A flood control optimization scheduling method based on integrated multi-level water conservancy projects

Abstract

The present invention discloses an integrated flood control optimization scheduling method based on multi-level water conservancy projects, and relates to the technical field of water conservancy projects. Compared with the previous reservoir flood control scheduling method, the present invention solves the problem that the existing reservoir flood control scheduling method is single, cannot effectively deal with the complexity of the scheduling decision-making process of large-scale mixed reservoir groups, and cannot scientifically and uniformly allocate the control proportion of each flood control point, so that the flood control effect of the joint scheduling of the reservoir group cannot be fully exerted; by constructing a multi-level water conservancy project flood control optimization scheduling model, and decomposing the integrated system applied by it into multiple unrelated subsystems, repeatedly solving the optimal decision of each subsystem, and realizing the update and iteration of the overall goal, until the overall goal value converges to the optimal overall goal; thus, a variety of flood control conservancy node projects can be reasonably planned to solve the problem of optimized peak-shifting scheduling between reservoir groups, gate pump groups and main flood processes, so as to reduce the risk of urban flooding.

Description

Technical Field

[0001] The invention relates to the technical field of water conservancy projects, and in particular to an integrated flood control optimization scheduling method based on multi-level water conservancy projects. Background Art

[0002] Under the current complex and changeable climate environment, major river basins continue to build and improve reservoir flood control systems to reduce flood disasters; reservoir flood control scheduling is one of the important non-engineering flood control measures, and reservoir operation scheduling is a process of using system engineering theory and optimization technology to seek optimal operation strategies and corresponding decisions. At present, reservoir scheduling methods include linear programming, integer programming, optimal control theory, particle swarm algorithm, artificial neural network algorithm, genetic algorithm, etc. Among them, dynamic programming (DP) is the most widely used optimization method in reservoir group optimization scheduling. This method does not have strict requirements on objective functions and constraints. The mathematical model and solution method are relatively flexible. Whether the system is continuous or discrete, linear or nonlinear, deterministic or random, as long as it can constitute a multi-stage decision-making process, this method can be used to solve it. The key to using DP to solve is to correctly write the basic recursive relationship and appropriate boundary conditions; the process of the problem must be divided into several interrelated stages, which can be referred to Figure 1 , appropriately select state variables and decision variables and define the optimal value function, so as to transform a large problem into a group of sub-problems of the same type, and then solve them one by one, that is, starting from the boundary conditions, recursively searching for the best solution step by step. In the solution of each sub-problem, the optimization results of the previous sub-problem are used, and the optimal solution obtained from the last sub-problem is the optimal solution to the entire problem.

[0003] However, most of the existing conventional flood control scheduling rules for reservoirs are generally based on a single discriminant and command scheduling. There is a lack of scientific and unified coordinated scheduling between reservoir groups and sluice pump groups, which makes it difficult to meet the needs of joint flood control in the basin.

[0004] The reservoir-sluice pump group joint flood control optimization dispatch involves dozens of reservoirs and hundreds of sluice pump stations. The dispatch scale is huge and there are tens of thousands of constraints. It is a typical large-scale optimization dispatch problem. The typical characteristics of large-scale optimization problems are that the number of decision variables increases, the optimization search space expands or even becomes infinite, which makes it difficult to solve the problem quickly. At the same time, due to the expansion of the search space, there must be many local optimal solutions. These solutions easily cause the algorithm to fall into the local optimum, thus losing the opportunity to search for the global optimum.

[0005] A large number of studies have been conducted at home and abroad on the joint flood control dispatching problem of reservoir groups and sluice station groups, and dispatching models with different objectives and solution algorithms have been proposed. Relevant technicians have established a dynamic programming model for the joint dispatching of reservoir groups based on the situation of tributaries affecting floods, with the goal of maximizing the downstream peak reduction. They also proposed an aggregation-decomposition method suitable for the joint flood control dispatching of parallel reservoir groups, which decomposes the original n-order random variable optimization problem into n second-order variable optimization sub-problems and obtains a global suboptimal solution. Other scholars have established a standard system for identifying effective reservoirs in the joint flood control dispatching of reservoir groups, and constructed a real-time flood control joint dispatching adaptive model of reservoir groups with dynamic topological structure, which has been applied to the reservoir group system containing 14 reservoirs in the Huaihe River Basin; however, the existing reservoir flood control dispatching methods usually only consider the flood control dispatching decision of a single flood control control point. In the process of large-scale mixed reservoir group dispatching decision-making, it is impossible to effectively deal with the complexity of the hydraulic connection between upstream and downstream sections, and it is impossible to scientifically and uniformly allocate the control proportion of each flood control point, so that the flood control function of the joint dispatching of reservoir groups cannot be fully exerted.

[0006] In order to solve the above problems, the present invention proposes an integrated flood control optimization scheduling method based on multi-level water conservancy projects. Summary of the invention

[0007] The purpose of the present invention is to propose a flood control optimization scheduling method based on multi-level water conservancy project integration to solve the problems raised in the background technology:

[0008] Existing reservoir flood control scheduling methods usually only consider the flood control scheduling decisions of a single flood control point. In the scheduling decision-making process of a large-scale hybrid reservoir group, they are unable to effectively deal with the complexity of the hydraulic connection between upstream and downstream sections, and are unable to scientifically and uniformly allocate the control proportion of each flood control point, making it impossible to fully play the flood control role of the joint scheduling of the reservoir group.

