Urban flood-oriented river network real-time control rule optimization method

Abstract

The invention discloses an urban flood-oriented river network real-time control rule optimization method. The method comprises the steps of S1, collecting river network data; s2, constructing an urban hydrological hydrodynamic coupling model; s3, simulating and calculating flood processes under different design rainstorm scenes according to a rainstorm intensity formula, and comprehensively identifying drainage ports which are easily jacked by a river network water level in the research area; s4, performing spatial clustering on the jacking water outlets by adopting a clustering algorithm to divide regulation and control zones, performing feature clustering on the jacking water outlets again on the basis of spatial clustering, and outputting a clustering center with the maximum influence as a river network regulation and control point; s5, by taking the river network regulation and control point as a regulation and control reference object, compiling an optimized gate pump control rule based on the warning water level and ecological water level data of the river channel; s6, inputting the optimized gate pump control rule into the urban hydrological and hydrodynamic coupling model, evaluating the performance of the method through a rainstorm event, and quantitatively analyzing the response of flood to the gate pump control rule; the river network water level in the rainstorm period can be regulated and controlled, and then the flood risk is relieved.

Description

Technical Field

[0001] The present invention relates to the technical field of water resource management and flood control and regulation, in particular to an optimization method for real-time control rules of river networks for urban floods. Background Art

[0002] As a key component of the urban drainage system, river networks undertake multiple tasks such as flood discharge and water resource regulation. Reasonable real-time control rules for water projects are related to the efficient operation of river networks, can prevent river overflow and urban waterlogging, especially during rainstorms and tidal surges, can effectively reduce the water level of inland rivers and alleviate flood disasters. Traditional real-time control technologies for water projects mostly take the upstream and downstream sections of the project location as the reference objects for regulation, resulting in the inactive response of the urban internal river network to real-time control rules. When the water level of the target section is controlled within the safe range, it cannot ensure that the urban internal drainage outlets are not affected by the river network water level. Therefore, traditional real-time control technologies for water projects are difficult to achieve precise regulation of urban internal river networks and are not conducive to the scientific management of urban river networks. Summary of the Invention

[0003] The purpose of the present invention is to overcome the above deficiencies, provide an optimization method for real-time control rules of river networks for urban floods, solve the problems that traditional real-time control technologies for water projects are difficult to achieve precise regulation of urban internal river networks and are not conducive to the scientific management of urban river networks, and regulate the river network water level during rainstorms to alleviate flood risks.

[0004] To solve the above technical problems, the technical solution adopted by the present invention is: an optimization method for real-time control rules of river networks for urban floods, including the following steps:

[0005] S1, collect river network data;

[0006] S2, construct an urban hydrological model according to detailed spatial division and different surface runoff characteristics, construct a river network hydrodynamic model considering the hydraulic conditions of pumping stations, gates, weir flows, reservoir and lake facilities, construct a pipe network hydrodynamic model considering the hydraulic conditions of inspection wells and outlet facilities, and couple the above models through the process of water volume exchange at section nodes to construct an urban hydrological-hydrodynamic coupling model;

[0007] S3, simulate and calculate the flood process under different design rainstorm scenarios according to the rainstorm intensity formula, and comprehensively identify the drainage outlets in the study area that are vulnerable to river network water level backwater;

[0008] S4, use a clustering algorithm to spatially cluster and divide the backwater drainage outlets into regulation zones, and on the basis of spatial clustering, further perform feature clustering on the backwater drainage outlets, and output the maximum influence clustering center as the river network regulation point;

[0009] S5. Taking the river network regulation points as the regulation reference objects, optimize the gate-pump control rules based on the warning water levels and ecological water level data of the river channels.

[0010] S6. Input the optimized gate-pump control rules into the urban hydrological-hydrodynamic coupling model, evaluate the performance of the rainstorm event assessment method, and quantitatively analyze the response of flood control to the gate-pump control rules.

[0011] Preferably, the urban hydrological-hydrodynamic coupling model in S2 is constructed based on the Infoworks ICM software, and the specific construction process is as follows:

[0012] S21. The runoff calculation for the impervious surfaces of houses and roads adopts the fixed runoff coefficient model in InfoWorks ICM, as shown in the following formula:

[0013] R n = CR P ;

[0014] Where R n is the net rainfall, mm; C is the runoff coefficient; P is the recurrence interval, years; R P is the rainfall corresponding to the recurrence interval, mm;

[0015] S22. The runoff calculation for the pervious surfaces of green spaces adopts the Horton infiltration model, as shown in the following formula:

[0016] f t = f c +(f0 - f c )e -kt ;

[0017] Where f t is the infiltration rate at time t, mm / h; f c is the stable infiltration rate, mm / h; f0 is the initial infiltration rate, mm / h; k is the reduction coefficient, with the dimension of h -1 ; t is the infiltration time, h;

[0018] S23. To simulate the surface runoff in the city, call the nonlinear reservoir model in the stormwater management module and solve it jointly by the continuity equation and the Manning formula, as shown in the following formula:

[0019] Manning formula:

[0020]

