Grey-green-blue measure multi-target segmentation optimization method for coupling wild dog optimization algorithm with SWMM

Abstract

The invention provides a multi-target segmentation optimization method for grey green blue measures by coupling a wild dog optimization algorithm with an SWMM, and relates to the technical field of urban inland inundation treatment. According to the method, an optimization model taking the minimization of the total investment cost and the maximization of the runoff reduction rate as targets is constructed by analyzing the accumulated water distribution and determining the measure type, scale and cost. Decision variables comprise pump station drainage flow and the like, the flow is analyzed by adopting a period-by-period water regulation and storage method, and a variable relation is established through multiple regression. In the first round, a Pareto solution set is generated by using a wild dog algorithm, and optimization is carried out by combining an SWMM model, so that a Pareto leading solution set suitable for a data-lacking region is obtained. The urban toughness can be improved.

Description

Technical Field

[0001] The invention relates to the technical field of urban waterlogging control, and in particular to a gray-green-blue measure multi-objective segmented optimization method of a wild dog optimization algorithm coupled with a SWMM. Background Art

[0002] In recent years, climate warming has led to changes in global precipitation patterns, rapid urbanization has changed the runoff infiltration process, and the combined effects of problems such as mismatched drainage system design have made urban waterlogging a challenge that needs to be addressed globally. Gray-green-blue measures are particularly important for urban waterlogging management, but due to the complexity of the operating mechanism and the uncertainty of the synergistic effect, gray measures, especially pumping stations, are not used as decision variables in traditional multi-objective optimization models. At present, research on urban drainage planning at the basin scale lacks systematicity, and does not fully consider the synergistic effects of pumping station drainage flow, regional water surface rate, and green measure layout rate on the runoff production and convergence process. How to achieve the organic combination of gray-green-blue measures and quantify the synergistic effect is still a problem that needs to be solved.

[0003] The gray-green-blue hybrid infrastructure framework involves multiple factors. Under the complex multi-dimensional coupling relationship, the effective allocation of resources is crucial. The location and scale of green measures and storage pump stations should be reasonably arranged to maximize the benefits of water management. Therefore, how to determine the optimal ratio of gray-green-blue measures depends on the balance between multiple objectives. The integrated hydrological model and optimization algorithm analyze the synergistic effect mechanism of gray-green-blue measures and find the optimal spatial combination solution. However, it is also challenging to obtain and process various high-precision basic data required for stormwater models. Summary of the invention

[0004] In order to overcome the shortcomings of the prior art, the purpose of the present invention is to provide a multi-objective segmented optimization method of gray-green-blue measures of wild dog optimization algorithm coupled with SWMM, which is applicable to both areas with insufficient data and areas with modeling conditions. The present invention reduces the constraint range of decision variables through segmented optimization strategy and seeks the optimal spatial layout solution. It provides a new idea for solving the optimization of drainage in areas with insufficient data, and at the same time, provides better technical support for improving urban resilience.

[0005] To achieve the above object, the present invention provides the following solutions: A multi-objective segmented optimization method of gray-green-blue measures of wild dog optimization algorithm coupled with SWMM, including: Determine the current status of waterlogging distribution and simulation results based on the collected and collated basic data of the study area, and determine the type, scale, parameters and unit investment cost of the gray-green-blue measures; Optimize the type and layout of the gray-green-blue measures based on the current water distribution status; A multi-objective optimization model was constructed, with the minimization of total investment cost and the maximization of total runoff reduction rate as the objective function, the drainage flow of the pump station, the water surface ratio, the sunken green space ratio and the permeable pavement ratio as the decision variables, and the constraint conditions were set by combining the area constraint, the effectiveness constraint and the fitting relationship between the decision variables. Based on the gray-green-blue measures and the multi-objective optimization model, the water storage calculation method for each period is used to analyze the drainage flow of the pump station under different working conditions, and the mathematical relationship expression between the decision variables is established as an equality constraint through multivariate regression analysis; The multi-objective optimization model is optimized for the first iteration using the wild dog optimization algorithm to generate the Pareto solution set and the range of decision variables; Building a SWMM model based on the basic data, and using historical rainfall data for calibration and verification to ensure the accuracy of the SWMM model; The SWMM model is coupled with the Dingo optimization algorithm, and the objective function and constraints are updated based on the range of decision variables of the first optimization. The second optimization is performed to generate the Pareto front solution set of the spatial layout of the gray-green-blue measures.