[0009] In order to achieve the above object, the present invention adopts the following technical solutions:

[0010] A flood control optimization scheduling method based on multi-level water conservancy project integration includes the following steps:

[0011] S1: Collect the water conservancy engineering measurement data in the area to be optimized and perform preprocessing;

[0012] S2: Setting the objective function of the multi-level integrated flood control optimization scheduling model for water conservancy projects;

[0013] S3: Set constraints according to the set objective function;

[0014] S4: Solve the integrated flood control optimization scheduling model of multi-level water conservancy projects.

[0015] Preferably, in S2, the integrated flood control system applied by the multi-level water conservancy project integrated flood control optimization scheduling model is decomposed into a number of unrelated subsystems to achieve the update iteration of the overall goal until the overall goal value converges to the optimal overall goal; specifically, the multi-level water conservancy project integrated flood control optimization scheduling problem is decomposed into a reservoir group optimization scheduling sub-problem and a gate pump station group drainage optimization scheduling sub-problem, an overall control target is set, and optimization scheduling is performed in two stages;

[0016] Phase I: With the optimization goal of lowering the water level in the station, the current dispatching status of the pump station group is maintained, and only the dispatching decision of the reservoir group is optimized to obtain the maximum value of the water level process in the station under the premise of satisfying the safety of the reservoir group dam. ;

[0017] The second stage: with the optimization goal of lowering the water level in the station, maintaining the optimal scheduling decision of the above reservoir, only optimizing the scheduling of the pump group for drainage, and obtaining the maximum value of the water level process in the station under the full cooperation of the pump station group .

[0018] Preferably, the reservoir group optimization scheduling subproblem in S2 establishes a reservoir group flood control optimization scheduling model based on a basin general flood control system composed of a mixed reservoir group and an interval river channel. In the basin general flood control system, the basin includes several flood control control points and a mixed reservoir group; the interval inflow of the flood control point is not ignored; and the evolution of the reservoir discharge in the river channel is considered at the same time;

[0019] The constraints of the optimal scheduling model of the reservoir group flood control optimal scheduling model include: flood control storage capacity constraints, water balance constraints, water level restrictions at the end of the control period, discharge flow constraints, reservoir storage capacity constraints, reservoir discharge capacity constraints, reservoir maximum discharge constraints and reservoir discharge stability constraints.

[0020] Preferably, the steps for solving the reservoir group flood control optimization scheduling model are as follows:

[0021] Starting from an initial solution, a feasible solution generator is used to continuously search for a feasible solution that is better than the current solution in the neighborhood of the current solution, and replace the current solution with it until no better solution is found in the neighborhood of the current solution;

[0022] It also searches by constructing an associated search pattern of feasible solutions in the neighborhood of the current solution, and modifies the local outbound flow process associated with the search initiation point;

[0023] The associated search in the associated search mode includes modifying the outflow of the search initiating reservoir and adjusting the outflow of each downstream reservoir to meet the control constraints of the water conservancy project;

[0024] When performing a single-step association search, the search is performed through four operations: initial search, influence range expansion, influence range edge correction, and in-and-out water volume difference correction; the details are as follows:

[0025] The reservoir is The process of single-step association search initiated by a time period is as follows:

[0026] Operation A: Initial Search:

[0027] Use the initial search mode: the reservoir is and The outbound traffic of the two time periods initiates the initial search and keeps the time period The final water level remains unchanged; When the outbound flow increases or decreases during the period, The outflow flow in the period is adjusted in the opposite direction by the same amount; if the reservoir discharge stability constraint is not considered, this mode is used for solution; otherwise, the other three operations are used to adjust the associated variables;

[0028] Operation B: Expansion of influence: Time period / The time period is adjusted towards the beginning of the control period: If the flood discharge near the starting point is greater than the preset fluctuation threshold due to operation 1, the The time period expands the flood discharge change range towards the beginning of the control period to meet the minimum number of time periods for the flood discharge to rise and fall;

[0029] Operation C: Correction of the edge of the affected area:

[0030] Step C.1: Correction of flow rate change speed at the edge of the impact range: Correct the flow rate change speed at the edge of the flood discharge flow change area after operation B is completed;

[0031] Step C.2: Adjustment of flow fluctuation frequency at the edge of the impact range: Determine whether there is unreasonable fluctuation at the edge of the flood discharge flow change area in step C.1, and make corresponding adjustments;

[0032] Operation D: Correction operation of water volume difference between inflow and outflow:

[0033] Step D.1: Keep the total inflow and outflow water difference of the search initiating reservoir unchanged;

[0034] Step D.2: Adjust the difference in water volume between the inlet and outlet of each downstream reservoir.

[0035] Preferably, the flood drainage optimization scheduling subproblem of the sluice pump station group in S2 simulates the water flow in the river network by a one-dimensional unsteady flow method represented by Saint-Venant partial differential equations, and builds a river network confluence model based on sluice simulation, pump simulation and lake simulation, and then builds a flood drainage optimization scheduling model of the sluice pump station group;

[0036] The constraints of the flood drainage optimization scheduling model of the sluice pump station group include: water level constraints within the station, pumping flow constraints of the sluice pump station, and flood risk constraints in the built-up area.

[0037] Preferably, the flood drainage optimization scheduling model of the sluice pump station group converts multiple objectives into a single-objective optimization problem according to several optimization objectives, and the solution method is as follows:

[0038] S2.1: List the sequence of single-objective optimization problems according to their importance, and find the optimal solution of the next objective in the optimal solution set of the previous objective each time until a common optimal solution is found;

[0039] S2.2: transform multiple objectives into single or dual objectives for solution;

[0040] S2.3: Based on the penalty function method, the constrained optimization problem is transformed into an unconstrained optimization problem for optimization.

