Cascade reservoir optimization dynamic control method based on different scheduling targets of flood season stages

Abstract

A cascade reservoir optimization dynamic control method based on different scheduling objectives of flood season staging comprises the following steps: step 1, performing flood season staging on a cascade reservoir based on rainstorm flood characteristics, and establishing a pre / main flood season'flood control-power generation 'and post flood season'flood control-water storage' optimization scheduling model; step 2, constructing a cascade reservoir dynamic control domain; step 3, based on the flood season staging result in the step 1 and the cascade reservoir dynamic control domain constructed in the step 2, constructing an optimization model taking the maximum cascade generating capacity and the minimum flood control risk as targets; setting constraint conditions of the optimization model; and solving the optimization model through a multi-target intelligent optimization algorithm to obtain a dynamic scheduling scheme. According to the method, optimal scheduling methods such as flood season staging and'aggregation-decomposition 'are adopted, flood control risks of drainage basin reservoir group scheduling are fully considered, and efficient utilization of cascade reservoir water resources is realized by utilizing a cascade reservoir flood control capacity complementation mechanism.

Description

Technical Field

[0001] The present invention relates to the technical field of efficient utilization of water resources in cascade reservoirs during flood season, and in particular to an optimized dynamic control method for cascade reservoirs based on different dispatching targets in different flood season phases. Background Art

[0002] During the real-time operation of reservoirs, rainstorms and floods change dynamically. There are many reservoirs in the basin, with multiple flood control targets and multiple economic development goals. Therefore, how to identify typical floods and combine forecast information to define the upper and lower thresholds of the operating water levels of the reservoir group during the flood season, and determine the control domain of each reservoir using forecast information of different forecast periods are key technical issues to be solved.

[0003] Combining the flood season classification method and historical scheduling process, it is a feasible idea to formulate scheduling rules according to the water inflow conditions in different stages.

[0004] For example, Zhang Yongbo, Tang Li, Zhu Xueping, et al. published the impact of climate change on reservoir flood season staging and scheduling [J]. Hydropower Energy Science, 2019, 37(02): 67-69: The impact of different flood season staging periods on reservoir scheduling was analyzed. The results showed that the main flood season was shortened by 11 days after the climate change, which can effectively improve the water resource utilization conditions downstream of the reservoir and alleviate the water supply pressure of the reservoir during the reservoir scheduling process.

[0005] Li Xiaoying, Zhang Yan, Tong Zechun et al. published a study on multi-objective collaborative decision-making for reservoir flood limit water level control [J]. Journal of Hydroelectric Engineering, 2022, 41(2): 31-42: The circular distribution method is used to divide the flood season of a reservoir into stages, and based on the "flood control risk-power generation benefit" collaborative decision-making, the optimal flood limit water level control scheme for the pre-flood season and post-flood season under different preferences is determined.

[0006] However, current research techniques ignore the competitive storage phenomenon that occurs in cascade reservoirs during the dry season at the end of the flood season. This results in low reservoir storage rates at the end of the flood season and inadequate utilization of flood resources. These research results often provide conservative estimates of flood control risks during the flood season, which contradicts actual operational principles and is difficult to apply in practice. Summary of the Invention

[0007] In response to the problem that existing cascade reservoir design and scheduling schemes fail to efficiently utilize flood resources during the flood season, the present invention provides a cascade reservoir optimization dynamic control method based on different scheduling objectives in different flood season phases. It is based on the "aggregation-decomposition" concept and the pre-storage and pre-discharge method, and determines the upper limit relationship of the dynamic control of the reservoir according to the lower limit results of the cascade reservoir's phased operation water level. It also uses a multi-objective intelligent optimization algorithm to implement optimization control on the dynamic operation space of the cascade reservoir, and generate an optimization scheme that meets the different scheduling objectives of "flood prevention-power generation-water storage".

[0008] In order to solve the above technical problems, the present invention adopts the following technical solutions:

[0009] A method for optimizing dynamic control of cascade reservoirs based on different flood season scheduling objectives includes the following steps:

[0010] Step 1: Divide the flood season of cascade reservoirs into stages based on the characteristics of rainstorms and floods

[0011] Use comprehensive analysis methods to determine the phased plan for the pre-flood season, main flood season, and post-flood season;

[0012] The flood samples were obtained by cross-period sampling method, with the sampling time of each period extended forward or backward by 5 days. The flood frequency was fitted by P-Ⅲ curve, and the parameters were determined by visual estimation of the fitting method.

