Three gorges reservoir operation period staging design flood and staging water storage level determination method

Abstract

The invention provides a three gorges reservoir operation period staging design flood and staging storage level determination method, which is characterized in that a flood season is divided into a previous flood season, a main flood season and a later flood season by considering a flood seasonal change rule on the basis of a long-series rainfall flow data series; taking the corrected reservoir coefficient as a covariable to consider the regulation and storage influence of the upstream reservoir group, and adopting a P-III distribution time-varying moment model to deduce stage design flood and flood control water level in the reservoir operation period; a Jinsha river downstream cascade and Three Gorges reservoir combined water storage multi-target scheduling model is constructed, and on the premise of keeping flood control standards unchanged, the three Gorges reservoir staged water storage level is determined through a parameterization-simulation-optimization framework and an intelligent algorithm. The reservoir operation period staging design flood and staging water storage level determination method provided by the invention is reasonable and feasible in technology, and provides an important and operable reference basis for three gorges reservoir operation scheduling management.

Description

Technical Field

[0001] The present invention relates to the technical field of water conservancy and hydropower engineering, and particularly relates to a method for determining the staged design flood and staged storage levels during the operation period of the Three Gorges Reservoir. Background Art

[0002] More than 40,000 reservoirs have been built in the upper reaches of the Yangtze River (above Yichang), including 117 large reservoirs, with a total regulating storage capacity of 84.2 billion m³ and a reserved flood control storage capacity of 50.1 billion m³. In recent years, with global warming and changes in the underlying surface conditions of the basin, especially affected by the construction and operation of water conservancy projects, the runoff generation and concentration laws in the upper reaches of the Yangtze River have changed significantly. According to statistics, in recent years, the water inflow at the Yichang Station of the Yangtze River has decreased by nearly 10%, and the sediment inflow has decreased by 80%-90%. At the same time, the role of the reservoir group in storing floodwater during the flood season and releasing it during the dry season is obvious, and the annual distribution law of the incoming water has changed significantly. The dilemma of "no water to store" for the reservoir group during the end-of-flood-season water storage period has become prominent. The development and construction of main-stream hydropower in China's basins are basically carried out in the order from bottom to top. The design flood and the determined flood limit water level in the initial design stage of the Three Gorges Reservoir in early 1993 neither considered the seasonal variation law of basin floods nor the regulation influence of the reservoir group in the upper reaches of the Yangtze River. After the operation of the Three Gorges Reservoir, the water levels at the outlets of Dongting Lake and Poyang Lake during the water storage period have dropped by 1.2 m and 1.09 m respectively, exacerbating the severity of drought and water shortage in the lake areas under the background of extreme weather. At the same time, the functions of the Three Gorges Reservoir and social demands have also undergone major adjustments in recent years. In addition to ensuring flood control safety, power generation, shipping, water supply, and ecological goals have been placed in a prominent position. Currently, China's reservoirs basically follow the design flood results during the construction period and use the flood limit water level to guide the reservoir operation and dispatch, often resulting in the situation of forced water abandonment when floods occur during the flood season and difficulty in filling up at the end of the flood season, affecting the "peak shaving and filling the dry season" function of the reservoir group; the low water level operation of the reservoir during the flood season not only affects the comprehensive utilization benefits but also restricts the ecological environment protection of the basin.

[0003] Under the background of global climate change, climate variability has been very large in recent years. It is necessary not only to study the staged design flood during the reservoir operation period but also to determine the staged storage levels in combination with the changes in the comprehensive dispatching requirements such as power generation, water supply, and shipping. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a method for determining the staged design flood and staged storage levels during the operation period in view of the deficiencies in the existing design and operation dispatching of the Three Gorges Reservoir, so as to provide a theoretical reference for the operation and dispatching during the reservoir operation period.

[0005] To solve the above technical problems, the technical solution adopted by the present invention is: A method for determining the staged design flood and staged storage levels during the operation period of the Three Gorges Reservoir, including the following steps:

[0006] Step 1: Collect the construction and operation information of key large reservoirs in the upper reaches of the Yangtze River and the observation data of basin hydrological stations, and restore the natural runoff process;

[0007] Step 2: Based on the flood season staging method of mathematical statistics, use the measured historical flow data, select statistical indicators, analyze the variation law of the indicators during the flood season, and adopt the mean change point analysis method to determine the staging node time;

[0008] Step 3: Deduce the staged design flood during the operation period of the Three Gorges Reservoir;

[0009] According to the staging results of Step 2, obtain the flood data sequence by the staged annual maximum sampling method, consider the regulation influence of the upstream reservoir group by correcting the reservoir coefficient, adopt the time-varying P-III type curve fitting method to deduce the staged design flood during the operation period of the Three Gorges Reservoir, select the typical year's staged flood hydrograph and amplify the staged design flood hydrograph by the same frequency method, and obtain the staged flood control water level through the flood regulation calculation of the Three Gorges Reservoir;

[0010] Step 4: Determine the staged water storage levels of the Three Gorges Reservoir;

[0011] Construct a multi-objective operation model for the joint water storage of the cascades in the lower reaches of the Jinsha River and the Three Gorges Reservoir, solve it by using the "parameterization - simulation - optimization" framework structure and intelligent algorithms, comprehensively consider the flood control and beneficial operation requirements, and finally determine the staged water storage levels of the Three Gorges Reservoir through the simulation and regulation of flood control of the cascade reservoirs on the premise of keeping the flood control standard unchanged.

[0012] Preferably, Step 1 specifically includes:

[0013] Step 1.1: Collect the construction and operation scheduling information of large key reservoirs in the upper reaches of the Yangtze River and the data series of basin rainfall and hydrological observation stations;

[0014] Step 1.2: Invert the flow process under the natural state of each control section according to the flow observation data of the hydrological station.

[0015] Preferably, in Step 1.2, the inversion adopts the water balance equation to deduce the inflow.

[0016] Preferably, Step 2 specifically includes:

[0017] Step 2.1: Statistically describe the frequency distribution of the rainstorm amount or flood peak flow during the flood season by ten-day periods, plot the scatter diagram of the indicators and take the envelope line to determine the dividing points of the flood season staging;

[0018] Step 2.2: The start and end times of the flood season are determined based on the rising situation of the rivers, specifically using the relative coefficient of the ten-day precipitation change as the measurement index, that is:

[0019]

[0020] Where: C R is used to represent the relative amount of dekadal precipitation; R is the multi-year average precipitation of a certain dekad in a year at a certain station; is the average dekadal precipitation within a year at a certain station; when C of a certain station for 2 or more consecutive dekads R ≥1.5, it is called the flood season;

[0021] Step 2.3, use the mean change model to describe the change law of the flood season.