[0006] Preferably, the basic data includes: rainfall data, land use type, drainage network information, pump station parameters, elevation data and river water level data.

[0007] Preferably, the types and layouts of the gray-green-blue measures are optimized based on the current water distribution status, including: Deploy storage pump stations near rivers and flood-prone areas as gray measures; According to the suitability of the land use type, sunken green spaces are laid out in ordinary green areas, and permeable pavements are rebuilt in residential areas, commercial areas and road areas as green measures; The blue measure is to expand the water surface rate of the original rivers and lakes in the study area.

[0008] Preferably, the decision variables of the multi-objective optimization model are: ;in, X A vector representing the decision variables; q represents the flow rate of the drainage pumping station in the study area; p i Representative i Water surface ratio of water body type; x j Representative j The area ratio of green measures, m Represents the number of water body types, n represents the number of green measures; The calculation formula of the objective function is: ;in, f 1 represents the total investment cost,A It is the sum of the present value of investment per unit flow and the present value of operation and management of the drainage pump station; B i For each cubic meter of excavation i the cost of planting water bodies; h i For the i The depth of the water body; F is the area of ​​the study area; P 0,i It is i Current water surface rate; C j For the j The total construction cost and maintenance and operation cost of the LID measures; f 2 It represents the total runoff reduction rate; α A is the runoff coefficient before any measures are implemented; α is the runoff coefficient after the implementation of blue-green measures, △t It is the time for the pump station to drain water; The constraint condition is expressed as: ;in, P i * For the i Minimum water surface ratio of the water body; x i * For the i Minimum deployment area ratio of green measures; P is the rainfall, β is the percentage of discharge flow from the pump station; T Design duration for drainage.

[0009] Preferably, the formula for calculating the water storage regulation in each period is: in, V 1 , V 2 are the river water storage at the beginning and end of the time period respectively; Q 1 , Q 2 are the inflow water volume at the beginning and end of the time period respectively; P 1 , P 2 are the rainfall in each period, φ is the rainfall runoff coefficient for permeable pavement; Δt is the time interval between the beginning and end of the period, F green ,F harden They are the area of ​​green space and the area of ​​hardened ground in the study area respectively.

[0010] Preferably, the wild dog optimization algorithm comprises: Initialize the population and generate random solutions; Evaluate fitness through non-dominated sorting and select the optimal solution; Simulate wild dog hunting behavior, including siege, pursuit, scavenging and survival rules, and update the solution position; Iterate until the termination condition is met and output the Pareto solution set.

[0011] Preferably, a SWMM model is constructed based on the basic data, and historical rainfall data is used for calibration and verification to ensure the accuracy of the SWMM model, including: ArcGIS was used to convert the CAD files of different land use types in the study area into SHP files. The study area was divided into several sub-catchments, and the roads were divided separately, taking into account the elevation changes, land use types, street distribution, and stormwater wells in the study area. ArcGIS was used to split and merge the drainage pipelines in the study area, divide the drainage zones according to the north-south direction of the drainage main pipeline, export the SHP file, and convert the SHP file into an inp file that can be recognized by SWMM through the inpPINS plug-in; The parameters in the SWMM model are divided into deterministic parameters and calibrated parameters; the deterministic parameters include the characteristic width, area, slope, permeability, pipe diameter, burial depth, and rainwater well depth of the sub-catchment area; the area of ​​each sub-catchment area is statistically calculated through the "Computational Geometry" function in ArcGIS, and the area proportion of different land use types is calculated; the elevation data is converted and the average slope is calculated through the 3DAnalyst and Spatial Analyst functions; the calibrated parameters are determined by referring to the SWMM model manual and the parameter value range of the study area; The model parameters of the SWMM model are preliminarily adjusted using the comprehensive runoff coefficient, the parameter values ​​of different regions are selected according to the empirical range, and the accuracy of the SWMM model is improved through multiple adjustments; The measured data of two historical rainfall events were used for calibration and verification. The maximum overflow of the simulated flood-prone point was evenly distributed over the road area. The maximum simulated surface water depth was calculated and the relative error between the simulated value and the measured water depth was calculated to verify the accuracy of the SWMM model. The formula for the relative error is as follows: ; In the formula, H 1 is the measured value of water accumulation at the monitoring point; H 2 It is the simulated value of water accumulation at the monitoring point.