[0013] Flood control calculations are used to derive the flood control water levels of each reservoir in different phases. The flood control water level during the pre-flood season is used as the lower limit of operation, and the flood control water level during the post-flood season is used as the initial boundary of water storage operation.

[0014] By amplifying the flood process line using the same frequency method, an optimized scheduling model for "flood control-power generation" in the pre-flood season and "flood control-water storage" in the post-flood season was established;

[0015] Step 2: Construct the dynamic control domain of cascade reservoirs

[0016] Aggregate cascade reservoirs into virtual reservoirs and determine the maximum allowable pre-storage capacity of the aggregated reservoirs based on flood control constraints and water and rainfall forecasts;

[0017] Decompose the pre-storage capacity of each reservoir according to the hydraulic connection and establish the upper and lower limit relationship of dynamic water level;

[0018] Adopting a pre-storage and pre-release strategy to allocate additional allocable storage for small and medium flood events;

[0019] Step 3: Multi-objective optimization scheduling

[0020] Based on the flood season phase results of step 1 and the dynamic control domain of the cascade reservoirs constructed in step 2, an optimization model is constructed with the goal of maximizing cascade power generation and minimizing flood control risks.

[0021] Set water balance constraints, water balance constraints, reservoir discharge flow constraints, power station output constraints and boundary constraints;

[0022] The optimization model is solved by a multi-objective intelligent optimization algorithm to obtain a dynamic scheduling solution.

[0023] Furthermore, the comprehensive analysis method described in step 1 includes causal analysis method, mathematical statistics method, fuzzy analysis method, fractal analysis method, change point analysis method, system clustering method, vector statistics method and relative frequency method.

[0024] Furthermore, the phased flood sample selection in step S1 includes:

[0025] The empirical frequency of extreme floods is calculated using the formula:

[0026] (1)

[0027] The empirical frequency of measured floods is calculated using the formula:

[0028] (2)

[0029] Where, M is the serial number of the major floods arranged from large to small; P M is the empirical frequency of the Mth serial number of the major flood; N is the number of years from the earliest survey year to the present; a is the number of major floods; m is the serial number of the measured series arranged from large to small; P m is the empirical frequency of the mth item in the measured series; n is the number of years in the measured series; l is the number of items in the measured series that are treated as extremely large values.

[0030] Furthermore, the design flood process line amplification in step S1 includes:

[0031] The peak magnification ratio is:

[0032] (3)

[0033] The maximum 1d flood magnification ratio is:

[0034] (4)

[0035] The magnification ratios of the remaining 2 days, except for the maximum 1 day, in the maximum 3 day flood volume are:

[0036] (5)

[0037] The magnification ratios of the remaining 4 days, except for the maximum 3 days, in the maximum 7 days of flood volume are:

[0038] (6)

[0039] Where Q p Design peak; Q d is the typical annual flood peak; W p is the design flood volume; W d This is a typical annual flood volume.

[0040] Furthermore, the maximum pre-storage capacity allowed by the aggregated reservoir in step 2 is calculated as follows:

[0041] (7)

[0042] In formula (7), and are the inflow and outflow of the aggregated reservoir at time t, respectively; The effective forecast period; is the reservoir water level-storage capacity function relationship; and are the upper and lower limit water levels of the reservoir respectively; By using short-term forecast information at time t, the cascade reservoirs can store more water.

[0043] Furthermore, the relationship between the upper and lower limits of the dynamic water level in step 2 is:

[0044] f B ( Z B ' ( t )) ≤ f B ( Z B ( t )) + Q max B ⋅ T f − ∫ t c t c + T f Q in B ( t ) dt = f B ( Z B ( t )) + [ Q max B − K ( t )] ⋅ T f − C 0 ⋅ [ f A ( Z A ' ( t )) − f A ( Z A ( t )) + ∫ t c t c + T f Q in A ( t ) dt ] (8)

[0045] Where K(t) is the Muskegon parameter of river channel evolution, Q max B It is the maximum safe discharge volume of the downstream flood control point of Reservoir B.

