A method for predicting and simulating regional surface water resources and optimizing configuration thereof

Abstract

The application discloses a regional surface water resource prediction simulation and optimization configuration method, which comprises the following steps: step one, collecting data for a research area, and constructing a watershed hydrological simulation model, namely, a SWAT model; step two, collecting future meteorological data, and inputting the data into the calibrated and verified SWAT model to predict and simulate the regional surface water resource quantity under the background of climate change in the research area; step three, integrating the regional surface water resource prediction simulation model under the background of climate change into an interval multi-objective optimization model; and step four, solving the interval multi-objective optimization configuration model of the regional surface water resource under the background of climate change; the beneficial effect is that, on the basis of predicting the future regional surface water resource quantity, the water resource configuration scheme of a reservoir under different climate change scenarios, which meets the power generation, water supply and water abandonment targets, can be obtained.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of regional surface water resources optimal allocation, in particular to a regional surface water resources prediction simulation and optimal allocation method. BACKGROUND

[0002] Regional surface water resources are of great significance in promoting social and economic development and protecting the ecological system; reservoirs, as an effective engineering measure, play a key role in the supply and regulation of surface water resources; however, climate change has a significant impact on surface water resources (such as reservoir basin available water resources) by affecting meteorological elements, and the impact has become more significant in recent years, bringing great pressure to the prediction of reservoir basin available water resources, and even causing water shortage risks; how to predict and simulate reservoir basin available water resources under climate change and optimize the allocation has become an important problem in water resources management research;

[0003] At the same time, the management process of regional surface water resources in the reservoir basin is extremely complex, with multiple objectives and uncertainties; for example, a reservoir usually has multiple service functions such as power generation, water supply, and ecological regulation, which results in the need to meet multiple objectives during the operation of the reservoir; at the same time, due to incomplete information collection and system complexity, it is difficult to give a certain value for multiple parameters; which significantly affects the reliability of the optimal allocation scheme; in addition, climate change may exacerbate these uncertainties and complexities, which may lead to some social and economic problems, such as unstable regional water supply and deterioration of downstream ecosystems;

[0004] Therefore, effectively dealing with the uncertainties and complexities of the system under climate change has become the focus and difficulty of reservoir basin surface water resources research; it is necessary to propose some advanced models to support the prediction simulation and optimal allocation of water resources, which is of great significance to ensure the safety of regional water supply and the sustainable development of social economy. SUMMARY

[0005] The purpose of the present application is to solve the above problems, and a regional surface water resources prediction simulation and optimal allocation method is designed.

[0006] To achieve the above purpose, the technical scheme of the present application is a regional surface water resources prediction simulation and optimal allocation method, comprising the following steps:

[0007] Step 1, collect data for the study area, and build a watershed hydrological simulation model-SWAT model;

[0008] Step 2, collect future meteorological data and input it into the calibrated and verified SWAT model to predict and simulate the regional surface water resources under the climate change background in the study area;

[0009] Step three, integrate the regional surface water resources prediction and simulation model under the background of climate change into the interval multi-objective optimization model, and form the interval multi-objective optimization allocation model of regional surface water resources under the background of climate change;

[0010] Step four, solve the interval multi-objective optimization allocation model of regional surface water resources under the background of climate change, and obtain the optimal allocation scheme of regional surface water resources considering future climate change, system uncertainty and multi-objective.

[0011] Further supplement to the technical solution, in step one, the data of the research area includes: digital elevation data, hydrological data, meteorological data, land use data, and soil type data of the research area, and the steps for constructing the SWAT model include: sub-basin division, hydrological response unit division, input meteorological data, parameter sensitivity analysis, model calibration and verification.

[0012] Further supplement to the technical solution, in step two, the constructed SWAT model is calibrated and verified based on the measured hydrological data, and the planning year daily rainfall, daily minimum temperature and daily maximum temperature under different greenhouse gas emission scenarios are collected to input the calibrated and verified SWAT model, and the regional surface water resources under the background of climate change are predicted and simulated.

[0013] Further supplement to the technical solution, in step three, the RCP meteorological scenario, the SWAT model, the interval linear programming and the multi-objective programming are coupled in one framework to construct the interval multi-objective optimization allocation model of regional surface water resources under the background of climate change.

[0014] Further supplement to the technical solution, the constructed interval multi-objective optimization allocation model includes: objective function, decision variable and constraint condition, and the water allocation to power generation, the water intake in the reservoir and the abandoned water quantity are taken as the decision variables, and the maximum hydropower generation, the maximum economic benefit generated by the water intake in the reservoir and the minimum abandoned water quantity are taken as the objective function.

[0015] Further supplement to the technical solution, in step four, the membership function is introduced, the fuzzy geometric weighted algorithm is adopted to convert the multi-objective problem in the interval multi-objective optimization allocation model of regional surface water resources into a single objective problem with the maximum satisfaction degree, the interval number is converted into a certain value by introducing an auxiliary variable, and the single objective problem with the maximum satisfaction degree is converted into an optimal sub-model and a worst sub-model by using the optimal worst model, and the optimal solution and the worst solution of the reservoir power generation water quantity, the water intake in the reservoir and the abandoned water quantity are calculated, wherein the optimal solution and the worst solution of each decision variable can form an interval solution of the decision variable; the decision variable is substituted into the corresponding objective function to obtain the interval solution of the reservoir power generation and the interval solution of the economic benefit generated by the water intake in the reservoir.

