time and a water storage scheme, and extracting index data by adopting a KNN nearest neighbor algorithm;

s30, constructing a flood end water storage strategy optimization model based on an NSGA-II multi-objective optimization algorithm.

The hydrologic hydrodynamic model for combined use of the lake and the flood storage areas thereof constructed in the step S10 specifically comprises the following steps:

S11, manufacturing parameter files such as river network files, section files, boundary files, river course roughness, scheduling rules of related hydraulic buildings and the like of a research area, and establishing a one-dimensional river network hydrodynamic model by adopting MIKE 11;

based on MIKE11, a one-dimensional river network hydrodynamic model is established, six-point implicit differential format is adopted for dispersion, water level or flow of grids arranged in sequence is calculated alternately, a chasing method is adopted for solving a discrete equation set, and basic equations are as follows:

Wherein: z is the average water level of the section; q is the section flow; a is the cross-sectional area; c is the thank you coefficient; q is the lateral inflow; g is gravity acceleration; r is the hydraulic radius; alpha is a momentum correction coefficient.

S12, dividing a research area into a plurality of unstructured triangular grids according to a topography map of the research area and digital elevation data, manufacturing mesh irregular triangular grid files, underlying surface roughness files, initial water depth files and the like, and establishing a two-dimensional flood area hydrodynamic model by adopting MIKE;

Based on MIKE & lt 21 & gt, a two-dimensional flooding area hydrodynamic model is established, the assumption of Boussinesq and hydrostatic pressure is obeyed, an incompressible Reynolds average Navier-Stokes equation is taken as a theoretical basis, a two-step Runge-Kutta method with a finite volume method and a second-order precision is adopted to carry out space and time dispersion on a control equation, and the basic equation is as follows:

Wherein: x and y are Cartesian coordinate systems; t is time; h is the water depth; u and v are average flow velocity in x and y directions respectively; s is a source sink item; g is gravity acceleration; η is the elevation of the river bottom; ρ is the density of water; ρ ₀ is the relative density of water; p _a is atmospheric pressure; τ _ax、τ_ay is wind load force ,τ_ax＝ρ_aC_Ds|ω_ax|ω_ax,τ_ay＝ρ_aC_Ds|ω_ay|ω_ay;ρ_a is air density; omega _ax、ω_ay is the wind speed at 10m above the water surface; c _Ds is the drag coefficient; τ _bx、τ_by is the resistance of the river bed, N is the river bed roughness; f is the Coriolis force coefficient,/>Omega is the angular velocity of the earth rotation 0.729 x 10 ^-4,Is latitude; s _xx、s_xy、s_yx、s_yy is the radiation stress component; t _xx、T_xy、T_yx、T_yy is the horizontal viscous stress component; u _s、v_s is the source sink stream flow rate.

S13, adopting MIKEFLOOD to establish a one-dimensional and two-dimensional coupling flood evolution model.

Establishing a one-dimensional and two-dimensional coupled flood evolution model based on MIKEFLOOD, including:

Establishing standard connection, mapping connecting lines to boundaries of one or more grid units to form coupled lines, providing boundary flow for a two-dimensional model by a one-dimensional model, and returning an average water level value on the coupled lines to the one-dimensional model by the two-dimensional model;

Establishing lateral connection, connecting grid cells of the two-dimensional model to a river reach of the one-dimensional model from the side, determining coupling lines by coordinates, wherein the grid cells mapped to the coupling lines need to participate in coupling, and calculating the water flow through the lateral connection by adopting a flow formula of the hydraulic building.

And S14, parameter calibration and verification are carried out on the one-dimensional river course roughness and the two-dimensional flood area underlying surface roughness.

And selecting actual measurement rainfall runoff and water level data in the live year, comparing the average water level difference absolute value and the maximum water level difference absolute value of the actual measurement water level and the simulated water level of the hydrologic station in the research area, and observing whether the water level change process and the peak time are basically consistent.

Furthermore, the input conditions of the hydrologic hydrodynamic model used by the combination of the lake and the flood storage area are as follows:

one-dimensional river network hydrodynamic model: river network files, section files, boundary files and parameter files;

two-dimensional flood area hydrodynamic model: mesh irregular triangle mesh file, roughness file and initial water depth file;

1. two-dimensional coupled flood evolution model: a link mode.

