A plain river network large-scale drainage and waterlogging prevention engineering group optimal scheduling method and system

Abstract

The present application relates to the technical field of urban drainage and waterlogging prevention, and particularly relates to a plain river network large-scale drainage and waterlogging prevention engineering group optimization scheduling method and system. The method comprises determining a typical composite waterlogging event of the plain river network, establishing a hydrodynamic model to analyze the typical composite waterlogging event, identifying key nodes of waterlogging, determining a linkage relationship between a waterlogging facility scheduling behavior and a hydrological response of the key nodes, establishing a multi-objective optimization scheduling model with waterlogging risk minimization and operation cost optimization, gradually decomposing a global scheduling target into a local scheduling target, establishing a mapping relationship between the local scheduling target and a waterlogging facility scheduling decision, obtaining a real solution of the multi-objective optimization scheduling model for potential individuals predicted in different candidate scheduling scheme populations, constructing an expected improvement matrix to select an optimal solution to obtain an optimal scheduling scheme. The present application effectively meets the scheduling timeliness requirement by combining machine learning prediction capability and an optimization algorithm, and realizes the optimization scheduling of the plain river network large-scale drainage and waterlogging prevention engineering group.

Description

Technical Field

[0001] This invention relates to the field of urban drainage and flood control technology, and in particular to an optimized scheduling method and system for large-scale drainage and flood control engineering projects in plain river networks. Background Technology

[0002] Urban areas in plain river networks are characterized by intricate water systems and low-lying terrain, making drainage and flood control reliant on the coordinated operation of water storage systems, sluice gates, and pumping stations. The combined effects of heavy rainfall, backwater effects from external rivers, and rising water levels in inland rivers exacerbate the problem of complex urban flooding. The core of integrated drainage and flood control systems lies in constructing an optimized scheduling model and finding the optimal solution for scheduling strategies. However, the scale of drainage and flood control projects in plain river network urban areas is enormous, and under conditions of complex urban flooding, the scheduling relationships between various engineering facilities are complex and their mutual influences are significant. Furthermore, drainage and flood control systems must consider various constraints such as tiered management of facilities, engineering failures, and operating costs, further increasing the complexity of scheduling model construction and application difficulty. Therefore, there is an urgent need to adopt new technologies for precise and efficient systematic scheduling of large-scale drainage and flood control projects in plain river network urban areas. Summary of the Invention

[0003] The technical problem to be solved by the embodiments of the present invention is to provide an optimized scheduling method and system for large-scale drainage and flood control projects in plain river networks, so as to solve the problem of difficult scheduling of large-scale drainage and flood control projects in urban areas of plain river networks in the prior art.

[0004] This invention discloses an optimized scheduling method for large-scale drainage and flood control engineering projects in plain river networks, comprising:

[0005] Collect multi-dimensional flood-causing factor data of plain river network and construct a composite flood-causing probability model based on it. Use the composite flood-causing probability model to determine typical composite waterlogging events in plain river network.

[0006] A hydrodynamic model coupled with a large-scale drainage and flood control engineering group was established to simulate and analyze the process of the typical complex waterlogging event and identify the key nodes of waterlogging.

[0007] Establish the correlation matrix of the impact of each drainage facility on the key node in the hydrodynamic model, and determine the linkage between the scheduling behavior of each drainage facility and the hydrological response of the key node;

[0008] Based on the joint scheduling of various drainage facilities, a multi-objective optimization scheduling model is established with the dual objectives of minimizing urban flooding risk and optimizing operating costs, so as to obtain the global scheduling objective at the upper-level city scale.

[0009] Based on the linkage between the drainage facilities and the key nodes, the global scheduling objective is gradually decomposed into local scheduling objectives for multiple sub-regions, and a mapping relationship is established between the local scheduling objectives within the sub-regions and the scheduling decisions of the drainage facilities.

[0010] A Latin hypercube sampling is used to generate a population of different candidate scheduling schemes, and a cross-scale nonlinear mapping relationship of all individuals in the population is predicted, a true solution of the multi-objective optimization scheduling model of the predicted potential individual is obtained, an optimal solution is selected by constructing an expected improvement matrix, and an optimal scheduling scheme is obtained.

[0011] Optionally, the large-scale drainage and waterlogging prevention engineering group optimization scheduling method of the plain river network also includes a method for constructing a composite waterlogging causing probability model and determining a typical composite waterlogging event, comprising:

[0012] Around the composite waterlogging problem induced by multiple factors in the plain river network urban area, multi-dimensional waterlogging causing factor data is collected and integrated;

[0013] The time evolution characteristics of the multi-dimensional waterlogging causing factor data are analyzed, and a three-dimensional Copula function is used to construct a composite waterlogging causing probability model;

[0014] According to the constructed composite waterlogging causing probability model, the occurrence probability of a composite waterlogging event is quantified, and a typical composite waterlogging event under the mutual induction of multiple factors is determined according to the occurrence probability greater than a preset threshold.

[0015] Optionally, the process of the typical composite waterlogging event is simulated and analyzed by establishing a water dynamics model coupled with the large-scale drainage and waterlogging prevention engineering group, and the key nodes of waterlogging are identified, comprising:

[0016] A water dynamics model coupled with the large-scale drainage and waterlogging prevention engineering group is established to simulate and analyze the process of the typical composite waterlogging event, and the scheduling range of waterlogging is divided into different influence areas;

[0017] According to the control effect and the inhibition ability of the waterlogging risk of different drainage facilities in different disaster situations in the influence area, the influence of the regulation and control parameters of each drainage facility on the waterlogging index is quantified by using a sensitivity analysis method, and the disaster reduction contribution degree of each drainage facility under complex waterlogging situations is obtained;

[0018] The mutation characteristics of hydrological parameters in the process of the typical composite waterlogging event are analyzed, the nonlinear influence of the scheduling behavior of each drainage facility on the hydrological parameters is captured according to the disaster reduction contribution degree of each drainage facility, and the key nodes of waterlogging control are identified.

[0019] Optionally, the large-scale drainage and waterlogging prevention engineering group optimization scheduling method of the plain river network also includes a method for establishing a correlation matrix and determining a linkage relationship, comprising:

[0020] A set of all the drainage facilities in the water dynamics model and a set of all the identified key nodes are obtained;

[0021] traversing the set of drainage facilities and the set of key nodes, if the drainage facility and the key node have a hydrological strong coupling relationship, determining a matrix element , if the drainage facility and the key node do not have a hydrological strong coupling relationship, determining a matrix element ;

[0022] According to the determined all matrix elements, a correlation matrix of the influence of each drainage facility on the key node is established.

[0023] From the hydrodynamic model, the linkage path of the drainage facility associated with the key node is obtained, and each linkage path is assigned a weight according to the flow contribution rate or water level response strength of the key node.

[0024] Obtain the set of all linkage paths, and according to the set of drainage facilities, key nodes and linkage paths, the hydrodynamic model is abstracted into a directed graph by introducing a graph theory method, and the function expression of the hydrodynamic model directed graph is:

[0025]

[0026] In the formula, G is a directed graph, V is a set of drainage facilities and key nodes, and E is a set of linkage paths.