[0016] Further supplements to this technical solution include: land use type data including: construction land, cultivated land, forest land, grassland and water area in the study area; soil type data including: spatial distribution data of soil types in the study area and soil attribute database; meteorological data including: daily rainfall, temperature, wind speed, solar radiation and humidity in the study area over the years; and hydrological data including: measured runoff data of hydrological monitoring stations in the study area over the years.

[0017] The construction of a SWAT model to simulate historical runoff mainly includes the following steps: using digital elevation data to analyze digital terrain and define river networks, dividing sub-basins, and calculating sub-basin parameters; based on the sub-basins, further dividing the sub-basins into multiple hydrological response units according to soil data, land use data, and slope data; inputting meteorological data, calculating the runoff on each hydrological response unit step by step, and then obtaining the total runoff of the basin through confluence calculation.

[0018] To further supplement this technical solution, the constructed SWAT model is calibrated and validated. Hydrological monitoring stations within the watershed are selected, and the simulated data of the SWAT model is corrected using the measured hydrological data from these stations. This includes the following steps: Calibration and validation are performed using SWAT-CUP software; the SUFI-2 algorithm is used for iterative calculations to determine the optimal parameter values; the SWAT model is adjusted based on the optimal parameter values, and the results are substituted into the model for simulation validation; the Nash coefficient (…) is selected. NSE ) and correlation coefficient ( R 2 ) as two indicators for evaluating the applicability of the SWAT model, and NSE >0.5, R 2 The SWAT model is applicable when the value is greater than 0.6.

[0019] As a further supplement to this technical solution, in step three, the regional surface water resource interval multi-objective optimization allocation model includes:

[0020] The objective function is:

[0021] a) Maximum power generation:

[0022]

[0023] =

[0024] In the formula, for t Total power generated by the reservoir during the period, in kWh; For the reservoir in t Average power generation during the period, in MW (megawatts per hour). This indicates the reservoir's output coefficient; The water head of the reservoir is expressed in meters (m). Indicates that the reservoir is t Total water consumption for power generation during the period, in m³ 3 ; express t The duration of each time period is in days; T represents the total number of time periods.

[0025] b) Water supply generates the greatest economic benefits:

[0026]

[0027] In the formula, for t The economic benefits generated by reservoir water supply during the period, in RMB; The economic benefit generated per unit of water supplied is expressed in RMB / m³. 3 ; The total water supply from the reservoir during time period t is expressed in cubic meters (m³). 3 ;

[0028] c) Minimize water waste: Make full use of water resources

[0029]

[0030] In the formula, Total water volume to be discarded, in cubic meters (m³). 3 ; for t The amount of water released from the reservoir during the specified period, in cubic meters (m³). 3 ; T This represents the total number of time periods.

[0031] Constraints:

[0032] a) Reservoir water balance constraints:

[0033]

[0034] In the formula, and They are t+ 1 and t Reservoir capacity during a given period, in meters (m). 3 ; I t yes t Reservoir inflow rate during a given period, in cubic meters (m³) 3 / s;

[0035] b) Turbine flow constraint:

[0036]

[0037]

[0038] wherein, is the power generation flow of the reservoir at t time period, in m 3 ; Q max,p is the maximum power generation water consumption of the reservoir at t time period, in m 3 ; is the maximum turbine flow of the hydropower station, in m 3 / s, wherein the maximum turbine flow of the Xinfengjiang Reservoir is 490 m 3 / s;

[0039] c) Reservoir discharge water volume constraint:

[0040]

[0041]

[0042] wherein, is the discharge water volume of the reservoir at t time period, in m 3 , wherein the discharge water volume is equal to the sum of the power generation water consumption and the abandoned water volume; is the minimum discharge flow of the reservoir at t time period, in m 3 ; is the maximum discharge flow of the reservoir at t time period, in m 3 ;

[0043] d) Power generation constraint:

[0044]

[0045] wherein, is the total power generation of the reservoir in the planning period, in kWh; is the minimum power generation of the reservoir in the planning year, in kWh;

[0046] e) Reservoir in-reservoir water intake constraint:

[0047]

[0048] wherein, is the total water supply volume of the reservoir at 3 time period, i.e., the total in-reservoir water intake, in m 3 ; is the minimum in-reservoir water intake of the reservoir in the planning year, in m 3 ; Maximum reservoir water intake in the planning year, unit: m 3 ;

[0049] f) Reservoir storage constraints:

[0050]

[0051] where, is t End-of-month storage in the time period, unit: m 3 ; and are the minimum and maximum storage allowed in the reservoir t+1 during the time period, respectively;

[0052] g) Variable non-negativity constraints

[0053]

[0054]

[0055]

[0056] where, , , , , , , , , , , , , , , , , , , are interval parameters and variables.

[0057] Further supplement to the technical solution, in step four, solving the interval multi-objective optimization allocation model of regional surface water resources includes the following specific steps:

[0058] First step: introduce auxiliary variables to convert interval numbers in the objective function and constraint conditions into a determined form; the overall framework of the interval multi-objective optimization allocation model can be expressed as:

[0059]

[0060]

[0061]

[0062] By introducing auxiliary variables , and , can be interval number , , Transform it into a deterministic form; the objective function and constraints can be transformed into:

[0063]

[0064]

[0065]

[0066]

[0067]

[0068]

[0069] Step 2: Employ the fuzzy geometric weighted algorithm to transform the multi-objective programming problem into a single-objective programming problem with the goal of maximizing overall satisfaction. Introduce the membership function and calculate the membership degree of each objective. The formula for calculating the membership function is shown below:

[0070]

[0071] In the formula, For the first k Membership function of each target; and for Lower and upper bounds for single-objective linear programming, assuming Not equal to , and The objective functions are the same, but the constraints are different. The specific formulas are as follows:

[0072] :

[0073]

[0074]

[0075]

[0076]

[0077]

[0078]

[0079] :

[0080]

[0081]

[0082]

[0083]

[0084]

[0085]

[0086] Third step: according to the decision maker's decision preference for each objective function to give each objective membership function different weight wherein and ; and the original interval multi-objective optimization configuration model is converted into a single objective optimization configuration model with the goal of maximizing satisfaction, and the calculation formula is as follows:

[0087]

[0088]

[0089]

[0090]

[0091]

[0092]

[0093]

[0094]

[0095] Fourth step: using the optimal-worst model to solve the above interval single objective programming model with the goal of maximizing satisfaction; the optimal and worst sub-models are represented as follows:

[0096] Optimal model:

[0097]

[0098]

[0099]

[0100]

[0101]

[0102]

[0103] Worst model:

[0104]

[0105]

[0106]

[0107]

[0108]

[0109]

[0110] Fifth step: the optimal model is obtained, that is, the optimal optimal solution is substituted into the interval single-objective optimization configuration model with the maximum satisfaction as the target to obtain the maximum satisfaction , similarly, the solution obtained by the worst model, that is, the optimal worst solution is substituted into the interval single-objective optimization configuration model with the maximum satisfaction as the target to obtain the minimum satisfaction ; and satisfaction interval ]; the optimal optimal solution is substituted into , and is obtained, the optimal worst solution is substituted into , and is obtained, and then the interval solution of each target function ] is obtained; in addition, for the single objective with the maximum solving as the target, it needs to be multiplied by-1 to become the minimum solving as the target.

[0111] The beneficial effects are that: 1. The application collects data such as digital elevation, land use, soil type, historical meteorology, hydrological elements, etc. in the research area, establishes a SWAT model database for quantitative evaluation of regional surface water resources, and calibrates the model parameters by using the measured runoff data to verify the model, improve the simulation accuracy of runoff, and evaluate the applicability, which can provide data support for the management of surface water resources in the research area; the coupling of the SWAT model and future meteorological data can predict the future water resources in the research area on the basis of reflecting the hydrological dynamic characteristics;

[0112] 2, The present application considers the uncertainty and multi-objective in the regional surface water resources management process, forms a regional surface water resources interval multi-objective optimization allocation model by integrating the regional surface water resources quantitative evaluation simulation model into the interval multi-objective programming model, and solves the influence caused by climate change, system uncertainty and complexity in the regional surface water resources management process;The method can obtain the reservoir water resources allocation scheme meeting the power generation, water supply and water abandonment targets under different climate change and reservoir initial reservoir capacity scenarios on the basis of predicting the future regional surface water resources. BRIEF DESCRIPTION OF DRAWINGS

[0113] Figure 1 It is the whole working principle schematic flow chart in the present application;

[0114] Figure 2 It is the optimal and worst solution of monthly power generation water consumption and power generation capacity under different climate change scenarios;

[0115] Figure 3 It is the monthly water supply quantity and economic benefit optimization result under different climate change scenarios;

[0116] Figure 4 It is the monthly water abandonment quantity optimization result under different climate change scenarios. DETAILED DESCRIPTION

[0117] In order to make the technical personnel in the art more clearly understand the present technical solution, the following will combine the attached Figures 1-4 The technical scheme of the present application will be described in detail:

[0118] As Figure 1 shown in a kind of climate change background under regional surface water resources prediction simulation and its optimization allocation method, comprising the following steps:

[0119] Step one, data collection is carried out for the study area, and a watershed hydrological simulation model-SWAT (Soil and Water Assessment Tool) model is constructed;

[0120] The relevant data of the study area include: digital elevation data;Land use type data includes: construction land, farmland, forest land, grassland and water area in the study area;Soil type data includes: soil type spatial distribution data and soil attribute database in the study area;Meteorological data include: daily rainfall, temperature, wind speed, solar radiation, humidity in the study area in each year;Hydrological data include: measured data of runoff in each year in the hydrological monitoring station in the study area.

[0121] The steps of constructing the SWAT model to simulate historical runoff mainly include the following: using digital elevation data to analyze digital terrain and define river network, dividing sub-basins, and calculating sub-basin parameters; on the basis of the sub-basins, further dividing the sub-basins into a plurality of hydrological response units according to soil data, land use data and slope data; inputting meteorological data, and gradually calculating the runoff on each hydrological response unit, and then calculating the total runoff of the basin through confluence.

[0122] The calibration and verification of the SWAT model are performed, a hydrological monitoring site in the basin is selected, and the measured data of the site are used to correct the simulation data of the SWAT model, including the following steps: using the SWAT-CUP software to perform calibration and verification operation, and selecting the SUFI-2 algorithm to perform iterative operation to determine the optimal value of the parameters; adjusting the parameters of the SWAT model according to the optimal value of the parameters, and inputting the model for simulation verification; selecting the Nash coefficient (EF) and the correlation coefficient (R) as two indexes for evaluating the applicability of the SWAT model, and when the EF and the R are greater than 0.5 and 0.6 respectively, the SWAT model has applicability. NSE )and the correlation coefficient (R) as two indexes for evaluating the applicability of the SWAT model, and when the EF and the R are greater than 0.5 and 0.6 respectively, the SWAT model has applicability. R 2 )and the correlation coefficient (R) as two indexes for evaluating the applicability of the SWAT model, and when the EF and the R are greater than 0.5 and 0.6 respectively, the SWAT model has applicability. NSE >0.5, R 2 >0.6, the SWAT model has applicability.