[0046] Furthermore, the additional allocable water storage capacity in step 2 is:

[0047] (9)

[0048] Where N is the number of cascade reservoirs, Allocate additional storage capacity for reservoir i at time t; and The total water storage capacity allocation coefficient for each reservoir satisfies ;

[0049] For predicted small and medium-sized flood events, each reservoir dynamically controls the target water level Determined by formula (10):

[0050] (10).

[0051] Furthermore, the objective function expression of the optimization model in step 3 with the goal of maximizing cascade power generation and minimizing flood control risk is:

[0052] Minimizing flood risks:

[0053] (11)

[0054] The average total power generation over many years is the largest:

[0055] (12)

[0056] Among them, FCR and THG are the multi-year flood control risk and power generation of the cascade reservoirs, respectively;

[0057] FCRi (t) and HG i (t) are the flood control risk and power generation of the i-th reservoir at time t, respectively; is the ratio of the flood control storage capacity of the i-th reservoir to the total storage capacity of the cascade;

[0058] Vi(t) is the storage capacity of reservoir i at time t;

[0059] V i max (t) and V i min (t) are the normal water storage capacity and flood control limit water storage capacity of the i-th reservoir at time t;

[0060] K i is the output coefficient of reservoir i, QG i (t) is the power generation flow of reservoir i in period t, H i (t) is the average power generation head of reservoir i in period t;

[0061] △t is the time interval;

[0062] Y is the number of years in the scheduling period;

[0063] T is the total number of scheduling periods per year;

[0064] M is the number of cascade reservoirs.

[0065] Furthermore, the water balance constraints, water balance constraints, reservoir discharge flow constraints, power station output constraints, and boundary constraints in step 3 are as follows:

[0066] (a) Water balance constraints

[0067] (13)

[0068] In formula (13), and is the inflow and outflow of reservoir i during period t;

[0069] (b) Reservoir water level constraints:

[0070] (14)

[0071] In formula (14), and are the maximum and minimum allowable operating water levels of the i-th reservoir during period t, respectively;

[0072] (c) Reservoir discharge flow constraints:

[0073] (15)

[0074] In formula (15), and are the minimum and maximum allowable discharge flows of the i-th reservoir during period t, respectively;

[0075] (d) Power station output constraints:

[0076] (16)

[0077] In formula (16), and are the minimum and maximum output limits of the i-th reservoir during period t, respectively;

[0078] (e) Boundary constraints:

[0079] (17)

[0080] In formula (17), and are the initial and final water levels of the i-th reservoir respectively.

[0081] Furthermore, it also includes:

[0082] Step 4: Comprehensive performance evaluation

[0083] Calculate the water storage rate and power generation guarantee rate, and compare and analyze the comprehensive benefits of flood control, power generation, and water storage between the dynamic scheduling scheme and the original design scheme. The water storage rate calculation formula is:

[0084] (18)

[0085] The calculation formula for power generation guarantee rate is:

[0086] (19)

[0087] In formulas (18) and (19), and They are the water storage rate and power generation guarantee rate indicators of cascade reservoirs; It is a Boolean variable. When the output of the cascade reservoir is greater than the guaranteed output When , its value is 1; otherwise it is 0.

[0088] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0089] 1. Scientific and reasonable, close to engineering practice:

[0090] This paper adopts a flood season-based strategy, combined with my country's actual reservoir scheduling needs and the current status of reservoir expansion, to conduct targeted optimization research on the scheduling targets of cascade reservoirs in different flood seasons. For example, the cascade reservoirs in the pre-flood season and the main flood season are optimized for "flood control-power generation", while the cascade reservoirs in the post-flood season are optimized for "flood control-water storage". Compared with traditional flood control scheduling technologies, this method is more in line with the engineering scheduling decision-making process.