[0041] Preferably, the solution method of the multi-level water conservancy project integrated flood control optimization scheduling model is as follows:

[0042] S3.1: Calculate the optimal dispatching sub-problem of the reservoir group: take the set objective function as the standard, use the optimal dispatching method of the reservoir group, maintain the current pump station dispatching status unchanged, and obtain the optimal dispatching decision of the reservoir group and on-site optimization goals ;

[0043] S3.2: Then calculate the optimal scheduling sub-problem of flood discharge of the sluice pump station group: take the set objective function as the standard to maintain the current scheduling status of the reservoir group The optimal drainage scheduling decision of the pump station group is obtained , and update the on-site optimization goals ;

[0044] S3.3: Repeat S3.1 and S3.2 to perform Iterations to obtain the optimal dispatching decision of the reservoir group Optimal drainage scheduling decision for pumping station groups , so that the site optimization goals further reduce;

[0045] S3.4: When completed After iterations, the optimization goal of the site If the change is controlled within a given accuracy range, the overall goal is determined to have converged, and the optimal scheduling decision for the reservoir group is output. Optimal drainage scheduling decision for pumping station groups .

[0046] Compared with the prior art, the present invention provides a flood control optimization scheduling method based on multi-level water conservancy project integration, which has the following beneficial effects:

[0047] The present invention constructs a multi-level water conservancy project flood control optimization scheduling model, decomposes the integrated system it applies into multiple unrelated subsystems, repeatedly solves the optimization decision of each subsystem, and realizes the update and iteration of the overall goal until the overall goal value converges to the optimal overall goal; thereby, various flood control water conservancy node projects such as reservoir groups, pump station groups, and sluice station groups can be reasonably planned to solve the problem of optimized peak-shifting scheduling between reservoir groups, sluice pump groups and main flood processes, so as to achieve the purpose of reducing the risk of urban flooding. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 It is a flow chart of the DP solution method mentioned in the background technology of the present invention;

[0049] Figure 2 A technical roadmap for solving the multi-level water conservancy project integrated flood control optimization scheduling model mentioned in Example 1 of the present invention;

[0050] Figure 3 This is a generalized diagram of the universal flood control system mentioned in Example 1 of the present invention;

[0051] Figure 4 This is a schematic diagram of the expansion operation mentioned in Example 1 of the present invention;

[0052] Figure 5 This is a schematic diagram of the edge flow speed change correction operation mentioned in Example 1 of the present invention;

[0053] Figure 6 This is a schematic diagram of the edge flow fluctuation correction operation mentioned in Example 1 of the present invention;

[0054] Figure 7 This is a schematic diagram of the water difference balance operation mentioned in Example 1 of the present invention;

[0055] Figure 8 This is a schematic diagram of the single-step association search mentioned in Example 1 of the present invention;

[0056] Fig. 9 This is a schematic diagram of setting the water level control point of the downstream reservoir mentioned in Example 1 of the present invention;

[0057] Fig.10 Schematic diagram of the Preissmann four-point implicit difference format mentioned in Example 1 of the present invention;

[0058] Fig.11 It is a diagram of the measured water level process of a certain station mentioned in Example 2 of the present invention;

[0059] Fig.12 The measured water level process subsection of a station mentioned in Example 2 of the present invention is Figure 1 ;

[0060] Fig.13 The measured water level process subsection of a station mentioned in Example 2 of the present invention is Figure 2 ;

[0061] Fig.14 It is a structural diagram of the integrated flood control optimization scheduling method for a multi-level water conservancy project at a certain station mentioned in Example 2 of the present invention;

[0062] Fig.15 This is a schematic diagram of the combined algorithm mentioned in Example 2 of the present invention;

[0063] Fig.16 This is a schematic diagram of the water level control process during a typhoon transit decision mentioned in Example 2 of the present invention. DETAILED DESCRIPTION

[0064] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0065] The present invention constructs a multi-level water conservancy project flood control optimization scheduling model, decomposes the integrated system it applies into multiple unrelated subsystems, repeatedly solves the optimization decision of each subsystem, and realizes the update and iteration of the overall goal until the overall goal value converges to the optimal overall goal; thus, various flood control water conservancy node projects such as reservoir groups, pump station groups, and gate station groups can be reasonably planned to solve the problem of optimized peak-shifting scheduling between reservoir groups, gate pump groups and main flood processes, so as to achieve the purpose of reducing the risk of urban flooding. Specifically, it includes the following contents.

[0066] Embodiment 1:

[0067] See also Figure 1-Figure 10 The present invention provides a method for optimizing flood control and scheduling based on multi-level integrated water conservancy projects, comprising the following steps:

[0068] S1: Collect the water conservancy project measurement data in the area to be optimized and pre-process it; the details are as follows:

[0069] Taking a certain city’s water conservancy project as an example, we collected water conditions, water conservancy infrastructure, and basic hydrological data in the basin. The water conservancy project in a certain city mainly included hydrological data of all 55 large, medium and small reservoirs in the city, and measurement data of precipitation stations in a river basin. We also preprocessed the collected data to ensure its accuracy.

[0070] S2: Set the objective function of the multi-level water conservancy project integrated flood control optimization scheduling model; the details are as follows:

[0071] Based on the existing hydrological station layout, according to the forecast needs of small watersheds and reservoirs, the hydrological calculation divisions are further refined to establish a basin flood forecasting model system integrating precipitation, runoff generation and confluence. To support the smooth implementation of the functions involved in the design of a flood forecasting application scenario for a city's flood control and dispatching decision support system.

