Cascade reservoir water storage period joint power generation optimization scheduling method, system and storage medium

Abstract

The present application relates to a method, system and storage medium for optimizing the joint power generation during the impoundment period of cascade reservoirs. The method includes collecting basin hydrological data; determining basic calculation parameters; determining the optimization scheduling target of power generation during the impoundment period of cascade reservoirs, setting the impoundment scheduling constraints of cascade reservoirs, and establishing a reservoir impoundment period scheduling model; obtaining an initial feasible solution; based on the number of reservoirs M and the number of discrete states F, using a key sampling method to construct a small number of all possible state combinations that are representative, and making each reservoir state meet the feasible domain range; using a parallel dynamic programming algorithm to calculate the target optimization function subproblem until the convergence condition of the optimization algorithm is reached, and the final optimized impoundment scheduling plan is obtained. The present application can not only ensure that the optimization result is close to the global optimal solution, but also effectively reduce the state variable search space compared to the traditional dynamic programming algorithm, so as to reduce the storage and computation required by the algorithm.

Description

Technical Field

[0001] The present application relates to the field of cascade reservoir scheduling, and in particular to a method, system and storage medium for optimizing scheduling of joint power generation during the water storage period of cascade reservoirs based on dimensionality reduction dynamic programming. Background Art

[0002] In recent years, as cascade reservoirs in large river basins have been built and put into use, the proportion of these reservoirs' beneficial storage capacity relative to the average annual runoff of the river basin has increased significantly. The impact of post-flood water storage scheduling on the natural flow of the river channel has become increasingly significant, and the contradiction between the water storage demand of upstream reservoirs and the water demand of downstream reservoirs has become increasingly severe. According to traditional designs, these reservoirs usually store water in a centralized manner within 1 to 2 months after the flood season, resulting in competition between the reservoirs during the water storage process; if there is insufficient water after the flood season, it will be more difficult to reach the normal water storage level, which will in turn affect the realization of the reservoir's beneficial goals. At the same time, the increase in the proportion of water storage during the water storage period has caused the downstream areas to face a significant reduction in water volume, which has also brought inconvenience to the ecological protection and public welfare scheduling of the river basin. In view of the scale of hydropower development, scientifically formulating water storage plans for cascade reservoirs in the river basin to efficiently utilize flood resources has become an engineering problem that needs to be solved urgently.

[0003] Although domestic and foreign scholars have achieved some results in optimizing the scheduling of advance water storage at the end of the flood season for a single reservoir, there is currently no complete set of theories for water storage scheduling at the end of the flood season to guide the joint water storage scheduling of cascade reservoirs. The key problem is that it is difficult to formulate unified water storage scheduling rules to maximize the benefits of water storage (such as the maximum total power generation during the water storage period) while ensuring the safety of flood control in the basin and meeting the needs of downstream water supply and river navigation. For the single-objective optimization scheduling problem of reservoir water storage period, numerical optimization algorithms represented by dynamic programming have achieved a series of outstanding achievements and have been widely used in engineering practice. However, with the increase in the number of cascade reservoirs, the problem of "dimensionality explosion" has gradually emerged in the optimization process. How to overcome the "dimensionality disaster" problem in the optimization process of cascade reservoir water storage period has become a new technical problem.

[0004] The parallel dynamic programming algorithm can alleviate the dimensionality curse problem to a certain extent. In a multi-core environment, it can use the "fork / merge" framework to implement parallel programming calculations, effectively shortening the calculation time of traditional dynamic programming methods. However, when it comes to cascade reservoirs (the number of reservoirs 4) When it comes to daily-scale water storage scheduling problems, as the number of reservoirs increases, the number of state combinations in the reservoir capacity state space grows exponentially, and the traditional parallel dynamic programming algorithm loses its versatility because it cannot traverse all state combinations. In fact, most state combinations in the state space violate complex hydraulic constraints and do not need to be traversed. Under such conditions, this application introduces a key sampling method, which can effectively screen a small number of highly representative state combinations through Manhattan distance. Combining key sampling with the parallel dynamic programming algorithm, an optimization algorithm of dimensionality reduction dynamic programming is proposed for the first time, and it is applied to the optimal water storage scheduling of cascade reservoirs to obtain the optimized water storage scheduling process, analyze and evaluate the algorithm performance and its impact on the water storage optimization results. Summary of the invention