[0027] According to the weight assigned to the linkage path, the time series linkage characteristics corresponding to each linkage path are verified from the directed graph of the hydrodynamic model based on Granger causality test, and the linkage relationship between the drainage facility scheduling behavior and the hydrological response of the key node associated with the key node is determined according to the verification result. The function expression of the directed graph verification of the hydrodynamic model is:

[0028]

[0029] In the formula, is the scheduling behavior of the i-th upstream drainage facility at time t, is the water level or flow response of the j-th downstream key node at time t, is a constant term, and are to be estimated coefficients, is a random disturbance term at time t, and p and q are autoregressive and lag orders.

[0030] Optionally, the multi-objective optimization scheduling model is established with the dual objectives of minimizing the waterlogging risk and optimizing the operation cost, including:

[0031] obtain operation parameters of each drainage facility and water conservancy parameters of each key node, to establish a multi-objective optimization scheduling model for minimizing waterlogging risk and optimizing operation cost, and a function expression of the multi-objective optimization scheduling model is:

[0032]

[0033] in the formula, F1 is the objective of minimizing waterlogging risk, I is the total number of waterlogging points in the influence range of each drainage facility in the project, is the maximum water accumulation of the ith key node, F2 is the objective of optimizing operation cost, M is the total number of drainage facilities in the project, is the operation power consumption of the mth drainage facility at the tth time period, is the total operation time of the drainage facility, N is the total number of drainage facilities that need to be maintained, is the maintenance cost of the nth drainage facility.

[0034] Optionally, the large-scale drainage and waterlogging prevention engineering group optimization scheduling method of the plain river network further comprises a method of gradually decomposing a global scheduling objective and obtaining a multi-scale collaborative optimization scheduling mode, comprising:

[0035] According to the linkage relationship between the drainage facility and the key node associated with the key node, the global scheduling objective is gradually decomposed into local scheduling objectives of a plurality of sub-regions in the upper city scale according to each linkage path;

[0036] According to the identified waterlogging key nodes, the associated nodes that can take into account the hydrological coupling characteristics of the remaining sub-regions are selected from the key nodes of all sub-regions;

[0037] According to the water conservancy parameters of the associated nodes, the local scheduling objectives in the sub-region in the upper city scale are established, and a cross-scale nonlinear mapping relationship between the scheduling decision of the drainage facility in the corresponding sub-region in the lower storage and drainage scale is obtained, to obtain a multi-scale collaborative optimization scheduling mode, and a function expression of the cross-scale nonlinear mapping relationship is:

[0038]

[0039] in the formula, is the scheduling decision of the drainage facility in the sub-region in the lower storage and drainage scale, is the ith coefficient, is the local scheduling objective in the sub-region in the upper city scale, and n is the order of the polynomial.

[0040] Optionally, the plain river network large-scale drainage flood control engineering group optimization scheduling method further comprises a method of generating a population of different candidate scheduling schemes, obtaining a real solution of the multi-objective optimization scheduling model of the predicted potential individuals, and selecting an optimal solution to obtain an optimal scheduling scheme, comprising:

[0041] According to the multi-scale collaborative optimization scheduling mode, Latin hypercube sampling is used to generate an initial population composed of different candidate scheduling schemes, and to cover the decision space under the constraints of the multi-objective optimization scheduling model;

[0042] All individuals in the initial population are used as initial samples, and the distribution range of the initial population in the decision space is updated based on an evolutionary algorithm to dynamically shrink the search space of an adaptive search space algorithm, and an agent database is constructed according to individual samples close to the current population in the decision space, and the function expression of the adaptive search space algorithm is:

[0043]

[0044]

[0045] In the formula, is the maximum value of the d-dimensional search space, is the minimum value of the d-dimensional search space, is the maximum coordinate of the individual in the current population, is the minimum coordinate of the individual in the current population, is the maximum allowed value of the search space, is the minimum allowed value of the search space, is the diffusion coefficient;

[0046] A Kriging surrogate model is established according to the agent database, and the Kriging surrogate model is used to predict the cross-scale nonlinear mapping relationship of all individuals in the current population;

[0047] Potential individuals with excellent performance are selected from the prediction results, and a hydrodynamic model is called for the multi-objective optimization scheduling model of the potential individuals to obtain Pareto solutions;

[0048] According to the multi-objective optimization scheduling model and the Pareto solutions of the potential individuals, the expected improvement value of each potential individual in the multi-objective space is calculated and obtained, and an expected improvement matrix is constructed therefrom, and the function expression of the expected improvement matrix is:

[0049]

[0050] In the formula, is the expected improvement matrix, For the candidate scheduling scheme, WEI is the expected improvement value of the potential individual in the multi-objective space, J is the number of multi-objective optimization scheduling models, and J is the number of points on the Pareto front.

[0051] The expected improvement matrix is combined into a scalar function by the Euler distance, and the Pareto solution with the highest comprehensive potential is evaluated from all Pareto solutions by the scalar function, and the candidate scheduling scheme corresponding to the optimal solution is determined as the optimal scheduling scheme.

[0052] The application also discloses a scheduling system adopting the large-scale drainage and waterlogging prevention engineering group optimization scheduling method for a plain river network.

[0053] The waterlogging event determination module is used for collecting multi-dimensional waterlogging-causing factor data of the plain river network, and constructing a composite waterlogging-causing probability model based on the data, and determining a typical composite waterlogging event of the plain river network through the composite waterlogging-causing probability model.

[0054] The waterlogging node identification module is used for establishing a hydrodynamic model coupled with the large-scale drainage and waterlogging prevention engineering group to simulate and analyze the process of the typical composite waterlogging event, and identifying key nodes of waterlogging.

[0055] The linkage relationship determination module is used for establishing a correlation matrix of the influence of each drainage facility on the key nodes in the hydrodynamic model, and determining the linkage relationship between the scheduling behavior of each drainage facility and the hydrological response of the key nodes.

[0056] The scheduling target establishment module is used for establishing a multi-objective optimization scheduling model with the double targets of minimizing waterlogging risk and optimizing operation cost according to the joint scheduling of each drainage facility, and obtaining a global scheduling target at the upper city scale.

[0057] The mapping relationship establishment module is used for gradually decomposing the global scheduling target into local scheduling targets of multiple sub-regions according to the linkage relationship between the drainage facilities and the key nodes, and establishing the mapping relationship between the local scheduling targets in the sub-regions and the scheduling decisions of the drainage facilities.

[0058] The optimal scheduling determination module is used for generating a population of different candidate scheduling schemes by using Latin hypercube sampling, predicting the cross-scale nonlinear mapping relationship of all individuals in the population, obtaining real solutions of the multi-objective optimization scheduling model of the predicted potential individuals, constructing an expected improvement matrix, selecting an optimal solution, and determining an optimal scheduling scheme.

[0059] The application also discloses a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the large-scale drainage and waterlogging prevention engineering group optimization scheduling method for a plain river network.