[0012] Preferably, the SWMM model is coupled with the wild dog optimization algorithm, and the objective function and constraints are updated based on the range of decision variables of the first optimization, and a second optimization is performed to generate a Pareto front solution set of the spatial layout of the gray-green-blue measures, including: Based on the range of decision variables under the corresponding scheme of the first optimized Pareto front, the average drainage flow of all schemes and the average water surface rate are used as known variables to update the multi-objective optimization function, and the constraint conditions are updated with the extreme values ​​of the sunken green space rate and the permeable pavement rate; The updated optimization model of multi-objective function is constructed through Python, and the iteration is realized by the wild dog optimization algorithm; The SWMM model was automatically established through Python, and the wild dog optimization algorithm was coupled to achieve automatic optimization of the sunken green space and permeable pavement layout area; Based on the obtained optimization model of multi-objective functions, the wild dog optimization algorithm is adopted for iteration to obtain the Pareto frontier of the second optimization and determine the solution set of spatial layout under green measures.

[0013] A multi-objective segmented optimization system of gray-green-blue measures coupled with SWMM by wild dog optimization algorithm, including: The data collection unit is used to determine the current status of waterlogging distribution and simulation results based on the basic data of the study area collected and sorted, and to determine the type, scale, parameters and unit investment cost of the gray-green-blue measures; A measure optimization unit, used for optimizing the type and layout of the gray-green-blue measures based on the current water distribution status; The target construction unit is used to construct a multi-objective optimization model, with the minimization of total investment cost and the maximization of the total runoff reduction rate as the objective function, the pump station drainage flow, water surface ratio, sunken green space ratio and permeable pavement ratio as decision variables, and the constraint conditions are set in combination with the area constraint, effectiveness constraint and the fitting relationship between the decision variables; A multi-objective optimization unit, which is used to analyze the drainage flow of the pump station under different working conditions by using a time-period water storage calculation method based on the gray-green-blue measures and the multi-objective optimization model, and to establish a mathematical relationship expression between decision variables as an equality constraint through multivariate regression analysis; The first optimization unit is used to perform the first iteration optimization on the multi-objective optimization model by using the wild dog optimization algorithm to generate a Pareto solution set and a decision variable range; A model building unit, used to build a SWMM model based on the basic data, and use historical rainfall data for calibration and verification to ensure the accuracy of the SWMM model; The second optimization unit is used to couple the SWMM model with the wild dog optimization algorithm, update the objective function and constraints based on the range of decision variables of the first optimization, perform the second optimization, and generate the Pareto front solution set of the spatial layout of the gray-green-blue measures.

[0014] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects: The present invention fully considers the synergistic effect of gray-green-blue measures, especially the runoff reduction and control effect of pump stations. Traditional models usually focus on a single flood control measure, or take gray measures and water surface rate as known conditions rather than decision variables of the model. The present invention creatively proposes a multi-objective segmented optimization model for urban waterlogging prevention and control that takes into account the drainage flow, water surface area, and LID measures of pump stations. This segmented optimization strategy creatively solves the bottleneck of data acquisition and processing in hydrological models. It is suitable for areas with scarce data and areas with modeling conditions, and expands the scope of application of the optimization model. It provides new ideas for solving drainage optimization in areas with a lack of data, and also provides better technical support for improving urban resilience. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0016] Figure 1 A flowchart of steps provided for an embodiment of the present invention; Figure 2 A schematic diagram of a technical route provided for an embodiment of the present invention; Figure 3 A generalized schematic diagram of a SWMM model for a main urban area of ​​a city provided by an embodiment of the present invention; Figure 4 The error between the simulated waterlogging value and the measured waterlogging value of the SWMM model of the study area provided in the embodiment of the present invention; Figure 5 The Pareto front set obtained by secondary optimization of a main urban area of ​​a city provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0017] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0018] The purpose of this invention is to provide a multi-objective segmented optimization method of gray-green-blue measures of a dingo optimization algorithm coupled with SWMM, which takes into account the runoff reduction and control effect of pumping stations, is suitable for areas lacking data, and improves urban resilience.