[0005] In view of the shortcomings of the prior art, the present application provides a method, system and storage medium for optimizing the joint power generation scheduling of cascade reservoirs during the water storage period based on dimensionality reduction dynamic programming. This application aims at the power generation optimization problem of cascade reservoirs during the water storage period, comprehensively considers the actual constraint boundary conditions faced in the operation of water storage scheduling, establishes a single-objective optimization scheduling model with the maximum power generation during the water storage period, and proposes a dimensionality reduction dynamic programming algorithm that combines parallel computing, key sampling and successive approximation techniques to solve the "curse of dimensionality" problem in the joint optimization scheduling of cascade reservoirs, and obtains the optimal scheduling plan for joint water storage.

[0006] To achieve the above objectives, this application provides the following technical solutions:

[0007] In a first aspect, an embodiment of the present application provides a method for optimizing the joint power generation scheduling during the water storage period of cascade reservoirs, comprising the following steps:

[0008] Step 1, collecting basin hydrological data;

[0009] Step 2: Determine the basic calculation parameters and the number of reservoirs , the discrete number of reservoir capacity states F, the number of dynamic programming calculation processors P, and the algorithm convergence accuracy ;

[0010] Step 3, determine the optimization scheduling target of power generation during the water storage period of the cascade reservoirs, set the water storage scheduling constraints of the cascade reservoirs, and establish a scheduling model for the reservoir water storage period;

[0011] Step 4: Use a dynamic programming algorithm to gradually optimize the cascade reservoirs from upstream to downstream to obtain an initial feasible solution;

[0012] Step 5, based on the number of reservoirs and discrete number F, use the key sampling method to construct a small number of representative combinations of all possible states, and make each reservoir state meet the feasible domain range;

[0013] Step 6: Use the parallel dynamic programming algorithm to calculate the target optimization function sub-problem until the convergence condition of the optimization algorithm is reached to obtain the final optimized water storage scheduling plan.

[0014] In step 3, the idea of ​​penalty function is introduced. For constraints that are difficult to satisfy the feasible solution range, a Lagrangian objective function based on penalty function is constructed by adding penalty function factors.

[0015] The specific objectives of optimized power generation dispatching during the impoundment period of cascade reservoirs are:

[0016] (1)

[0017] In formula (1), is the total power generation of the cascade reservoirs during the impoundment period; For the Reservoir Average output at any time; is the scheduling time step; for The average output coefficient of the reservoir; For the Reservoir The power generation flow at the moment; For the Reservoir Average net water head of power generation at the time; is the number of reservoirs in the cascade reservoir group; is the total number of scheduling time steps.

[0018] The constraints for the cascade reservoir water storage scheduling are:

[0019] (a) Water balance constraints

[0020] (2)

[0021] In formula (2), , Indicates Reservoir Initial and final storage capacity at each time; , , The water storage period Reservoir Inbound, outbound and loss flows at any given moment;

[0022] (b) Water level upper and lower limit constraints and water level range constraints

[0023] (3)

[0024] (4)

[0025] In formula (3)-(4), , , Respectively Reservoir The lower limit water level, operating water level and upper limit water level allowed at all times; For the The allowable water level fluctuation of the reservoir at adjacent moments;

[0026] (c) Reservoir outflow and flow range constraints

[0027] (5)

[0028] (6)

[0029] In formulas (5)-(6), and Respectively Reservoir Minimum and maximum outbound flow at each time m 3 / s, It is determined by the maximum discharge capacity of the reservoir and the downstream flood control task; For the Reservoir Amount of water discarded at any time m 3 / s; For the The maximum fluctuation of the reservoir’s daily discharge;

[0030] (d) Power station output constraints

[0031] (7)

[0032] In formula (7), and For the The reservoir guarantees output and installed capacity;

[0033] (e) Boundary Adjustment Constraints

[0034] (8)

[0035] Where: and Respectively The reservoir's initial water level and final water level generally correspond to the flood control water limit level and the reservoir's normal water storage level;

[0036] (f) The shape constraint of the water storage scheduling line means that the scheduling lines do not cross each other and are as smooth as possible to ensure that the water level does not fluctuate significantly.