[0060] The application further discloses a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, and characterized in that the processor implements the plain river network large-scale drainage and waterlogging prevention engineering group optimal scheduling method when executing the computer program.

[0061] Compared with the prior art, the plain river network large-scale drainage and waterlogging prevention engineering group optimal scheduling method and system provided by the embodiments of the application has the beneficial effects that:

[0062] By constructing a composite waterlogging-causing probability model to identify a typical composite waterlogging event, an association matrix of drainage facilities and key nodes is established based on a hydrodynamic model to determine the linkage relationship thereof. A multi-objective optimization model is used to realize the double-objective coordination of waterlogging risk minimization and operation cost optimization, the global scheduling target is converted into a local target of a sub-region through target decomposition, and a mapping relationship between the local target and the drainage facility scheduling decision is established. A Latin hypercube sampling is used to generate a candidate scheduling scheme population, and an optimal scheduling scheme is obtained by constructing an expected improvement matrix through real solving of potential individuals. BRIEF DESCRIPTION OF DRAWINGS

[0063] The technical solutions of the application will be further described in detail below with reference to the drawings and embodiments. In the drawings:

[0064] Figure 1 A step schematic diagram of the plain river network large-scale drainage and waterlogging prevention engineering group optimal scheduling method provided by the embodiments of the application is shown in the drawings.

[0065] Figure 2 A scheduling flowchart schematic diagram of the plain river network large-scale drainage and waterlogging prevention engineering group optimal scheduling method provided by the embodiments of the application is shown in the drawings.

[0066] Figure 3 A flowchart schematic diagram of real solving of the potential individual multi-objective optimization scheduling model provided by the embodiments of the application is shown in the drawings.

[0067] Figure 4 A Pareto solution distribution schematic diagram of calling the hydrodynamic model to solve the potential individual multi-objective optimization scheduling model provided by the embodiments of the application is shown in the drawings.

[0068] Figure 5 A plain river network urban area waterlogging risk comparison schematic diagram before and after scheduling provided by the embodiments of the application is shown in the drawings. DETAILED DESCRIPTION

[0069] It should be noted that the embodiments in the present application and the features in the embodiments can be combined with each other without conflict. The preferred embodiments of the present application will be described in detail with reference to the drawings.

[0070] The application discloses a plain river network large-scale drainage and waterlogging prevention engineering group optimization scheduling method, which comprises the following steps: Figure 1 and Figure 2 as shown in the drawings, comprising:

[0071] S1, collecting multi-dimensional waterlogging causing factor data of the plain river network, and constructing a composite waterlogging causing probability model based on the data, and determining a typical composite waterlogging event of the plain river network through the composite waterlogging causing probability model;

[0072] S2, establishing a water dynamics model coupled with the large-scale drainage and waterlogging prevention engineering group to simulate and analyze the process of the typical composite waterlogging event, and identifying a key node of waterlogging;

[0073] S3, establishing a correlation matrix of the influence of each drainage facility on the key node in the water dynamics model, and determining the linkage relationship between the scheduling behavior of each drainage facility and the hydrological response of the key node;

[0074] S4, establishing a multi-objective optimization scheduling model according to the joint scheduling of each drainage facility, with the double objectives of minimizing the waterlogging risk and optimizing the operation cost, to obtain a global scheduling target at the upper city scale;

[0075] S5, according to the linkage relationship between the drainage facility and the key node, gradually decomposing the global scheduling target into local scheduling targets of multiple sub-regions, and establishing a mapping relationship between the local scheduling targets in the sub-regions and the scheduling decisions of the drainage facility;

[0076] S6, generating a population of different candidate scheduling schemes by using Latin hypercube sampling, predicting the cross-scale nonlinear mapping relationship of all individuals in the population, obtaining the real solution of the multi-objective optimization scheduling model of the predicted potential individuals, constructing an expected improvement matrix, selecting the optimal solution to obtain an optimal scheduling scheme.

[0077] By implementing the above-mentioned optimization scheduling method embodiment, a composite waterlogging probability model is constructed, a typical composite waterlogging event of a plain river network is identified based on multi-dimensional waterlogging factor data, and accurate disaster scenario input is provided for subsequent optimization scheduling. By establishing a water power model coupled with a large-scale drainage and waterlogging prevention engineering group, the whole process of the composite waterlogging event is simulated, and the key nodes of the waterlogging are accurately identified, laying a foundation for optimization scheduling. Based on the analysis results of the water power model, a correlation matrix of the influence of each drainage facility on the key nodes is established, the linkage relationship between the drainage facility scheduling behavior and the hydrological response of the key nodes is clarified, and the precise correlation between the engineering group and the waterlogging nodes is realized. In terms of optimization model construction, the minimization of waterlogging risk and the optimization of operation cost are taken as dual objectives, a global scheduling target at the city scale is established, and the dual requirements of disaster prevention and reduction and economic operation are ensured. By decomposing the global scheduling target into local targets of sub-regions and establishing a mapping relationship between the local targets and the drainage facility scheduling decisions, multi-scale collaborative optimization is realized. Latin hypercube sampling is used to generate a candidate scheduling scheme population, the cross-scale nonlinear mapping relationship of the population individuals is predicted, potential individuals are selected for real solving, and an expected improvement matrix is constructed to obtain the optimal scheduling scheme, which significantly improves the optimization efficiency. The method of the present application combines machine learning prediction with optimization algorithm, ensures the accuracy of the scheduling scheme, meets the timeliness requirement of emergency response, realizes the systematic and precise scheduling of the large-scale drainage and waterlogging prevention engineering group, and provides reliable technical support for urban waterlogging disaster management. The scheduling result is verified by a high-precision waterlogging process model and has good robustness, can maintain stable disaster prevention effect under composite waterlogging conditions, and takes into account the economy of engineering operation, which is a practical and efficient optimization scheduling method for drainage and waterlogging prevention engineering.

[0078] Further, the large-scale drainage and waterlogging prevention engineering group optimization scheduling method of the plain river network further comprises a method for constructing a composite waterlogging probability model and determining a typical composite waterlogging event, comprising:

[0079] Collect and integrate multi-dimensional waterlogging factor data around the composite waterlogging problem induced by multiple factors in the urban area of the plain river network;

[0080] Analyze the time evolution characteristics of the multi-dimensional waterlogging factor data, and construct a composite waterlogging probability model using a three-dimensional Copula function (Copula Function, Connection Function);

[0081] According to the constructed composite waterlogging probability model, the occurrence probability of the composite waterlogging event is quantified, and the typical composite waterlogging event induced by multiple factors is determined according to the occurrence probability greater than a preset threshold.

[0082] Through the implementation of the above-mentioned optimal scheduling method embodiment, multi-dimensional waterlogging factor data of the plain river network urban area are collected and integrated, and multiple factors such as heavy rainfall, external river flood jacking and internal river water rise are comprehensively considered to provide a data basis for subsequent optimal scheduling. The composite waterlogging probability model constructed by the three-dimensional Copula function can accurately describe the nonlinear correlation and time evolution characteristics between each waterlogging factor, breaking through the limitations of traditional single-factor analysis methods. By quantifying the occurrence probability of composite waterlogging events and determining typical composite waterlogging events based on a preset threshold, the complex waterlogging scenarios of the plain river network urban area are accurately identified and probabilistically described. Further, the synergistic effect and coupling mechanism between different waterlogging factors can be effectively captured, providing reliable input conditions for subsequent water dynamics model simulation and optimal scheduling. Copula is a Latin word, and the Copula function describes the correlation between variables. In fact, it is a class of functions that connect joint distribution functions with their respective marginal distribution functions, so it is also called a connection function.