[0021] Where Q is the flow rate, m 3 / s; W is the overland flow width of the sub-catchment, m; n is the surface Manning coefficient; d is the water depth, m; d p is the maximum depression storage depth, m; S is the average slope of the sub-catchment;

[0022] Continuity equation:

[0023]

[0024] where V is the surface water accumulation volume, m 3 ; t is time, s; A1 is the surface area, m 2 ; i* is the net rainfall, mm / s;

[0025] For S24, the hydrodynamic module in Infoworks ICM is adopted, and the river network runoff is calculated by completely solving the Saint-Venant equations using the dynamic wave method, as shown in the following formula:

[0026] Continuity equation:

[0027]

[0028] Momentum equation:

[0029]

[0030] where A2 is the cross-sectional area, m 2 ; Q is the flow rate, m 3 / s; g is the acceleration due to gravity, m / s 2 ; θ is the angle between the centerline of the pipeline and the horizontal line, degree; K is the water conveyance rate; h is the water depth, m; t is time, s; S0 is the bottom slope of the channel;

[0031] For S25, the surface runoff and river network water flow are coupled by the method of water volume exchange at the cross-section nodes.

[0032] Preferably, in the S25, the formula for coupling the surface runoff and river network water flow by the method of water volume exchange at the cross-section nodes is as follows:

[0033]

[0034] where is the flow rate of the i-th river network cross-section at time t, is the flow rate of the i-th river network cross-section at time t-1, is the surface runoff corresponding to the i-th river network cross-section at time t.

[0035] Preferably, the rainfall intensity formula in S3 is:

[0036]

[0037] where q is the average rainfall intensity, mm / min; P is the recurrence interval, a; t is the rainfall duration, min; the rainfall process is calculated using the Chicago rainfall pattern, with the design rainfall peak coefficient r = 0.39, the rainfall process time interval set to 1 min, and the rainfall time set to 120 min.

[0038] Preferably, the AGNES algorithm is used for spatial clustering of the apex support drainage outlets in S4. It belongs to the bottom-up agglomerative hierarchical clustering, and the calculation process is as follows:

[0039] S41, Initialization: Each sample point forms a single cluster; if there are m samples in the population, there are m clusters in the initial state;

[0040] S42, Calculate the distance matrix: Initially, use the selected distance metric, and choose the Euclidean distance to calculate the distance matrix between all clusters;

[0041] S43, Merge the closest clusters: Find the two closest clusters, merge them into a new cluster, and update the distance matrix to reflect the distances between the new cluster and other clusters; The merging process is based on the selected linkage method, and the Ward method is chosen to minimize the within-cluster sum of squares after merging:

[0042]

[0043] Where C i and C j are two clusters to be merged; ∣C i ∣ and ∣C j ∣ are the numbers of data points contained in clusters C i and C j respectively; μ i and μ j are the mean vectors of clusters C i and C j respectively; d(μ i , μ j ) is the distance between the mean vectors of the two clusters; d ward (C i , C j ) is the "merging cost" or "distance" between the two clusters calculated based on the Ward method;

[0044] S44, Repeat step S43 until all sample points are merged into one cluster or the predetermined number of clusters is reached;

[0045] S45, Generate the hierarchical tree: Represent the clustering process with a dendrogram, and each merge corresponds to a node on the tree.

[0046] Preferably, in S42, the calculation method of the Euclidean distance between two n-dimensional vectors is as follows:

[0047]

[0048] where d is the Euclidean distance; x 1i and x 2i are any two n-dimensional vectors.

[0049] Preferably, the K-Medoids algorithm is used for feature clustering of the apex support drain outlet in S4. It belongs to the partitioning clustering that directly divides data points into different clusters. The calculation process is as follows:

[0050] S46, Initialization: Randomly select k points from the total m sample points as the initial clustering centers;

[0051] S47, Cluster assignment: Assign the remaining m - k points to the cluster where the nearest clustering center is located according to the minimum distance principle to achieve initial clustering; The distance between each sample point is calculated using the Euclidean distance;

[0052] S48, Update clustering centers: Calculate the distance from the points in the cluster except the clustering centers to all other points, and select the point corresponding to the minimum value as the new clustering center, and recalculate the clustering result;

[0053] S49, Repeat processes S47 and S48 until all the clustering centers no longer change or the set maximum number of iterations is reached to obtain the final clustering result.

[0054] Preferably, the basis for writing the real-time control rules of the water project in S5 is the warning water level and ecological water level data of the river channel, specifically as follows:

[0055] Z(i) min =H(i) 生态水位 ;

[0056] Z(i) max =H(i) 警戒水位 ;

[0057] B(j) min =H(j) 警戒水位 ;

[0058] B(j) max =H(j) 警戒水位 +Δh(j);

[0059] Where Z(i) max is the upper limit water level regulated by the i-th gate; Z(i) min is the lower limit water level regulated by the i-th gate; B(j) max is the upper limit water level regulated by the j-th pumping station; B(j) min is the lower limit water level regulated by the j-th pumping station; H(i) 警戒水位 、H(i) 生态水位 are the warning water level and ecological water level of the regulation reference section of the i-th gate; H(j) 警戒水位 is the warning water level of the regulation reference section of the j-th pumping station; Δh(j) is the regulation threshold of the regulation reference section of the j-th pumping station.