[0022] Preferably, the specific implementation method of the mean change point model in Step 2.3 is as follows:

[0023] Let the sequence X = [x1, x2,... x n represent the multi-year daily average flow sequence of the flood season at a certain main and tributary control station. Assuming it follows a normal distribution, judge the staging point according to the following steps:

[0024] Step 2.3.1, assume the number of change points q, take 1 or 2, corresponding to two periods and three periods respectively. When q = 2, take all (m1, m2) combinations where 1 < m1 < m2 < n; among them,

[0025] Step 2.3.2, for each group (m1, m2), calculate the statistic S according to the following formula:

[0026]

[0027] Step 2.3.3, take the group (m1, m2) combination with the smallest statistic S as the final change point staging result, that is, the (m2 + 1)-th to n-th days of the flood season are the post-flood season;

[0028] Repeat the above steps for each main and tributary control station to obtain the start time of the post-flood season in each area:

[0029] M2 = (m2(1), m2(2),..., m2(k)) + 1;

[0030] Where: k is the number of main and tributary control stations.

[0031] Preferably, the specific method of Step 3 is as follows:

[0032] Step 3.1, collect and analyze the measured flood data series at the reservoir dam site, and directly use the staged annual maximum sampling to further organize it into a flood peak and flood volume data series;

[0033] Step 3.2, collect the catchment area and total reservoir capacity values of all large reservoirs upstream of the reservoir. According to the construction and operation conditions of key large reservoirs upstream of the reservoir, use the corrected reservoir coefficient RI to measure the regulation effect of the upstream reservoir group on the design flood of the reservoir:

[0034]

[0035] Where: N is the number of large reservoirs above the study section; A i and C i are the catchment area and regulation storage capacity of the i-th reservoir, respectively; A T and C T are the catchment area of the study section and the average annual runoff, respectively;

[0036] Step 3.3: Taking the reservoir coefficient RI as the covariate, let Y t be the response variable at time t (t = 1, 2,..., n), and y t be its observed value, and construct a time-varying P-III frequency distribution model:

[0037]

[0038] Where: μ t , σ t and ν t are the location parameter, scale parameter and shape parameter of the time-varying P-III distribution, respectively;

[0039] Step 3.4: According to the generalized linear additive model, the parameters [μ t , σ t , ν t of the time-varying P-III distribution can be further expressed as:

[0040]

[0041] Where: α = [α 10 , α 11 , α 20 , α 30 are the parameters of the generalized linear model in the time-varying P-III distribution, and g i (·) (i = 1, 2, 3) are monotonic continuous link functions, and their specific forms are determined by the domain of their independent variables;

[0042] Step 3.5: Taking the minimum sum of squared deviations as the criterion for fitting the curve, fit the empirical frequency quantiles of the measured flood samples with the theoretical distribution quantiles of the time-varying P-III distribution under the same samples, and estimate the statistical parameters and design values of the time-varying P-III distribution;

[0043] Step 3.6: Select the typical hydrograph of the stage design flood, which has the characteristics of continuous multi-peaks and concentrated flood volumes. The flood peak shape and its time course distribution are relatively severe for flood control in the middle and lower reaches, and it has a certain representativeness in terms of the flood regional composition;

[0044] Step 3.7: Use the same-frequency amplification method to separately derive the flood hydrographs for the 1000-year and 100-year return periods for different stages; obtain the flood control water levels for each stage through iterative solution of the reservoir flood routing calculation.

[0045] Preferably, the specific method for step 4 is as follows:

[0046] Step 4.1: Construct a multi-objective operation model for the cascade reservoirs in the lower reaches of the Jinsha River and the Three Gorges Reservoir for joint water storage, and solve it using the "parameterization - simulation - optimization" framework and multi-objective intelligent algorithms;

[0047] Step 4.2: Based on long-term measured data and the flood hydrographs for different stages of design floods, conduct a multi-objective operation simulation study on the cascade reservoirs in the lower reaches of the Jinsha River and the Three Gorges Reservoir for joint water storage, and finally determine the staged water storage levels of the Three Gorges Reservoir through comparison of multiple scenarios.

[0048] Preferably, the specific steps for step 4.1 are as follows:

[0049] Step 4.1.1: Determine the objective function:

[0050] The main objective during the water storage period is to fill the reservoir as much as possible at the end of the water storage period while meeting the existing flood control standards, and also to generate as much power generation benefit as possible during this period. The objective function is selected as the maximum full storage rate and power generation, and the minimum flood control risk;

[0051] Step 4.1.2: Determine the constraint conditions:

[0052] The cascade reservoirs still need to comply with some basic constraint conditions, specifically including water balance constraints, reservoir storage capacity constraints, reservoir discharge flow constraints, power station output constraints, and boundary conditions;

[0053] Step 4.1.3: Solve using an optimization algorithm:

[0054] Use Borg MOEA to optimize and solve the multi-objective problems of the cascade reservoirs.

[0055] Preferably, the objective function in step 4.1.1 specifically includes:

[0056] (1) The maximum full storage rate IE, expressed as a percentage of the storage capacity:

[0057]

[0058] In the formula: IE k,i represents the average annual full storage rate of the i-th reservoir in the k-th sub-region; VE k,i (y) is the storage capacity of the i-th reservoir in the k-th sub-region at the end of the water storage period; SU k,i and SD k,i represent the normal beneficial storage capacity and dead storage capacity of the i-th reservoir in the k-th sub-region respectively; K and M krespectively represent the total number of partitions and the number of reservoirs in the k-th partition; Y represents the number of water storage years;

[0059] (2) Minimum flood control risk (FCR):

[0060] min FCR = min{max{FCR(t)}},(0 < t ≤ T·Y);

[0061]

[0062] In the formula: FCR(t) represents the flood control risk of the reservoir group system at time t; FCR k,i (t) corresponds to the flood control risk of the i-th reservoir in the k-th partition at time t; V k,i (t) and SS k,i (t) are the reservoir capacity of the i-th reservoir in the k-th partition at time t and the reservoir capacity corresponding to the highest safe water level in front of the dam respectively; T is the total number of time periods in the annual water storage period;

[0063] (3) The power generation E of the cascade reservoirs in each partition k is maximized:

[0064]

[0065] In the formula: N k,i (t) is the average output of the i-th reservoir in the k-th partition at time t; △t is the step interval of the water storage period; A k,i is the average power generation efficiency coefficient of the i-th reservoir in the k-th partition; Q k,i (t) and H k,i (t) are the power generation flow rate and power generation head of the i-th reservoir in the k-th partition at time t respectively.