[0019] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0020] Figure 1 A flowchart of the steps provided in the embodiment of the present invention is as follows: Figure 1 As shown, the present invention provides a multi-objective segmented optimization method of gray-green-blue measures of a wild dog optimization algorithm coupled with a SWMM, comprising: Step 100: Determine the current status of waterlogging distribution and simulation results based on the collected and collated basic data of the study area, and determine the type, scale, parameters and unit investment cost of the gray-green-blue measures; Step 200: Optimize the type and layout of gray-green-blue measures based on the current water distribution and simulation results; Step 300: construct a multi-objective optimization model, with minimization of total investment cost and maximization of total runoff reduction rate as objective functions, pump station drainage flow, water surface ratio, sunken green space ratio and permeable pavement ratio as decision variables, and set constraint conditions in combination with area constraints, effectiveness constraints and fitting relationships between decision variables; Step 400: Based on the gray-green-blue measures and the multi-objective optimization model, the water storage calculation method for each period is used to analyze the drainage flow of the pump station under different working conditions, and a mathematical relationship expression between decision variables is established through multivariate regression analysis as an equality constraint; Step 500: Perform the first iteration optimization on the multi-objective optimization model using the wild dog optimization algorithm to generate a Pareto solution set and a decision variable range; Step 600: constructing a SWMM model based on basic data, and using historical rainfall data for calibration and verification to ensure the accuracy of the SWMM model; Step 700: Couple the SWMM model with the wild dog optimization algorithm, update the objective function and constraints based on the range of decision variables of the first optimization, perform a second optimization, and generate a Pareto front solution set for the spatial layout of the gray-green-blue measures.

[0021] Specifically, Figure 2 As shown, a wild dog optimization algorithm coupled with a gray-green-blue measure multi-objective segmented optimization method of SWMM in this embodiment includes: Step 1: Collection and collation of basic data and information of the study area; mainly including basic data such as rainfall, waterlogging distribution, elevation, drainage network, land use, river water level, and drainage flow of existing pumping stations required for building the SWMM model; real-time monitoring of regional rainfall and waterlogging to obtain the current status of waterlogging distribution. And determine the scale, parameters and unit investment cost of various gray-green-blue measures in combination with local reality and planning design. Specifically: Go to the intersection of the main urban area of ​​the study area to carry out waterlogging monitoring, and record the waterlogging process and the maximum waterlogging depth.

[0022] Step 2: Selection of gray-green-blue measures: Based on the regional waterlogging distribution status in step 1, determine the specific gray-green-blue measure type, design parameters and layout.

[0023] Step 2.1, select appropriate gray-green-blue measures based on the planning and design of the study area. Pump stations are selected as gray measures to achieve rapid drainage, and the location and number of pump stations are expanded and newly built according to the current situation of the study area. Regulating and storing pump stations are selected near rivers and flood-prone areas with serious waterlogging.

[0024] Step 2.2, green measures include sunken green space and permeable pavement. The layout area, location and design parameters of various green measures are determined according to the planning and design of the study area. Considering the differences in the suitability of different green measures according to the land use type and the difficulty of construction in old urban areas, it is chosen to lay out sunken green space on ordinary green land, and to reconstruct permeable pavement in residential areas, commercial areas and road land.

[0025] In step 2.3, the water surface ratio in the study area is used to represent the blue measures, which involves excavation and expansion on the original rivers and lakes in the study area.

[0026] Step 3, establishment of a multi-objective optimization model; based on the gray-green-blue measures selected in step 3, determine the lowest total investment cost and the maximum runoff reduction rate as the objective function of the study area, use the pump station drainage flow, lake surface ratio, river water surface ratio, sunken green space ratio and permeable pavement ratio as decision variables, and use the regional area constraint, effectiveness constraint and the fitting relationship between the decision variables as constraints to construct a regional multi-objective optimization model.

[0027] Step 3.1, based on the gray-green-blue measures selected in step 2, determine the decision variables as follows: In the formula, X A vector representing the decision variables; q Represents the flow rate of the drainage pump station in the study area (m 3 / s); p i Representative i Water surface ratio of each water body type (lakes, rivers); xj Representative j The area ratio of various green measures (sunken green space, permeable pavement).

[0028] Step 3.2, optimize the gray-green-blue measures with the minimum total investment cost and the maximum runoff reduction rate as the objective function to seek a balance between economic and environmental benefits. The specific calculation formula is as follows: In the formula, f 1 It represents the total cost of investment. A The sum of the present value of investment per unit flow and the present value of operation and management of the drainage pump station (ten thousand yuan); B i For each cubic meter of excavation i Cost of water bodies (rivers, lakes) (ten thousand yuan); h i For the i Depth of water body (river, lake) (m); F is the area of ​​the study area; P 0,i It is i Current water surface rate; C j For the j The total construction cost and maintenance and operation cost of the LID measures. f 2 It represents the total runoff reduction rate; α A is the runoff coefficient before any measures are implemented; α is the runoff coefficient after the implementation of blue-green measures, △t It is the time for the pump station to drain water. The other symbols are the same as those that appeared before.