Step S20, constructing a water storage strategy sample library formed by water storage time and a water storage scheme, and extracting index data by adopting a KNN nearest neighbor algorithm, wherein the method specifically comprises the following steps:

S21, designing a water storage opportunity scheme set in a later flood period; designing a water storage scheduling rule scheme set in a later flood period; the water storage opportunity scheme set and the water storage scheduling rule scheme set are arranged and combined to obtain a water storage strategy sample library;

setting a water storage opportunity scheme set in a later flood period: t= { T ₀+t,T₀+2t,…,T_e };

The water storage time is the starting time of water storage scheduling, T is the flood duration corresponding to 95% frequency of the annual flood duration frequency discharge fitting curve, T ₀ is the first water storage time, T _e is the last water storage time, and T _e is smaller than the end date of the later flood period;

Designing a water storage scheduling rule scheme set in a later flood period, and assuming M down-leakage gates, setting a control interval of the lake water level as [ Z _l,Z_n],Z_l is a flood limit water level, and Z _n is a normal water storage level; setting a change interval of the total leakage amount of the gate as [ Q _d,Q_u],Q_u、Q_d is respectively the upper limit and the lower limit of the total leakage amount; setting the total amount of the leakage between the lake water levels Z _l,Z_n to be increased according to the double ratio Q _b; and setting the total gate leakage amount as an arithmetic series with Q _c as tolerance increment, thereby obtaining N water storage scheduling rule scheme sets:

The water storage scheduling rule refers to a permutation and combination form of a plurality of down-leakage gates for combining the down-leakage flow under different water level conditions;

j is a down-flow gate, j=1, 2, …, M; i is a water storage scheme, i=1, 2, …, N;

For the downward leakage flow rate of the ith scheme of the jth gate when the lake water level is at the flood limit water level,/> The method is characterized in that the method is the discharging flow of the jth gate in the ith scheme when the lake water level is at the normal water storage level; /(I)Q _u、Q_d is the upper and lower limits of the total amount of leakage respectively;

And (3) arranging and combining the water storage opportunity scheme set and the water storage scheduling rule scheme set to obtain a water storage strategy sample library.

S22, inputting a water storage strategy sample library into the hydrologic and hydrokinetic model constructed in the step S10, simulating the flood process in the later flood season, and counting two indexes of the water level overrun R _e and the highest water storage level Z _max of the lake representative hydrologic station in the water storage process in the later flood season, wherein the two indexes reflect flood control risks, and the two indexes of the total water discard W _d and the final water storage level Z _t reflect water storage benefits;

inputting a water storage strategy sample library into a constructed hydrologic hydrodynamic model, and simulating a flood process in a later flood season:

Z_t＝z_i,i＝t_end (9)

Wherein: r _e is the water level overrun of the lake representative hydrologic station in the water storage process of the later flood season, Z _max is the highest water storage level of the lake representative hydrologic station in the water storage process of the later flood season, W _d is the total amount of abandoned water of the lake representative hydrologic station in the water storage process of the later flood season, Z _t is the water storage level of the lake representative hydrologic station at the flood end in the water storage process of the later flood season, wherein two items R _e and Z _max reflect flood control indexes, and two items W _d and Z _t reflect water storage indexes;

t _i is the time period number that the lake water level exceeds the flood limit water level in the later flood period; z _i is the mean lake level of the ith period; t ₁ is the time period number of the later flood period; t ₂ is the time period number of the main flood season; q _ij is the average down-flow rate of the jth gate in the ith period; t _end is the end date of the later flood season; z _l is the flood limit water level process line.

S23, forming a time-quantity-scale three-dimensional mapping relation and an image of a water storage time (time), a water storage scheduling rule (quantity) and an evaluation index (scale) by adopting a uniform distribution method;

S24, extracting index values of the 'time-quantity-standard' mapping image by adopting a two-dimensional interpolation technology based on a KNN nearest neighbor algorithm, so as to obtain a large number of data samples under different water storage time and water storage scheduling rules.

Performing data interpolation on the acquired index data based on a KNN nearest neighbor algorithm to acquire data samples of flood control indexes and water storage indexes under any water storage time and water storage scheduling rules; the Euclidean distance formula between two water storage strategy samples is as follows:

Wherein: d is the Euclidean distance between two samples; t ₁、t₂ is the start-up time; q ₁、q₂ is the total amount of leakage.