[0066] Preferably, the constraint conditions in step 4.1.2 specifically include:

[0067] (1) Water balance constraint:

[0068]

[0069] In the formula: I k,i (t) and O k,i (t) are the inflow and outflow flows of the i-th reservoir in the k-th partition at time t; Φ j,k,i represents the set of reservoirs upstream of the i-th reservoir in the k-th partition that have hydraulic connections with it; O j (t) is the outflow flow of the j-th reservoir in the Φ j,k,i set;

[0070] (2) Reservoir capacity constraint:

[0071] SL k,i (t) ≤ V k,i (t) ≤ SSk,i (t);

[0072] Where: SL k,i (t) and SS k,i (t) are the minimum and maximum allowable water storage volumes of the i-th reservoir in the k-th sub-region at time t, respectively;

[0073] (3) Constraint on the discharge flow of the reservoir:

[0074] OL k,i (t) ≤ O k,i (t) ≤ OU k,i (t), O k,i (t) = Q k,i (t) + QS k,i (t);

[0075] Where: QL k,i (t) and QU k,i (t) are the minimum and maximum allowable discharge flows of the i-th reservoir in the k-th sub-region at time t, respectively, which are generally jointly determined by the downstream ecological flow, shipping requirements and flood control tasks, and QS k,i (t) is the waste water flow of the reservoir;

[0076] (4) Constraint on the power generation of the power station:

[0077] PL k,i (t) ≤ N k,t (t) ≤ PU k,i (t);

[0078] Where: PL k,i and PU k,i are the minimum and maximum power generation limits of the i-th reservoir in the k-th sub-region at time t, respectively;

[0079] (5) Boundary conditions:

[0080]

[0081] Where: Z k,i,begin and Z k,i,end are the initial and final water levels of the i-th reservoir in the k-th sub-region, respectively.

[0082] The present invention has the following beneficial effects:

[0083] 1. The method for determining the staged design flood and staged water storage levels during the operation period of the Three Gorges Reservoir of the present invention first proposes a method for determining the staged design flood and staged water storage levels during the operation period of the reservoir considering the regulation and storage effects of the reservoir group, which can provide an important and highly operable reference basis for the actual operation and scheduling of the Three Gorges Reservoir.

[0084] 2. For the first time, the present invention combines the changing comprehensive dispatching requirements such as power generation, water supply, and shipping, sets relevant demand targets, and uses a cascade reservoir optimal dispatching model to solve, determines the staged storage levels, and can consider the comprehensive utilization benefits of the basin. BRIEF DESCRIPTION OF THE DRAWINGS

[0085] The present invention will be further described below in conjunction with the drawings and embodiments.

[0086] Figure 1 Specific flowchart of the method of the present invention.

[0087] Figure 2 Generalized schematic diagram of 30 key large cascade reservoirs and hydrological stations in the upper reaches of the Yangtze River.

[0088] Figure 3 Relationship diagram between the change point staging of the mean of the annual maximum 7-day flood volume at the Cuntan Hydrological Station and the ten-day rainfall time series of the basin water system.

[0089] Figure 4 Framework diagram of the main loop and restart mechanism of the Borg MOEA.

[0090] Figure 5 (a)(b) Comparison of Pareto fronts of multi-objective optimization schemes for cascade reservoirs. Among them, (a) compares the Pareto fronts of multi-objective optimization schemes, and (b) shows the optimized Pareto front with a flood control risk of 0.

[0091] Figure 6 Schematic diagram of the staged designed flood control levels and staged storage levels during the operation period of the Three Gorges Reservoir. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0092] The present invention will be further described in detail below in conjunction with the embodiments of the drawings. The examples of the embodiments are shown in the drawings, and the same or similar reference numerals represent the same or similar elements or elements with the same or similar functions throughout. The embodiments described below with reference to the drawings are exemplary and are intended to explain the present invention and should not be construed as limiting the present invention.

[0093] Embodiment 1:

[0094] A method for determining the staged designed flood and staged storage levels during the operation period of a reservoir, the method comprising the following main steps:

[0095] Step 1: Restore and extend the natural daily flow data series (1882 - 2023). Among them, the observation data of the Yichang Hydrological Station is used for the period of 1882 - 2003; after the Three Gorges Reservoir started to store water in June 2003, according to the outflow discharge and dam site water level observation data of the Three Gorges Reservoir, the inflow discharge is inversely calculated using the water balance equation, and the natural flow series data from 2004 to 2023 is obtained through reduction calculation considering the regulation effect of upstream large reservoirs.

[0096] Step 2: Flood season staging based on mathematical statistics method. Using the measured historical flow (rainfall) data, select statistical indicators, analyze the variation law of the indicators during the flood season, and use the mean change point analysis method, etc. to determine the staging node time;

[0097] Step 3: Deduce the staged design flood during the operation period of the Three Gorges Reservoir. According to the staging results of Step 2, obtain the flood data sequence by using the staged annual maximum sampling method, consider the regulation influence of the upstream reservoir group by using the corrected reservoir coefficient, use the time-varying P-III curve fitting method to deduce the staged design flood during the operation period of the Three Gorges Reservoir, select the staged flood hydrograph of the typical year and use the same frequency method to amplify and deduce the staged design flood hydrograph, and then obtain the staged flood control water level through the reservoir flood regulation calculation;

[0098] Step 4: Determine the staged water storage levels of the Three Gorges Reservoir. Construct a multi-objective operation model for joint water storage of the cascades in the lower reaches of the Jinsha River and the Three Gorges Reservoir, use the "parameterization - simulation - optimization" framework and multi-objective intelligent algorithms for solution, comprehensively consider the flood control and water utilization operation requirements, and on the premise of keeping the flood control standard unchanged, determine the staged water storage levels of the Three Gorges Reservoir through the simulation operation of the cascade reservoirs.

[0099] Furthermore, Step 1 includes the following sub-steps:

[0100] Step 1.1: Collect the construction operation scheduling information of large key reservoirs in the upper reaches of the Yangtze River, and the data series of basin rainfall and hydrological observation stations;

[0101] Step 1.2: Invert the flow process of each control section under natural conditions according to the flow observation data of the hydrological station.

[0102] Furthermore, Step 2 includes the following sub-steps:

[0103] Step 2.1: Statistically describe the frequency distribution of the storm rainfall or flood peak flow during the flood season by ten-day periods, plot the scatter diagram of the indicators and take the envelope line to determine the division points of the flood season staging;

[0104] Step 2.2: The division of the start and end times of the flood season is a qualitative regulation based on the river rising situation. The meteorological department in China defines a relative coefficient reflecting the ten-day precipitation change as a measurement index, that is:

[0105]

[0106] In the formula: C R is used to represent the relative amount of ten-day precipitation; R is the multi-year average precipitation of a certain ten-day period at a certain station within a year; is the average precipitation per ten-day period at a certain station within a year. Generally, it is considered that when C of a certain station is R ≥1.5 for two consecutive ten-day periods or more, it is called the flood season.