[0029] Step 3.3, the constraint range of parameters mainly considers area constraints and effectiveness constraints. The maximum value of blue-green measures is determined according to the planning and design of the study area and land use restrictions; the minimum pump station drainage flow is determined according to the target of regional external discharge flow. In addition, the water storage and infiltration of blue-green measures will affect the runoff production and convergence process, and the fitting relationship between the drainage pump station and the water surface rate and the green measure rate needs to be used as a constraint condition. The formula is as follows: In the formula, P i * For the i Minimum water surface ratio of each water body (measured as a percentage of the total area of ​​the planning area); x i * For the i Minimum deployment area ratio of green measures;P is the rainfall (mm), β It is the percentage of the discharge flow of the pump station. Considering that part of the rainfall is consumed and stored by other facilities, 80% of the rainfall is used to calculate the runoff; T is the designed drainage duration (h); other symbols are the same as before.

[0030] Step 4, fitting the relationship between decision variables; based on the gray-green-blue measures selected in step 3 and the multi-objective optimization model established in step 4, the water storage calculation method for each period is used to calculate the drainage flow of the pump station corresponding to different water surface rates and sunken green space rates, and the multivariate regression analysis method is used to fit the mathematical relationship expression between the decision variables as the equality constraint condition.

[0031] Step 4.1: When calculating the drainage water storage in the planning area, the storage and regulation effects of various measures should be considered. The storage and regulation calculations should be carried out in each period according to the normal water level, drainage level, minimum control water level and maximum control water level of the river. The formula is as follows: In the formula, V 1 , V 2 is the water storage volume of the river at the beginning and end of the time period (m 3 ); Q 1 , Q 2 is the inflow volume at the beginning and end of the time period (m 3 ); P 1 , P 2 The rainfall in each period is taken as 80% of the rainfall to calculate the design rainfall (mm). φ is the rainfall runoff coefficient of permeable pavement, generally 0.08~0.45, taken as 0.2; Δt is the time interval between the beginning and the end of a time period (h); the other symbols have the same meanings as above.

[0032] Step 5, the first optimization of the data-deficient area; based on the regional multi-objective optimization model obtained in steps 4 and 5, the wild dog optimization algorithm is used for iterative calculation to obtain the first optimized Pareto front and the range of decision variables.

[0033] Step 5.1, use Python to write the wild dog optimization algorithm, define the multi-objective function based on steps 3 and 4, and initialize the population to generate a set of random solutions.

[0034] Step 5.2, evaluate the fitness of each solution, perform non-dominated sorting according to the value of the objective function, and select the optimal solution.

[0035] In the formula, , The current iteration number Searchers are substituted into the objective function , The value of .

[0036] Step 5.3, consider the problem space as the predation space, simulate the siege, pursuit, scavenging and survival process of wild dogs hunting large prey, and iteratively improve the solution by updating the rules. The formula is as follows: ① Siege: Multiple searchers approach the location of the prey, that is, optimize the position towards our goal Search and get a new location: In the formula, For the new location of the searcher; For the interval A randomly generated integer in; is the initial population size; for randomly generated wild dog subgroups; is the searcher's current location; is the global optimal individual in the previous iteration; : A random scale factor within , which changes the size and direction of the motion trajectory; ② Hunting: Hunting small prey, that is, individuals at the global optimal position of the current iteration number Search nearby and get a new location: In the formula, for A random number within ③ Scavenging: The behavior of eating carrion found during random walking, that is, choosing a new position between the current position and any searcher: In the formula, , For two randomly selected searchers, ; ④Survival: calculate the survival probability of each dog. When the survival probability is less than a certain value, it is necessary to move to the global optimal individual Nearby, update the "leader" according to the new fitness function value, that is, the global optimal individual position ; In the formula, A binary number that is either 0 or 1; Step 5.4, repeat step 5.3 until the iteration stop condition is met; multiple iterations are performed to obtain the first optimized Pareto front and the range of decision variables.

[0037] Step 6, construction and verification of SWMM model: construct the SWMM model based on the data in step 1, and calibrate and verify it using two historical rainfall data.

[0038] Step 6.1: Use ArcGIS to convert the CAD files of different land use types in the study area into SHP files. Considering the actual conditions such as elevation changes, land use types, street distribution, and rainwater wells in the study area, the study area is divided into several sub-catchments, and the roads are divided separately.