[0091] 2. It can provide important and highly operational reference for reservoir operation:

[0092] The present invention adopts optimization scheduling methods such as flood season staging and "aggregation-decomposition", fully considering the flood control risks of basin reservoir group scheduling, and utilizing the complementary mechanism of flood control storage capacity of cascade reservoirs to achieve efficient utilization of water resources of cascade reservoirs, which has important reference value for the optimization management of cascade reservoirs. BRIEF DESCRIPTION OF THE DRAWINGS

[0093] Figure 1 This is a flow chart of a method for optimizing dynamic control of cascade reservoirs based on different scheduling objectives during flood seasons according to an embodiment of the present invention;

[0094] Figure 2 This is a flow chart of reservoir flood control calculation according to an embodiment of the present invention;

[0095] Figure 3 This is a design drawing of a water storage scheme for a reservoir in the post-flood season according to an embodiment of the present invention;

[0096] Figure 4 This is a topological diagram of the hydraulic relationship between two cascade reservoirs according to an embodiment of the present invention;

[0097] Figure 5 This is a diagram showing the relationship between the upper and lower limits of dynamic control of two cascade reservoirs in an embodiment of the present invention. DETAILED DESCRIPTION

[0098] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0099] The present invention addresses the problem that existing cascade reservoir design and scheduling schemes fail to efficiently utilize flood resources during the flood season. First, based on the historical flood distribution characteristics, a flood season staging method is used to divide the flood season into pre-flood season, main flood season, and post-flood season, and the scheduling objectives for each stage are determined: the pre-flood and main flood seasons are mainly "flood control-power generation", while the post-flood season is mainly "flood control-water storage". Through flood control calculations and designed water storage schemes, the lower limit of the operating water level of each reservoir in each stage is determined. Subsequently, based on the "aggregation and decomposition" concept, the cascade reservoir flood control storage capacity compensation mechanism is utilized, and a pre-storage and pre-discharge method is adopted to determine the dynamic control upper limit relationship of each cascade reservoir. Finally, a multi-objective intelligent optimization algorithm is used to implement dynamic optimization control of the cascade reservoirs, generating multiple optimization schemes that meet different scheduling objectives.

[0100] The specific implementation process of the present invention is shown in Figure 1 , the steps are as follows:

[0101] Step 1: Divide the flood season of cascade reservoirs into stages based on the characteristics of rainstorms and floods

[0102] A comprehensive analysis method is used to determine the phased plan for the pre-flood season, main flood season and post-flood season; the comprehensive analysis method includes cause analysis method, mathematical statistics method, fuzzy analysis method, fractal analysis method, change point analysis method, system clustering method, vector statistics method and relative frequency method.

[0103] The flood samples were obtained by cross-period sampling method, with the sampling time of each period extended forward or backward by 5 days. The flood frequency was fitted by P-Ⅲ curve, and the parameters were determined by visual estimation of the fitting method.

[0104] Flood control calculations are used to derive the flood control water levels of each reservoir in different phases. The flood control water level during the pre-flood season is used as the lower limit of operation, and the flood control water level during the post-flood season is used as the initial boundary of water storage operation.

[0105] The flood process line is amplified according to the same frequency method, and the optimized scheduling models of "flood control-power generation" in the pre-flood / main flood season and "flood control-water storage" in the post-flood season are established.

[0106] Step 1 further includes the following sub-steps:

[0107] Step 1.1: Based on the deterministic, random and transitional characteristics of the runoff flood season variation pattern, a comprehensive analysis method including causal analysis, mathematical statistics, fuzzy analysis, fractal analysis, change point analysis, system clustering, vector statistics and relative frequency method is used to obtain a reasonable and feasible pre-flood season, main flood season and post-flood season stage division plan.

[0108] Step 1.2: Based on the historical daily measured flow data, the sampling of flood samples for each phase is completed by selecting the maximum sampling method for each phase and conducting appropriate cross-period sampling according to the three phases of pre-flood season, main flood season and post-flood season. The sampling time for each phase is appropriately extended forward or backward by 5 days so that the maximum flood that is greater than the current phase in the short period before and after the phase can also be included in the statistical series. It is expected that the design results will be safer.

[0109] The flood frequency curve adopts the P-Ⅲ curve. The annual maximum parameters of different periods are all calculated using the moment method as the initial value, and the parameters are determined by the eye-estimation fitting method.

[0110] The sampling of flood samples by stage includes:

[0111] The empirical frequency of extreme floods is calculated using the formula:

[0112] (1)

[0113] The empirical frequency of measured floods is calculated using the formula:

[0114] (2)

[0115] In formulas (1) and (2), M is the serial number of the major floods from large to small; P M is the empirical frequency of the Mth serial number of the major flood; N is the number of years from the earliest survey year to the present; a is the number of major floods (including the major floods in actual measurement); m is the serial number of the actual measurement series arranged from large to small; P m is the empirical frequency of the mth item in the measured series; n is the number of years in the measured series; l is the number of items in the measured series that are treated as extremely large values.