The optimal k value step for calculating four indexes of the water level overrun, the highest water storage level, the water discard amount and the water storage level at the tail of the flood based on the KNN nearest neighbor algorithm is as follows:

p×p (P should be a larger value) water policy samples are divided into training sets (X groups) and test sets (Y groups):

① Calculating the Euclidean distance between each training sample j in the training set and each target sample i in the test set;

② Selecting k training sample points nearest to the target sample i, setting an initial value k ₀ of k as 1, constructing a weight matrix according to Euclidean distance between the k training sample points and the target sample i by adopting an inverse distance method, and calculating predicted values of all the target samples;

③ The error between the target sample and its predicted value is denoted as E _i,k, the total error value of the target sample is denoted as E _k,

Let k=k+1, repeat steps ② to k to X, k when E _k obtains the minimum value is the best near point;

④ And extracting interpolation samples of four indexes in the time-quantity-standard three-dimensional space by adopting the KNN spatial interpolation parameters obtained by training. And calculating the optimal k values of four indexes of the water level overrun, the highest water storage level, the water discarding amount and the water storage level at the tail of the flood according to the steps.

Step S30 is to construct a flood end water storage strategy optimization model based on an NSGA-II multi-objective optimization algorithm, and the method specifically comprises the following steps:

S31, defining two objective functions of a flood control objective F ₁ and a water storage objective F ₂;

defining two objective functions of a flood control objective and a water storage objective, wherein the formula is as follows:

Wherein: f ₁ is a flood control target; for the t day, discharging the highest water storage level when the total amount is q; z _max、Z_minZ_min is the upper and lower limits Z ^max-gap＝Z_max-Z_min of the highest water storage level under all possible water storage strategies; /(I) For the t day, the water level overrun when the total amount of the leakage is q; /(I)The upper limit and the lower limit of the water level overrun rate are used for all the feasible water storage schemes; alpha ₁、β₁ is a weight factor of the flood control index;

f ₂ is a water storage target; For the t day, discharging the flood end water storage level when the total amount is q; z' _tZ_target is the target water level of the water storage; z _t,max,/> Upper and lower limits of last flood water level for all feasible water storage strategies For the t day, discharging the water discarding amount when the total amount is q; w _d,max,/>Upper and lower limits for water reject volume for all viable water storage strategiesAlpha ₂、β₂ is a weight factor of the water storage index;

S32, defining constraint conditions:

The flood end water storage strategy optimization model is also provided with constraint conditions:

T_min≤t≤T_max (13)

Q_min≤q≤Q_max (14)

Z_max,tq≤Z_max≤Z_w (15)

Z_r≤Z_t,tq≤Z′_r (16)

R_min≤R_e,tq≤R_max (17)

wherein: equation 13 is a water storage time constraint, equation 14 is a control flow constraint, equation 15 is a highest water level constraint, equation 16 is a target water level constraint, and equation 17 is a water level overrun constraint;

t _min、T_max is the start and stop time of the later flood period respectively; q _max、Q_min is the maximum value and the minimum value of the total leakage amount of the gate when the lake water level reaches the normal water storage level respectively; z _w is the flood-stagnation warning water level of the broken leves; z _r、Z'_r is the flood limit water level before and after the flood period stage control adjustment respectively; r _max、R_min is the upper limit and the lower limit of the water level overrun constraint respectively.

S33, analyzing a replacement relation and comparing and selecting a water storage strategy:

Taking the water storage opportunity t and the total drainage q as decision variables, taking data extracted by a 'time-quantity-standard' three-dimensional mapping relation as calculation basis of targets, carrying out optimal solution search iteration by adopting an NSGA-II intelligent optimization algorithm to obtain a non-inferior leading edge curve between two targets of a flood end water storage strategy problem, and obtaining a displacement relation curve between a flood control target and a water storage target by deriving a Pareto leading edge fitting function so as to compare and select the water storage strategy.

Example 2

Based on the embodiment 1, this embodiment describes a specific example of the intelligent optimization method of the lake water storage strategy based on the hydrographic model in embodiment 1.