[0060] Preferably, the real-time control rules of the water project input into the urban hydrological-hydrodynamic coupling model in S6 are expressed as follows:

[0061] S61. The gate regulation rule is as follows:

[0062] If NODE A < Z(i) min ;

[0063] THEN ORIFICE A = 0;

[0064] If NODE A > Z(i) max ;

[0065] THEN ORIFICE A = 1.0;

[0066] Where NODE A is the ID of the regulation reference section of the gate; 0.0 indicates that the gate is closed, and 1.0 indicates that the gate is fully open;

[0067] S62. The pump station regulation rule:

[0068] If NODE A < B(j) min ;

[0069] THEN PUMP ASTATUS = OFF;

[0070] If NODE A > B(j) max ;

[0071] THEN PUMP ASTATUS = ON;

[0072] Where NODE A is the ID of the regulation reference section of the pump station; OFF indicates that the pump station is closed, and ON indicates that the pump station is open;

[0073] S63. Using the performance of the rainstorm event assessment method, the indicators for quantitatively analyzing the response of flood control to the gate-pump control rules are the number of stormwater well overflows and the water level process at the key river cross-sections.

[0074] Preferably, in S1, the collected river network data includes: the river network topology relationship, cross-section data, underlying surface, gate and pump station data required for constructing the river network water system. The gate and pump station data include design parameters and distribution locations, measured rainfall data, measured river network water levels corresponding to the measured rainfall data, and flow data.

[0075] Advantages of the present invention:

[0076] 1. In view of the problem that the response of the sluice pumps control rules in the urban internal river network is relatively lagging, the present invention constructs an urban hydrological and hydrodynamic model, identifies the backwater drainage outlets at the regional top with different designed rainstorms, clusters the characteristics of the backwater drainage outlets based on the clustering algorithm to determine the river network regulation points, takes the river network regulation points as the new regulation reference objects to optimize the real-time control rules of the river network, ensures that the urban internal drainage outlets are not backwatered by the river network water level, improves the response speed of the internal river network to the real-time control rules, improves the efficiency and accuracy of the regulation of the existing river network water projects, takes into account the ecological environment requirements while ensuring flood control safety, and has good real-time performance and applicability.

[0077] 2. The present invention solves the problem that the traditional real-time control technology of water projects is difficult to achieve the precise regulation of the urban internal river network and is not conducive to the scientific management of the urban river network, and can regulate the river network water level during rainstorms to alleviate the flood risk. BRIEF DESCRIPTION OF THE DRAWINGS

[0078] Figure 1 It is a schematic diagram of the research area in Cangshan District, Fuzhou City;

[0079] Figure 2 It is a statistical chart of the number of backwater drainage outlets;

[0080] Figure 3 It is a distribution map of the positions of the river network regulation points;

[0081] Figure 4 It is the rainstorm and tidal process of Typhoon Soudelor;

[0082] Figure 5 It is a comparison chart of the water head processes of the river network regulation points between the original sluice pump rules and the optimized sluice pump rules;

[0083] Figure 6 It is a comparison of the simulated and observed water level processes of the rainwater wells and river cross-section samples. DETAILED DESCRIPTION OF THE INVENTION

[0084] The present invention will be further described in detail below with reference to the drawings and specific embodiments.

[0085] Embodiment 1:

[0086] A certain area in Cangshan District, Fuzhou City, Fujian Province has an area of 39.17 km 2 , and the rainfall is concentrated from April to October in the main flood season. The internal river network in the area is rich in water systems, with 13 rivers and a total length of 47.69 km. The north and south sides of the area are surrounded by branches of the Minjiang River, belonging to a two-way tidal area, and the river network water level is easily backwatered by the rising tide level. Fuzhou has been hit by typhoons many times, including but not limited to "Haitang", "Longwang", and "Megi". Typhoon weather and heavy rainfall have led to frequent flood disasters in the research area.

[0087] A method for optimizing real-time control rules of river networks for urban flood control. The method steps for substituting data are as follows:

[0088] S1: Collect river network data. The collected river network data includes the river network topology required for constructing the river network water system, 175 river reach data, 7 sluices (including Changpu Sluice, Shengli Sluice, and Jiangbian Sluice located on the north side, with a bottom elevation of 2m for all, and Yangqi Regulating Sluice, Jimuyu Sluice, Wufeng Sluice, and Yixu Sluice located on the south side, with bottom elevations of 2m, 1.8m, 2m, and 1.5m respectively), 4 pumping stations (including Changpu Pumping Station and Shengli Pumping Station located on the north side, with scales of 6m 3 / s and 16m 3 / s respectively, and Yangqi Pumping Station and Yixu Pumping Station located on the south side, with scales of 8m 3 / s for both), and the total duration of the actual typhoon rainstorm event "Soudelor" is 3360 min, with a rainfall of 408 mm.