[0083] Further, a water dynamics model coupled with a large-scale drainage and waterlogging prevention engineering group is established to simulate and analyze the process of a typical composite waterlogging event, and the key nodes of waterlogging are identified, including:

[0084] A water dynamics model coupled with a large-scale drainage and waterlogging prevention engineering group is established to simulate and analyze the process of a typical composite waterlogging event, and the scheduling range of waterlogging is divided into different influence areas;

[0085] According to the control effect and waterlogging risk suppression ability of different drainage facilities in different disaster scenarios, the influence of the regulation parameters of each drainage facility on the waterlogging index is quantified by using a sensitivity analysis method, and the disaster reduction contribution of each drainage facility in a complex waterlogging scenario is obtained;

[0086] The mutation characteristics of hydrological parameters in the process of a typical composite waterlogging event are analyzed, the nonlinear influence of the scheduling behavior of each drainage facility on the hydrological parameters is captured according to the disaster reduction contribution of each drainage facility, and the key nodes of waterlogging control are identified.

[0087] By implementing the above-mentioned optimization scheduling method embodiment, a water power model coupled with a large-scale drainage and waterlogging prevention engineering group is established, and the accurate simulation analysis of the whole process of a typical composite waterlogging event can scientifically divide the waterlogging scheduling range into different influence areas, providing a spatial basis for subsequent targeted scheduling. For example, the waterlogging scheduling range can be divided into the following three types of influence areas: (1) rainfall influence area: a waterlogging area dominated by rainfall intensity and duration, which is greatly affected by terrain slope and pipe network drainage capacity; (2) jacking influence area: a waterlogging area dominated by external river tide jacking, which shows obvious water level rise and water flow backflow characteristics; (3) joint influence area: a waterlogging area jointly affected by rainfall and external river jacking, which has complex waterlogging mechanism and high disaster risk. In each type of influence area, around different drainage facilities (such as drainage pumping stations, flood diversion gates, and storage reservoirs), the control effect and risk suppression ability of each drainage facility under different disaster-causing situations are evaluated.

[0088] The sensitivity analysis method is used to quantify the influence of each drainage facility regulation parameter (such as opening and closing time, pump flow, and storage volume) on the waterlogging water depth, waterlogging time, and other indicators, accurately evaluate the control effect and risk suppression ability of each drainage facility under different disaster-causing situations, and obtain the disaster reduction contribution degree of each facility under complex waterlogging-causing situations. This quantitative result provides key data support for optimization scheduling. By analyzing the mutation characteristics of hydrological parameters in the process of a typical composite waterlogging event, combining the disaster reduction contribution degree of each drainage facility, capturing the nonlinear influence of pumping station start-stop, gate regulation, and other scheduling behaviors on the hydrological parameters of the river network and pipe network system, the complex action mechanism between facility scheduling and hydrological response is revealed. Therefore, based on the above analysis results, the key nodes of waterlogging control are accurately identified to clarify the relative importance of each drainage facility in the disaster reduction process, laying a solid foundation for the subsequent establishment of a multi-objective optimization scheduling model.

[0089] Further, the large-scale drainage and waterlogging prevention engineering group optimization scheduling method for plain river networks further includes a method for establishing a correlation matrix and determining a linkage relationship, including:

[0090] Obtaining a set of all drainage facilities in the water power model and a set of all identified key nodes;

[0091] Traversing the set of drainage facilities and the set of key nodes, if the drainage facility and the key node have a hydrological strong coupling relationship, determining the matrix element , if the drainage facility and the key node do not have a hydrological strong coupling relationship, determining the matrix element ;

[0092] Establishing a correlation matrix of the influence of each drainage facility on the key nodes according to all the determined matrix elements;

[0093] The linkage path of the drainage facility and the associated key node is obtained from the water power model, and each linkage path is weighted according to the flow contribution rate or water level response intensity of the key node;

[0094] The set of all linkage paths is obtained, and the water power model is abstracted into a directed graph by introducing a graph theory method according to the set of drainage facilities, key nodes and the set of linkage paths, and the function expression of the water power model directed graph is:

[0095]

[0096] In the formula, G is a directed graph, V is a set of drainage facilities and key nodes, and E is a set of linkage paths;

[0097] According to the weight given to the linkage path, the time sequence linkage characteristics corresponding to each linkage path are tested from the directed graph of the water power model based on Granger causality test, and the linkage relationship between the drainage facility scheduling behavior and the hydrological response of the associated key node is determined according to the test result, and the function expression of the water power model directed graph test is:

[0098]

[0099] In the formula, is the scheduling behavior of the i-th upstream drainage facility at time t, is the water level or flow response of the j-th downstream key node at time t, is a constant term, and are to be estimated coefficients, is a random disturbance term at time t, and p and q are autoregressive and lag orders.

[0100] Through the implementation of the above optimization scheduling method embodiment, the association matrix of the set of drainage facilities and the set of key nodes is established, and the topological structure of the engineering group and the waterlogging node is accurately described by using a binary assignment rule: the matrix element represents that the drainage facility and the key node have a hydrological strong coupling relationship, and the matrix element represents that the drainage facility and the key node have no significant coupling, so that the association matrix can clearly describe the linkage structure of each drainage facility and key node, thereby providing a system-level association framework for subsequent optimization scheduling.

[0101] On the basis of the association matrix, the water power model is abstracted into a directed graph by introducing a graph theory method wherein the vertex set V contains all drainage facilities and key nodes, the edge set E represents the linkage path with hydrological strong coupling, the path weight is quantified by the flow contribution rate or the water level response intensity, and the physical system is digitally modeled. The Granger Causality Test model (Granger Causality Test) is used to analyze the time sequence characteristics of each linkage path, wherein characterizes the scheduling behavior of the drainage facility, reflects the hydrological response of the key node, and the significance test (Granger causality test) =0) verifies the causal influence intensity of the facility action on the node state, so as to determine whether the scheduling behavior of the upstream facility significantly causes the hydrological response of the downstream node, and to extract the lag order q to determine the scheduling response time lag, and to reveal the complex linkage relationship in the system. If the test result is not zero, it is determined that the scheduling behavior of the drainage facility has Granger causality on the hydrological response of the key node . The method realizes the three-level analysis of the two-dimensional modeling of “spatial topology + time linkage” by constructing the system framework through the correlation matrix, quantifying the coupling strength through the directed graph, and analyzing the dynamic causality through the Granger test, and provides the linkage rule library of the drainage facility-key node for the multi-objective optimization. The correlation matrix selects the key control relationship, the graph theory weight calibrates the influence degree, and the Granger parameter (p, q) determines the control time and intensity, which together constitute the basis for the linkage and scheduling of the drainage and flood control system.