In this example, hongze lake was chosen as an example. The geographical location of the Hongze lake, the in-and-out lake river and the controlled site distribution are shown in FIG. 2. The Hongze lake is positioned at the joint part of the middle and the downstream of the Huai river, and is used for discharging 15.8 ten thousand km ² of the water coming from the middle and the upstream of the Huai river. The elevation of the lake bottom is 10.0m, the dead water level is 11.3m, the flood limit water level is 12.5-13.5 m, the normal water storage level is 13.5m, the warning water level of the broken levee is 14.5m, and the total reservoir capacity is 37.3 hundred million m ³. The influence of east Asia monsoon is obvious, the distribution is uneven in the rainfall year, the distribution is concentrated for 5-8 months, and the river basin type storm flood is easy to form. The main river channel of Hongze lake is the dry river, other river channels include Xu Honghe in the north, hong Xin in the west, new and old Sui river, and the pool in the new and south, and the lake area mainly has 5 hydrologic stations, which are respectively a Jiang dam station (S1, representative hydrologic station), an old mountain station (S2), a Xiangcheng station (S3), a Linghuai head station (S4), and a Shangzui station (S5). The drainage channel is concentrated at the eastern part of the lake region, and flood water is respectively drained into a river channel, a Huai Shu new river, a Subei irrigation main channel and a sea water channel by a Sanhe gate (R1), a Gaowan mountain gate (R2) and a two-river gate (R3). The regional flood control engineering consists of a lake inlet and outlet control building, a flood lake dyke, a flood stagnation area, a flood discharge area and the like. For a long time, the flood storage areas around the Hongze lake are used as places for flood inundation and regulation, and play an important role in a downstream flood control system in the Huaihe river.

Basic topography and engineering data required by modeling flood lakes and peripheral flood storage areas are selected, wherein the basic topography and engineering data comprise 5m multiplied by 5m DEM topography, river water systems and river channel sections, hydraulic buildings, embankments, land utilization classifications and the like.

The hydrologic hydrodynamic model of combined use of the lake and the flood storage areas constructed in the step S10 specifically comprises the following steps:

s11, adopting MIKE to establish a one-dimensional river network hydrodynamic model:

128 river ditches and 322 polder inland channels with larger influence on the lake regulation effect are constructed in the one-dimensional river network model (figure 3-1); 2135 river cross sections (fig. 3-2), river roughness, scheduling rules for the associated hydraulic structure, and model boundary files (fig. 3-3) containing time series are described.

S12, adopting MIKE to establish a two-dimensional flood area hydrodynamic model:

The study area is divided into 85894 unstructured triangular grids in a two-dimensional flooding area model, the section 1634 of the road is generalized (figure 4-1), different roughness values are set according to the specific type of the underlying surface (figure 4-2), and the initial water depth is set based on measured water level data (figure 4-3).

S13, adopting MIKE FLOOD to establish a one-dimensional and two-dimensional coupling flood evolution model:

1. The two-dimensional coupling flood evolution model adopts an actual inflow process as an upper boundary, takes a water level-flow relation of a lake outlet as a lower boundary, establishes standard connection between a river channel entering the lake and a flood lake area, and between a water gate of a flood and a flood storage area, and maps connecting lines to the boundary of one or more grid units to form coupling lines; and a lateral connection is established between a river channel and a flood area in the polder, grid cells of the two-dimensional model are connected to a river reach of the one-dimensional model from the side, the coupling lines are determined by coordinates, the grid cells mapped to the coupling lines need to participate in coupling, a flow formula of the hydraulic building is adopted to calculate the water flow through the lateral connection, and the coupling lines are shown in figure 5.

S14, parameter calibration and verification are carried out on the one-dimensional river course roughness and the two-dimensional flood area underlying surface roughness:

and selecting actual measurement rainfall runoff and water level data in the live year, comparing the average water level difference absolute value and the maximum water level difference absolute value of the actual measurement water level and the simulated water level of the hydrologic station in the research area, and observing whether the water level change process and the peak time are basically consistent.

And selecting 1991 and 2006 as live years, and carrying out parameter calibration on the one-dimensional river course roughness and the two-dimensional flood area underlying surface roughness, wherein the results are shown in the following table. Model verification adopts typical flood processes of 25 days to 7 months and 31 days in 6 months in 2003 and 1 day to 8 months and 3 days in 7 months in 2007, actual water level processes and simulated water level processes of five hydrologic stations S1 to S5 are compared, the comparison results in 2003 are shown in figures 6-1 to 6-5, and the comparison results in 2007 are shown in figures 6-6 to 6-10. The water level change process and the peak time are basically consistent, so that the model has higher precision, and is reasonable and reliable.

Step S20, constructing a water storage strategy sample library formed by water storage time and a water storage scheme, and extracting index data by adopting a KNN nearest neighbor algorithm, wherein the method specifically comprises the following steps:

S21, designing a water storage opportunity scheme set in a later flood period; designing a water storage scheduling rule scheme set in a later flood period; the water storage opportunity scheme set and the water storage scheduling rule scheme set are arranged and combined to obtain a water storage strategy sample library:

The later flood season of Hongze lake is usually 8 months 20 days to 9 months 31 days, the initial time of the later flood season is taken as the first water storage time, and in view of the obvious influence of the water storage time on the lake water level and the water discard amount in the dispatching process, a plurality of times are set as a water storage time scheme set at equal intervals of 5d, wherein T= {8 months 21 days, 8 months 26 days, 8 months 31 days, 9 months 5 days, 9 months 10 days, 9 months 15 days, 9 months 20 days }.