[0037] The Lagrangian objective function based on the penalty function is:

[0038] The storage capacity of each reservoir is taken as the state variable, the outflow flow is taken as the decision variable, the water balance constraint that is easy to achieve is taken as the state transfer equation, and the water storage start and end boundary constraints are taken as the initial and final states; for the constraints that are difficult to meet the feasible solution range, the penalty function factor is introduced to form a penalty function. The new Lagrangian optimization function is shown in formula (9):

[0039] (9)

[0040] (10)

[0041] (11)

[0042] (12)

[0043] In formulas (9)-(12), is the Lagrangian gain function; As the penalty function factor, the corresponding penalty function is [ ], which are used to penalize the infeasible solutions that violate the reservoir water level fluctuation constraint, the downstream flow constraint and the unit output limit respectively.

[0044] In a second aspect, an embodiment of the present application provides a system for optimizing the joint power generation and scheduling during the water storage period of cascade reservoirs, comprising:

[0045] Data collection module, collecting basin hydrological data;

[0046] Calculation parameter determination module, determine the basic calculation parameters, number of reservoirs , the discrete number of reservoir capacity states F, the number of dynamic programming calculation processors P, and the algorithm convergence accuracy ;

[0047] The model building module determines the optimization scheduling target of power generation during the impoundment period of cascade reservoirs, sets the impoundment scheduling constraints of cascade reservoirs, and establishes the scheduling model of reservoir impoundment period;

[0048] The initial feasible solution calculation module uses a dynamic programming algorithm to gradually optimize the cascade reservoirs from upstream to downstream to obtain the initial feasible solution;

[0049] State combination building block, based on the number of reservoirs and discrete number F, use the key sampling method to construct a small number of representative combinations of all possible states, and make each reservoir state meet the feasible domain range;

[0050] The water storage scheduling scheme determination module uses a parallel dynamic programming algorithm to calculate the target optimization function sub-problem until the convergence condition of the optimization algorithm is reached, and the final optimized water storage scheduling scheme is obtained.

[0051] In a third aspect, an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores program code, and when the program code is executed by a processor, the steps of the method for optimizing the scheduling of joint power generation during the water storage period of cascade reservoirs as described above are implemented.

[0052] Compared with the prior art, the beneficial effects of this application are:

[0053] 1. Fast optimization efficiency, meeting the needs of optimizing water storage calculation for cascade reservoirs:

[0054] This application uses key sampling technology to effectively reduce the number of candidate state variables of reservoir capacity at each moment, and achieves efficient solution through parallel computing methods, thereby solving problems such as "dimensionality curse" in the optimal scheduling of cascade reservoirs. The more processors involved in parallel computing, the less time it takes. Compared with traditional dynamic programming algorithms, the larger the scale of cascade reservoirs, the more obvious the advantages of this method, which can provide a feasible idea for large-scale single-objective optimization problems.

[0055] 2. It can provide important and highly operable reference for the water storage scheduling of cascade reservoirs:

[0056] Make full use of the limited water resources during the impoundment period of the cascade reservoirs to optimize the comprehensive benefits of the cascade reservoirs. Compared with the single-reservoir simulation scheduling of the original designed impoundment plan, it can ensure that the power generation benefits of the reservoir group are maximized without changing the current water conservancy project pattern. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments of the present application will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0058] Figure 1 A specific flow chart of the application method;

[0059] Figure 2 This is a schematic diagram of the key sampling of five levels in two reservoirs;

[0060] Figure 3 It is the flow chart of parallel dynamic programming algorithm;

[0061] Figure 4 The principle of parallel dynamic programming for M reservoir F level;

[0062] Figure 5 This is a system block diagram of this application. DETAILED DESCRIPTION

[0063] The technical solutions in the embodiments of the present application will be described below in conjunction with the drawings in the embodiments of the present application. It should be noted that similar reference numerals and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further defined and explained in the subsequent drawings.

[0064] The terms "comprises," "comprising," or any other variation thereof are intended to encompass non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of additional identical elements in the process, method, article, or apparatus that includes the element.

[0065] The terms "first", "second", etc. are only used to distinguish one entity or operation from another entity or operation, and should not be understood as indicating or implying relative importance, nor should they be understood as requiring or implying any such actual relationship or order between these entities or operations.