[0102] Further, a multi-objective optimization scheduling model is established with the dual objectives of minimizing the waterlogging risk and optimizing the operation cost, including:

[0103] The operation parameters of each drainage facility and the water conservancy parameters of each key node are obtained, and a multi-objective optimization scheduling model is established with the dual objectives of minimizing the waterlogging risk and optimizing the operation cost. The function expression of the multi-objective optimization scheduling model is:

[0104]

[0105] In the formula, F1 is the objective of minimizing the waterlogging risk, I is the total number of waterlogging points in the scheduling influence range of each drainage facility in the project, is the maximum water accumulation of the i-th key node, F2 is the objective of optimizing the operation cost, M is the total number of drainage facilities in the project, is the operation power consumption of the m-th drainage facility at the t-th time period, is the total operation time length of the drainage facility, N is the total number of drainage facilities that need to be maintained, is the maintenance cost of the n-th drainage facility.

[0106] ​By implementing the above-mentioned optimization scheduling method embodiment, for the joint scheduling of engineering facilities undertaking the drainage and waterlogging prevention task, the utilization rate of various types of projects should be improved as much as possible, while reducing the cost on the basis of ensuring drainage. Therefore, by comprehensively considering the mutual influence of urban surface waterlogging and the operation benefit of drainage and waterlogging prevention projects, a multi-objective optimization scheduling model with the double objectives of minimizing the waterlogging risk F1 and optimizing the operation cost F2 is established, so as to realize the collaborative optimization scheduling of the drainage and waterlogging prevention project group in the plain river network city. In the model construction, F1 quantifies the waterlogging risk by minimizing the sum of the maximum water accumulation of all key nodes, and the objective function ensures that the scheduling scheme can effectively control the water accumulation depth of each key node; F2 comprehensively considers the operation power consumption and maintenance cost of the drainage facilities, and the objective function ensures the economic feasibility of the scheduling scheme. By simultaneously optimizing the two mutually restrictive objective functions, the model can seek the best balance point between reducing the waterlogging risk and controlling the operation cost, so that the Pareto optimal solution set of the objective function obtained in the subsequent solution can provide multiple optional scheduling strategies under different risk-cost trade-offs. This double-objective optimization method fully considers the actual operation parameters of the drainage facilities and the water conservancy parameters of the key nodes, and converts the complex engineering scheduling problem into a solvable multi-objective optimization problem through mathematical modeling, thereby providing a scientific decision basis for the joint scheduling of the drainage and waterlogging prevention project group in the plain river network city. Among them, the constraint conditions of the multi-objective optimization scheduling model include the basic design and operation limitations of the project itself, such as river water level, gate dam change amount, pump flow, and storage capacity, etc. In addition, the scheduling rule constraint requires to follow the existing scheduling classification to ensure hierarchical control for different rainstorm grades and river flow, thereby ensuring the safety of the system scheduling.

[0107] Further, the large-scale drainage and waterlogging prevention project group optimization scheduling method in the plain river network further includes a method of gradually decomposing the global scheduling target and obtaining a multi-scale collaborative optimization scheduling mode, including:

[0108] According to the linkage relationship between the drainage facilities and the associated key nodes, the global scheduling target is gradually decomposed into local scheduling targets of multiple sub-regions in the upper city scale according to each linkage path;

[0109] According to the identified all key nodes of waterlogging, the associated nodes that can take into account the hydrological coupling characteristics of the remaining sub-regions are selected from the key nodes of all sub-regions;

[0110] According to the water conservancy parameters of the associated nodes, the local scheduling target in the sub-region in the upper city scale is established, and a cross-scale nonlinear mapping relationship between the scheduling decision of the drainage facilities in the corresponding sub-region in the lower storage scale is obtained, to obtain a multi-scale collaborative optimization scheduling mode, and the function expression of the cross-scale nonlinear mapping relationship is:

[0111]

[0112] wherein, is the scheduling decision of the drainage facilities in the lower storage and drainage scale sub-region, is a constant term, is the i-th coefficient, is the local scheduling target in the upper city scale sub-region, and n is the order of the polynomial.

[0113] Through the implementation of the above-mentioned optimization scheduling method embodiment, a cross-scale nonlinear mapping relationship is established, which can realize the precise coupling of the upper city scale global scheduling target and the lower storage and drainage scale drainage facility scheduling decision. Specifically, Y represents the specific scheduling decision parameters (such as gate opening, pump station power, etc.) of the drainage facilities in the lower storage and drainage scale sub-region, and X represents the local scheduling target quantitative value (such as the lowest water level of the river control section, the minimum pipe network node flow, the optimal pump station power consumption, etc.) in the upper city scale sub-region. is the polynomial coefficient fitted by the least square method, to ensure that the mapping function has high fitting accuracy and robustness. The polynomial order n value is dynamically adjusted according to the hydrological characteristics of the sub-region. The above method first decomposes the global scheduling target into local scheduling targets of multiple sub-regions according to the hydrological influence path based on the linkage relationship between the drainage facilities and the key nodes, to ensure that the scheduling requirements of each sub-region are consistent with the overall flood control target; then, the related nodes with cross-regional hydrological coupling characteristics are selected from all the key nodes, which respond to the scheduling of the drainage facilities in the region and also affect the hydrological state of the adjacent region, so as to serve as the link between the upper and lower layers, and further reveal the transmission mechanism of the scheduling decision (such as gate opening, pump station control water level, etc.) and the local scheduling target (such as pipe network and storage capacity, river limited water level) between different scales, to ensure that the scheduling requirements of each sub-region are consistent with the overall flood control target, to realize multi-level dynamic optimization from global to local, to ensure the global consistency and regional coordination of the scheduling instructions; finally, the quantitative conversion relationship between X and Y is established through the polynomial mapping, so that the abstract local scheduling target X is converted into the executable facility operation parameter Y. Among them, the coefficient The determination needs to comprehensively consider the lag parameter of Granger causality test, the hydrological influence strength of graph theory weight and historical optimal scheduling data, to ensure that the mapping relationship meets the physical law of hydrological system and adapts to the actual operation condition. Thus, through target decomposition, the complex global problem is converted into a sub-regional optimization problem which can be solved in parallel, the associated nodes are used to build the coordination link across regions, the scale conversion of scheduling instructions is realized by means of polynomial mapping, and a complete scheduling chain of “global target → local target → facility operation” is formed. This multi-scale collaborative optimization scheduling mode can effectively solve the problems of global target and local execution, and time and space coordination between different sub-regions in the scheduling of large-scale drainage and waterlogging prevention engineering group in plain river network urban areas, and provides a standardized and quantifiable decision conversion method for large-scale engineering system scheduling, thereby reducing the complexity of the optimization problem.