The specific design principle of the invention for each gate scheduling rule is as follows:

① And determining that the lake water level control range taking the dam station as the representative water level station is 12.50-13.50 m according to the upper limit and the lower limit of the flood control water level dynamic control of the Hongze lake.

② According to the 'Huaihe river basin flood control planning', the change range of the total leakage amount is 600-9600 m3/s, and the two river gates are closed in the range. Considering the water use requirement of the irrigation main canal and the high-yield hydropower station, the high-yield river gate is used for discharging preferentially, the discharging capacity is 600-800 m ³/s, and the three-river gate is kept closed before the discharging capacity of the gate is reached; if the water level rises to 14.0m or above, the flood control scheduling rule of the main flood season is adopted, namely the total drainage amount corresponding to the water level of 14.0m is 9600m ³/s.

③ And (3) controlling the total discharge amount of the gate when the lake water level is 13.0m to be 600-3000 m ³/s by taking the scheme I (the scheme I is a preferred strategy for enumerating the water storage strategy and simulating flood evolution) as a reference, wherein the discharge amount of the three-river gate of the schemes II-V III is an arithmetic series which grows with the tolerance of 400m ³/s.

④ Setting the total discharge amount of two gates of 13.5m and 13.0m of the lake water level to be increased by 1.3 times by referring to the water level-discharge amount relation of the flood lake in the main flood period. In summary, 7 sets of water storage scheduling rule schemes in the later flood season are set up to form a scheme set Q, as shown in the following table.

S22, inputting a water storage strategy sample library into the hydrologic and hydrokinetic model constructed in the step S10, simulating the flood process in the later flood season, and counting two indexes of the water level overrun R _e and the highest water storage level Z _max of the lake representative hydrologic station in the water storage process in the later flood season, wherein the two indexes reflect flood control risks, and the two indexes of the total water discard W _d and the final water storage level Z _t reflect water storage benefits:

In the embodiment, the daily flow data of the main river channel entering and exiting the lake in 2003, the daily precipitation, evaporation and water level data around the lake area are selected, and the hydrologic hydrodynamics model constructed in the step S10 is adopted to simulate the flood process of 49 groups of water storage strategies. The combination of 7 water storage opportunities and 7 water storage scheduling rules forms a search space with the optimal solution as a core, and four indexes of the overrun of the water level of the building dam, the highest water storage level, the water discarding amount and the water storage level at the flood end in the water storage process of each group of schemes are counted.

S23, forming a time-quantity-scale three-dimensional mapping relation and an image of a water storage time (time), a water storage scheduling rule (quantity) and an evaluation index (scale) by adopting a uniform distribution method:

And 4 groups of time-quantity-standard three-dimensional mapping relations of 100 multiplied by 100 are formed by adopting a uniform distribution method, wherein three-dimensional mapping relation diagrams of flood control risk indexes are shown in fig. 7-1 and 7-2, and three-dimensional mapping relation diagrams of water storage benefit indexes are shown in fig. 8-1 and 8-2.

S24, extracting index values of the 'time-quantity-standard' mapping image by adopting a two-dimensional interpolation technology based on a KNN nearest neighbor algorithm, so as to obtain a large number of data samples under different water storage time and water storage scheduling rules:

Two-dimensional interpolation is performed on four indexes in the implementation case (3) by adopting a KNN algorithm, 100×100 water storage strategy samples are divided into 8000 training sets and 2000 test sets for each index, and according to the implementation step ①～④ of the KNN index value extraction method in embodiment 1, the optimal k values of 4 indexes of the dam water level overrun, the highest water storage level, the water discarding amount and the flood end water storage level are calculated to be 4, 3, 4 and 3 respectively, as shown in fig. 9-1-9-4.

Step S30 is to construct a flood end water storage strategy optimization model based on an NSGA-II multi-objective optimization algorithm, and the method specifically comprises the following steps:

S31, defining two objective functions of a flood control objective F ₁ and a water storage objective F ₂;

The model takes the water storage time and the total drainage amount as decision variables, and calculates an index mapping value under each decision variable based on the optimal k value acquired in the step S24 and the time-quantity-standard three-dimensional relationship established in the step S23. Based on the objective function and constraint conditions of the above embodiment 1, in the formula (11), let In formula (12), α ₂＝0.8、β₂ =0.2; in formula (16), Z _r＝12.5m、Z'_r =13.5m; in formula (17), R _max＝1、R_min =0.