[0072] Corresponding to the flood control dispatch of a certain city, the multi-level water conservancy project integrated flood control optimization dispatch problem is decomposed into the reservoir group optimization dispatch sub-problem and the gate pump station group drainage optimization dispatch sub-problem. Figure 2 , taking the water level of a certain station as the overall control target, the optimization dispatch is carried out in two stages:

[0073] Phase 1: With the optimization goal of lowering the water level of a station, the current dispatching status of the pump station group is maintained, and only the dispatching decision of the reservoir group is optimized. The maximum value of the water level process of a station can be obtained under the premise of satisfying the safety of the reservoir group dam. ;

[0074] The second stage: with the optimization goal of lowering the water level of a certain station, maintaining the best scheduling decision of the above reservoir, only optimizing the scheduling of the pump group for drainage, the maximum value of the water level process of a certain station under the full cooperation of the pump station group can be obtained. .

[0075] In the first stage, a reservoir group flood control optimization dispatching model is first established. This stage more comprehensively considers the general flood control system of the basin composed of the hybrid reservoir group and the interval river channel, and considers the evolution process of the water flow and river channel. Its basic structural characteristics include: ① The basin has several flood control points; ② The basin includes a hybrid reservoir group; ③ The interval flow of the flood control point is not ignored; ④ Considering the evolution of the reservoir discharge in the river channel, a general flood control system structure diagram that meets the above characteristics is constructed. For details, please refer to Figure 3 , where A, B, C, D, and E are mixed reservoir groups, and a, b, and c are flood control points. , , is the water flow rate of the interval; It is the water flow rate between the hybrid reservoirs A and B.

[0076] For the flood control optimization scheduling problem of a reservoir group, starting from an initial solution, a feasible solution generator is used to continuously search for a feasible solution that is better than the current solution in the neighborhood of the current solution, and replace the current solution with it until a better solution is found in the neighborhood of the current solution. Due to the constraints such as the stability of water discharge and the number of duration periods of gate opening and closing, when a variable is assigned a value, the feasible value range of other variables will also change. It is necessary to design an association search mode that can construct a feasible solution in the neighborhood of the current solution, that is, to ensure the feasibility of the search mode by correcting the local outflow flow process associated with the search initiation point. The association search includes correcting the outflow of the search initiating reservoir to meet constraints such as time-related constraints and final water level control, and also includes adjusting the outflow of each downstream reservoir to meet its final water level control. The single-step association search process consists of four basic operations: initial search, expansion of the influence range, correction of the edge of the influence range, and correction of the inflow and outflow water volume difference. A reservoir consists of The search process initiated by the time period is as follows:

[0077] Operation A: Initial Search:

[0078] Adopt the initial search mode: take a reservoir and The outbound traffic of the two time periods initiates the initial search and keeps the time period The final water level remains unchanged; When the outbound flow increases or decreases during the period, The outflow flow in the period is adjusted in the opposite direction by the same amount; if the water discharge stability constraint is not considered, this mode is used for solution; otherwise, the other three operations are used to adjust the associated variables;

[0079] Operation B: Expansion of influence: Time period / The time period is adjusted towards the beginning of the control period. Take the adjustment of time period as an example. Time period and The adjustment of the time period is similar: if operation 1 causes the flood discharge near the starting point to fluctuate too quickly, The time period expands the flood discharge change range towards the beginning of the control period to meet the minimum number of time periods for the flood discharge to rise and fall; Figure 4 for There are several basic modes of expansion operation when the period flood discharge increases, the expansion mode when the flow decreases, and the expansion mode towards the end of the control period.

[0080] Operation C: Correction of the edge of the affected area:

[0081] Step C.1: Correction of flow change speed at the edge of the affected area: After operation B is completed, the edge of the flood discharge change area is adjusted according to Figure 5 The flow rate change rate can be corrected in this way.

[0082] Step C.2: Adjust the frequency of flow fluctuation at the edge of the impact range: Determine whether there is unreasonable fluctuation at the edge of the flood discharge flow change area in step C.1, and make corresponding adjustments; the specific adjustment mode can be referred to Figure 6 .

[0083] Operation D: Correction operation of water volume difference between inflow and outflow:

[0084] Step D.1: Keep the total inflow and outflow water difference of the search initiating reservoir unchanged;

[0085] In the process of operating AD, The total inflow and outflow difference of the No. 1 reservoir may change. In order to keep the final water level unchanged during the outflow change period, the flood discharge process needs to be further adjusted. The period is from the beginning of the control period to is the number of allocation periods, and the water volume difference change is evenly distributed in each period. If the flow change speed and fluctuation constraints are violated, the flow correction end period will be advanced or the number of allocation periods will be increased. , until the change in the difference between the outflow and inflow is fully allocated or the start time of the control period is reached. Figure 7 As shown, ,exist The correction results at this time all violate the constraint. If it cannot be completely allocated, then in a similar way The change in water volume difference is distributed in the direction of the end of the control period. If it still cannot be fully distributed, it is considered that the construction of the feasible solution has failed.

[0086] Reference Figure 8 , the figure shows , , An example of a complete association search process for a single station at . Step 1 is the original feasible solution; Step 2 is the initial search mode; Steps 3-4 are The flow correction process from the time period to the start of the control period; step 5 is The flow correction result of the time period towards the end of the control period; step 6 is the correction result of the water volume difference change.