[0114] Further, as shown in Figure 3 and Figure 4 The large-scale drainage and waterlogging prevention engineering group optimization scheduling method in plain river network further includes the method of generating a population of different candidate scheduling schemes, obtaining real solution of the multi-objective optimization scheduling model of the predicted potential individual, and selecting the optimal solution to obtain the optimal scheduling scheme, including:

[0115] According to the multi-scale collaborative optimization scheduling mode, Latin hypercube sampling is used to generate an initial population composed of different candidate scheduling schemes, and to cover the decision space under the constraints of the multi-objective optimization scheduling model;

[0116] All individuals in the initial population are used as initial samples, and the distribution range of the initial population in the decision space is updated based on the evolutionary algorithm, the adaptive search space algorithm is used to dynamically shrink the search space, the agent database is constructed according to the individual samples close to the current population in the decision space, and the function expression of the adaptive search space algorithm is:

[0117]

[0118]

[0119] In the formula, is the maximum value of the d-dimensional search space, is the minimum value of the d-dimensional search space, is the maximum coordinate of the individual in the current population, is the minimum coordinate of the individual in the current population, is the maximum allowed value of the search space, is the minimum allowed value of the search space, is the diffusion coefficient;

[0120] A Kriging surrogate model is established according to the agent database, and the cross-scale nonlinear mapping relationship of all individuals in the current population is predicted through the Kriging surrogate model;

[0121] Potential individuals with excellent performance are selected from the prediction results, and a hydrodynamic model is called for the multi-objective optimization scheduling model of the potential individuals to obtain Pareto solutions;

[0122] According to the multi-objective optimization scheduling model of the potential individuals and the Pareto solutions, the expected improvement value of each potential individual in the multi-objective space is calculated and obtained, and an expected improvement matrix is constructed therefrom, and the functional expression of the expected improvement matrix is:

[0123]

[0124] In the formula, is the expected improvement matrix, is a candidate scheduling scheme, WEI is the expected improvement value of the potential individual in the multi-objective space, is the number of multi-objective optimization scheduling models, and J is the number of points on the Pareto front;

[0125] The expected improvement matrix is combined into a scalar function through the Euler distance, the Pareto solution with the highest comprehensive potential is evaluated from all Pareto solutions through the scalar function, and the candidate scheduling scheme corresponding to the optimal solution is determined as the optimal scheduling scheme.

[0126] Through the implementation of the above-described optimized scheduling method, an initial population is generated using Latin hypercube sampling, ensuring that candidate scheduling schemes are uniformly distributed within the decision space constrained by the multi-objective optimized scheduling model. Each individual represents a complete set of drainage facility scheduling parameters (such as gate opening and pumping station power). The population distribution range is updated based on an evolutionary algorithm, and an adaptive search space algorithm is used to dynamically shrink the search space to calculate the next generation search boundary. A Kriging surrogate model is used to predict the cross-scale nonlinear mapping relationship of all individuals in the population. This model utilizes a surrogate database to establish the statistical relationship between decision variables and the objective function, and selects high-performing potential individuals from the prediction results. For these individuals, the hydrodynamic model is called to obtain accurate Pareto solutions (Pareto Optimal Solutions), significantly reducing the number of hydrodynamic model calls. This greatly improves the optimization computation efficiency by reducing the number of solver calls and ensures the physical authenticity of the optimization results. Since the optimal individuals in each generation are calculated using the hydrodynamic model, the authenticity of the optimization is guaranteed, thus avoiding evolutionary misleading caused by insufficient model accuracy. Furthermore, individuals obtained through real-world solutions are added to the surrogate database as new sample data. Online updates to the surrogate database gradually improve the accuracy of the surrogate model as the optimization process progresses, enabling precise and efficient optimization solutions to drainage and flood control system scheduling problems under limited resources.

[0127] In multi-objective optimization problems, traditional single-objective filling criteria have limitations in selecting multi-dimensional candidate solutions, such as the MSP (Multi-objective Sampling Criterion) and EI (Expected Improvement Criterion). Directly extending single-objective criteria to multi-objective problems usually requires complex decomposition and integration of irregular non-dominated regions, resulting in cumbersome and time-consuming computation. Therefore, the method in this invention introduces an expected improvement matrix to quantify the improvement potential of potential individuals in the multi-objective space, treating the approximate Pareto solution set of the multi-objective optimization problem as an extension of the single-objective optimal solution in two directions. Through this improved criterion, only one-dimensional integration is required, and it has an explicit analytical expression, significantly reducing computational complexity. Furthermore, an expected improvement matrix is ​​constructed for each approximation point beyond the Pareto front: matrix elements... Indicate candidate scheduling schemes Relative to the J-th Pareto solution at the th The expected improvement value on a multi-objective optimization scheduling model is obtained by introducing weighting coefficients to balance the global optimum and local optimum. indicates the number of multi-objective optimization scheduling models (inundation risk minimization F1 and optimal operation cost F2). Wherein, each element in the expected improvement matrix is a one-dimensional expected improvement function, and finally the multi-dimensional expected improvement value is combined into a scalar function by the Euclidean distance, and the optimal solution with the highest comprehensive potential is evaluated from all Pareto solutions, and the candidate scheduling scheme corresponding to the solution is the optimal scheduling scheme considering the effect of inundation prevention and control and economic operation cost. The above method constructs an optimization framework combining Latin hypercube sampling, adaptive search space algorithm, Kriging surrogate model prediction, potential individual precise solution and multi-objective decision-making, realizes the online coupling of the hydrodynamic model and the optimization algorithm by constructing a machine learning model as a surrogate model, greatly improves the calculation efficiency while ensuring the calculation accuracy, and introduces the matrix idea to improve the multi-objective expected improvement filling criterion, improves the selection efficiency of sample points in the multi-dimensional space in the optimization iteration process, so as to realize the efficient and accurate scheduling of large-scale drainage and flood control engineering group of plain river network. The adaptive search space algorithm ensures the rapid convergence of the optimization process, the Kriging surrogate model balances the calculation efficiency and accuracy, and the expected improvement matrix provides a scientific multi-objective decision-making basis, so that the scheduling scheme generated finally reaches the best trade-off between risk and cost on the Pareto frontier.

[0128] Figure 5 The figure is a comparison diagram of the inundation risk before and after the plain river network city scheduling by using the method of the embodiment of the present application. From Figure 5 It can be seen that the high-risk area after scheduling is significantly reduced, and the proportion of dark gray blocks is obviously reduced, verifying the effectiveness of the multi-objective optimization scheduling model in the embodiment of the present application. Moreover, the high-risk area in the figure shifts from the lower middle position and shrinks, and the high-risk distribution changes from "concentrated and contiguous" to "scattered and punctiform", which meets the design expectation of "decomposing the global target into local targets of sub-regions". The white low-risk area after scheduling is expanded, further proving the coordination ability of the embodiment of the present application in the spatial scale.