Introducing NSGA-II algorithm to perform optimal solution search, setting iteration times to 1000, obtaining non-inferior leading edge between two targets of flood lake flood end water storage strategy problem when iteration reaches evolution algebra, and adopting curve fitting function expression of non-inferior leading edge, namelyAs shown in fig. 10.

S32, analyzing the replacement relation and comparing and selecting the water storage strategy

Obtaining a displacement relation expression between the flood control target and the water storage target by deriving a Pareto front fitting function, namelyAs shown in fig. 11, when the flood control target value is greater than 0.1, the water storage requirement of the lake is basically satisfied, and excessive flood control risks are not needed. Thus, a flood control target of 0.1 is selected as an alternative to the optimal impoundment strategy, and the corresponding Pareto non-inferior solution set is shown in fig. 12. From the figure, the approximate position of the optimal solution of the water storage strategy can be known: the water storage time is 28-30 d after entering the later flood period, the water storage scheduling rule is between the scheme VII and the scheme VIII, and the water level change process line and the total drainage process line corresponding to different optimization strategies are shown in figures 13-1 and 13-2. And finally determining the global optimal solution of the water storage strategy in the late flood season of the Hongze lake in 2003 through comparative analysis, namely, starting the water storage for 9 months and 20 days, and increasing the total drainage amount corresponding to the water level of 13.0m in the water storage scheduling rule from 1600m ³/s to 2213m ³/s in the planning scheme.

Example 3

This embodiment describes a computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements the intelligent optimization method for a lake end water storage strategy based on a hydrographic hydrodynamic model described in embodiment 1 or 2.

It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.

The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

The foregoing is merely a preferred embodiment of the present invention, and it should be noted that modifications and variations could be made by those skilled in the art without departing from the technical principles of the present invention, and such modifications and variations should also be regarded as being within the scope of the invention.

Claims (10)

1. An intelligent optimization method for a lake flood end water storage strategy based on a hydrologic hydrodynamic model is characterized by comprising the following steps:

acquiring a water storage strategy sample library containing water storage time and water storage scheduling rules, inputting the water storage strategy sample library into a hydrological hydrodynamic model which is jointly applied to a pre-constructed lake and a water storage diapause area thereof, and acquiring a three-dimensional mapping relation of the water storage time, the water storage scheduling rules and evaluation indexes of each water storage strategy sample;

extracting characteristic value data of an evaluation index in the three-dimensional mapping relation by adopting a KNN nearest neighbor algorithm according to the three-dimensional mapping relation;

And taking the extracted evaluation index characteristic value data as a calculation basis, and adopting an NSGA-II intelligent optimization algorithm to perform optimal solution search iteration on the water storage time and the water storage scheduling rule so as to obtain an optimal water storage strategy.

2. The intelligent optimization method for the water storage strategy at the tail of the lake based on the hydrographic model as claimed in claim 1, wherein the water storage strategy sample library comprising water storage time and water storage scheduling rules comprises the following steps:

The scheme set of water storage opportunity in the later flood season: t= { T ₀+t,T₀+2t,…,T_e };

The water storage time is the starting time of water storage scheduling, T is the flood duration corresponding to 95% frequency of the annual flood duration frequency discharge fitting curve, T ₀ is the first water storage time, T _e is the last water storage time, and T _e is smaller than the end date of the later flood period;

A water storage scheduling rule scheme set in the later flood period:

The water storage scheduling rule refers to a permutation and combination form of a plurality of down-leakage gates for combining the down-leakage flow under different water level conditions;

j is a down-flow gate, j=1, 2, …, M; i is a water storage scheme, i=1, 2, …, N;

For the downward leakage flow rate of the ith scheme of the jth gate when the lake water level is at the flood limit water level,/> The method is characterized in that the method is the discharging flow of the jth gate in the ith scheme when the lake water level is at the normal water storage level; /(I)Q _u、Q_d is the upper and lower limits of the total amount of leakage respectively;

And (3) arranging and combining the water storage opportunity scheme set and the water storage scheduling rule scheme set to obtain a water storage strategy sample library.