[0116] Step 1.3, use flood control calculation method (such as Figure 2 ) Calculate the flood control water levels for each reservoir by stage. The flood control water levels calculated for the pre-flood season and the main flood season serve as the lower limit of the reservoir water level, while the flood control water levels for the post-flood season serve as the initial boundary of the reservoir water storage and scheduling level. When selecting the reservoir's staged design flood hydrographs, the principle that high flood peaks and large flood volumes are detrimental to reservoir flood control is followed. The design flood hydrographs for each flood season are amplified using the same frequency method.

[0117] The design flood process line amplification includes:

[0118] The peak magnification ratio is:

[0119] (3)

[0120] The maximum 1d flood magnification ratio is:

[0121] (4)

[0122] The magnification ratios of the remaining 2 days, except for the maximum 1 day, in the maximum 3 day flood volume are:

[0123] (5)

[0124] The magnification ratios of the remaining 4 days, except for the maximum 3 days, in the maximum 7 days of flood volume are:

[0125] (6)

[0126] In formulas (3)-(6), Q p Design peak; Q d is the typical annual flood peak; W p is the design flood volume; W d This is a typical annual flood volume.

[0127] Step 1.4: Determine the dispatching targets for each stage of the flood season. The first and main flood seasons are still based on the traditional “flood control-power generation” strategy, while the post-flood season is based on “flood control-water storage” strategy. Figure 3 As shown in the figure, establishing a "flood control-water storage" optimization scheduling model and conducting a feasibility analysis can achieve the goal of meeting water storage requirements without increasing flood control risks.

[0128] Step 2: Construct the dynamic control domain of cascade reservoirs

[0129] Aggregate cascade reservoirs into virtual reservoirs and determine the maximum allowable pre-storage capacity of the aggregated reservoirs based on flood control constraints and water and rainfall forecasts;

[0130] Decompose the pre-storage capacity of each reservoir according to the hydraulic connection and establish the upper and lower limit relationship of the dynamic water level (Equation 8);

[0131] The pre-storage and pre-release strategy is used to allocate the additional allocable storage capacity for small and medium flood events (Equations 9-10).

[0132] Step 2 further includes the following sub-steps:

[0133] Step 2.1: Treat the cascade reservoirs as an aggregate reservoir and determine the maximum allowable pre-storage capacity of the aggregate reservoir based on the cascade flood control constraints and the forecast runoff information. Figure 4 For the two series-connected reservoirs (reservoirs A and B) and their respective downstream flood control points (F1 and F2), the maximum allowable pre-storage capacity of the aggregated reservoir is calculated as follows:

[0134] (7)

[0135] In formula (7), and are the inflow and outflow of the aggregated reservoir at time t, respectively; The effective forecast period; is the reservoir water level-storage capacity function relationship; and are the upper and lower limit water levels of the reservoir respectively; By using short-term forecast information at time t, the cascade reservoirs can store more water.

[0136] Step 2.2, according to the hydraulic connection between the cascade reservoirs and the system objectives, coordinate and decompose the pre-storage capacity of each reservoir, that is, determine the upper and lower limit relationship of the water level of each cascade reservoir. Figure 4 Take the two reservoirs in the middle as an example, the control area of the cascade reservoirs is as follows: Figure 5 The shaded part is shown in the figure, and its mathematical expression is shown in formula (8).

[0137] f B ( Z B ' ( t )) ≤ f B ( Z B ( t )) + Q max B ⋅ T f − ∫ t c t c + T f Q in B ( t ) dt = f B ( Z B ( t )) + [ Q max B − K ( t )] ⋅ T f − C 0 ⋅ [ f A ( Z A ' ( t )) − f A ( Z A ( t )) + ∫ t c t c + T f Q in A ( t ) dt ] (8)

[0138] In formula (8), K(t) is the Muskegon parameter of river channel evolution, Q max B is the maximum safe discharge at the downstream flood control point of reservoir B. Based on this aggregation and decomposition method, the dynamic upper and lower limit water level relationship of each reservoir in the cascade reservoir can be determined according to formula (8).

[0139] In step 2.3, for small and medium-sized flood events within the limited forecast period, the forecast and pre-discharge dynamic control strategy is adopted to determine the additional allocable water storage capacity of each reservoir within the dynamic control domain.