[0107] Step 2.3: In the phased calculation of the reservoir flood season, people generally focus on the periods of larger floods rather than their dispersion. Therefore, the mean change model can be used to describe the variation law of the flood season. The specific implementation method of the mean change point model is as follows:

[0108] Let the sequence X = [x1, x2, …, x n represent the sequence of daily average flow of the flood season over the years at the control stations of a certain main and tributary. Assume it follows a normal distribution. Determine the phased points according to the following steps:

[0109] (1) Assume the number of change points q, generally take 1 or 2, corresponding to two periods and three periods respectively. Taking q = 2 as an example, take all combinations of (m1, m2) where 1 < m1 < m2 < n;

[0110] (2) For each group of (m1, m2), calculate the statistic S according to the following formula:

[0111]

[0112] (3) Take the group of (m1, m2) combinations with the minimum statistic S as the final phased point result, that is, the (m2 + 1)-th to n-th days of the flood season are the later flood season.

[0113] Repeat the above steps for each control station of the main and tributary, and the start time of the later flood season for each sub-region can be obtained:

[0114] M2 = (m2(1), m2(2), …, m2(k)) + 1(3)

[0115] Where: k is the number of control stations of the main and tributary.

[0116] Furthermore, the specific method of Step 3 of the present invention is as follows:

[0117] Step 3.1: Collect and analyze the measured flood data series at the dam site of the reservoir, and directly use the phased annual maximum sampling to further organize it into a flood peak and flood volume data series.

[0118] Step 3.2: Collect the catchment area and total reservoir capacity values of all large reservoirs upstream of the reservoir. According to the construction and operation conditions of the key large reservoirs upstream of the reservoir, use the corrected reservoir coefficient RI to measure the regulation effect of the upstream reservoir group on the design flood of the reservoir:

[0119]

[0120] Where: N is the number of large reservoirs above the study section; A i and C i are the catchment area and regulation storage capacity of the i-th reservoir respectively; A T and C TThey are the catchment area of the research section and the annual average runoff respectively.

[0121] Step 3.3: Taking the reservoir coefficient RI as the covariate, let Y t be the response variable at time t (t = 1, 2, …, n), and y t be its observed value, and construct a time-varying P-III frequency distribution model:

[0122]

[0123] In the formula: μ t , σ t and ν t are the location parameter, scale parameter and shape parameter of the time-varying P-III distribution respectively.

[0124] Step 3.4: According to the generalized linear additive model, the parameters [μ t , σ t , ν t of the time-varying P-III distribution can be further expressed as:

[0125]

[0126] In the formula: α = [α 10 , α 11 , α 20 , α 30 are the parameters of the generalized linear model in the time-varying P-III distribution, and g i (·) (i = 1, 2, 3) are monotonic continuous link functions, and the specific form is determined by the domain of their independent variables.

[0127] Step 3.5: Taking the minimum sum of squared deviations as the criterion for curve fitting, fit the quantiles of the empirical frequency points of the measured flood samples with the quantiles of the theoretical distribution of the time-varying P-III distribution under the same samples, and estimate the statistical parameters and design values of the time-varying P-III distribution.

[0128] Step 3.6: Select the typical hydrograph of the stage design flood, which has the characteristics of continuous multi-peaks and concentrated flood volumes. The flood peak shape and its time course distribution are relatively severe for flood control in the middle and lower reaches, and it has a certain representativeness in terms of the flood regional composition.

[0129] Step 3.7: Use the same-frequency amplification method to respectively deduce the hydrographs of the 1000-year return period and 100-year return period stage design floods; obtain the flood control control levels of each stage through iterative solution of the reservoir flood routing calculation.

[0130] Furthermore, the specific method of step 4 of the present invention is:

[0131] Step 4.1: Build a multi-objective operation model for the joint water storage of cascaded reservoirs in the lower reaches of the Jinsha River and the Three Gorges Reservoir, and solve it using the "parameterization - simulation - optimization" framework and multi-objective intelligent algorithms. The specific steps are as follows:

[0132] (1) Objective function. During the water storage period, the main objectives are to fill the reservoir as much as possible at the end of the water storage period while meeting the existing flood control standards, and also to generate more power generation benefits during this period. The objective function is selected to maximize the full storage rate and power generation, and minimize the flood control risk.

[0133] 1) Maximize the full storage rate IE, expressed as a percentage of the reservoir capacity:

[0134]

[0135] In the formula: IE k,i represents the average full storage rate of the i-th reservoir in the k-th partition over the years; VE k,i (y) is the reservoir capacity at the end of the water storage period of the i-th reservoir in the k-th partition; SU k,i and SD k,i represent the normal beneficial storage capacity and dead storage capacity of the i-th reservoir in the k-th partition respectively; K and M k represent the total number of partitions and the number of reservoirs in the k-th partition respectively; Y represents the number of water storage years.

[0136] 2) Minimize the flood control risk FCR:

[0137] min FCR = min {max {FCR(t)}}, (0 < t ≤ T·Y) (9)

[0138]

[0139] In the formula: FCR(t) represents the flood control risk of the reservoir group system at time t; FCR k,i (t) corresponds to the flood control risk of the i-th reservoir in the k-th partition at time t; V k,i (t) and SS k,i (t) are the reservoir capacity of the i-th reservoir in the k-th partition at time t and the reservoir capacity corresponding to the highest safe water level in front of the dam respectively; T is the total number of time periods during the annual water storage period.

[0140] 3) Maximize the power generation E k of the cascade reservoirs in each partition:

[0141]

[0142] In the formula: N k,i (t) is the average output of the i-th reservoir in the k-th partition at time t; △t is the time step interval of the water storage period; A k,i is the average power generation efficiency coefficient of the i-th reservoir in the k-th partition; Q k,i (t) and Hk,i Let \(Q_{gi}(t)\) and \(H_{gi}(t)\) be the power generation discharge and the power generation head of the \(i\)-th reservoir in the \(k\)-th partition at time \(t\), respectively.

[0143] (2) Constraints. The cascade reservoirs still need to comply with some basic constraints:

[0144] 1) Water balance constraint:

[0145]

[0146] In the formula: \(I_{ki}(t)\) and \(O_{ki}(t)\) are the inflow and outflow discharges of the \(i\)-th reservoir in the \(k\)-th partition at time \(t\); \(\varPhi_{ki}\) represents the set of reservoirs upstream of the \(i\)-th reservoir in the \(k\)-th partition that have a hydraulic connection with it; \(O_{kj}(t)\) is the outflow discharge of the \(j\)-th reservoir in the set \(\varPhi_{ki}\). k,i \(I_{ki}(t)\) and \(O_{ki}(t)\) k,i are the inflow and outflow discharges of the \(i\)-th reservoir in the \(k\)-th partition at time \(t\); \(\varPhi_{ki}\) j,k,i represents the set of reservoirs upstream of the \(i\)-th reservoir in the \(k\)-th partition that have a hydraulic connection with it; \(O_{kj}(t)\) j is the outflow discharge of the \(j\)-th reservoir in the set \(\varPhi_{ki}\). j,k,i

[0147] 2) Reservoir storage capacity constraint:

[0148] \(SL_{ki}(t)\leq V_{ki}(t)\leq SS_{ki}(t)\) (14) k,i \(SL_{ki}(t)\leq V_{ki}(t)\leq SS_{ki}(t)\) k,i k,i

[0149] In the formula: \(SL_{ki}(t)\) and \(SS_{ki}(t)\) are the minimum and maximum allowable storage volumes of the \(i\)-th reservoir in the \(k\)-th partition at time \(t\), respectively. k,i \(SL_{ki}(t)\) and \(SS_{ki}(t)\) k,i are the minimum and maximum allowable storage volumes of the \(i\)-th reservoir in the \(k\)-th partition at time \(t\), respectively.