[0039] Step 6.2: Use ArcGIS to reasonably split and merge the drainage pipelines in the study area, divide the drainage zones according to the north-south direction of the drainage main pipeline, and export the SHP file. Use the inpPINS plug-in to convert the SHP file into an inp file that can be recognized by SWMM.

[0040] Step 6.3, the parameters in the model are divided into deterministic parameters and calibrated parameters. Deterministic parameters include the characteristic width, area, slope, permeability of the sub-catchment area, the diameter, burial depth, and depth of the rainwater well of the pipe network. Among them, the area of ​​each sub-catchment area can be counted through the "Calculate Geometry" function in ArcGIS, and the area proportion of different land use types can be calculated. The elevation data can be converted and the average slope can be calculated through the 3DAnalyst and Spatial Analyst functions, and other parameters are obtained based on the planning and design data of the area. Calibrated parameters such as the Manning coefficient of the catchment area, infiltration rate, and depression storage capacity need to refer to the SWMM model manual and the parameter value range of the study area and be empirically determined.

[0041] Step 6.4, use the comprehensive runoff coefficient to make preliminary adjustments to the model parameters, select the parameter values ​​in different regions according to the empirical range, and improve the accuracy of the model through multiple adjustments. Then use the measured data of two historical rainfalls for calibration and verification, spread the maximum overflow of the simulated flood-prone point over the road area, and calculate the maximum surface water depth simulation value. Calculate the relative error between the simulated value and the measured water depth to verify the accuracy of the model. The formula is as follows: In the formula, H 1 is the measured value of water accumulation at the monitoring point, cm; H 2 is the simulated value of water accumulation at the monitoring point, cm.

[0042] Step 7, coupling of the wild dog optimization algorithm and the SWMM model; based on the Pareto frontier and the range of decision variables of the first optimization obtained in step 5, the mean of the drainage flow rate and water surface rate of the pump station are taken as known variables, and the extreme values ​​of the sunken green space rate and the permeable pavement rate are used to update the objective function and the range of decision variables. The SWMM model is automatically established, run and analyzed through Python programming, and the wild dog optimization algorithm is coupled to realize the automatic optimization of the sunken green space and permeable pavement layout area, and the second optimized Pareto frontier is obtained as the solution set for the spatial layout of the gray-green-blue measures.

[0043] Step 7.1, based on the range of decision variables under the corresponding scheme of the first optimized Pareto front obtained in steps 4 and 5, the average pump station drainage flow and average water surface rate of all schemes are used as known variables to update the multi-objective optimization function, and the constraints are updated with the extreme values ​​of the sunken green space rate and the permeable pavement rate.

[0044] Step 7.2, construct the updated multi-objective function optimization model through Python, and implement iteration with the wild dog optimization algorithm.

[0045] Step 7.3, the SWMM model is automatically established through Python, and the dingo optimization algorithm is coupled to realize the automatic optimization of the sunken green space and permeable pavement layout area. At this time, the model calculates the surface runoff and compares it with the surface runoff before any measures are implemented to obtain the total runoff reduction rate.

[0046] Step 7.4, based on step 7.2 and step 7.3, a multi-objective optimization model is constructed, and the wild dog optimization algorithm is used for iteration to obtain the Pareto frontier of the second optimization and the solution set of the spatial layout under the green measures.

[0047] Example: The optimization method of the present invention is applied to the main urban area of ​​Handan City, Hebei Province, and the technical process includes: The collection and collation of basic data and information of the study area, the selection of gray-green-blue measures, the establishment of a multi-objective optimization model, the fitting of the relationship between decision variables, the first optimization of areas with insufficient data, the construction and verification of the SWMM model, and the coupling of the wild dog optimization algorithm with the SWMM model. The specific process is as follows: (1) Collect and organize basic data related to the study area. Including land use, 20-year design rainfall, unit cost of gray-green-blue measures, unit price of green measures, etc. Among them, land use data includes the area of ​​the study area and the area proportion of each land use type. The study area covers an area of ​​approximately 24.204 km 2, 2.2% of the current water surface is river, 0.3% is lake, and the average water depth of the main river is about 3.16m. The land use types include 12.63% roads, 40.14% residential areas, 18.82% green land, 22.66% commercial areas, and 3.26% industrial land.