[0066] This application aims at the power generation optimization problem of cascade reservoirs during the water storage period. It comprehensively considers the actual constraint boundary conditions faced in the water storage scheduling operation, establishes a single-objective optimization scheduling model with the maximum power generation during the water storage period, and proposes a dimensionality reduction dynamic programming algorithm using techniques such as collective parallel computing, key sampling, and successive approximation to solve the "curse of dimensionality" problem in the joint optimization scheduling of cascade reservoirs and obtain the optimal scheduling plan for joint water storage.

[0067] The specific implementation process of this application is shown in Figure 1 , the steps are as follows:

[0068] Step 1: Collect basin hydrological data.

[0069] Collect reservoir hydrological data, including inflow data observed at upstream stations of the reservoir, characteristic curves such as reservoir water level~storage capacity, reservoir water level~maximum downstream flow capacity, downstream flow~reservoir tail water level, etc.

[0070] Step 2: Determine basic calculation parameters, such as the number of reservoirs , the discrete number of reservoir capacity states F, the number of dynamic programming calculation processors P, and the algorithm convergence accuracy .

[0071] Step 3: Determine the optimal dispatching target for power generation during the impoundment period of the cascade reservoirs, set the impoundment dispatching constraints for the cascade reservoirs, and establish a dispatching model for the impoundment period of the reservoirs.

[0072] This step belongs to the conventional technology in this technical field. For the sake of easy understanding, this step will be described in detail below.

[0073] In this specific implementation, the goal of cascade reservoir water storage scheduling is to ensure that the reservoir is as full as possible at the end of water storage while generating as much power generation benefit as possible at the time of reservoir water storage while satisfying various scheduling constraints.

[0074] (1)

[0075] In formula (1), is the total power generation of the cascade reservoirs during the impoundment period; For the Reservoir Average output at any time; is the scheduling time step; for The average output coefficient of the reservoir; For the Reservoir The power generation flow at the moment; For the Reservoir Average net water head of power generation at the time; is the number of reservoirs in the cascade reservoir group; is the total number of scheduling time steps.

[0076] The constraints considered in this specific implementation are as follows.

[0077] (a) Water balance constraints

[0078] (2)

[0079] In formula (2), , Indicates Reservoir Initial and final storage capacity at each time; , , The water storage period Reservoir Inbound, outbound and loss flows at any given moment;

[0080] (b) Water level upper and lower limit constraints and water level range constraints

[0081] (3)

[0082] (4)

[0083] In formula (3)-(4), , , Respectively Reservoir The lower limit water level, operating water level and upper limit water level allowed at all times; For the The allowable water level fluctuation of the reservoir at adjacent moments;

[0084] (c) Reservoir outflow and flow range constraints

[0085] (5)

[0086] (6)

[0087] In formulas (5)-(6), and Respectively Reservoir Minimum and maximum outbound flow at each time m 3 / s, It is determined by the maximum discharge capacity of the reservoir and the downstream flood control task; For the Reservoir Amount of water discarded at any time m 3 / s; For the The maximum fluctuation of the reservoir’s daily discharge;

[0088] (d) Power station output constraints

[0089] (7)

[0090] In formula (7), and For the The reservoir guarantees output and installed capacity;

[0091] (e) Boundary Adjustment Constraints

[0092] (8)

[0093] Where: and Respectively The reservoir's initial water level and final water level generally correspond to the flood control water limit level and the reservoir's normal water storage level;

[0094] (f) The shape constraint of the water storage scheduling line means that the scheduling lines do not cross each other and are as smooth as possible to ensure that the water level does not fluctuate significantly.

[0095] For the single-objective optimization model with constraints, a feasible idea is to introduce the penalty function idea. The storage capacity of each reservoir is taken as the state variable, the outflow flow is taken as the decision variable, the water balance constraint that is easy to achieve is taken as the state transfer equation, and the water storage start and end boundary constraints are taken as the initial and end states; for the constraints that are difficult to meet the feasible solution range, the penalty function factor is introduced to form the penalty function, and the new Lagrangian optimization function is shown in formula (9).