[0123] Step two, collect future meteorological data and input it into the calibrated and verified SWAT model to predict and simulate the regional surface water resources under the background of climate change in the study area;

[0124] Specifically, it includes collecting the future daily rainfall, daily minimum temperature and daily maximum temperature of the meteorological station in the study area under different greenhouse gas emission (RCP) scenarios in the regional climate model.

[0125] Step three, integrate the regional surface water resources prediction and simulation model under the background of climate change into the interval multi-objective optimization model to form a regional surface water resources interval multi-objective optimization configuration model under the background of climate change;

[0126] Collect the basic information of the current water resources configuration in the study area, and construct a regional surface water resources interval multi-objective optimization configuration model, which includes a target function, decision variables and constraint conditions, and takes the water allocation to power generation, the water intake of the reservoir (hereinafter referred to as the water supply) and the abandoned water as the decision variables, and takes the maximum power generation, the maximum economic benefit of the water supply and the minimum abandoned water as the target function.

[0127] The target function is:

[0128] a) Maximum power generation:

[0129]

[0130] =

[0131] In the formula, for t Total power generated by the reservoir during the period, in kWh; For the reservoir in t Average power generation during the period, in MW (megawatts per hour). express t The duration of the time period is in days; T represents the total number of time periods. This indicates the reservoir's output coefficient; The water head of the reservoir is expressed in meters (m). Indicates that the reservoir is t Total water consumption for power generation during the period, in m³ 3 .

[0132] b) Water supply generates the greatest economic benefits:

[0133]

[0134] In the formula, for t The economic benefits generated by reservoir water supply during the period, in RMB; The economic benefit generated per unit of water supplied is expressed in RMB / m³. 3 ; The total water supply from the reservoir during time period t is expressed in cubic meters (m³). 3 .

[0135] c) Minimize water waste: Make full use of water resources

[0136]

[0137] In the formula, Total water volume to be discarded, in cubic meters (m³). 3 ; for t The amount of water released from the reservoir during the specified period, in cubic meters (m³). 3 ; T This represents the total number of time periods.

[0138] Constraints:

[0139] a) Reservoir water balance constraints:

[0140]

[0141] In the formula, and They are t+ 1 and tReservoir capacity during a given period, in meters (m). 3 ; I t yes t Reservoir inflow rate during a given period, in cubic meters (m³) 3 / s.

[0142] b) Turbine flow constraint:

[0143]

[0144]

[0145] In the formula, For the reservoir in t Power generation flow rate during a given period, in m³ 3 ; Q max,p for t Maximum water consumption for power generation in the reservoir during the specified time period, in cubic meters. 3 ; The maximum flow rate through the turbine of the hydropower station is expressed in cubic meters per second (m³). 3 / s, of which the maximum flow rate through the turbine of Xinfengjiang Reservoir is 490 m³ / s. 3 / s.

[0146] c) Reservoir discharge constraints:

[0147]

[0148]

[0149] In the formula, for t The amount of water discharged from the reservoir during the specified period, expressed in cubic meters (m³). 3 The amount of water discharged is equal to the sum of the water used for power generation and the amount of water wasted. for t Minimum discharge flow rate of the reservoir during a given period, in m³. 3 ; for t The maximum outflow from the reservoir during a given period, expressed in cubic meters per second (m³). 3 .

[0150] d) Power generation constraints:

[0151]

[0152] In the formula, The total power generation of the reservoir during the planning period is expressed in kWh. The minimum annual power generation of the reservoir is planned, expressed in kWh.

[0153] e) Reservoir water intake constraints:

[0154]

[0155] In the formula, The total water supply from the reservoir during time period t is the total water intake from the reservoir, expressed in cubic meters (m³). 3 ; This represents the minimum water intake from the reservoir in the planned year, expressed in cubic meters (m³). 3 ; The planned annual maximum water intake from the reservoir, in cubic meters. 3 ;

[0156] f) Reservoir capacity constraints:

[0157]

[0158] In the formula, for t+1 Month-end storage capacity for a given period, in meters (m). 3 ; and Reservoirs t Minimum and maximum storage capacity allowed within the time period.

[0159] g) Variable nonnegation constraint

[0160]

[0161]

[0162]

[0163] In the formula, , , , , , , , , , , , , , , , , , , These are interval parameters and variables.

[0164] Step 4: Solve the multi-objective optimization allocation model of regional surface water resources under the background of climate change to obtain the regional surface water resources optimization allocation scheme that takes into account future climate change, system uncertainty and multi-objectives;

[0165] The main steps for solving include: introducing membership function, using fuzzy geometric weighted algorithm to convert the multi-objective problem in the interval multi-objective optimal allocation model of regional surface water resources into a single objective problem of solving the maximum satisfaction, converting the interval number into a determined value by introducing auxiliary variables, and converting the single objective problem of solving the maximum satisfaction into optimal sub-model and worst sub-model by using optimal worst model, and calculating the optimal optimal solution and optimal worst solution of the reservoir power generation water consumption, reservoir water intake, and abandoned water, wherein the optimal optimal solution and optimal worst solution of each decision variable can form the interval solution of the decision variable. The interval solution of the reservoir power generation and the interval solution of the economic benefits of the reservoir water intake are obtained by substituting the decision variable into the corresponding objective function.