[0048] (2) The drainage zones in the study area are relatively independent, and the ends of the drainage pipe networks are connected to the Qin River and the Fuyang River. The outlets of the existing gravity-flowing rainwater pipes are mostly close to the riverbed, resulting in slow flow in the pipes and prone to backflow during the flood season. Therefore, this embodiment considers the construction and reconstruction of 5 forced drainage pumping stations along the river as gray measures, and the widening of rivers and lakes as blue measures. Considering the differences in the suitability of different green measures according to land use types and the difficulty of construction in old urban areas, it is chosen to lay out sunken green spaces on ordinary green land, and to reconstruct permeable pavements in residential areas, commercial areas, and road land.

[0049] (3) The objective function of the study area is to minimize the total investment cost and maximize the total runoff reduction rate, and seek a balance between economic and environmental benefits. The total investment cost mainly includes annual infrastructure costs and management and maintenance costs. The unit cost of various measures obtained by referring to relevant literature and Handan City Construction Plan is shown in Table 1. The annual runoff reduction rate is based on the ratio of the sum of the pump station discharge flow and the reduction under the action of blue and green measures to the surface runoff before any measures are implemented. The formula of the multi-objective function is summarized as follows: In the formula, f 1 represents the total investment cost, f 2 It represents the total runoff reduction rate. q Represents the flow rate of the drainage pump station in the study area (m 3 / s); p 1 and p 2 represent the water surface ratios of lakes and rivers, respectively; x 1 and x 2 Represents the area ratio of sunken green space (measured as % of the green space in the study area) and permeable pavement.

[0050] Table 1 Basic costs and operating management expenses corresponding to different measures

[0051] Referring to the planning and design of the study area, the minimum pump station drainage flow is calculated with the goal of discharging 40% of the flood volume within 24 hours due to the runoff generated by rainfall, while the water surface rate and green measures are mainly constrained by the area. In order to ensure that the study area has a significant runoff reduction effect, the minimum water surface rate is set to the existing water surface rate; referring to the relevant indicators of other sponge city pilot projects, the proportion of sunken green space area is set to be no less than 30% (in terms of % of the green space area in the planning area), and the permeable pavement rate is set to be no less than 10% (in terms of % of the hardened area in the planning area). The maximum values ​​of various parameters are limited by the area of ​​the study area. After sorting out, the constraint condition formulas of the decision variables are as follows:

[0052] (4) The pump station discharge flow corresponding to different water surface ratios, sunken green spaces and permeable pavement ratios is calculated by using water storage and regulation in each period. The initial pre-drainage level of the river in this embodiment is set to 3.4m, and the highest water surface line of the river in the middle and lower reaches of the river network area is controlled within 5.1m. It is believed that the pump station discharge flow obtained at this time is more reasonable. After rainfall, the pumping and drainage are started to control the water level of the inner river between the highest control water level and the lowest control water level. In order to ensure that the storage capacity in the dike is emptied in time and prevent the next rainstorm and flood, the flood volume stored in the river and pond needs to be discharged on the same day, and the river returns to the normal water level before the rain. The minimum discharge flow is calculated accordingly. According to the basic data of Handan's main urban area and the constraints in the mathematical model, the approximate working range of the water surface rate, sunken green space rate and permeable pavement rate is determined. The calculation results of drainage flow under different water surface rates (2.5%~5%), sunken green space rates (30%~70%) and permeable pavement rates (10%~26.4%) are shown in Table 2. The fitting relationship formula between the parameters obtained by multiple regression analysis is as follows:

[0053] Table 2 Pumping station drainage flow corresponding to different water surface ratios, sunken green space ratios and permeable pavement

[0054] (5) The wild dog optimization algorithm is used to iterate the multi-objective function optimization model to obtain the first optimized Pareto solution set and the corresponding decision variable range. From the Pareto solution set, it can be seen that the cost is 293.72 million yuan to 431.74 million yuan, but the reduction rate is 44.2%34%~44.2%. From the distribution of decision variables, the range of pump station drainage flow is from 18.18 to 28.08m 3 / s, the lake rate and water surface rate ranged from 2.63% to 2.79% and 4.5% to 4.7% respectively, the sunken green area rate ranged from 31.6% to 69% (in terms of green area), and the permeable pavement rate ranged from 10% to 12.4%. Considering the uniform flow distribution of the drainage pump station, the concentrated water surface rate, and the relatively complex control in the SWMM model in the second stage of optimization, the average values ​​of 23%, 2.7%, and 4.65% were selected as fixed values ​​respectively.