[0096] (9)

[0097] (10)

[0098] (11)

[0099] (12)

[0100] In formulas (9)-(12), is the Lagrangian gain function; As the penalty function factor, the corresponding penalty function is [ ], which are used to penalize the infeasible solutions that violate the reservoir water level fluctuation constraint, the downstream flow constraint and the unit output limit respectively.

[0101] Step 4: Use the dynamic programming algorithm to gradually optimize the cascade reservoirs from upstream to downstream to obtain the initial feasible solution, as shown in part (1a).

[0102] Step 5, based on the number of reservoirs and the discrete number F, the key sampling method is used to construct a small but representative combination of all possible states, and each reservoir state satisfies the feasible domain.

[0103] In this specific implementation, each reservoir in the system is regarded as a full factor, the number of discrete states of each reservoir is regarded as the number of levels of the factor, each state combination is regarded as a possible experiment, and the optimization target size is regarded as the quality of each experimental result. By introducing the Manhattan distance (the difference between adjacent state variables of the same reservoir at the same time in this application represents a unit of Manhattan distance), the principle of key sampling is used to greatly reduce the candidate solutions of each iteration, saving the optimization running time, thereby making it possible to optimize the water storage scheduling of cascade reservoirs on a daily scale.

[0104] The dispatching result obtained by gradually optimizing from the upstream reservoir to the downstream reservoir in step (4) is used as the initial feasible solution. The state vector closer to the initial solution is easier to meet the hydraulic and power dispatching constraints of the complex water storage system, that is, the smaller its Manhattan distance is, the greater its importance is, and the more it should be extracted for the experiment in this application.

[0105] like Figure 2 For example, two reservoirs and five levels, the full sampling algorithm has 5 5 = 25 state combinations, but the focused sampling method only screens out 6 representative state combinations. Figure 2 It can be seen that the key sampling method requires a given initial solution trajectory. Generally speaking, the initial feasible solution needs to satisfy all constraints as much as possible. The closer the state variable is to the initial trajectory, the greater the possibility that it satisfies a series of complex constraints. Quantitatively speaking, the shorter the Manhattan distance from the initial solution, the greater its importance. Figure 2 For the five levels of the two reservoirs in the middle, the initial Manhattan distance has five different values ​​from 0 to 4, and the corresponding numbers are 1, 4, 8, 8 and 4 (the total is 25); according to the corresponding importance sampling theory, the values ​​can be extracted as 1, 2, 1, 1 and 1 (the total is 6). It can be seen that the key sampling method can effectively reduce the calculation time and obtain a more satisfactory target solution in practical engineering applications. The larger the scale of the research problem, the more obvious its excellent performance.

[0106] Step 6: Use the parallel dynamic programming algorithm to calculate the target optimization function subproblem until the optimization algorithm converges to the desired accuracy. , and obtain the final optimized water storage scheduling plan.

[0107] In this specific implementation, the parallel dynamic programming algorithm inherits the idea of ​​the traditional dynamic programming algorithm. In essence, it still decomposes a multi-stage decision problem into several two-stage sub-problems, and then solves these sub-problems in sequence. The algorithm flow is as follows: Figure 3 As shown in the figure, the discrete form of the parallel dynamic programming algorithm enumerates all state combinations to ensure the global optimality of the solution to the problem. At the same time, the time complexity of the algorithm is minimized, that is, the calculation tasks of the algorithm at the same calculation time are increased and handed over to multiple processors or multiple computers to complete the calculation tasks at the same time. The total time of the algorithm operation is reduced by increasing the space complexity, thereby achieving the purpose of improving the calculation performance.

[0108] In this application Taking the reservoirs as an example, the feasible domain space of each reservoir's storage capacity is evenly divided into F equal parts, the storage capacity of each reservoir is regarded as a state variable, and the corresponding outflow is regarded as a decision variable. The objective function and the state transfer equation are shown in Formula (13).

[0109] (13)

[0110] In formula (13), From the initial time to The maximum cumulative function value at the time, initial ; and are the discrete state variable set and the decision variable set respectively; Cascade reservoirs The state vector at the moment, ; for The decision vector of cascade reservoirs at time .