[0089] S2: Construct an urban hydrological model according to detailed spatial division and different surface runoff characteristics, construct a river network hydrodynamic model considering the hydraulic conditions of facilities such as pumping stations, gates, weir flows, and reservoir lakes, construct a pipe network hydrodynamic model considering the hydraulic conditions of facilities such as inspection wells and outfalls, and couple the above models through the process of water volume exchange at cross-section nodes based on Infoworks ICM software to construct an urban hydrological-hydrodynamic coupling model. The regional hydrological model consists of multiple sub-catchments, and the parameter settings of each sub-catchment are different. Therefore, only the parameter value ranges or a certain sub-catchment and river cross-section are shown for illustration below. The specific construction process is as follows:

[0090] Taking sub-catchment Y3501540059 as an example to illustrate the calculation process of the hydrological model.

[0091] S21, The runoff calculation for impervious surfaces such as houses and roads uses the fixed runoff coefficient model in InfoWorks ICM, as shown in the following formula:

[0092] R n =CR P

[0093] where R n is the net rainfall, in mm; P is the recurrence interval, in years. The following involves 9 recurrence intervals of 1a, 2a, 3a, 5a, 10a, 20a, 30a, 50a, and 100a; C is the runoff coefficient, with an empirical value of 0.95; R P is the rainfall corresponding to the corresponding recurrence interval, in mm; The 2h net rainfall for different rainfall recurrence intervals is shown in Table 1.

[0094] Table 1 2h Net Rainfall Table for Different Rainfall Recurrence Intervals

[0095]

[0096] S22. For the runoff calculation of the pervious surface of the green space, the Horton infiltration model is adopted as follows:

[0097] f t = f c +(f0 - f c )e -kt

[0098] where f t is the infiltration rate at time t, in mm / h; f c is the steady infiltration rate, with an empirical value range of 2.5 - 5.3 mm / h; f0 is the initial infiltration rate, with an empirical value range of 55 - 76 mm / h; k is the reduction coefficient, with a dimension of h -1 , and the empirical value is 3.79 - 5.29; t is the infiltration time, in h. Taking the 2-hour 20-year return period design rainfall as an example, the infiltration rate at t = 1 h is 2.5+(76 - 2.5)e -3.8×1 = 4.18 mm / h.

[0099] S23. To simulate the surface runoff in the city, the nonlinear reservoir model in the stormwater management module is called and solved jointly by the continuity equation and the Manning formula.

[0100] Manning formula:

[0101]

[0102] where Q is the outflow, in m 3 / s; W is the overland flow width of the sub-catchment, with a value of 34.6 m; n is the surface Manning coefficient, with a value of 0.03; d is the water depth, with an initial value of 0.05 m; d p is the maximum depression storage depth, with a value of 5 m; S is the average slope of the sub-catchment, with a value of 0.003.

[0103] Continuity equation:

[0104]

[0105] where V is the surface water accumulation, in m 3 ; t is the time, in s; A1 is the surface area, with a value of 36670 m 2 ; i* is the net rainfall, and the net rainfall for the 2-hour 20-year return period is 0.0118 mm / s.

[0106] Substitute the relevant parameter values of the sub-catchment Y3501540059 to obtain the outflow Q, and dynamically update Q using the Manning formula to obtain the change in surface water accumulation over time, that is, the runoff. The flow process of this sub-catchment is shown in Table 2.

[0107] Table 2 Flow process of sub-catchment Y3501540059 under the design rainfall with a return period of 20a for 2 hours

[0108]

[0109] Taking the yjh-wmh-2_4 section of the river reach yqh-wmh-2_us.1 as an example, the calculation process of the hydrodynamic model is described.

[0110] S24. The hydrodynamic module in Infoworks ICM is adopted, and the river network runoff is calculated by completely solving the Saint-Venant equations using the dynamic wave.

[0111] Continuity equation:

[0112]

[0113] Momentum equation:

[0114]

[0115] Where Q is the flow rate, m 3 / s; A2 is the cross-sectional area, with a value of 76.824 m 2 ; g is the acceleration due to gravity, 9.8 m / s 2 ; θ is the angle between the center line of the pipeline and the horizontal line, with a value of 0 degrees; K is the water conveyance rate, with a value of 10, and S0 is the bottom slope of the channel, with a value of 0.002.

[0116] S25. The surface runoff and river network water flow are coupled by means of water volume exchange at the cross-sectional nodes, as follows:

[0117]

[0118] Where is the flow rate of the i-th river network cross-section at time t, is the flow rate of the i-th river network cross-section at time t-1, is the surface runoff corresponding to the i-th river network cross-section at time t.

[0119] Substituting the relevant parameter values, the flow process of the river channel cross-section yjh-wmh-2_4 is obtained as shown in Table 3.