[0087] Step D.2: Adjust the difference in water volume between the inlet and outlet of each downstream reservoir.

[0088] The end of the last period of the possible change range of the downstream reservoir inflow flow is used as the water level control point. Fig. 9 If the search is initiated Reservoir No. exists in the downstream reservoir, and its direct downstream reservoir is labeled By the reservoir Starting from the first, the final water level control within the possible range of inflow changes of each downstream reservoir is carried out in turn. No. reservoir, set its direct upstream The discharge range of reservoir No. ,but The possible range of inflow to reservoir No. , , Since the number of lag periods for outbound flows of different levels is different, and Not necessarily equal. Internal outbound flow remains unchanged and The water level at the end of the period remains unchanged. Due to changes in the storage within the range, two situations will lead to the first The No. 1 reservoir cannot maintain the original outflow process: First, The water level in some time periods reaches the upper and lower limits and cannot be adjusted according to the original outflow flow. This may cause unreasonable fluctuations in flood discharge. If this still exists after the optimization calculation is completed, feasibility corrections need to be made; second, Part of the control period is exceeded. Due to the influence of water flow lag, the total amount of water entering the reservoir during the control period changes (such as Fig. 9 In and Reservoir No.), the same method as in step D.1 is used to At the end of the time period, excess or insufficient outflow water is allocated to the water level control point. If it cannot be fully allocated, it is also considered that the construction of the feasible solution has failed. The above two situations can lead to The discharge process of reservoir No.1 changes. The discharge flow of reservoir No.1 after adjustment calculation or water volume difference correction varies in the range of , and the inflow flow of reservoir No.1 directly downstream may vary in the range of , , , proceed in the same way The control of the final water level of the reservoir inflow change time range is stopped if one of the following three situations occurs: a downstream reservoir can be completely regulated according to the original outflow flow, the affected range is completely beyond the control period (such as Fig. 9 In Reservoir No.), reaching the last level of the cascade.

[0089] The main advantages of this associative search mode are that on the one hand, it overcomes the limitations of the correlation constraints between time periods on the optimization process, maximizes the feasibility of the intermediate results, and reduces the damage of feasibility corrections to the optimization results; on the other hand, it expands the scope of influence of local search and improves the "transitivity" when variables change.

[0090] In the second stage, a drainage optimization dispatching model for a river sluice pump station group was built:

[0091] Water flows in the river network, and its flow, water level, flow velocity, and water-passing cross-section vary with time and location. Its movement elements can be expressed as: , where S represents position and t represents time. It means that the physical quantity related to water flow (such as flow rate, water level, flow velocity and water flow section) is a function of position and time. The water flow represented belongs to the open channel non-steady flow, which is often simplified to one-dimensional non-steady flow in actual calculation. It is considered that the water level, flow rate, flow velocity and other elements are the same in the section, and the lateral flow of the water flow is not considered. Only the change process of the water level and flow rate when the water flows in different sections is concerned. The calculation adopts the one-dimensional non-steady flow method, and its basic equation Saint-Venant partial differential equation group is:

[0092]

[0093]

[0094]

[0095]

[0096] in, , are water level and flow rate respectively; , , , They are river width, water-flowing area, hydraulic radius and flow velocity; is the flow modulus, is Xie Cai coefficient, is the roughness; g is the gravitational constant;

[0097] For the water balance equation, the change of the tank storage volume of a certain micro-segment is reflected by the difference between the water volume of the upstream and downstream sections and the difference between the side inflow. For the momentum balance equation, it reflects the external force acting on the movement of a certain micro-segment, where: is the friction term, reflecting the influence of riverbed resistance; The water pressure term reflects the pressure caused by the different water depths in the upstream and downstream of the river section; is the inertia term, which is an imaginary force.

[0098] Discrete solution of Saint-Venant partial differential equation based on finite difference method:

[0099] The one-dimensional unsteady flow calculation model uses the Preissmann four-point eccentric implicit format; under certain conditions, this difference format has quite good compatibility, stability and convergence. Fig.10The Preissmann four-point space-time eccentric implicit format performs Tailor first-order polynomial expansion of the function at a certain point in the two-dimensional coordinate system composed of space and time, and obtains the relationship between the function value and partial derivative value at this point and the function value at the surrounding four points, and discretizes it according to the following formula:

[0100]

[0101] in, is the weight parameter; Represents the distance between two adjacent calculation points in space; Represents the interval between two adjacent calculation moments in time; Indicates the calculation point number in space; Indicates the calculation moment number in time; Indicated in Spatial position The function value at a certain moment;

[0102] Then, a river network confluence model is built based on the gate simulation, pump simulation and lake simulation. By simulating the gates, pumps and lakes, the water flow changes of a river gate pump station group can be more accurately grasped, as follows:

[0103] Gate simulation:

[0104] The water flow through the sluice on the river is weir flow (broad-top weir or practical weir).

[0105] Flood discharge from weirs The calculation is as follows:

[0106]

[0107] in, is the overflow weir discharge coefficient; is the water head above the weir; is the width of the weir crest;

[0108] The flood discharge form of the sluice gate is the flow rate during free outflow. The calculation is as follows:

[0109]

[0110] in, is the free outflow coefficient of the gate hole; is the water head before the gate;

[0111] Flow rate when flooding outflow The calculation is as follows:

[0112]

[0113] in, is the flooded outflow coefficient of the gate hole; is the water level before the gate; is the water level behind the gate; is the water head behind the gate; is the width of the gate hole;

[0114] The water level of the water flowing through the weir may not be continuous, but the flow through the sluice gate is continuous, and the water volume regulation near the sluice gate can be ignored. The continuity equation is:

[0115]

[0116] in, , is the flow rate of the sections above and below the gate;

[0117] The calculation formula (dynamic equation) for non-steady flow through the gate is:

[0118]

[0119] in, , The water level above and below the gate; is the flooding coefficient, is the lateral contraction coefficient, is the flow coefficient, is the overflow width, is the effective water head, It is a sign function; when the number in the brackets is less than 0, equal to 0, and greater than 0, the function values ​​are -1, 0, and +1 respectively.