[0140] (9)

[0141] In formula (9), N is the number of cascade reservoirs, Allocate additional storage capacity for reservoir i at time t; and The total water storage capacity allocation coefficient for each reservoir satisfies .

[0142] Therefore, for the predicted small and medium-sized flood events, each reservoir dynamically controls the target water level. It can be determined by formula (10).

[0143] (10)

[0144] Step 3: Multi-objective optimization scheduling

[0145] Establish an optimization model with the goal of maximizing cascade power generation and minimizing flood control risks (Equations 11-12);

[0146] Set water balance constraints, water balance constraints, reservoir discharge flow constraints, power station output constraints, and boundary constraints (Equations 13-17);

[0147] The optimization model is solved by a multi-objective intelligent optimization algorithm to obtain a dynamic scheduling solution.

[0148] Specifically, taking the maximum average total power generation of cascade reservoirs over many years and the minimum flood control risk rate as the optimization objective function, an optimal scheduling model for cascade reservoirs is established, and its objective function expression is:

[0149] Minimizing flood risks:

[0150] (11)

[0151] The average total power generation over many years is the largest:

[0152] (12)

[0153] Among them, FCR and THG are the multi-year flood control risk and power generation of the cascade reservoirs, respectively;

[0154] FCR i (t) and HG i (t) are the flood control risk and power generation of the i-th reservoir at time t, respectively; is the ratio of the flood control storage capacity of the i-th reservoir to the total storage capacity of the cascade;

[0155] Vi(t) is the storage capacity of reservoir i at time t;

[0156] V i max (t) and V i min (t) are the normal water storage capacity and flood control limit water storage capacity of the i-th reservoir at time t;

[0157] K i is the output coefficient of reservoir i, QG i (t) is the power generation flow of reservoir i in period t, H i (t) is the average power generation head of reservoir i in period t;

[0158] △t is the time interval;

[0159] Y is the number of years in the scheduling period;

[0160] T is the total number of scheduling periods per year;

[0161] M is the number of cascade reservoirs.

[0162] The following constraints are considered in this embodiment:

[0163] (a) Water balance constraints

[0164] (13)

[0165] In formula (13), and is the inflow and outflow of reservoir i during period t.

[0166] (b) Reservoir water level constraints:

[0167] (14)

[0168] In formula (14), and are the maximum and minimum allowable operating water levels of the i-th reservoir during period t.

[0169] (c) Reservoir discharge flow constraints:

[0170] (15)

[0171] In formula (15), and are the minimum and maximum allowable discharge flows of the i-th reservoir during period t, respectively.

[0172] (d) Power station output constraints:

[0173] (16)

[0174] In formula (16), and are the minimum and maximum output limits of the i-th reservoir during period t, respectively.

[0175] (e) Boundary constraints:

[0176] (17)

[0177] In formula (17), and are the initial and final water levels of the i-th reservoir respectively.

[0178] A multi-objective intelligent optimization method is used to find the optimal solution for the optimization model and obtain the final dynamic operation plan for the cascade reservoirs to guide the real-time operation of the cascade reservoirs. The multi-objective intelligent optimization algorithm is a non-dominated sorting genetic algorithm (NSGA-II) with an elite strategy.

[0179] Step 4: Comprehensive performance evaluation

[0180] The water storage rate (Equation 18) and power generation guarantee rate (Equation 19) were calculated, and the comprehensive benefits of flood control, power generation, and water storage of the dynamic scheduling scheme and the original design scheme were compared and analyzed.

[0181] In this specific implementation, the optimization scheduling objectives other than power generation and flood control, such as water storage rate and power generation guarantee rate, are analyzed, and their definitions are shown in Equations (18)-(19).

[0182] (18)

[0183] (19)

[0184] In formulas (18) and (19), and They are the water storage rate and power generation guarantee rate indicators of cascade reservoirs; It is a Boolean variable. When the output of the cascade reservoir is greater than the guaranteed output When , its value is 1; otherwise, it is 0. The other symbols are defined in the above formula.