[0120] Table 3 Flow process of the river channel cross-section yjh-wmh-2_4

[0121]

[0122]

[0123] S3: Use the model to simulate and calculate the flood processes under 9 scenarios with recurrence intervals of 1a, 2a, 3a, 5a, 10a, 20a, 30a, 50a, and 100a, comprehensively identify the drainage outlets vulnerable to backwater from river network water levels within the study area, and the statistical data are shown in Figure 2 , and the following is the rainstorm intensity formula used:

[0124]

[0125] where q is the average rainstorm intensity, in mm / min; P is the recurrence interval, in years; t is the rainfall duration, set to 120 min. The Chicago rain pattern is used to calculate the rainfall process, with the design rain peak coefficient r = 0.39, and the time interval of the rainfall process set to 1 min. The rainfall intensities obtained by calculating for 1-year, 2-year, 3-year, 5-year, 10-year, 20-year, 30-year, 50-year, and 100-year recurrence intervals are shown in Table 4, and only the first 15 min are presented.

[0126] Table 4 Rainfall Intensity Table for Design Rainfall Scenarios

[0127]

[0128] S4: Use the clustering algorithm to conduct spatial clustering of the backwater drainage outlets to divide the regulation areas. Based on the spatial clustering, conduct feature clustering of the backwater drainage outlets again, and output the clustering center with the greatest influence as the river network regulation point. There are 198 drainage outlets in the entire study area. Under 9 design rainstorm scenarios, a total of 105 are affected by the backwater of the river network water level, and there are 78 river cross-sections where backwater occurs.

[0129] The AGNES algorithm is used for spatial clustering, which belongs to the bottom-up agglomerative hierarchical clustering. The calculation process is as follows:

[0130] S41, Initialization: Each sample point forms a separate cluster. There are 78 samples in total, so there are 78 clusters in the initial state;

[0131] S42, Calculate the distance matrix: Initially, use the selected distance metric. In this paper, the Euclidean distance is selected to calculate the distance matrix between all clusters. The calculation method of the Euclidean distance between two n-dimensional vectors is as follows:

[0132]

[0133] where d is the Euclidean distance; x 1i and x 2i are any two n-dimensional vectors. Taking point 04047_2630 and point 04001_2490 as examples, x 1i and x 2i represent the spatial coordinates of the points, which are two two-dimensional vectors, x 1iThe coordinates are (431781.408, 2880105.636), x 2i The coordinates are (433116.133, 2878371.127), then the distance between the two is:

[0134]

[0135] S43, Merge the nearest clusters: Find the two nearest clusters, merge them into a new cluster, and update the distance matrix to reflect the distances between the new cluster and other clusters. The merging process is based on the selected linkage method. In this paper, the Ward method is selected to minimize the within-cluster sum of squares after merging:

[0136]

[0137] Where C i and C j are two clusters to be merged; ∣C i ∣ and ∣C j ∣ are the numbers of data points contained in clusters C i and C j respectively; μ i and μ j are the mean vectors of clusters C i and C j respectively; d(μ i , μ j ) is the Euclidean distance between the mean vectors of the two clusters; d ward (C i , C j ) is the "merging cost" or "distance" between the two clusters calculated based on the Ward method. There are currently two clusters C1 and C2 to be merged, with 15 and 8 sample points in the clusters respectively, and the corresponding mean vectors μ i = (4, 6), μ j = (7, 2), Euclidean distance The "merging cost" between the two clusters

[0138] S44, Repeat step S43 until all sample points are merged into one cluster or the predetermined number of clusters is reached. The regulation area is partitioned through spatial clustering, and the final clustering result is 3, divided into the north area, the southwest area, and the southeast area;

[0139] The K-Medoids algorithm is used for feature clustering. Clustering is performed based on the spatial clustering results. Taking the spatial clustering clusters in the north area as an example, the calculation process is described as follows:

[0140] S46, Initialization: Randomly select 9 points from 39 sample points in the north area as the initial clustering centers;

[0141] S47, Cluster Assignment: Assign the remaining 30 points to the cluster where the nearest cluster center is located according to the principle of minimum distance to achieve initial clustering. The distance between each sample point is calculated using the Euclidean distance, and the calculation formula is the same as that in S42;

[0142] S48, Update Cluster Center: Calculate the distances from the points in the cluster except the cluster center to all other points, and take the point corresponding to the minimum value as the new cluster center, and recalculate the clustering result;

[0143] S49, Repeat the processes of S47 and S48 until all cluster centers no longer change or the set maximum number of iterations is reached. Finally, the clustering result is 4, and the cluster center with the greatest influence is output as the river network regulation point. The corresponding river channel section ID is 04047_2630. Using the same method, the river network regulation point IDs for the southwest area and the southeast area are 04053_1500 and 04001_2490 respectively, as shown in Figure 3 。

[0144] S5: Taking the river network regulation point as the regulation reference section, write the optimized gate-pump control rules based on the warning water level and ecological water level data of the river channel:

[0145] Z(i) min = H(i) 生态水位

[0146] Z(i) max = H(i) 警戒水位

[0147] B(j) min = H(j) 警戒水位

[0148] B(j) max = H(j) 警戒水位 + Δh(j)

[0149] where Z(i) max is the upper limit water level regulated by the i-th gate; Z(i) min is the lower limit water level regulated by the i-th gate; B(j) max is the upper limit water level regulated by the j-th pump station; B(j) min is the lower limit water level regulated by the j-th pump station; H(i) 警戒水位 、H(i) 生态水位 are the warning water level and ecological water level of the regulation reference section of the i-th gate; H(j) 警戒水位 is the warning water level of the regulation reference section of the j-th pump station; Δh(j) is the regulation threshold of the regulation reference section of the j-th pump station, and the empirical value is 0.3m.