[0150] 3) Reservoir outflow discharge constraint:

[0151] \(QL_{ki}(t)\leq O_{ki}(t)\leq QU_{ki}(t)\), \(O_{ki}(t)=Q_{gi}(t)+QS_{ki}(t)\) (15) k,i \(QL_{ki}(t)\leq O_{ki}(t)\leq QU_{ki}(t)\), \(O_{ki}(t)=Q_{gi}(t)+QS_{ki}(t)\) k,i k,i \(O_{ki}(t)\), \(Q_{gi}(t)\) k,i \(O_{ki}(t)=Q_{gi}(t)+QS_{ki}(t)\) k,i k,i

[0152] In the formula: \(QL_{ki}(t)\) and \(QU_{ki}(t)\) are the minimum and maximum allowable outflow discharges of the \(i\)-th reservoir in the \(k\)-th partition at time \(t\), respectively, which are generally determined jointly by the downstream ecological flow, shipping requirements, and flood control tasks. \(QS_{ki}(t)\) is the spillway discharge of this reservoir. k,i \(QL_{ki}(t)\) and \(QU_{ki}(t)\) k,i are the minimum and maximum allowable outflow discharges of the \(i\)-th reservoir in the \(k\)-th partition at time \(t\), respectively, which are generally determined jointly by the downstream ecological flow, shipping requirements, and flood control tasks. \(QS_{ki}(t)\) k,i is the spillway discharge of this reservoir.

[0153] 4) Power station output constraint:

[0154] \(PL_{ki}(t)\leq N_{gi}(t)\leq PU_{ki}(t)\) (16) k,i \(PL_{ki}(t)\leq N_{gi}(t)\leq PU_{ki}(t)\) k,t k,i

[0155] Where: PL k,i and PU k,i are respectively the minimum and maximum output limits of the ith reservoir in the kth partition at time t.

[0156] 5) Boundary conditions:

[0157]

[0158] Where: Z k,i,begin and Z k,i,end are respectively the initial and final water levels of the ith reservoir in the kth partition.

[0159] (3) Optimization algorithm

[0160] In this study, BorgMOEA is used to optimize the multi-objective problem of cascade reservoirs. The initial population size is set to 200. After normalizing the three objectives, the ε value is uniformly set to 0.001. The maximum number of steps for BorgMOEA optimization is set to 200,000, which is equivalent to approximately 5.6 million calculations of the objective function.

[0161] Step 4.2: Based on the long-term measured data and the stage design flood hydrograph, carry out a multi-objective operation simulation study on the joint water storage of the cascade reservoirs in the lower reaches of the Jinsha River and the Three Gorges Reservoir, and finally determine the staged water storage levels of the Three Gorges Reservoir through comparison of multiple schemes.

[0162] Example 2:

[0163] For the method for determining the stage design flood and the staged water storage levels during the operation period of the Three Gorges Reservoir in the embodiments of the present invention, first, collect and organize the data series of rainfall and flow observation stations in the upper reaches of the Yangtze River, and carry out flood season division based on the mathematical statistics method; then, consider the regulation effect of the upstream reservoir to calculate the stage design flood during the operation period of the Three Gorges Reservoir and deduce the stage flood control levels; finally, establish a multi-objective joint operation model for cascade reservoirs, and determine the staged water storage levels of the Three Gorges Reservoir through simulation and scheduling. The overall research plan and technical route are as Figure 1 shown.

[0164] The technical solution of the present invention will be further specifically described below through examples and in combination with the accompanying drawings:

[0165] Step 1: Collect the construction and operation information of key large reservoirs in the upper reaches of the Yangtze River and the observation data of basin hydrological stations, and restore the natural runoff process, which specifically includes the following sub-steps.

[0166] Step 1.1: According to the operation and scheduling data of the Three Gorges Reservoir, based on the observed data of the discharge flow and the reservoir water level above the dam, use the water balance model to back-calculate the inflow.

[0167] Step 1.2: Obtain the flow series at the dam site based on the relationship between the inflow to the Three Gorges Reservoir and the flood at the dam site, and restore the flow process at the dam site section under natural conditions considering the regulation effect of upstream reservoirs.

[0168] Step 1.3: Construct the long-term daily flow series data of the Three Gorges Reservoir from 1982 to 2023.

[0169] Figure 2 Figure 1 shows a generalized schematic diagram of 30 key large-scale cascade reservoirs and hydrological stations in the upper reaches of the Yangtze River, including 6 in the middle reaches of the Jinsha River, 3 in the Yalong River, 4 in the lower reaches of the Jinsha River, 5 in the Min River, 4 in the Jialing River, 7 in the Wu River, and the Three Gorges - Gezhouba cascade.

[0170] Step 2: Based on the flood season staging method of mathematical statistics, use the measured historical flow (rainfall) data, select statistical indicators, analyze the variation law of the indicators during the flood season, and determine the staging node time by using the mean change point analysis method, etc.

[0171] Step 2 includes the following sub-steps:

[0172] Step 2.1: Division of the start and end times of the flood season;

[0173] Adopt a relative coefficient reflecting the change of dekadal precipitation as a measurement index, that is:

[0174]

[0175] In the formula: C R is used to represent the relative amount of dekadal precipitation; R is the multi-year average precipitation of a certain dekad at a certain station within a year; is the average dekadal precipitation of a certain station within a year. Generally, when C of a certain station is R ≥1.5 for two consecutive dekads or more, it is called the flood season.

[0176] Step 2.2: Use the mean change point model to judge the staging point;

[0177] Let the sequence X = [x1, x2, …, x n represent the multi-year average daily flow sequence during the flood season of a control station on a certain main or tributary river, assuming it follows a normal distribution. Judge the staging point according to the following steps:

[0178] (1) Assume the number of change points q, generally take 1 or 2, corresponding to two periods and three periods respectively. Taking q = 2 as an example, take all combinations of (m1, m2) where 1 < m1 < m2 < n;

[0179] (2) For each group of (m1, m2), calculate the statistic S according to the following formula:

[0180]

[0181] (3) Take the combination of (m1, m2) with the smallest statistic S as the final breakpoint staging result, that is, the (m2 + 1)-th to n-th days of the flood season are the late flood season.

[0182] Repeat the above steps for each main and tributary control station to obtain the start time of the late flood season for each sub-region:

[0183] M2 = (m2(1), m2(2), …, m2(k)) + 1(3)

[0184] In the formula: k is the number of main and tributary control stations.