[0120] Pump simulation:

[0121] In When there are 1 pump station, the design flow rate of the pump station , then the flow rate of the pump exchange is :

[0122]

[0123] in, , Respectively The pumping capacity and water supply of each pumping station.

[0124] Lake simulation:

[0125] A generalized lake is a river network unit in the field divided according to natural boundaries (such as roads, polder areas, dams, etc.). It has a certain volume and can drain water out of the lake or draw water into the lake. In other words, a generalized lake is a river network unit with the ability to exchange water with the outside world and regulate water.

[0126] According to the "Improved Pinghu Method" which takes into account the slope of the lake surface, and The change of form is related to this. That is:

[0127]

[0128] in, It is the intersection of the average water level of the static volume equal to the volume of the eastern reservoir of the lake and the dynamic water surface line Distance to the exit; is the longitudinal length of the lake; is a function of change; is the water delivery width of the lake, which varies along the longitudinal length of the lake.

[0129] The expression of the regulation and storage capacity of generalized lakes is the water level-volume relationship. The area of ​​generalized lakes at different elevations is calculated based on topographic survey data and converted into a water level-volume relationship curve or relationship.

[0130] Design a generalized lake exist The effective precipitation during this period is , the volume change is The traffic exchanged with the outside world is (Entering the lake is positive, leaving the lake is negative), according to the law of conservation of mass:

[0131] ,

[0132] Assume that during the ∆t period, the generalized lake The volume from becomes , then according to the integral mean value theorem, we can integrate the above formula to get:

[0133]

[0134] The amount of water exchanged between a general lake and the outside world depends on the water level difference between them and the water conservancy facilities at the junction. The exchange methods can be summarized into three types: river type, pump type, and weir type.

[0135] Boundary conditions:

[0136] Then set the river network boundary conditions:

[0137] The boundaries of a river network generally include rivers and river boundary sections where floods converge, as well as rivers, lakes, and seas connected by river inlet (outlet) sluices, pumping stations, etc. Combining the characteristics of these boundaries, they can be summarized into three boundary conditions:

[0138] Water level process line of boundary section: Z=Z(t);

[0139] Flow process line of boundary section: Q=Q(t);

[0140] Water level-discharge relationship curve of the boundary section: Z=Z[Q(t)];

[0141] For the river boundary where the flood flows into, since the flood process line is the basic data necessary for calculation, the flow process line Q=Q(t) is mostly used as the boundary condition.

[0142] For the boundaries of rivers, lakes and seas at the drainage outlets of river networks, it is generally advisable to obtain their water level data, so the water level process line Z=Z(t) is often used as the boundary condition.

[0143] The method for determining the initial conditions of the model includes linear interpolation along the process flow according to the measured data of the measuring station, selecting the time of water level and flow rate, and the moment with small rate of change as the initial moment, and using the river network constant flow method to calculate the initial value. For the tree-like river network with more measuring stations, the first method is often used, and for the complex plain river network, the second method is often used. This embodiment can use the second method.

[0144] Then analyze and calibrate other parameters of the river network confluence model:

[0145] The parameters of the river network confluence model mainly include river length, cross-sectional dimensions, cross-sectional shape, land width, river roughness, water surface ratio, sluice characteristics and outflow coefficient, storage capacity, etc. The analysis and calibration of the parameters are as follows:

[0146] 1. River length: Calculated based on the electronic map of the river basin and using the functions of the geographic information system.

[0147] 2. Cross-sectional dimensions: Calculated based on the measured or planned river cross-sectional diagram, generally generalized into a trapezoidal cross-sectional shape.

[0148] 3. Land width: Based on the electronic map of the river basin, the average drainage area per unit length of each river channel is calculated using the functional quantity of the geographic information system as the land width.

[0149] 4. River roughness: The roughness of the main river is determined based on the measured river water level and flow data. The roughness of other rivers is determined by analogy analysis.

[0150] 5. Sluice gate outflow coefficient: The value is determined based on the measured water levels before and after the sluice gate and the diversion and drainage flow data at the same time.

[0151] 6. Storage capacity: Derived from the survey statistics and analysis of the planned river characteristics and water surface area data.

[0152] 7. Characteristics of sluice gates: For the more important sluice gates along rivers and coasts, they are mainly analyzed based on engineering design data; for other sluice gates, they are analyzed based on water conservancy design data or through analogy analysis.