[0150] The ecological water levels and warning water levels of river cross-section 04047_2630 serve as the regulation references for the calamus sluice, victory sluice, riverside sluice, calamus pumping station, and victory pumping station in the northern area. The ecological water levels and warning water levels of river cross-section 04053_1500 serve as the regulation references for the Yangqi regulating sluice, Jimuyu sluice, and Yangqi pumping station in the southwest area. The ecological water levels and warning water levels of river cross-section 04001_2490 serve as the regulation references for the Wufeng sluice, Yixu sluice, and Yixu pumping station in the southeast area. The warning water level and ecological water level data of the regulation reference cross-sections provided by relevant units are shown in Table 5, and the optimized gate-pump control rules are shown in Table 6.

[0151] Table 5 Ecological water levels and warning water levels corresponding to the regulation reference cross-sections

[0152]

[0153] Table 6 Optimized gate-pump control rules

[0154]

[0155] S6: Input the optimized gate-pump control rules into the urban hydrological-hydrodynamic coupling model to evaluate the performance of the rainstorm event assessment method and quantitatively analyze the response of flood control to the gate-pump control rules.

[0156] The real-time control rules of water projects input into the urban hydrological-hydrodynamic coupling model are expressed as:

[0157] S61, the gate regulation rule is:

[0158] If NODE A < Z(i) min

[0159] THEN ORIFICE A = 0

[0160] If NODE A > Z(i) max

[0161] THEN ORIFICE A = 1.0

[0162] Where NODE A is the ID of the regulation reference cross-section of the gate; 0.0 indicates that the gate is closed, and 1.0 indicates that the gate is fully open.

[0163] S62, the pumping station regulation rule:

[0164] If NODE A < B(j) min

[0165] THEN PUMP ASTATUS = OFF

[0166] If NODE A > B(j) max

[0167] THEN PUMP A STATUS=ON

[0168] Among them, NODE A is the ID of the regulation reference section of the pump station; OFF indicates that the pump station is closed, and ON indicates that the pump station is open;

[0169] S63. To evaluate the performance of the rainstorm event assessment method, the indicators for quantitatively analyzing the response of flood control to the gate-pump control rules are the number of overflows in the stormwater wells and the water level process at the key cross-sections of the river channels;

[0170] Considering under the regulation of the original gate-pump control rules (Table 7) (denoted as Scenario A), the urban hydrological-hydrodynamic coupling model is used to simulate the flood process of the "Soudelor" typhoon rainstorm event ( Figure 4 ). The inundation depth of the stormwater well refers to the difference between the water level of the stormwater well and the ground elevation where the stormwater well is located. Under the regulation of the original gate-pump control rules, the number of overflows in the stormwater wells is 2203, and the overflow ratio is 43.15%. The maximum water heads at the three river network regulation points are 6.475m, 7.193m, and 5.935m respectively.

[0171] Table 7 Original gate-pump control rules

[0172]

[0173] Furthermore, considering under the regulation of the optimized gate-pump control rules (Table 3) (Scenario B), the urban hydrological-hydrodynamic coupling model is used to simulate the flood process of the "Soudelor" typhoon rainstorm event. Compared with Scenario A, the number of overflows in the stormwater wells in Scenario B is 2125, and the overflow ratio is 41.63%. The overflow quantity has decreased by 1.52%, and the overflow of the drainage system has been improved; the maximum water heads at the three river network regulation points are 6.367m, 6.726m, and 5.866m respectively, with a decrease of 1.67%, 6.49%, and 1.16%, as shown in Figure 5 .

[0174] It can be seen from this that the optimized gate-pump control rules during the rainstorm play an important role in flood control regulation, the water head of the river channel is significantly reduced, the overflow of the drainage system around the river network is significantly reduced, and the flood disaster is alleviated.

[0175] Example 2:

[0176] The urban hydro - hydrological dynamic coupling model was calibrated and verified using two heavy rain events, from 14:00 on March 26 to 4:00 on March 27, 2022 and from 0:00 to 12:00 on April 27, 2022, denoted as Event 1 and Event 2 respectively. Event 1 was used to calibrate the model, and Event 2 was used to verify the model. The rainfall amounts of Event 1 and Event 2 were 68.3 mm and 126.0 mm respectively. Three stormwater wells (M1, M2, M3) and three river cross - sections (S1, S2, S3) were randomly selected from the observation points as samples. The average NSE of the sample water heads under Event 1 and Event 2 was higher than 0.85 and 0.79 respectively, and the model simulation effect was good. The values of the calibrated model parameters are shown in Table 8. The comparison of the simulated water heads and the observed water heads of the stormwater well samples and river cross - section samples during the rainfall of Event 1 can be seen in Figure 6 , as can be seen from the figure, there were certain fluctuations in the simulated water head process of some stormwater well samples. For example, for stormwater well 1, it was mainly manifested in the advance of the water head peak and the larger amplitude of the water head change. However, the errors between the simulated water heads and the measured water heads at the three nodes were all within the allowable range; the simulated water head process of the river cross - section samples was basically consistent with the measured water head process, and the overall simulation accuracy was higher than that of the stormwater well samples. The simulated water head processes of the selected nodes and cross - sections were basically consistent with the measured water head processes, indicating that the model could largely reflect the actual situation and could be used for river network regulation calculation and analysis.