3. The intelligent optimization method for the water storage strategy at the tail end of a lake based on a hydrographic model as claimed in claim 1, wherein the hydrographic model jointly applied by the lake and the flood storage area thereof is constructed by the following steps:

based on MIKE 11, a one-dimensional river network hydrodynamic model is established, and the basic equation is as follows:

Wherein: x and t are space coordinates and time coordinates, respectively; z is the average water level of the section; q is the section flow; a is the cross-sectional area; c is the thank you coefficient; q is the lateral inflow; g is gravity acceleration; r is the hydraulic radius; alpha is a momentum correction coefficient; |q| is the absolute value of Q;

Based on MIKE, a two-dimensional flood area hydrodynamic model is established, and the basic equation is as follows:

Wherein: x and y are Cartesian coordinate systems; t is time; h is the water depth; u and v are average flow velocity in x and y directions respectively; s is a source sink item; g is gravity acceleration; η is the elevation of the river bottom; ρ is the density of water; ρ ₀ is the relative density of water; p _a is atmospheric pressure; τ _ax、τ_ay is wind load force ,τ_ax＝ρ_aC_Ds|ω_ax|ω_ax,τ_ay＝ρ_aC_Ds|ω_ay|ω_ay;ρ_a is air density; omega _ax、ω_ay is the wind speed at 10m above the water surface; c _Ds is the drag coefficient; τ _bx、τ_by is the resistance of the river bed, N is the river bed roughness; f is the Coriolis force coefficient,/>Omega is the angular velocity of the earth rotation 0.729 x 10 ^-4,Is latitude; s _xx、s_xy、s_yx、s_yy is the radiation stress component; t _xx、T_xy、T_yx、T_yy is the horizontal viscous stress component; u _s、v_s is the source sink stream flow rate.

Establishing a one-dimensional and two-dimensional coupled flood evolution model based on MIKE FLOOD, including:

Establishing standard connection, mapping connecting lines to boundaries of one or more grid units to form coupled lines, providing boundary flow for a two-dimensional model by a one-dimensional model, and returning an average water level value on the coupled lines to the one-dimensional model by the two-dimensional model;

Establishing lateral connection, connecting grid cells of the two-dimensional model to a river reach of the one-dimensional model from the side, determining coupling lines by coordinates, wherein the grid cells mapped to the coupling lines need to participate in coupling, and calculating the water flow through the lateral connection by adopting a flow formula of the hydraulic building.

4. The intelligent optimization method for the water storage strategy at the tail end of a lake based on a hydrologic hydrodynamic model according to claim 1 or 3, wherein the hydrologic hydrodynamic model used by the lake and the water storage areas thereof in combination further comprises the steps of parameter calibration and verification of river course roughness of a one-dimensional river network hydrodynamic model and mat surface roughness under a two-dimensional flood area hydrodynamic model:

and selecting actual measurement rainfall runoff and water level data in the live year, comparing the average water level difference absolute value and the maximum water level difference absolute value of the actual measurement water level and the simulated water level of the hydrologic station in the research area, and observing whether the water level change process and the peak time are basically consistent.

5. The intelligent optimization method for the water storage strategy at the tail end of a lake based on a hydrographic hydrodynamic model according to claim 1 or 3, wherein the hydrographic hydrodynamic model jointly applied by the lake and the flood storage area thereof further comprises the following input conditions:

The input conditions of the one-dimensional river network hydrodynamic model comprise river network files, section files, boundary files and parameter files;

The input conditions of the two-dimensional flooding area hydrodynamic model comprise a mesh irregular triangular mesh file, a roughness file and an initial water depth file;

1. the input conditions of the two-dimensional coupled flood model comprise a link mode.

6. The intelligent optimization method for the water storage strategy at the tail of the lake based on the hydrographic model according to claim 1, wherein the method is characterized in that a water storage strategy sample library containing water storage time and water storage scheduling rules is obtained and is input into a hydrographic model which is jointly applied to a pre-constructed lake and a water storage flood area thereof, and a three-dimensional mapping relation among the water storage time, the water storage scheduling rules and evaluation indexes of each water storage strategy sample is obtained, and the method comprises the following steps:

inputting a water storage strategy sample library into a constructed hydrologic hydrodynamic model, and simulating a flood process in a later flood season:

Z_t＝z_i,i＝t_end (9)