[0153] The flood drainage optimization dispatching model of the sluice pump station group combines the actual needs of a city, observes the water level changes of a river station, establishes a prediction model based on the water balance principle in the sluice pump group optimization dispatching model, and adopts different dispatching strategies. The prediction model can deduce the amount of water that can be discharged into a river through the pump station in the future according to the current water level of a river station, reservoir discharge, interval inflow, interval outflow, and storage capacity of the main stream of a river. The sampling interval of the prediction model is 1h, and the water level forecast can be 1h, 3h, 6h, and 12h. The prediction calculation formula is as follows:

[0154]

[0155]

[0156] in, For the pump station group; For a certain station; For reservoirs; is the storage capacity of a river main stream at a certain moment; is the storage capacity of a river main stream at a certain time; is the interval inflow; is the interval outflow; is the unit time;

[0157] The optimization model for flood drainage optimization of the pump station group along the main stream of a river constructed this time has four optimization objectives, including minimizing the flooding risk of built-up areas in towns, minimizing the impact of water discharged from the pump station into the main stream of a river on the embankment, minimizing the flooding risk of farmland, and minimizing the difference in flooding duration of each town (street). It is a multi-objective optimization problem.

[0158] The method for solving multi-objective problems is generally to transform them into single-objective optimization problems. Methods for solving multi-objective programming: (1) List the objects in sequence according to their importance, and find the optimal solution for the next object in the optimal solution set of the previous object each time until a common optimal solution is found. The typical method is the hierarchical sequence method; (2) Transform multiple objectives into single objectives or dual objectives for solution. Common methods include linear weighted method, main objective method, etc.; (3) Penalty function method: transform the constrained optimization problem into an unconstrained optimization problem. Please refer to the following formula:

[0159]

[0160] in, Represents the sequence corresponding to the multi-objective problem; is the objective function; For constrained optimization; Represents the transpose of a vector.

[0161] S3: Set constraints according to the set objective function; the details are as follows:

[0162] The constraints of the optimal dispatching model of the reservoir group flood control optimal dispatching model include:

[0163] Flood control storage capacity constraints:

[0164] Water balance constraints:

[0165] Water level limit at the end of control period:

[0166] Downflow flow restriction:

[0167] Reservoir capacity constraints:

[0168] Reservoir discharge capacity constraints:

[0169] Reservoir maximum discharge constraints:

[0170] Reservoir discharge stability constraints:

[0171] The constraints of the optimal scheduling model for flood discharge of sluice pump stations include:

[0172] Water level constraint of a station:

[0173] Pumping flow constraints of the sluice pump station:

[0174] Flood risk constraints in built-up areas:

[0175]

[0176] S4: Solve the multi-level integrated flood control optimization scheduling model of water conservancy projects. The details are as follows:

[0177] Finally, the multi-level integrated flood control optimization dispatching model of water conservancy projects is summarized and solved, and the coordination process of the first and second stages is repeated continuously, and the objective function value is continuously iterated until the target change is less than a certain threshold, and the dispatching decision of reservoirs and pumping stations is output. Figure 2 , the specific steps are as follows:

[0178] S3.1. Calculate the optimal dispatching sub-problem of the reservoir group: take the set objective function as the standard, use the optimal dispatching method of the reservoir group, maintain the current pump station dispatching status unchanged, and obtain the optimal dispatching decision of the reservoir group And a site optimization goal ;

[0179] S3.2, then calculate the sub-problem of optimal scheduling of flood discharge of the sluice pump station group: take the set objective function as the standard, maintain the current scheduling state of the reservoir group The optimal drainage scheduling decision of the pump station group is obtained , and update the optimization goal of a certain site ;

[0180] S3.3, repeat S3.1 and S3.2, and proceed to The optimal dispatching decision of the reservoir group can be obtained by iteration. Optimal drainage scheduling decision for pumping station groups , so that a station optimizes the goal further reduce;

[0181] S3.4. When completed After iterations, the optimization goal of a station If the change is controlled within a given accuracy range, the overall goal is determined to have converged, and the optimal scheduling decision for the reservoir group is output. Optimal drainage scheduling decision for pumping station groups .

[0182] Embodiment 2:

[0183] First, collect relevant data on a certain site, you can refer to Figure 11-13 and Table 1, which is as follows:

[0184] Table 1 Statistics of characteristic water levels at major representative stations

[0185]

[0186] Please refer to Fig.14 The multi-level integrated flood control optimization dispatching method of water conservancy projects uses POA to transform the multi-stage, multi-decision reservoir group optimization problem into multiple two-stage sub-problems. DPSA controls the number of reservoir optimizations, and DDDP is responsible for rapid optimization in a small range. Fig.15 ,

[0187] Based on a typhoon and rainstorm passing through a city, the changes in the water level process were simulated using the combinatorial optimization algorithm and without using the combinatorial optimization algorithm.

[0188] Take the first and third reservoirs in a certain city as an example, and refer to a typhoon as a calculation scenario. Fig.16 The results show that compared with the highest water level of 3.42m at a certain station in actual scheduling, the highest water level at a certain station was 3.35m when the reservoir group optimization scheduling model was used for decision-making control, which was reduced by 0.07m, a decrease of 2%, effectively alleviating the flood control pressure in the urban area of ​​a certain city.