[0111] For the water storage scheduling of the cascade reservoirs, the state variables of each reservoir at the start and stop of storage are fixed, and their values ​​are determined by the combination of the flood control water limit level and the normal water level state of each reservoir. At the rest of the water storage period, the state variables of the storage capacity of each reservoir are evenly and discretely distributed within the feasible domain, and form a series of state combinations. For a certain state combination at a certain moment, its calculation is only related to its own state and It is related to all state combinations at the same time, but has nothing to do with other state combinations at the current time, and has no effect on the optimization calculation of other state combinations at the current time. In other words, at the same time of the cascade reservoirs, the calculations of the storage capacity state combinations are independent of each other. Figure 4 As shown in the figure, the parallel dynamic programming algorithm makes good use of this independence, and through advanced computer multi-core configuration technology, a large number of computing tasks are distributed as evenly as possible to multiple pre-set computer processors. Under the same computing volume, the computing efficiency of the parallel mode dynamic programming algorithm is much higher than that of the traditional serial dynamic programming algorithm.

[0112] In summary, this application aims at the power generation optimization problem of cascade reservoirs during the water storage period, comprehensively considers the actual constraint boundary conditions faced in the water storage scheduling operation, establishes a single-objective optimization scheduling model with the maximum power generation during the water storage period, and proposes a dimensionality reduction dynamic programming algorithm that combines parallel computing, key sampling, and successive approximation techniques to solve the "curse of dimensionality" problem in the joint optimization scheduling of cascade reservoirs, and obtains the optimal scheduling plan for joint water storage. The proposed method can not only ensure that the optimization result is close to the global optimal solution, but also effectively reduce the state variable search space compared to the traditional dynamic programming algorithm, so as to reduce the storage and computation required by the algorithm. The larger the scale of the cascade reservoir, the more obvious the advantage of this method, which can provide a feasible idea for large-scale single-objective optimization problems.

[0113] like Figure 5 As shown, the embodiment of the present application provides a system for optimizing the joint power generation during the water storage period of cascade reservoirs, comprising:

[0114] Data collection module 1, collecting basin hydrological data;

[0115] Calculation parameter determination module 2, determine the basic calculation parameters, the number of reservoirs , the discrete number of reservoir capacity states F, the number of dynamic programming calculation processors P, and the algorithm convergence accuracy ;

[0116] Model building module 3, determines the optimization scheduling target of power generation during the water storage period of the cascade reservoirs, sets the constraints of the water storage scheduling of the cascade reservoirs, and establishes a scheduling model for the water storage period of the reservoirs;

[0117] The initial feasible solution calculation module 4 uses a dynamic programming algorithm to gradually optimize the cascade reservoirs from upstream to downstream to obtain an initial feasible solution;

[0118] State combination construction module 5, based on the number of reservoirs and discrete number F, use the key sampling method to construct a small number of representative combinations of all possible states, and make each reservoir state meet the feasible domain range;

[0119] The water storage scheduling scheme determination module 6 uses a parallel dynamic programming algorithm to calculate the target optimization function sub-problem until the optimization algorithm convergence condition is reached to obtain the final optimized water storage scheduling scheme.

[0120] An embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores program code, and when the program code is executed by a processor, the steps of the method for optimizing the scheduling of joint power generation during the water storage period of cascade reservoirs as described above are implemented.