[0129] The present application also discloses a scheduling system adopting the large-scale drainage and flood control engineering group optimization scheduling method of plain river network, and the scheduling system comprises:

[0130] The inundation event determination module is used for collecting multi-dimensional waterlogging factor data of the plain river network, and constructing a composite waterlogging probability model based on the data, and determining a typical composite inundation event of the plain river network through the composite waterlogging probability model;

[0131] The inundation node identification module is used for establishing a hydrodynamic model coupled with the large-scale drainage and flood control engineering group to simulate and analyze the process of the typical composite inundation event, and identifying the key nodes of the inundation;

[0132] The linkage relationship determining module is used for establishing a correlation matrix of influences of each drainage facility on the key node in the hydrodynamic model, and determining a linkage relationship between scheduling behaviors of each drainage facility and hydrological responses of the key node.

[0133] The scheduling target establishing module is used for establishing a multi-objective optimization scheduling model according to joint scheduling of each drainage facility, with double targets of minimizing the waterlogging risk and optimizing the operation cost, to obtain a global scheduling target at an upper city scale.

[0134] The mapping relationship establishing module is used for gradually decomposing the global scheduling target into local scheduling targets of multiple sub-regions according to the linkage relationship between the drainage facility and the key node, and establishing a mapping relationship between the local scheduling target in the sub-region and a scheduling decision of the drainage facility.

[0135] The optimal scheduling determining module is used for generating a population of different candidate scheduling schemes by using Latin hypercube sampling, predicting a cross-scale nonlinear mapping relationship of all individuals in the population, obtaining a real solution of a multi-objective optimization scheduling model of a predicted potential individual, constructing an expected improvement matrix to select an optimal solution to obtain an optimal scheduling scheme.

[0136] The application further discloses a computer readable storage medium, which stores a computer program, and the computer program is characterized in that the computer program is executed by a processor to realize the large-scale drainage and waterlogging prevention engineering group optimization scheduling method for the plain river network.

[0137] The application further discloses a computer device, which comprises a memory, a processor and a computer program stored in the memory and capable of running on the processor, and the computer device is characterized in that the processor executes the computer program to realize the large-scale drainage and waterlogging prevention engineering group optimization scheduling method for the plain river network.

[0138] The application is described according to flowcharts and / or block diagrams of the method, device (system) and computer program product according to the specific embodiments. It should be understood that each flow and / or block in the flowcharts and / or block diagrams and the combination of the flows and / or blocks in the flowcharts and / or block diagrams can be realized by computer program instructions. The 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 produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device realize the functions specified in the flowcharts and / or block diagrams. Figure 1 The apparatus for realizing the functions specified in one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0139] These computer program instructions can also be stored in a computer readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer readable memory produce an article of manufacture including instructions which implement the flow Figure 1 The functions of a flow or multiple flows and / or a block or multiple blocks in conjunction with the disclosed methods can be implemented on a single device or distributed across several devices. Figure 1

[0140] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flow Figure 1 The functions of a flow or multiple flows and / or a block or multiple blocks in conjunction with the disclosed methods can be implemented on a single device or distributed across several devices. Figure 1

[0141] It should be understood that the above embodiments are only used to illustrate the technical solutions of the present application, rather than limit them. Modifications or equivalent replacements to the technical solutions recorded in the above embodiments, or equivalent replacements to part of the technical features, can be made by those skilled in the art; all these modifications and replacements should belong to the protection scope of the present application.​​

Claims

1. A method for optimizing the scheduling of large-scale drainage and flood control engineering groups in a plain river network, characterized in that, The method comprises the following steps: Collecting plain river network multi-dimensional waterlogging factor data and constructing a composite waterlogging probability model to determine a typical composite waterlogging event of the plain river network through the composite waterlogging probability model; Establishing a water dynamics model coupled with a large-scale drainage and waterlogging prevention engineering group to simulate and analyze the process of the typical composite waterlogging event and identifying key nodes of waterlogging; Establishing a correlation matrix of the influence of each drainage facility in the water dynamics model on the key nodes and determining the linkage relationship between the scheduling behavior of each drainage facility and the hydrological response of the key nodes; The establishment of the correlation matrix and the determination of the linkage relationship comprise: Obtaining a set of all the drainage facilities in the water dynamics model and a set of all the identified key nodes; traversing the set of drainage facilities and the set of key nodes, if the drainage facility and the key node have a hydrological strong coupling relationship, determining a matrix element , if the drainage facility and the key node do not have a hydrological strong coupling relationship, determining a matrix element ; According to all the determined matrix elements, a correlation matrix of the influence of each drainage facility on the key nodes is established; Obtaining the linkage paths of the drainage facilities and the key nodes from the water dynamics model and assigning a weight to each linkage path according to the flow contribution rate or water level response strength of the key nodes; Obtaining a set of all the linkage paths, and abstracting the water dynamics model into a directed graph by introducing a graph theory method according to the set of the drainage facilities, the key nodes and the linkage paths, and the function expression of the directed graph of the water dynamics model is: In the formula, G is a directed graph, V is a set of drainage facilities and key nodes, and E is a set of linkage paths; According to the weight assigned to the linkage path, the time series linkage characteristics corresponding to each linkage path are tested from the directed graph of the water dynamics model based on Granger causality test, and the linkage relationship between the scheduling behavior of the drainage facility and the hydrological response of the key nodes is determined according to the test result, and the function expression of the directed graph test of the water dynamics model is: wherein, is the dispatching behavior of the ith upstream flood control facility at time t, is the water level or flow response of the jth downstream critical node at time t, is a constant term, and are coefficients to be estimated, is a random disturbance term at time t, and p and q are the autoregressive and lag orders. According to the joint scheduling of each drainage facility, a multi-objective optimization scheduling model is established with the dual objectives of minimizing waterlogging risk and optimizing operation cost to obtain a global scheduling target at an upper city scale; According to the linkage relationship between the drainage facilities and the key nodes, the global scheduling target is gradually decomposed into local scheduling targets of multiple sub-regions, and a mapping relationship between the local scheduling target and the scheduling decision of the drainage facility in the sub-region is established; Latin hypercube sampling is used to generate a population of different candidate scheduling schemes, and the cross-scale nonlinear mapping relationship of all individuals in the population is predicted, the multi-objective optimization scheduling model is solved to obtain a predicted potential individual, an expected improvement matrix is selected to obtain an optimal solution, and an optimal scheduling scheme is constructed.

2. The method according to claim 1, wherein, The plain river network large-scale drainage and waterlogging prevention engineering group optimization scheduling method further comprises a method for constructing a composite waterlogging probability model and determining a typical composite waterlogging event, which comprises: Collecting and integrating multi-dimensional waterlogging factor data around the composite waterlogging problem induced by multiple factors in the urban area of the plain river network; Analyzing the time evolution characteristics of the multi-dimensional waterlogging factor data and constructing a composite waterlogging probability model by using a three-dimensional Copula function; According to the constructed composite waterlogging probability model, the occurrence probability of a composite waterlogging event is quantified, and a typical composite waterlogging event under the mutual induction of multiple factors is determined according to an occurrence probability greater than a preset threshold.