[0189] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A flood control optimization scheduling method based on multi-level integrated water conservancy project, characterized in that: The steps include: S1: Collect the water conservancy engineering measurement data in the area to be optimized and perform preprocessing; S2: Setting the objective function of the multi-level integrated flood control optimization scheduling model for water conservancy projects; S3: Set constraints according to the set objective function; S4: Solve the integrated flood control optimization scheduling model of multi-level water conservancy projects; In S2, the integrated flood control system applied by the multi-level water conservancy project integrated flood control optimization scheduling model is decomposed into several unrelated subsystems to achieve the update iteration of the overall goal until the overall goal value converges to the optimal overall goal; specifically, the multi-level water conservancy project integrated flood control optimization scheduling problem is decomposed into reservoir group optimization scheduling subproblems and gate pump station group drainage optimization scheduling subproblems, the overall control target is set, and the optimization scheduling is carried out in two stages; The reservoir group optimization scheduling sub-problem in S2 establishes a reservoir group flood control optimization scheduling model based on a basin general flood control system consisting of a hybrid reservoir group and an interval river channel. In the basin general flood control system, the basin includes several flood control control points and a hybrid reservoir group; the interval inflow of the flood control point is not ignored; and the evolution of the reservoir discharge in the river channel is considered at the same time; The sub-problem of optimal scheduling of flood discharge of the sluice pump station group is to simulate the water flow in the river network by the one-dimensional unsteady flow method represented by the Saint-Venant partial differential equations, and to build a river network confluence model based on sluice simulation, pump simulation and lake simulation, and then to build an optimal scheduling model of flood discharge of the sluice pump station group; The constraints of the flood drainage optimization scheduling model of the sluice pump station group include: water level constraints within the station, pumping flow constraints of the sluice pump station, and flood risk constraints in the built-up area; The flood drainage optimization scheduling model of the sluice pump station group transforms multiple objectives into a single-objective optimization problem according to several optimization objectives. The solution method is as follows: S2.1: List the sequence of single-objective optimization problems according to their importance, and find the optimal solution of the next objective in the optimal solution set of the previous objective each time until a common optimal solution is found; S2.2: transform multiple objectives into single or dual objectives for solution; S2.3: Based on the penalty function method, the constrained optimization problem is transformed into an unconstrained optimization problem for optimization; The constraints of the reservoir group flood control optimization scheduling model include: flood control storage capacity constraints, water balance constraints, water level restrictions at the end of the control period, discharge flow constraints, reservoir storage capacity constraints, reservoir discharge capacity constraints, reservoir maximum discharge constraints and reservoir discharge stability constraints; Phase I: With the optimization goal of lowering the water level in the station, the current dispatching status of the pump station group is maintained, and only the dispatching decision of the reservoir group is optimized to obtain the maximum value of the water level process in the station under the premise of satisfying the safety of the reservoir group dam. ; The second stage: With the optimization goal of lowering the water level in the station, the optimal scheduling decision of the reservoir group is maintained, and only the pump group drainage is optimized to obtain the maximum value of the water level process in the station under the full cooperation of the pump station group. ; The specific steps for solving the reservoir group flood control optimization scheduling model are as follows: Starting from an initial solution, a feasible solution generator is used to continuously search for a feasible solution that is better than the current solution in the neighborhood of the current solution, and replace the current solution with it until no better solution is found in the neighborhood of the current solution; It also searches by constructing an associated search pattern of feasible solutions in the neighborhood of the current solution, and modifies the local outbound flow process associated with the search initiation point; The associated search in the associated search mode includes modifying the outflow of the search initiating reservoir and adjusting the outflow of each downstream reservoir to meet the control constraints of the water conservancy project; When performing a single-step association search, the search is performed through four operations: initial search, influence range expansion, influence range edge correction, and in-and-out water volume difference correction; the details are as follows: The reservoir is The process of single-step association search initiated by a time period is as follows: Operation A: Initial Search: Use the initial search mode: the reservoir is and The outbound traffic of the two time periods initiates the initial search and keeps the time period The final water level remains unchanged; When the outbound flow increases or decreases during the period, The outflow flow in the period is adjusted in the opposite direction by the same amount; if the reservoir discharge stability constraint is not considered, this mode is used for solution; otherwise, the other three operations are used to adjust the associated variables; Operation B: Expansion of influence: Time period / The time period is adjusted towards the beginning of the control period: If the flood discharge near the starting point is greater than the preset fluctuation threshold due to operation 1, the The time period expands the flood discharge change range towards the beginning of the control period to meet the minimum number of time periods for the flood discharge to rise and fall; Operation C: Correction of the edge of the affected area: Step C.1: Correction of flow rate change speed at the edge of the impact range: Correct the flow rate change speed at the edge of the flood discharge flow change area after operation B is completed; Step C.2: Adjustment of flow fluctuation frequency at the edge of the impact range: Determine whether there is unreasonable fluctuation at the edge of the flood discharge flow change area in step C.1, and make corresponding adjustments; Operation D: Correction operation of water volume difference between inflow and outflow: Step D.1: Keep the total inflow and outflow water difference of the search initiating reservoir unchanged; Step D.2: Adjust the water volume difference between the inflow and outflow of each downstream reservoir; The solution method of the multi-level water conservancy project integrated flood control optimization scheduling model is as follows: S3.1: Calculate the optimal dispatching sub-problem of the reservoir group: take the set objective function as the standard, use the optimal dispatching method of the reservoir group, maintain the current pump station dispatching status unchanged, and obtain the optimal dispatching decision of the reservoir group and on-site optimization goals ; S3.2: Then calculate the optimal scheduling sub-problem of flood discharge of the sluice pump station group: take the set objective function as the standard to maintain the current scheduling status of the reservoir group The optimal drainage scheduling decision of the pump station group is obtained , and update the on-site optimization goals ; S3.3: Repeat S3.1 and S3.2 to perform Iterations to obtain the optimal dispatching decision of the reservoir group Optimal drainage scheduling decision for pumping station groups , so that the site optimization goals further reduce; S3.4: When completed After iterations, the optimization goal of the site If the change is controlled within a given accuracy range, the overall goal is determined to have converged, and the optimal scheduling decision for the reservoir group is output. Optimal drainage scheduling decision for pumping station groups .

Images