[0185] In summary, the present invention addresses the problem that existing cascade reservoir design and scheduling schemes fail to efficiently utilize flood resources during the flood season. First, based on the historical flood distribution characteristics, the flood season is divided into pre-flood season, main flood season, and post-flood season using a flood season staging method, and the scheduling objectives for each stage are determined: the pre-flood and main flood seasons are mainly "flood control-power generation", while the post-flood season is mainly "flood control-water storage". Through flood control calculations and designed water storage schemes, the lower limit of the operating water level of each reservoir in each stage is determined. Subsequently, based on the idea of "aggregation and decomposition", the flood control storage capacity compensation mechanism of cascade reservoirs is utilized, and the pre-storage and pre-discharge method is adopted to determine the dynamic control upper limit relationship of each cascade reservoir. Finally, a multi-objective intelligent optimization algorithm is used to implement dynamic optimization control of the cascade reservoirs, generating multiple optimization schemes that meet different scheduling objectives.

[0186] Compared with existing scheduling technologies, the method of the present invention can promote the further utilization of flood water resources in the flood season within the basin and has important reference value for the optimized management of cascade reservoirs. The technical solution of the present invention has been verified in a certain reservoir and can increase power generation by 3.16% relative to the design solution without increasing flood control risks.

[0187] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present invention should be included in the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

Claims

1. A method for optimizing dynamic control of cascade reservoirs based on different flood season scheduling objectives, characterized in that: The steps include: Step 1: Divide the flood season of cascade reservoirs into stages based on the characteristics of rainstorms and floods Use comprehensive analysis methods to determine the phased plan for the pre-flood season, main flood season, and post-flood season; The flood samples were obtained by cross-period sampling method, with the sampling time of each period extended forward or backward by 5 days. The flood frequency was fitted by P-Ⅲ curve, and the parameters were determined by visual estimation of the fitting method. Flood control calculations are used to derive the flood control water levels of each reservoir in different phases. The flood control water level during the pre-flood season is used as the lower limit of operation, and the flood control water level during the post-flood season is used as the initial boundary of water storage operation. Using the same frequency method to amplify the flood process line, an optimal scheduling model for "flood control-power generation" in the pre-flood / main flood season and "flood control-water storage" in the post-flood season was established; Step 2: Construct the dynamic control domain of cascade reservoirs Aggregate cascade reservoirs into virtual reservoirs and determine the maximum allowable pre-storage capacity of the aggregated reservoirs based on flood control constraints and water and rainfall forecasts; Decompose the pre-storage capacity of each reservoir according to the hydraulic connection and establish the upper and lower limit relationship of dynamic water level; Adopting a pre-storage and pre-release strategy to allocate additional allocable storage for small and medium flood events; Step 3: Multi-objective optimization scheduling Based on the flood season stage results of step 1 and the dynamic control domain of the cascade reservoirs constructed in step 2, an optimization model is constructed with the goal of maximizing cascade power generation and minimizing flood control risks. Set water balance constraints, water balance constraints, reservoir discharge flow constraints, power station output constraints and boundary constraints; The optimization model is solved by a multi-objective intelligent optimization algorithm to obtain a dynamic scheduling solution.

2. The method for optimizing dynamic control of cascade reservoirs based on different flood season scheduling objectives according to claim 1, characterized in that: The comprehensive analysis method described in step 1 includes causal analysis method, mathematical statistics method, fuzzy analysis method, fractal analysis method, change point analysis method, system clustering method, vector statistics method and relative frequency method.

3. The method for optimizing dynamic control of cascade reservoirs based on different flood season scheduling objectives according to claim 1, characterized in that: The phased flood sample selection in step S1 includes: The empirical frequency of extreme floods is calculated using the formula: （1）; The empirical frequency of measured floods is calculated using the formula: （2）; Where, M is the serial number of the major floods arranged from large to small; P M is the empirical frequency of the Mth serial number of the major flood; N is the number of years from the earliest survey year to the present; a is the number of major floods; m is the serial number of the measured series arranged from large to small; P m is the empirical frequency of the mth item in the measured series; n is the number of years in the measured series; l is the number of items in the measured series that are treated as extremely large values.

4. The method for optimizing dynamic control of cascade reservoirs based on different flood season scheduling objectives according to claim 1, characterized in that: The design flood process line amplification in step S1 includes: The peak magnification ratio is: （3）; The maximum 1d flood magnification ratio is: （4）; The magnification ratios of the remaining 2 days, except for the maximum 1 day, in the maximum 3 day flood volume are: （5）; The magnification ratios of the remaining 4 days, except for the maximum 3 days, in the maximum 7 days of flood volume are: （6）; Where Q p Design peak; Q d is the typical annual flood peak; W p is the design flood volume; W d This is a typical annual flood volume.