3. The method according to claim 1, wherein, The process of the typical composite waterlogging event is simulated and analyzed by establishing a water dynamic model coupled with a large-scale drainage and waterlogging prevention engineering group, and key nodes of waterlogging are identified, including: The process of the typical composite waterlogging event is simulated and analyzed by establishing a water dynamic model coupled with a large-scale drainage and waterlogging prevention engineering group, and the dispatching range of waterlogging is divided into different influence areas; According to the control effect and the inhibition ability of the waterlogging risk of different drainage facilities in different disaster scenarios in the influence area, the influence of the regulation and control parameters of each drainage facility on the waterlogging index is quantified by using a sensitivity analysis method, and the disaster reduction contribution degree of each drainage facility under complex waterlogging scenarios is obtained; The mutation characteristics of hydrological parameters in the process of the typical composite waterlogging event are analyzed, the nonlinear influence of the dispatching behavior of each drainage facility on the hydrological parameters is captured according to the disaster reduction contribution degree of each drainage facility, and the key nodes of waterlogging control are identified.

4. The method according to claim 1, wherein, The multi-objective optimization scheduling model is established with the dual objectives of minimizing waterlogging risk and optimizing operation cost, including: The operation parameters of each drainage facility and the water conservancy parameters of each key node are obtained, and a multi-objective optimization scheduling model is established with the dual objectives of minimizing waterlogging risk and optimizing operation cost, and the function expression of the multi-objective optimization scheduling model is: In the formula, F1 is the target of minimizing the risk of waterlogging, I is the total number of waterlogging points in the influence range of each waterlogging drainage facility in the project, is the maximum waterlogging amount of the ith key node, F2 is the target of optimizing the operation cost, M is the total number of waterlogging drainage facilities in the project, is the operation power consumption of the mth waterlogging drainage facility in the tth time period, is the total operation time length of the waterlogging drainage facility, N is the total number of waterlogging drainage facilities that need to be maintained, is the maintenance cost required by the nth waterlogging drainage facility.

5. The method according to claim 4, wherein, The large-scale drainage and waterlogging prevention engineering group optimization scheduling method for the plain river network further includes a method of gradually decomposing global scheduling objectives and obtaining a multi-scale collaborative optimization scheduling mode, including: According to the linkage relationship between the drainage facilities and the key nodes, the global scheduling objectives are gradually decomposed into local scheduling objectives of multiple sub-regions in the upper city scale according to each linkage path; According to the identified key nodes of waterlogging, the associated nodes that can take into account the hydrological coupling characteristics of the remaining sub-regions are selected from the key nodes of all sub-regions; According to the water conservancy parameters of the associated nodes, the local scheduling objectives in the sub-region in the upper city scale are established, and a cross-scale nonlinear mapping relationship of the drainage facility scheduling decision in the corresponding sub-region in the lower storage and drainage scale is obtained, to obtain a multi-scale collaborative optimization scheduling mode, and the function expression of the cross-scale nonlinear mapping relationship is: wherein is the dispatching decision of the drainage facilities in the lower layer of the urban scale sub-region, is a constant term, is the i-th coefficient, is the local dispatching target in the upper layer of the urban scale sub-region, and n is the order of the polynomial.

6. The method according to claim 5, wherein, The large-scale drainage and waterlogging prevention engineering group optimization scheduling method for the plain river network further includes a method of generating a population of different candidate scheduling schemes, obtaining a real solution of the multi-objective optimization scheduling model of the predicted potential individual, and selecting an optimal solution to obtain an optimal scheduling scheme, including: According to the multi-scale collaborative optimization scheduling mode, a Latin hypercube sampling is used to generate an initial population combined by different candidate scheduling schemes, and to cover the decision space under the constraints of the multi-objective optimization scheduling model. All individuals in the initial population are taken as initial samples, and the distribution range of the initial population in the decision space is updated based on an evolutionary algorithm to establish an adaptive search space algorithm to dynamically shrink the search space, and an agent database is constructed according to individual samples close to the current population in the decision space, and a function expression of the adaptive search space algorithm is: wherein is the maximum value of the d-dimensional search space, is the minimum value of the d-dimensional search space, is the maximum coordinate of the individuals in the current population, is the minimum coordinate of the individuals in the current population, is the maximum allowed value of the search space, is the minimum allowed value of the search space, is the diffusion coefficient; A Kriging agent model is established according to the agent database, and a cross-scale nonlinear mapping relationship of all individuals in the current population is predicted through the Kriging agent model; Potential individuals with excellent performance are selected from the prediction results, and a hydrodynamic model is called for the multi-objective optimization scheduling model of the potential individuals to obtain Pareto solutions; According to the multi-objective optimization scheduling model and the Pareto solutions of the potential individuals, an expected improvement value of each potential individual in the multi-objective space is calculated and obtained, and an expected improvement matrix is constructed therefrom, and a function expression of the expected improvement matrix is: In the formula, is the expected improvement matrix, is the candidate scheduling scheme, WEI is the expected improvement value of the potential individual in the multi-objective space, is the number of multi-objective optimization scheduling models, and J is the number of points on the Pareto front. The expected improvement matrix is combined into a scalar function through the Euler distance, and the Pareto solution with the highest comprehensive potential is evaluated from all Pareto solutions through the scalar function, and a candidate scheduling scheme corresponding to the optimal solution is determined as the optimal scheduling scheme.

7. A dispatching system employing the optimization dispatching method for large-scale drainage and waterlogging prevention engineering groups of plain river networks according to any one of claims 1-6, characterized in that, The scheduling system comprises: A waterlogging event determination module is configured to collect multi-dimensional waterlogging factor data of the plain river network, and construct a composite waterlogging probability model based on the data, and determine a typical composite waterlogging event of the plain river network through the composite waterlogging probability model; A waterlogging node identification module is configured to simulate and analyze the process of the typical composite waterlogging event by establishing a hydrodynamic model coupled with a large-scale drainage and waterlogging prevention engineering group, and identify key nodes of waterlogging; A linkage relationship determination module is configured to establish a correlation matrix of the influence of each drainage facility on the key nodes in the hydrodynamic model, and determine the linkage relationship between the scheduling behavior of each drainage facility and the hydrological response of the key nodes; A scheduling target establishment module is configured to establish a multi-objective optimization scheduling model based on the joint scheduling of each drainage facility, with the dual objectives of minimizing waterlogging risk and optimizing operation cost, to obtain a global scheduling target at the upper city scale; A mapping relationship establishment module is configured to gradually decompose the global scheduling target into local scheduling targets of multiple sub-regions according to the linkage relationship between the drainage facilities and the key nodes, and establish a mapping relationship between the local scheduling targets and the scheduling decisions of the drainage facilities in the sub-regions; An optimal scheduling determination module is configured to generate a population of different candidate scheduling schemes by using Latin hypercube sampling, predict the cross-scale nonlinear mapping relationship of all individuals in the population, obtain real solutions of the multi-objective optimization scheduling model of the predicted potential individuals, construct an expected improvement matrix, and select an optimal solution to obtain an optimal scheduling scheme.

8. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program, when executed by a processor, implements the plain river network large-scale drainage and waterlogging prevention engineering group optimization scheduling method of any one of claims 1-6.

9. A computer device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The processor executes the computer program to implement the plain river network large-scale drainage and waterlogging prevention engineering group optimization scheduling method of any one of claims 1-6.

Images