[0185] As Figure 3 shown, a time series relationship diagram of the mean annual maximum 7-day flood volume breakpoint staging at the Cuntan Hydrological Station and the ten-day rainfall in the basin water system is given.

[0186] Step 3: Derive the staged design flood during the operation period of the Three Gorges Reservoir. According to the staging results of Step 2, obtain the flood data series using the staged annual maximum sampling method, consider the regulation effect of the upstream reservoir group by modifying the reservoir coefficient, use the time-varying P-III type curve fitting method to derive the staged design flood during the operation period of the Three Gorges Reservoir, select the staged flood hydrograph of the typical year and magnify the staged design flood hydrograph using the same frequency method, and obtain the staged flood control water level through the flood regulation calculation of the Three Gorges Reservoir.

[0187] Step 3 further includes the following sub-steps:

[0188] Step 3.1: Obtain the staged flood peak and flood volume data series from 1892 to 2023 using the staged annual maximum sampling.

[0189] Step 3.2: According to Figure 2 the catchment area and regulation storage capacity of the key large reservoirs shown, use the modified reservoir coefficient RI to measure the regulation effect of the upstream reservoir group:

[0190]

[0191] In the formula: N is the number of large reservoirs above the study section; A i and C i are the catchment area and regulation storage capacity of the i-th reservoir respectively; A T and C T are the catchment area of the study section and the average annual runoff respectively.

[0192] Step 3.3: Using the reservoir coefficient RI as the covariate, let Y t be the response variable at time t (t = 1, 2, …, n), y t be its observed value, and construct a time-varying P-III type frequency distribution model:

[0193]

[0194] Where: μ t , σ t and ν t are the location parameter, scale parameter, and shape parameter of the time-varying P-III distribution, respectively.

[0195] Step 3.4: According to the generalized linear additive model, the parameters [μ t , σ t , ν t of the time-varying P-III distribution can be further expressed as:

[0196]

[0197] Where: α = [α 10 , α 11 , α 20 , α 30 are the parameters of the generalized linear model in the time-varying P-III distribution, and g i (·) (i = 1, 2, 3) are monotonic continuous link functions, and their specific forms are determined by the domain of their independent variables.

[0198] Step 3.5: Taking the minimum sum of squared deviations as the criterion for curve fitting, the quantiles of the empirical frequency points of the measured flood samples are curve-fitted with the quantiles of the theoretical distribution of the time-varying P-III distribution under the same samples to estimate the statistical parameters and design values of the time-varying P-III distribution.

[0199] Step 3.6: Select the typical flood hydrograph for the staged design flood, which has the characteristics of continuous multiple peaks and concentrated flood volumes. The flood peak shape and its time course distribution are relatively severe for flood control in the middle and lower reaches, and it has a certain representativeness in terms of the flood regional composition.

[0200] Step 3.7: Use the same-frequency amplification method to separately derive the flood hydrographs for the 1000-year return period and 100-year return period staged design floods; the flood control water levels for each stage are obtained through iterative solution of the reservoir flood regulation calculation.

[0201] Step 4: Determine the staged water storage levels of the Three Gorges Reservoir. Construct a multi-objective operation model for the joint water storage of the cascades in the lower reaches of the Jinsha River and the Three Gorges Reservoir, and use the "parameterization - simulation - optimization" framework structure and intelligent algorithms to solve. Considering the flood control and water utilization operation requirements comprehensively, under the premise of maintaining the flood control standard unchanged, the staged water storage levels of the Three Gorges Reservoir are finally determined through the simulation of flood regulation of the cascade reservoirs.

[0202] Step 4 includes the following sub-steps:

[0203] Step 4.1: Construct a multi-objective operation model for the joint water storage of the cascades in the lower reaches of the Jinsha River and the Three Gorges Reservoir, and use the "parameterization - simulation - optimization" framework structure and multi-objective intelligent algorithms to solve. The specific steps are as follows:

[0204] (1) Objective function. During the water storage period, on the premise of meeting the existing flood control standards, the main objectives are to maximize the water storage rate and power generation, and minimize the flood control risk as the objective function, so that the reservoir can be filled as much as possible at the end of the water storage period, and more power generation benefits can be generated during this period.

[0205] ① Maximize the water storage rate (IE), expressed as a percentage of the reservoir capacity:

[0206]

[0207] In the formula: IE k,i represents the average annual water storage rate of the i-th reservoir in the k-th sub-region; VE k,i (y) is the reservoir capacity at the end of the water storage period of the i-th reservoir in the k-th sub-region; SU k,i and SD k,i represent the normal beneficial storage capacity and dead storage capacity of the i-th reservoir in the k-th sub-region respectively; K and M k represent the total number of sub-regions and the number of reservoirs in the k-th sub-region respectively; Y represents the number of water storage years.

[0208] ② Minimize the flood control risk FCR:

[0209] min FCR = min{max{FCR(t)}},(0 < t ≤ T·Y)(9)

[0210]

[0211] In the formula: FCR(t) represents the flood control risk of the reservoir group system at time t; FCR k,i (t) corresponds to the flood control risk of the i-th reservoir in the k-th sub-region at time t; V k,i (t) and SS k,i (t) are the reservoir capacity of the i-th reservoir in the k-th sub-region at time t and the reservoir capacity corresponding to the highest safety water level in front of the dam respectively; T is the total number of time periods during the annual water storage period.

[0212] ③ Maximize the power generation E of the cascade reservoirs in each sub-region k Maximum:

[0213]

[0214] In the formula: N k,i (t) is the average output of the i-th reservoir in the k-th sub-region at time t; △t is the step interval of the water storage period; A k,i is the average power generation efficiency coefficient of the i-th reservoir in the k-th sub-region; Q k,i (t) and H k,i (t) are the power generation flow rate and power generation head of the i-th reservoir in the k-th sub-region at time t respectively.

[0215] (2) Constraints. The cascade reservoirs still need to comply with some basic constraints:

[0216] 1) Water balance constraint:

[0217]

[0218] In the formula: I k,i (t) and O k,i (t) are the inflow and outflow discharges of the i-th reservoir in the k-th sub-region at time t; Φ j,k,i represents the set of reservoirs upstream of the i-th reservoir in the k-th sub-region that have hydraulic connections with it; O j (t) is the outflow discharge of the j-th reservoir in the Φ j,k,i set.

[0219] 2) Reservoir storage capacity constraint:

[0220] SL k,i (t) ≤ V k,i (t) ≤ SS k,i (t) (14)

[0221] In the formula: SL k,i (t) and SS k,i (t) are the minimum and maximum allowable storage volumes of the i-th reservoir in the k-th sub-region at time t, respectively.