[0166] The specific steps for solving include:

[0167] The first step is to introduce auxiliary variables to convert the interval numbers in the objective function and the constraint conditions into determined forms. The overall framework of the interval multi-objective optimal allocation model can be expressed as:

[0168]

[0169]

[0170]

[0171] By introducing auxiliary variables , and , the interval numbers , , can be converted into determined forms. The objective function and the constraint conditions can be converted into:

[0172]

[0173]

[0174]

[0175]

[0176]

[0177]

[0178] The second step is to use fuzzy geometric weighted algorithm to convert the multi-objective planning problem into a single objective planning problem with the goal of maximizing overall satisfaction. The membership function is introduced, and the membership of each objective is calculated. The membership function calculation formula is as follows:

[0179]

[0180] where, is the membership function of the kth objective; and is Lower and upper bounds of single-objective linear programming, assuming is not equal to , is the same as the objective function of , with different constraints, and the specific formula is as follows:

[0181] :

[0182]

[0183]

[0184]

[0185]

[0186]

[0187]

[0188] :

[0189]

[0190]

[0191]

[0192]

[0193]

[0194]

[0195] Step 3: Assign different weights to each objective membership function according to the decision maker's decision preference for each objective function where and . Further, the original interval multi-objective optimization configuration model is transformed into a single-objective optimization configuration model with the goal of maximizing satisfaction, and the calculation formula is as follows:

[0196]

[0197]

[0198]

[0199]

[0200]

[0201]

[0202]

[0203]

[0204] Step 4: The interval single-objective programming model above with the goal of maximizing satisfaction is solved by using the optimal-worst model. The optimal and worst sub-models are expressed as follows:

[0205] Optimal model:

[0206]

[0207]

[0208]

[0209]

[0210]

[0211]

[0212] Worst model:

[0213]

[0214]

[0215]

[0216]

[0217]

[0218]

[0219] Step 5: The solution obtained by the optimal model, i.e., the optimal optimal solution, is substituted into the interval single-objective programming model with the goal of maximizing satisfaction to obtain the maximum satisfaction . Similarly, the solution obtained by the worst model, i.e., the optimal worst solution, is substituted into the interval single-objective programming model with the goal of maximizing satisfaction to obtain the minimum satisfaction . and Composition satisfaction interval ]. The optimal and worst solutions are substituted into , and is obtained. The optimal and worst solutions are substituted into , and is obtained. Further, each objective function interval solution is obtained. In addition, for a single objective aiming at solving maximization, negative 1 is multiplied to become aiming at solving minimization.

[0220] The application is further described below by examples combined with the accompanying drawings: Figures 1-4

[0221] Optimal allocation of water resources in Xinfengjiang Reservoir Basin

[0222] 1. Overview of the study area

[0223] The Xinfengjiang Reservoir Basin is located between 23°40′-24°36′N and 113°57′-115°05′E, and is situated in the subtropics with high temperature and abundant rainfall. The regional vegetation coverage is high, and the water resources are significantly affected by climate change. The Xinfengjiang Reservoir Basin is located in the west of Heyuan City, Guangdong Province, and is distributed in the northeast of the Pearl River Basin in a diamond shape. It originates from the Jiulian Mountain area, and includes the Xinfengjiang River, the Lianping River, the Dizhi River, the Zhongxin River and other main water systems. From west to south, it flows into the East River through Heyuan City and Shaoguan City. The total length of the river is 163 km, and the water area is about 370 km 2 . The terrain in the basin slopes from northwest to southeast, and the shape of the basin is fan-shaped. The confluence time is short, and the length of the main stream and tributaries is small, the water surface area is large, the flood process is easy to rise and difficult to fall, and the average runoff coefficient is 58%. The Xinfengjiang Reservoir has a catchment area of 5813 km 2 , a total reservoir capacity of 13.9 billion m 3 , and an average annual inflow and outflow of 6.1 billion m 3 .The Xinfengjiang Reservoir is the seventh largest reservoir in China and the largest artificial lake in South China. It was built in 1958 and serves multiple purposes, including flood control, water supply, navigation, and power generation. In terms of water supply, the Xinfengjiang Reservoir not only directly supplies water to Heyuan City but also provides water to downstream cities like Huizhou and Dongguan through the Dongjiang River. Additionally, it supplies a large amount of water to cities outside the watershed, such as Shenzhen, Guangzhou, and Hong Kong, through inter-basin water transfer projects. As a result, the Xinfengjiang Reservoir is a crucial source of drinking water for Guangdong Province and is often referred to as "political water," "life water," and "economic water." Therefore, ensuring the safety of the Xinfengjiang Reservoir's water supply is of great importance to the production and daily life of the over 40 million people in the downstream area. However, in the past decade, the monthly water supply from the Xinfengjiang Reservoir to Heyuan City has been increasing, and the annual water supply has significantly increased. This indicates that the demand for water from the Xinfengjiang Reservoir has been increasing, and the reservoir is facing significant pressure. Since 2002, the Xinfengjiang Reservoir has shifted its focus from power generation to flood control and water supply, while still considering power generation, navigation, saltwater prevention, and irrigation. It has become one of the most important water supply rivers in Guangdong Province. The control of water quantity in the Xinfengjiang Reservoir plays a crucial role in regulating the temporal and spatial distribution of flood and low-flow water in the middle and lower reaches of the Dongjiang River. Therefore, a reasonable plan for the water resources of the Xinfengjiang Reservoir is of great strategic importance to the sustainable development of the cities in the middle and lower reaches of the Dongjiang River, as well as Shenzhen and Hong Kong. In conclusion, it is necessary to reasonably allocate the water resources of the Xinfengjiang Reservoir, especially considering the impact of climate change on water resources in this region. Therefore, it is necessary to study the optimal allocation of water resources in the Xinfengjiang Reservoir under the background of climate change.