Wherein: r _e is the water level overrun of the lake representative hydrologic station in the water storage process of the later flood season, Z _max is the highest water storage level of the lake representative hydrologic station in the water storage process of the later flood season, W _d is the total amount of abandoned water of the lake representative hydrologic station in the water storage process of the later flood season, Z _t is the water storage level of the lake representative hydrologic station at the flood end in the water storage process of the later flood season, wherein two items R _e and Z _max reflect flood control indexes, and two items W _d and Z _t reflect water storage indexes;

t _i is the time period number that the lake water level exceeds the flood limit water level in the later flood period; z _i is the mean lake level of the ith period; t ₁ is the time period number of the later flood period; t ₂ is the time period number of the main flood season; q _ij is the average down-flow rate of the jth gate in the ith period; t _end is the end date of the later flood season; z _l is a flood limit water level process line;

And obtaining the three-dimensional mapping relation and the image of the water storage time, the water storage scheduling rule and the evaluation index of each water storage strategy sample by adopting a uniform distribution method.

7. The intelligent optimization method for the lake water storage strategy at the tail end based on the hydrographic model according to claim 1, wherein the characteristic value data of the evaluation index in the three-dimensional mapping relation is extracted by adopting a KNN nearest neighbor algorithm according to the three-dimensional mapping relation, and the method comprises the following steps:

Performing data interpolation on the acquired index data based on a KNN nearest neighbor algorithm to acquire data samples of flood control indexes and water storage indexes under any water storage time and water storage scheduling rules; the Euclidean distance formula between two water storage strategy samples is as follows:

Wherein: d is the Euclidean distance between two samples; t ₁、t₂ is the start-up time; q ₁、q₂ is the total amount of leakage.

8. The intelligent optimization method for the water storage strategy at the tail of the lake based on the hydrographic model as claimed in claim 1, wherein the method is characterized in that the extracted characteristic value data of the evaluation index is used as a calculation basis, an NSGA-II intelligent optimization algorithm is adopted to perform optimal solution search iteration on the water storage opportunity and the water storage scheduling rule, and an optimal water storage strategy is obtained, and the method comprises the following steps:

setting an objective function as a flood control objective and a water storage objective;

Setting decision variables as water storage time and total drainage;

setting constraint conditions;

performing optimal solution search iteration by adopting an NSGA-II intelligent optimization algorithm to obtain a non-inferior leading edge curve between two targets of a flood end water storage strategy problem;

And deriving the Pareto front fitting function to obtain a displacement relation curve between the flood control target and the water storage target, thereby obtaining an optimal water storage strategy.

9. The intelligent optimization method for the lake water storage strategy at the tail of the flood, based on the hydrographic model, as claimed in claim 8, is characterized in that,

The objective function is:

Wherein: f ₁ is a flood control target; for the t day, discharging the highest water storage level when the total amount is q; z _max、Z_minZ_min is the upper and lower limits Z ^max-gap＝Z_max-Z_min of the highest water storage level under all possible water storage strategies; /(I) For the accumulation of the t day, the water level overrun rate/>, when the total amount of leakage is qThe upper limit and the lower limit of the water level overrun rate are used for all the feasible water storage schemes; alpha ₁、β₁ is a weight factor of the flood control index;

f ₂ is a water storage target; for the t day, discharging the flood end water storage level when the total amount is q; /(I) Is a target water level of the water storage; z _t,max,/>Upper and lower limits of last flood water level for all feasible water storage strategiesFor the t day, discharging the water discarding amount when the total amount is q; w _d,max,/>Upper and lower limits for water reject volume for all viable water storage strategiesAlpha ₂、β₂ is a weight factor of the water storage index;

The constraint conditions are as follows:

T_min≤t≤T_max (13)

Q_min≤q≤Q_max (14)

Z_max,tq≤Z_max≤Z_w (15)

Z_r≤Z_t,tq≤Z′_r (16)

R_min≤R_e,tq≤R_max (17)

wherein: equation 13 is a water storage time constraint, equation 14 is a control flow constraint, equation 15 is a highest water level constraint, equation 16 is a target water level constraint, and equation 17 is a water level overrun constraint;

t _min、T_max is the start and stop time of the later flood period respectively; q _max、Q_min is the maximum value and the minimum value of the total leakage amount of the gate when the lake water level reaches the normal water storage level respectively; z _w is the flood-stagnation warning water level of the broken leves; z _r、Z'_r is the flood limit water level before and after the flood period stage control adjustment respectively; r _max、R_min is the upper limit and the lower limit of the water level overrun constraint respectively.

10. A computer storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the intelligent optimization method for a flood and end water storage strategy based on a hydrographic hydrodynamic model as claimed in any one of claims 1-9.