[0177] Table 8 Range of model parameter values

[0178]

[0179] The above - mentioned embodiments are only the preferred technical solutions of the present invention and should not be regarded as limitations on the present invention. The protection scope of the present invention should be the technical solutions recorded in the claims, including the equivalent replacement solutions of the technical features in the technical solutions recorded in the claims. That is, the equivalent replacement improvements within this scope are also within the protection scope of the present invention.

Claims

1. A real-time control rule optimization method for urban river networks facing urban floods, characterized in that: Including the following steps: S1. Collect river network data; S2. Construct an urban hydrological model according to detailed spatial division and different surface runoff characteristics. Consider the hydraulic conditions of pumping stations, gates, weir flows, and reservoir-lake facilities to construct a river network hydrodynamic model. Consider the hydraulic conditions of inspection wells and outfall facilities to construct a pipe network hydrodynamic model. Couple the above models through the process of sectional node water volume exchange to construct an urban hydrological-hydrodynamic coupling model; S3. Simulate and calculate the flood process under different design rainstorm scenarios according to the rainstorm intensity formula, and comprehensively identify the drainage outlets vulnerable to river network water level backwater in the study area; S4. Use the clustering algorithm to conduct spatial clustering division of the backwater drainage outlets to regulate the partition. Based on the spatial clustering, conduct characteristic clustering of the backwater drainage outlets again, and output the maximum influence clustering center as the river network regulation point; S5. Take the river network regulation point as the regulation reference object, and write an optimized gate-pump control rule based on the warning water level and ecological water level data of the river channel; S6. Input the optimized gate-pump control rule into the urban hydrological-hydrodynamic coupling model, evaluate the performance by the rainstorm event assessment method, and quantitatively analyze the response of the flood to the gate-pump control rule.

2. The real-time control rule optimization method for river networks facing urban floods according to claim 1, characterized in that: The urban hydrological-hydrodynamic coupling model in S2 is constructed based on the Infoworks ICM software, and the specific construction process is as follows: S21. The runoff calculation of impervious surfaces of houses and roads adopts the fixed runoff coefficient model in InfoWorks ICM, as shown in the following formula: R n = CR P ; where R n is the net rainfall, in mm; C is the runoff coefficient; P is the recurrence interval, in years; R P is the rainfall corresponding to the recurrence interval, in mm; S22. The runoff calculation of pervious surfaces of green spaces adopts the Horton infiltration model, as shown in the following formula: f t = f c + (f0 - f c ) e -kt ; where f t is the infiltration rate at time t, in mm / h; f c is the steady infiltration rate, in mm / h; f0 is the initial infiltration rate, in mm / h; k is the reduction coefficient, with the dimension of h -1 ; t is the infiltration time, in h; S23. To simulate the urban surface runoff concentration, call the nonlinear reservoir model in the stormwater management module, and solve it jointly by the continuity equation and the Manning formula, as shown in the following formula: Manning formula: where Q is the flow rate, m 3 / s; W is the overland flow width of the sub-catchment, m; n is the surface Manning coefficient; d is the water depth, m; d p is the maximum depression storage depth, m; S is the average slope of the sub-catchment; Continuity equation: where V is the surface water accumulation, m 3 ; t is the time, s; A1 is the surface area, m 2 ; i* is the net rainfall, mm / s; S24. Adopt the hydrodynamic module in Infoworks ICM to calculate the river network runoff concentration by completely solving the Saint-Venant equations using the dynamic wave, as shown in the following formula: Continuity equation: Momentum equation: where A2 is the cross-sectional area, m 2 ; Q is the flow rate, m 3 / s; g is the acceleration due to gravity, m / s 2 ; θ is the angle between the pipeline centerline and the horizontal line, degree; K is the water conveyance rate; h is the water depth, m; t is the time, s; S0 is the bottom slope of the channel; S25. Couple the surface runoff and river network water flow by means of sectional node water volume exchange.

3. The real-time control rule optimization method for river networks facing urban floods according to claim 2, characterized in that: In S25, the formula for coupling the surface runoff and river network water flow by means of sectional node water volume exchange is as follows: wherein is the flow rate of the i-th river network section at time t, is the flow rate of the i-th river network section at time t-1, is the surface runoff corresponding to the i-th river network section at time t.