[0224] 2. Collection of basic data

[0225] A large amount of data is required to support the operation of the SWAT model, which can be broadly divided into spatial data and attribute data. Spatial databases mainly include DEM digital elevation databases, river system maps, land use databases, and soil type databases. Attribute databases mainly include soil attribute databases and meteorological databases. In this paper, the integrated version of SWAT2012 and ArcGIS10.2 is used to process the spatial data. To meet the requirements of spatial data for overlay analysis and unit division, and considering the detailed regional characteristics and spatial data of the river basin, the geographic coordinate system WGS-1984 and the projection coordinate system UTM-50 are used. The specific input data of the SWAT model and their sources are shown in the table below:

[0226]

[0227] 3. Calibration and validation of SWAT model parameters

[0228] This invention utilizes the SUFI-2 algorithm in SWAT-CUP for parameter sensitivity analysis. Through multiple iterations, sensitive parameters are selected and their value ranges are adjusted to ultimately determine the main parameters for calibration. Finally, this invention selects 19 watershed parameters that significantly impact runoff. The sensitivity analysis of these parameters employs the t-test (t-stat) and p-test (p-value) methods for global sensitivity analysis. In this process, the higher the parameter sensitivity, the larger the absolute value of the t-stat, and the closer the p-value is to zero. The final determined main sensitive parameters and their optimal value ranges are shown in the table below:

[0229]

[0230] The main steps of SWAT model calibration and validation are as follows: Import the SWAT output file of the SWAT model simulation results into SWAT-CUP; use sensitivity analysis to determine the parameters that significantly affect the simulation results and their value ranges; run the SWAT-CUP project; adjust the parameter values ​​according to the new parameter value range recommended by SWAT-CUP, and iterate back using the new parameter range until the calibration results are controlled within a reasonable range; finally, substitute the parameter value range determined after calibration back into the SWAT model, modify the optimal values ​​of the sensitive parameters, rerun the SWAT model, and then import the output results into SWAT-CUP for validation. Specifically, based on parameter sensitivity analysis, this invention uses measured data from Shuntian (located in sub-basin 16) and Yuecheng (located in sub-basin 18) from 2010 to 2014 to calibrate and validate the monthly-scale simulation results of the SWAT model using multi-site, multi-variable data. The period from 2010 to 2012 is the calibration period, and 2013 to 2014 is the validation period.

[0231] Based on the selected evaluation index range, it can be seen that the SWAT model has good applicability in the Xinfengjiang Reservoir watershed. During the calibration period (2010-2012), the coefficient of determination for the monthly runoff simulation at the Shuntian Hydrological Station was [data missing]. R 2 The Nash coefficient is 0.90. NSE The coefficient of determination for the monthly runoff simulation at Yuecheng Hydrological Station is 0.90. R 2 The Nash coefficient is 0.74. NSE The coefficient of determination for the monthly runoff simulation at the Shuntian Hydrological Station was 0.68. This was during the validation period (2013-2014). R 2 The Nash coefficient is 0.81. NSE The coefficient of determination for the monthly runoff simulation at Yuecheng Hydrological Station is 0.71. R 2 The Nash coefficient is 0.85.NSE is 0.55.In general, the SWAT model is suitable for follow-up studies.

[0232] 4. Runoff prediction of Xinfengjiang Reservoir under climate change scenarios

[0233] The present application selects 2025 as the prediction year, collects daily data in the HadGEM2-ES model under three different greenhouse gas scenarios of RCP2.6, RCP4.5 and RCP8.5 in 2025, reads the nc file by using Panoply software, and sorts out the data in the study area that meets the input data requirements of the SWAT model. Then, the SWAT model is input into the SWAT model which has been calibrated and verified to simulate the runoff in 2025. It can be seen that: there is no significant difference in annual runoff under RCP2.6, RCP4.5 and RCP8.5 scenarios; the annual runoff under RCP2.6, RCP4.5 and RCP8.5 scenarios in 2025 is 189.37, 238.16 and 217.77 m 3 / s respectively. Specifically, the runoff under RCP2.6, RCP4.5 and RCP8.5 scenarios is concentrated in May-September, and the runoff in other months is relatively small. The runoff peak under RCP2.6, RCP4.5 and RCP8.5 scenarios appears in July, May and August respectively.

[0234] 5. Multi-objective uncertain optimization allocation of reservoir water resources

[0235] The present application takes Xinfengjiang Reservoir as the research area, and optimizes the allocation of reservoir water resources in Xinfengjiang Reservoir in 2025: on the basis of predicting the amount of reservoir surface water resources under climate change scenarios, collecting the current situation of reservoir operation and the allocation scheme of reservoir water resources, taking the maximization of economic benefits of water supply, the maximization of power generation and the minimization of water abandonment as the optimization objectives, considering the water balance constraint, the reservoir capacity constraint, the water supply constraint, the power generation constraint, the discharge flow constraint, the maximum flow through the turbine constraint and the non-negative constraint, etc.; analyzing the uncertainty information in the water resources optimization allocation system, introducing interval mathematical programming and multi-objective programming method, and constructing an interval multi-objective uncertain optimization model of reservoir water resources. The model constructed in this paper has the following advantages: ① the benefits of multiple objectives can be weighed to maximize the comprehensive benefits; ② the parameters or variables that are difficult to determine in the system are represented as interval numbers to effectively represent the uncertainty of the system; ③ the water resources optimization allocation scheme under multiple climate change and initial reservoir capacity scenarios is discussed, thereby providing multiple decision-making suggestions for decision-makers.