[0222] 3) Reservoir outflow discharge constraint:

[0223] OL k,i (t) ≤ O k,i (t) ≤ OU k,i (t), O k,i (t) = Q k,i (t) + QS k,i (t) (15)

[0224] In the formula: QL k,i (t) and QU k,i (t) are the minimum and maximum allowable outflow discharges of the i-th reservoir in the k-th sub-region at time t, respectively, which are generally determined jointly by the downstream ecological flow, shipping requirements, and flood control tasks. QS k,i (t) is the waste water discharge of this reservoir.

[0225] 4) Power station output constraint:

[0226] PL k,i (t) ≤ N k,t (t) ≤ PU k,i (t) (16)

[0227] In the formula: PL k,i and PU k,iThey are the minimum and maximum output limits of the $i$-th reservoir in the $k$-th partition at time period $t$, respectively.

[0228] 5) Boundary conditions:

[0229]

[0230] In the formula: $Z$ k,i,begin and $Z$ k,i,end are the initial and final water levels of the $i$-th reservoir in the $k$-th partition, respectively.

[0231] (3) Optimization algorithm:

[0232] In this embodiment, BorgMOEA is used to optimize the multi-objective problem of cascade reservoirs. The initial population size is set to 200. After normalizing the three objectives, the $\epsilon$ value is uniformly set to 0.001, and the maximum number of steps for BorgMOEA optimization is set to 200,000, which is equivalent to approximately 5.6 million calculations of the objective function. Figure 4 The framework diagram of the main loop and restart mechanism of BorgMOEA is given.

[0233] Step 4.2: According to the long-term measured data, the stage design flood hydrograph, and the stage flood control water level, carry out the multi-objective simulation and simulation scheduling of the joint water storage of the cascade reservoirs in the lower reaches of the Jinsha River and the Three Gorges Reservoir. After comparing multiple schemes, the staged water storage levels of the Three Gorges Reservoir are finally determined.

[0234] As Figure 5 shown, the Pareto front comparison of the multi-objective optimization scheme for the joint water storage of the cascade reservoirs in the lower reaches of the Jinsha River and the Three Gorges Reservoir is given. It can be seen from Figure 5 (a) that the joint operation of cascade reservoirs can significantly increase power generation benefits and the reservoir fullness rate; in the optimization solution set with the same flood control standard (i.e., the flood control risk is 0), the fullness rate decreases with the increase of power generation, as shown in Figure 5 (b). Therefore, through simulation, the staged water storage levels of the Three Gorges Reservoir are finally determined as shown in Figure 6 .

[0235] Although the preferred embodiments of the present invention have been described, those skilled in the art can make additional changes and modifications to these embodiments once they learn the basic creative concepts. Therefore, the appended claims are intended to be construed as including the preferred embodiments and all changes and modifications falling within the scope of the present invention.

[0236] Although the embodiments of the present invention have been shown and described, those skilled in the art can understand that various changes, modifications, substitutions, and deformations can be made to these embodiments without departing from the principles and purposes of the present invention. The scope of the present invention is defined by the claims and their equivalents.

Claims

1. A method for determining the phased design flood and phased water storage level during the operation period of the Three Gorges Reservoir, characterized in that: The following steps are involved: Step 1: Collect the construction and operation information of key large reservoirs in the upper reaches of the Yangtze River and the observation data of hydrological stations in the basin to restore the natural runoff process; Step 2: Based on the mathematical statistics method, the flood season is divided into stages. The measured historical flow data are used to select statistical indicators, analyze the changing patterns of the indicators during the flood season, and use the mean change point analysis method to determine the stage node time; Step 3, deriving the phased design flood of the Three Gorges Reservoir during operation; According to the staged results of step 2, the flood data series is obtained by using the staged annual maximum sampling method, the regulation and storage effect of the upstream reservoir group is considered by using the modified reservoir coefficient, and the staged design flood of the Three Gorges Reservoir operation period is deduced by using the time-varying P-III type fitting method. The typical year staged flood process line is selected and the staged design flood process line is amplified by the same frequency method. The staged flood control control water level is obtained through the Three Gorges Reservoir flood regulation calculation; Step 4, determine the water storage level of the Three Gorges Reservoir in different stages; A multi-objective scheduling model for the joint water storage of the Jinsha River cascade and the Three Gorges Reservoir was constructed. The "parameterization-simulation-optimization" framework and intelligent algorithm were used to solve the model. The flood control and benefit-enhancing scheduling needs were comprehensively considered. Under the premise of keeping the flood control standards unchanged, the phased water storage level of the Three Gorges Reservoir was finally determined through flood control simulation of cascade reservoirs.

2. A method for determining the phased design flood and phased water storage level of the Three Gorges Reservoir during operation according to claim 1, characterized in that: The step 1 specifically includes: Step 1.1, collect the construction and operation dispatch information of large key reservoirs in the upper reaches of the Yangtze River, and the data series of basin rainfall and hydrological observation stations; Step 1.2: Invert the flow process of each control section under natural conditions based on the flow observation data of the hydrological station.

3. A method for determining the phased design flood and phased water storage level of the Three Gorges Reservoir during operation according to claim 2, characterized in that: In step 1.2, the water balance equation is used to invert the inflow flow.

4. A method for determining the phased design flood and phased water storage level of the Three Gorges Reservoir during operation according to claim 2, characterized in that: The step 2 specifically includes: Step 2.1, describe the frequency distribution of rainstorm or flood peak flow in the flood season by ten-day statistics, and take the outer envelope of the scatter diagram of the point-plotted index to determine the division points of the flood season; Step 2.2: The start and end time of the flood season is determined based on the flooding of rivers, with the relative coefficient of the change in ten-day precipitation as the measurement indicator, namely: Where: C R It is used to indicate the relative amount of precipitation in a decade; R is the multi-year average precipitation in a decade at a certain station in a year; The average precipitation per decade at a station in a year; when a station has continuous precipitation for 2 decades or more, R When ≥1.5, it is called the flood season; Step 2.3, use the mean change model to describe the changing pattern of the flood season.

5. A method for determining the phased design flood and phased water storage level of the Three Gorges Reservoir during operation according to claim 4, characterized in that: The specific implementation method of the mean change point model in step 2.3 is as follows: Let the sequence X = [x1, x2, …x n ] represents the multi-year average daily flow sequence of a main and tributary control station during the flood season. Assuming that it obeys the normal distribution, the stage points are determined according to the following steps: Step 2.3.1, assuming the number of change points q, take 1 or 2, corresponding to two periods and three periods respectively. When q = 2, take all (m1, m2) combinations of 1 < m1 < m2 < n; in, Step 2.3.2, for each group (m1, m2), calculate the statistic S as follows: Step 2.3.3, take the combination (m1, m2) with the smallest statistic S as the final change point periodization result, that is, the flood season m2+1~n days are the post-flood season; Repeat the above steps for each main and tributary control station to obtain the start time of the post-flood season in each sub-area: M2=(m2(1),m2(2),…,m2(k))+1; Where: k is the number of main and tributary flow control stations.