[0121] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0122] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0123] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0124] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0125] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0126] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0127] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0128] The above description is only an embodiment of the present application and is not intended to limit the protection scope of the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for optimizing joint power generation during the impoundment period of cascade reservoirs, characterized in that: The following steps are involved: Step 1, collecting basin hydrological data; Step 2: Determine the basic calculation parameters and the number of reservoirs , the discrete number of reservoir capacity states F, the number of dynamic programming calculation processors P, and the algorithm convergence accuracy ; Step 3, determine the optimization scheduling target of power generation during the water storage period of the cascade reservoirs, set the water storage scheduling constraints of the cascade reservoirs, and establish a scheduling model for the reservoir water storage period; Step 4: Use a dynamic programming algorithm to gradually optimize the cascade reservoirs from upstream to downstream to obtain an initial feasible solution; Step 5, based on the number of reservoirs and discrete number F, use the key sampling method to construct a small number of representative combinations of all possible states, and make each reservoir state meet the feasible domain range; Step 6, using a parallel dynamic programming algorithm to calculate the target optimization function sub-problem until the convergence condition of the optimization algorithm is reached, and the final optimized water storage scheduling plan is obtained; The specific objectives of optimized power generation dispatching during the impoundment period of cascade reservoirs are: (1) In formula (1), is the total power generation of the cascade reservoirs during the impoundment period; For the Reservoir Average output at any time; is the scheduling time step; for The average output coefficient of the reservoir; For the Reservoir The power generation flow at the moment; For the Reservoir Average net water head of power generation at the time; is the number of reservoirs in the cascade reservoir group; is the total number of scheduling time steps; The Lagrangian objective function based on the penalty function is: The storage capacity of each reservoir is taken as the state variable, the outflow flow is taken as the decision variable, the water balance constraint is taken as the state transfer equation, and the water storage start and end boundary constraints are taken as the initial and final states; for the constraints that are difficult to meet the feasible solution range, the penalty function factor is introduced to form a penalty function. The new Lagrangian optimization function is shown in formula (9): （9） （10） （11） （12） In formulas (9)-(12), is the Lagrangian gain function; As the penalty function factor, the corresponding penalty function is [ ], which are used to penalize the infeasible solutions that violate the reservoir water level fluctuation constraint, the downstream flow constraint and the unit output limit respectively.

2. The method for optimizing joint power generation during the impoundment period of cascade reservoirs according to claim 1, characterized in that: In step 3, the idea of ​​penalty function is introduced. For constraints that are difficult to satisfy the feasible solution range, a Lagrangian objective function based on penalty function is constructed by adding penalty function factors.

3. The method for optimizing joint power generation during the impoundment period of cascade reservoirs according to claim 1, characterized in that: The constraints for the cascade reservoir water storage scheduling are: (a) Water balance constraints (2) In formula (2), , Indicates Reservoir Initial and final storage capacity at each time; , , The water storage period Reservoir Inbound, outbound and loss flows at any given moment; (b) Water level upper and lower limit constraints and water level range constraints (3) (4) In formula (3)-(4), , , Respectively Reservoir The lower limit water level, operating water level and upper limit water level allowed at all times; For the The allowable water level fluctuation of the reservoir at adjacent moments; (c) Reservoir outflow and flow range constraints (5) (6) In formulas (5)-(6), and Respectively Reservoir Minimum and maximum outbound flow at each time m 3 / s, It is determined by the maximum discharge capacity of the reservoir and the downstream flood control task; For the Reservoir Amount of water discarded at any time m 3 / s; For the i The maximum fluctuation of the reservoir’s daily discharge; (d) Power station output constraints (7) In formula (7), and For the The reservoir guarantees output and installed capacity; (e) Boundary Adjustment Constraints (8) Where: and Respectively The reservoir's initial water level and final water level generally correspond to the flood control water limit level and the reservoir's normal water storage level; (f) The shape constraint of the water storage scheduling line means that the scheduling lines do not cross each other and are as smooth as possible to ensure that the water level does not fluctuate significantly.

4. A system for optimizing the joint power generation during the impoundment period of cascade reservoirs, for implementing the method according to any one of claims 1 to 3, characterized in that: include, Data collection module, collecting basin hydrological data; Calculation parameter determination module, determine the basic calculation parameters, number of reservoirs , the discrete number of reservoir capacity states F, the number of dynamic programming calculation processors P, and the algorithm convergence accuracy ; The model building module determines the optimization scheduling target of power generation during the impoundment period of cascade reservoirs, sets the impoundment scheduling constraints of cascade reservoirs, and establishes the scheduling model of reservoir impoundment period; The initial feasible solution calculation module uses a dynamic programming algorithm to gradually optimize the cascade reservoirs from upstream to downstream to obtain the initial feasible solution; State combination building block, based on the number of reservoirs and discrete number F, use the key sampling method to construct a small number of representative combinations of all possible states, and make each reservoir state meet the feasible domain range; The water storage scheduling scheme determination module uses a parallel dynamic programming algorithm to calculate the target optimization function sub-problem until the convergence condition of the optimization algorithm is reached, and the final optimized water storage scheduling scheme is obtained.

5. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores program codes, and when the program codes are executed by the processor, the steps of the method for optimizing the scheduling of joint power generation during the water storage period of cascade reservoirs as described in any one of claims 1 to 3 are implemented.

Images