[0236] The interval satisfaction of the corresponding scheme under the three climate change scenarios is [0.49, 0.99], indicating that the water resource allocation schemes under the three scenarios are applicable, and the decision maker can select the corresponding water resource allocation scheme of the reservoir according to the actual rainfall and temperature conditions of the reservoir in 2025. Overall, under the three schemes, the water used for power generation is the most, accounting for about 60% of the total water resource allocation. The water supply is the second, accounting for about 35%. Relatively speaking, the abandoned water is the least, not more than 5% of the total water resource allocation. In addition, different climate change scenarios have little effect on the water used for power generation and the water supply of the reservoir, but have a greater impact on the abandoned water of the reservoir.

[0237] In terms of water used for power generation, under the three climate change scenarios, the water allocated to power generation in April-September is more than that in other months, and the water used for power generation in April-September is 1 x 10 9 m 3 Above. For the power generation of the reservoir, combined with the power generation results under the three scenarios, the total power generation of the Xinfengjiang Reservoir in 2025 is predicted to be [12, 20] kWh. The optimal and worst solutions of the monthly power generation water and power generation under different climate change scenarios are shown in Figure 2 :

[0238] In terms of total water supply, the total water supply under different climate change scenarios is the same interval solution [5.48 x 108, 5.50 x 108] m3, which shows that the total water supply does not change with the change of climate change scenarios. In addition, the upper and lower bounds of the total water supply interval solution are not much different, leaving less adjustment space for the decision maker, which will be more helpful for the decision maker to make decisions. The optimal results of the monthly water supply and its economic benefits under different scenarios are shown in Figure 3 :

[0239] In terms of total abandoned water, with the change of climate scenarios, the total abandoned water and the monthly abandoned water have great differences. Under each scenario, the upper and lower bounds of the monthly abandoned water interval solution differ greatly, indicating that the adjustment space of the abandoned water is large. In May, July and August, the upper and lower bounds of the abandoned water interval solution differ by more than 5 orders of magnitude. Combined with the fact that the reservoir runoff under the three future climate change scenarios also concentrates in May-August, it can be concluded that if the actual inflow of the reservoir is too large, it will increase the possibility of abandoned water of the reservoir, threatening the safe operation of the hydropower station. Conversely, if the actual inflow of the reservoir is too small, when the water level of the reservoir breaks the lower limit, the hydropower station will be forced to reduce the output. Therefore, setting the abandoned water target in the optimization model is an effective way to improve the efficiency of water resource utilization and ensure the safe and stable operation of the power grid. The optimal results of the monthly abandoned water under different climate change scenarios are shown in Figure 4 .

[0240] The above technical solution only reflects the preferred technical solution of the present application, and some changes made by the skilled in the art to some parts thereof also reflect the principle of the present application and are within the protection scope of the present application.

Claims

1. A method for predicting and simulating regional surface water resources and optimizing its configuration, characterized in that, Comprise the following steps: Step one, data collection for the study area, build the watershed hydrological simulation model-SWAT model; Step two, collect future weather data and input it into the calibrated and verified SWAT model to predict and simulate the regional surface water resources under the background of climate change; Step three, integrate the regional surface water resources prediction and simulation model under the background of climate change into the interval multi-objective optimization model to form the regional surface water resources interval multi-objective optimization allocation model under the background of climate change; In step three, RCP meteorological scenarios, SWAT model, interval linear programming, and multi-objective programming are coupled in one framework to build the regional surface water resources interval multi-objective optimization allocation model under the background of climate change; Step four, solve the regional surface water resources interval multi-objective optimization allocation model under the background of climate change to obtain the optimal allocation scheme of regional surface water resources considering future climate change, system uncertainty and multi-objective; In step four, the membership function is introduced, the fuzzy geometric weighted algorithm is used to convert the multi-objective problem in the regional surface water resources interval multi-objective optimization allocation model into a single objective problem with the maximum satisfaction, auxiliary variables are introduced to convert interval numbers into deterministic values, and the optimal worst model is used to convert the single objective problem with the maximum satisfaction into optimal sub-model and worst sub-model to calculate the optimal solution and worst solution of the reservoir power generation water consumption, reservoir water intake, and abandoned water, wherein the optimal solution and worst solution of each decision variable can form an interval solution of the decision variable; the decision variables are substituted into the corresponding objective functions to obtain the interval solution of the reservoir power generation and the interval solution of the economic benefits generated by the reservoir water intake; The constructed interval multi-objective optimization allocation model includes objective functions, decision variables and constraint conditions, and the reservoir water allocation to power generation, reservoir water intake and abandoned water are used as decision variables, and the maximum reservoir hydroelectric power, the maximum economic benefits generated by the reservoir water intake and the minimum abandoned water are used as objective functions; The objective functions of the interval multi-objective optimization allocation model include: a) Maximum power generation: = In the formula, for t Total power generated by the reservoir during the period, in kWh; For the reservoir in t Average power generation during the period, in MW; This indicates the reservoir's output coefficient; The water head of the reservoir is expressed in meters (m). Indicates that the reservoir is t Total water consumption for power generation during the period, in m³ 3 ; express t The duration of each time period is in days; T represents the total number of time periods. b) Maximum economic benefits generated by water supply: In the formula, is t the economic benefit of water supply from the reservoir in the period t, in RMB; is the economic benefit of unit water supply, in RMB / m 3 ; is the total water supply from the reservoir in the period t, in m 3 ; c) Minimum abandoned water: fully utilize water resources: In the formula, is the total amount of abandoned water, in m 3 ; is t abandoned water in the period, in m 3 ; Each constraint condition: a) Reservoir water balance constraint: wherein and are t+ 1 and t the reservoir storage capacity in m 3 for the time period; I t are t the reservoir inflow in m 3 / s for the time period; b) Turbine flow constraint: In the formula, Q max,p is t the maximum power generation water volume of the reservoir in the period, in m 3 ; is the maximum turbine flow of the hydropower station, in m 3 / s, wherein the maximum turbine flow of the Xinfengjiang Reservoir is 490 m 3 / s; c) Reservoir discharge constraint: In the formula, is t the water release of the reservoir in the period, in m 3 wherein the water release is equal to the sum of the water for power generation and the abandoned water; is t the minimum water release of the reservoir in the period, in m 3 ; is t the maximum water release of the reservoir in the period, in m 3 ; d) Power generation constraint: In the formula, is the total power generation of the reservoir in the planning period, in kWh; is the minimum power generation required by the reservoir in the planning year, in kWh; e) Reservoir water intake constraint: In the formula, is the total water supply volume of the reservoir in the planning year, in m t ; 3 ; is the minimum reservoir water intake of the reservoir in the planning year, in m 3 ; is the maximum reservoir water intake of the reservoir in the planning year, in m 3 ; f) Reservoir storage constraint: In the formula, With Reservoir t+1 The minimum and maximum reservoir capacity allowed within the period, in m³; g) Variable non-negative constraint wherein , , , , , , , , , , , , , , , , , , are interval parameters and variables.