4. The real-time control rule optimization method for river networks facing urban floods according to claim 1, characterized in that: The rainstorm intensity formula in S3 is: Where q is the average rainstorm intensity, mm / min; P is the recurrence interval, a; t is the rainfall duration, min; The Chicago rainfall pattern is used to calculate the rainfall process, taking the design rain peak coefficient r = 0.39, the rainfall process time interval is set to 1 min, and the rainfall time is set to 120 min.

5. An optimization method for real-time control rules of river networks for urban flood control according to claim 1, characterized in that: In S4, the AGNES algorithm is used for spatial clustering of the backwater drainage outlets, which belongs to the bottom-up agglomerative hierarchical clustering, and the calculation process is as follows: S41. Initialization: Each sample point forms a cluster by itself; If there are m samples in the population, then there are m clusters in the initial state; S42. Calculate the distance matrix: Initially, use the selected distance metric, and choose the Euclidean distance to calculate the distance matrix between all clusters; S43. Merge the nearest clusters: Find the two nearest clusters, merge them into a new cluster, and update the distance matrix to reflect the distance between the new cluster and other clusters; The merging process is based on the selected linkage method. The Ward method is selected to minimize the within-cluster sum of squares after merging: Among them, C i and C j are two clusters to be merged; |C i | and |C j | are the numbers of data points contained in clusters C i and C j respectively; μ i and μ j are the mean vectors of clusters C i and C j respectively; d(μ i , μ j ) is the distance between the mean vectors of the two clusters; d ward (C i , C j ) is the "merging cost" or "distance" between the two clusters calculated based on the Ward method; S44. Repeat step S43 until all sample points are merged into one cluster or the predetermined number of clusters is reached; S45. Generate a hierarchical tree: Represent the clustering process using a dendrogram. Each merge corresponds to a node on the tree.

6. The real-time control rule optimization method for river networks facing urban floods according to claim 5, characterized in that: In S42, the calculation method of the Euclidean distance between two n-dimensional vectors is as follows: where d is the Euclidean distance; x 1i and x 2i are any two n-dimensional vectors.

7. An optimization method for real-time control rules of river networks for urban floods according to claim 1, characterized in that: In S4, the K-Medoids algorithm is used for feature clustering of the anti-carry drainage outlets. It belongs to the partitioning clustering that directly divides data points into different clusters. The calculation process is as follows: S46. Initialization: Randomly select k points from the total m sample points as the initial clustering centers; S47. Cluster assignment: Assign the remaining m - k points to the cluster where the nearest clustering center is located according to the minimum distance principle to achieve initial clustering; The distance between each sample point is calculated using the Euclidean distance; S48. Update the clustering centers: Calculate the distances from the points in the cluster except the clustering centers to all other points, and select the point corresponding to the minimum value as the new clustering center, and recalculate the clustering result; S49. Repeat processes S47 and S48 until all clustering centers no longer change or the set maximum number of iterations is reached to obtain the final clustering result.

8. An optimization method for real-time control rules of river networks for urban flood control according to claim 1, characterized in that: In S5, the basis for writing the real-time control rules of water projects is the warning water level and ecological water level data of the river channel, specifically as follows: Z(i) min = H(i) 生态水位 ; Z(i) max = H(i) 警戒水位 ; B(j) min = H(j) 警戒水位 ; B(j) max = H(j) 警戒水位 + Δh(j); where Z(i) max is the upper limit water level regulated by the i-th gate; Z(i) min is the lower limit water level regulated by the i-th gate; B(j) max is the upper limit water level regulated by the j-th pumping station; B(j) min is the lower limit water level regulated by the j-th pumping station; H(i) 警戒水位 、H(i) 生态水位 are the warning water level and ecological water level of the regulation reference section of the i-th gate; H(j) 警戒水位 is the warning water level of the regulation reference section of the j-th pumping station; Δh(j) is the regulation threshold of the regulation reference section of the j-th pumping station.

9. An optimization method for real-time control rules of river networks for urban flood control according to claim 1, characterized in that: In S6, the real-time control rules of water projects input into the urban hydrological-hydrodynamic coupling model are expressed as: S61. The gate regulation rule is: If NODE A<Z(i) min ; THEN ORIFICE A = 0; If NODE A>Z(i) max ; THEN ORIFICE A = 1.0; Where NODE A is the ID of the regulation reference section of the gate; 0.0 means the gate is closed, and 1.0 means the gate is fully open; S62. The pump station regulation rule: If NODE A<B(j) min ; THEN PUMP ASTATUS = OFF; If NODE A>B(j) max ; THEN PUMP A STATUS = ON; Where NODE A is the ID of the regulation reference section of the pump station; OFF means the pump station is closed, and ON means the pump station is open; S63. Evaluate the performance of the rainstorm event assessment method. The indicators for quantitatively analyzing the response of flood control to the gate and pump control rules are the number of rainwater well overflows and the water level process of the key river channel sections.

10. An optimization method for real-time control rules of river networks for urban flood control according to claim 1, characterized in that: In S1, the collected river network data includes: the river network topological relationship, cross-section data, underlying surface, gate and pump station data required for constructing the river network water system. The gate and pump station data include design parameters and distribution locations, measured rainfall data, measured river network water levels corresponding to the measured rainfall data, and flow data.

Images