5. The method for optimizing dynamic control of cascade reservoirs based on different flood season scheduling objectives according to claim 1, characterized in that: The calculation formula for the maximum pre-storage capacity allowed by the aggregate reservoir in step 2 is: （7）; In formula (7), and are the inflow and outflow of the aggregated reservoir at time t, respectively; The effective forecast period; is the reservoir water level-storage capacity function relationship; and are the upper and lower limit water levels of the reservoir respectively; By using short-term forecast information at time t, the cascade reservoirs can store more water.

6. The method for optimizing dynamic control of cascade reservoirs based on different flood season scheduling objectives according to claim 1, characterized in that: The relationship between the upper and lower limits of the dynamic water level in step 2 is: （8）; Where K(t) is the Muskegon parameter of river channel evolution, Q max B It is the maximum safe discharge volume of the downstream flood control point of Reservoir B.

7. The method for optimizing dynamic control of cascade reservoirs based on different flood season scheduling objectives according to claim 1, characterized in that: The additional allocable water storage capacity in step 2 is: （9）; Where N is the number of cascade reservoirs, Allocate additional storage capacity for reservoir i at time t; and The total water storage capacity allocation coefficient for each reservoir satisfies ; For predicted small and medium-sized flood events, each reservoir dynamically controls the target water level Determined by formula (10): （10）。 8. The method for optimizing dynamic control of cascade reservoirs based on different flood season scheduling objectives according to claim 1, characterized in that: The objective function expression of the optimization model in step 3, which aims to maximize cascade power generation and minimize flood control risk, is: Minimizing flood risks: （11）; The average total power generation over many years is the largest: （12）; Among them, FCR and THG are the multi-year flood control risk and power generation of the cascade reservoirs, respectively; FCR i (t) and HG i (t) are the flood control risk and power generation of the i-th reservoir at time t, respectively; is the ratio of the flood control storage capacity of the i-th reservoir to the total storage capacity of the cascade; Vi(t) is the storage capacity of reservoir i at time t; V i max (t) and V i min (t) are the normal water storage capacity and flood control limit water storage capacity of the i-th reservoir at time t; K i is the output coefficient of reservoir i, QG i (t) is the power generation flow of reservoir i in period t, H i (t) is the average power generation head of reservoir i in period t; △t is the time interval; Y is the number of years in the scheduling period; T is the total number of scheduling periods per year; M is the number of cascade reservoirs.

9. The method for optimizing dynamic control of cascade reservoirs based on different flood season scheduling objectives according to claim 8, characterized in that: The water balance constraints, water balance constraints, reservoir discharge flow constraints, power station output constraints, and boundary constraints in step 3 are as follows: (a) Water balance constraints （13）; In formula (13), and is the inflow and outflow of reservoir i during period t; (b) Reservoir water level constraints: （14）; In formula (14), and are the maximum and minimum allowable operating water levels of the i-th reservoir during period t, respectively; (c) Reservoir discharge flow constraints: （15）; In formula (15), and are the minimum and maximum allowable discharge flows of the i-th reservoir during period t, respectively; (d) Power station output constraints: （16）; In formula (16), and are the minimum and maximum output limits of the i-th reservoir during period t, respectively; (e) Boundary constraints: （17）; In formula (17), and are the initial and final water levels of the i-th reservoir respectively.

10. The method for optimizing dynamic control of cascade reservoirs based on different flood season scheduling objectives according to claim 1, characterized in that: Also includes: Step 4: Comprehensive performance evaluation Calculate the water storage rate and power generation guarantee rate, and compare and analyze the comprehensive benefits of flood control, power generation, and water storage between the dynamic scheduling scheme and the original design scheme. The water storage rate calculation formula is: （18）; The calculation formula for power generation guarantee rate is: （19）; In formulas (18) and (19), and They are the water storage rate and power generation guarantee rate indicators of cascade reservoirs; It is a Boolean variable. When the output of the cascade reservoir is greater than the guaranteed output When , its value is 1, otherwise it is 0.

Images