2. The method of claim 1, wherein, In step one, the data of the study area includes digital elevation data, hydrological data, meteorological data, land use data and soil type data of the study area, and the steps of building the SWAT model include sub-basin division, hydrological response unit division, input meteorological data, parameter sensitivity analysis, model calibration and verification.

3. The method according to claim 2, wherein, In step two, the constructed SWAT model is calibrated and verified based on the measured hydrological data, and the planning year daily rainfall, daily minimum temperature and daily maximum temperature under different greenhouse gas emission scenarios are collected and input into the calibrated and verified SWAT model to predict and simulate the regional surface water resources under the background of climate change.

4. The method according to claim 2, wherein, The land use type data includes construction land, cultivated land, forest land, grassland and water area in the study area; the soil type data includes the spatial distribution data of soil type in the study area and the soil attribute database; the meteorological data includes the daily rainfall, air temperature, wind speed, solar radiation and humidity in the study area in each year; the hydrological data includes the measured runoff data of the hydrological monitoring stations in the study area in each year; The steps of constructing the SWAT model to simulate the historical runoff are as follows: the digital terrain is analyzed and the river network is defined by using the digital elevation data, the sub-basins are divided and the parameters of the sub-basins are calculated; on the basis of the sub-basins, the sub-basins are further divided into a plurality of hydrological response units according to the soil data, the land use data and the slope data; the meteorological data is inputted, the runoff on each hydrological response unit is calculated gradually, and the total runoff of the basin is obtained through confluence calculation.

5. The method of claim 4, wherein the method further comprises: The calibrated and verified SWAT model is constructed, the hydrological monitoring station in the basin is selected, the measured hydrological data of the station is used to correct the simulation data of the SWAT model, including the following steps: using the SWAT-CUP software to calibrate and verify the operation, selecting the SUFI-2 algorithm to determine the optimal value of the parameter by iteration operation; the parameters of the SWAT model are adjusted according to the optimal value of the parameters, and the model is simulated and verified; Nash coefficient NSE and correlation coefficient R 2 as two indexes for evaluating the applicability of the SWAT model, and NSE > 0.5, R 2 SWAT model has applicability when > 0.

6.

6. The method of claim 1, wherein the method is characterized by: In step four, solving the interval multi-objective optimization allocation model of regional surface water resources includes the following specific steps: Step one: introducing auxiliary variables, converting the interval numbers in the objective function and the constraint conditions into a definite form; the overall framework of the interval multi-objective optimization allocation model can be expressed as: By introducing auxiliary variables , and , the interval numbers , , can be converted into a deterministic form; the objective function and the constraints are converted into: Step two: using fuzzy geometric weighting algorithm to convert the multi-objective programming problem into a single-objective programming problem with the maximum overall satisfaction as the target; introducing the membership function and calculating the membership degree of each target; The formula for calculating the membership function is as follows: In the formula, For the first k Membership function of each target; and for Lower and upper bounds for single-objective linear programming, assuming Not equal to , and The objective functions are the same, but the constraints are different. The specific formulas are as follows: ： ： Third step: according to the decision maker's decision preference for each objective function, give each objective membership function different weights wherein and ; and then the original interval multi-objective optimization configuration model is converted into a single objective optimization configuration model with the maximum satisfaction, and the calculation formula is as follows: Step four: using the optimal-worst model to solve the above interval single-objective optimization allocation model with the maximum satisfaction as the target; the optimal and worst sub-models are expressed as follows: Optimal model: Worst model: Fifth step: the optimal model to get the solution, that is, the optimal optimal solution to the maximum satisfaction of the interval single-objective optimization configuration model to get the maximum satisfaction , similarly, the worst model to get the solution, that is, the optimal worst solution to the maximum satisfaction of the interval single-objective optimization configuration model to get the minimum satisfaction ; and satisfaction interval ]; the optimal optimal solution is substituted into , get , the optimal worst solution is substituted into , get , and then get the interval solution of each objective function ]; in addition, for the single objective of solving the maximum, multiply by-1 to become the minimum.

Images