6. A method for determining the phased design flood and phased water storage level of the Three Gorges Reservoir during operation according to claim 5, characterized in that: The specific method of step 3 is: Step 3.1, collect and analyze the measured flood data series at the reservoir dam site, and directly use the annual maximum sampling of the phases to further organize it into a flood peak volume data series; Step 3.2, collect the catchment area and total storage capacity of all large reservoirs upstream of the reservoir, and use the modified reservoir coefficient RI to measure the regulation and storage effect of the upstream reservoir group on the design flood of the reservoir according to the construction and operation status of the key large reservoirs upstream of the reservoir: Where: N is the number of large reservoirs above the study section; A i and C i are the catchment area and regulating storage capacity of the i-th reservoir respectively; A T and C T are the catchment area and multi-year average runoff of the study section, respectively; Step 3.3, taking the reservoir coefficient RI as the covariate, let Y t is the response variable at time t (t=1,2,…,n), y t For its observed values, a time-varying P-III type frequency distribution model is constructed: Where: μ t , σ t and ν t are the location parameter, scale parameter, and shape parameter of the time-varying P-III distribution, respectively; Step 3.4: According to the generalized linear additive model, the parameters of the time-varying P-III distribution [μ t ,σ t ,ν t ] can be further expressed as: Where: α=[α 10 ,α 11 ,α 20 ,α 30 ] is the parameter of the generalized linear model in the time-varying P-III distribution, g i (·)(i=1,2,3) is a monotonically continuous link function, and its specific form is determined by the domain of its independent variable; Step 3.5, using the minimum sum of squared deviations as the fitting criterion, fit the empirical frequency point quantiles of the measured flood samples to the theoretical distribution quantiles of the time-varying P-III distribution under the same samples, and estimate the statistical parameters and design values ​​of the time-varying P-III distribution; Step 3.6, select the typical process line of staged design flood, which has the characteristics of continuous multi-peaks and concentrated flood volume. The flood peak shape and its time distribution are relatively severe for the flood control situation in the middle and lower reaches, and are representative in terms of the composition of the flood area; Step 3.7, use the same frequency amplification method to deduce the design flood process lines for 1000-year and 100-year floods respectively; obtain the flood control control water level for each period through the iterative solution of reservoir flood regulation calculation.

7. A method for determining the phased design flood and phased water storage level of the Three Gorges Reservoir during operation according to claim 6, characterized in that: The specific method of step 4 is: Step 4.1, construct a multi-objective scheduling model for the joint water storage of the Jinsha River downstream cascade and the Three Gorges Reservoir, using the "parameterization-simulation-optimization" framework and multi-objective intelligent algorithm to solve it; Step 4.2: Based on the long series of measured data and the phased design flood process line, a multi-objective scheduling simulation study on the joint water storage of the lower reaches of the Jinsha River cascade and the Three Gorges Reservoir was carried out, and the phased water storage level of the Three Gorges Reservoir was finally determined after comparing multiple schemes.

8. A method for determining the phased design flood and phased water storage level of the Three Gorges Reservoir during operation according to claim 7, characterized in that: The specific steps of step 4.1 are: Step 4.1.1, determine the objective function: The main goal of the water storage period is to fill the reservoir as much as possible at the end of the water storage period without exceeding the existing flood control standards, and to generate as much power generation benefits as possible during the period. The objective function is to maximize the storage rate and power generation and minimize the flood control risk. Step 4.1.2, determine the constraints: Cascade reservoirs still need to comply with some basic constraints, including water balance constraints, reservoir storage capacity constraints, reservoir discharge flow constraints, power station output constraints and boundary conditions; Step 4.1.3, optimization algorithm solution: Borg MOEA is used to optimize and solve the multi-objective problem of cascade reservoirs.

9. A method for determining the phased design flood and phased water storage level of the Three Gorges Reservoir during operation according to claim 8, characterized in that: The objective function in step 4.1.1 specifically includes: (1) The fullness factor IE is the largest and is expressed as a percentage of storage capacity: In the formula: IE k,i represents the average storage rate of the i-th reservoir in the k-th sub-area over many years; VE k,i (y) is the storage capacity of the i-th reservoir in the k-th partition at the end of the impoundment period; SU k,i and SD k,i They represent the normal storage capacity and dead storage capacity of the i-th reservoir in the k-th zone respectively; K and M k They represent the total number of zones and the number of reservoirs in the kth zone respectively; Y represents the number of years of water storage; (2) Flood risk FCR is the smallest: min FCR=min{max{FCR(t)}}, (0<t≤T·Y); Where: FCR(t) represents the flood control risk of the reservoir group system at time t; FCR k,i (t) corresponds to the flood risk of the i-th reservoir in the k-th zone at time t; V k,i (t) and SS k,i (t) are the reservoir capacity of the i-th reservoir in the k-th zone at time t and the reservoir capacity corresponding to the highest safe water level in front of the dam; T is the total number of water storage periods per year; (3) Power generation of cascade reservoirs in each sub-district E k maximum: Where: N k,i (t) is the average output of the i-th reservoir in the k-th sub-area during the t-th period; △t is the step interval during the water storage period; A k,i is the average power generation efficiency coefficient of the i-th reservoir in the k-th zone; Q k,i (t) and H k,i (t) are the power generation flow and power generation head of the i-th reservoir in the k-th partition during period t.

10. A method for determining the phased design flood and phased water storage level of the Three Gorges Reservoir during operation according to claim 9, characterized in that: The constraints in step 4.1.2 specifically include: (1) Water balance constraints: Where: I k,i (t) and O k,i (t) is the inflow and outflow flow of the i-th reservoir in the k-th zone during period t; Φ j,k,i represents the set of reservoirs that are hydraulically connected to the upstream of the i-th reservoir in the k-th sub-area; O j (t) is Φ j,k,i The outflow from reservoir j in the set; (2) Reservoir capacity constraints: SL k,i (t)≤V k,i (t)≤SS k,i (t); Where: SL k,i (t) and SS k,i (t) are the minimum and maximum permissible water storage capacity of the i-th reservoir in the k-th zone during period t, respectively; (3) Reservoir discharge flow constraints: OL k,i (t)≤O k,i (t)≤OU k,i (t),O k,i (t)=Q k,i (t)+QS k,i (t); Where: QL k,i (t) and QU k,i (t) are the minimum and maximum allowable discharge flows of the i-th reservoir in the k-th zone during period t, which are generally determined by the downstream ecological flow, navigation requirements and flood control tasks. k,i (t) is the discharge flow of the reservoir; (4) Power station output constraints: PL k,i (t)≤N k,t (t)≤PU k,i (t); Where: PL k,i and PU k,i They are the minimum and maximum output limits of the i-th reservoir in the k-th zone during period t, respectively; (5) Boundary conditions: Where: Z k,i,begin and Z k,i,end are the initial and final water levels of the ith reservoir in the kth subarea, respectively.

Images