[0055] (6) For the data required by the high-precision hydrological model, the SWMM model can be further coupled to optimize the spatial layout of green measures. The high-precision data required to build the hydrological model include elevation information, land type area, drainage network, historical rainfall data, waterlogging monitoring data, etc. Based on the above data, the study area is divided into 5 drainage zones, 1519 sub-catchments, 282 nodes, 282 pipelines, and 24 drainage outlets. Figure 3 .

[0056] There are two types of parameters in the model: deterministic parameters and empirical parameters. Deterministic parameters can be calculated or extracted directly, while empirical parameters require reference to relevant literature and empirical values ​​determined by standards. This embodiment uses the comprehensive runoff coefficient method to preliminarily determine the value range of the empirical parameters, and then uses the absolute error between the simulated water depth and the measured water depth of the water accumulation points of the two rainfall events on June 26, 2022 and July 5, 2022 for calibration and verification. The errors are all within 20%, see Figure 4 , indicating that the simulation effect of the model is good. The final values ​​of the calibration parameters are shown in Table 3.

[0057] Table 3 SWMM model calibration parameter values

[0058] (7) Update the objective function and constraints based on the first optimization result, and obtain the following formula: In the formula, f 3 Represents the total investment cost based on the first optimization update. f 4 represents the total runoff reduction rate based on the first optimization update; S is the surface flow rate obtained from the second optimization (m 3 ), S A is the surface runoff without any measures (m 3 ).

[0059] The LID parameters need to be set in the inp file, and the specific values ​​are based on the experience of the study area. Call the PySWMM library, and use Python programming to realize the automatic establishment, operation and analysis of the SWMM model. The DOA optimization algorithm is coupled to realize the automatic optimization of the layout area of ​​different LID measures, and the solution set of the optimal spatial layout of green measures after the second optimization of the study area is obtained. In the second stage of the optimization process, as the cost increases from 308 million to 345 million yuan, the reduction rate increases from 44.5% to 45.1%, see Figure 5 .

[0060] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0061] The principles and implementation methods of the present invention are described in this article using specific examples. The description of the above embodiments is only used to help understand the method and core idea of ​​the present invention. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. A multi-objective segmented optimization method of gray-green-blue measures coupled with a wild dog optimization algorithm in SWMM, characterized in that: include: Determine the current status of waterlogging distribution based on the basic data of the study area collected and collated, and determine the type, scale, parameters and unit investment cost of the gray-green-blue measures; Optimize the type and layout of the gray-green-blue measures based on the current water distribution status; A multi-objective optimization model was constructed, with the minimization of total investment cost and the maximization of total runoff reduction rate as the objective function, the drainage flow of the pump station, the water surface ratio, the sunken green space ratio and the permeable pavement ratio as the decision variables, and the constraint conditions were set by combining the area constraint, the effectiveness constraint and the fitting relationship between the decision variables. Based on the gray-green-blue measures and the multi-objective optimization model, the water storage calculation method for each period is used to analyze the drainage flow of the pump station under different working conditions, and the mathematical relationship expression between the decision variables is established as an equality constraint through multivariate regression analysis; The multi-objective optimization model is optimized for the first iteration using the wild dog optimization algorithm to generate the Pareto solution set and the range of decision variables; Building a SWMM model based on the basic data, and using historical rainfall data for calibration and verification to ensure the accuracy of the SWMM model; The SWMM model is coupled with the Dingo optimization algorithm, and the objective function and constraints are updated based on the range of decision variables of the first optimization. The second optimization is performed to generate the Pareto front solution set of the spatial layout of the gray-green-blue measures.

2. The multi-objective segmented optimization method of the wild dog optimization algorithm coupled with the gray-green-blue measure of SWMM according to claim 1 is characterized in that: The basic data include: rainfall data, land use type, drainage network information, pump station parameters, elevation data and river water level data.

3. The multi-objective segmented optimization method of the wild dog optimization algorithm coupled with the gray-green-blue measure of SWMM according to claim 1 is characterized in that: Based on the current water distribution situation, the types and layout of the gray-green-blue measures are optimized, including: Deploy storage pump stations near rivers and flood-prone areas as gray measures; According to the suitability of the land use type, sunken green spaces are laid out in ordinary green areas, and permeable pavements are rebuilt in residential areas, commercial areas and road areas as green measures; The blue measure is to expand the water surface rate of the original rivers and lakes in the study area.

4. The multi-objective segmented optimization method of the wild dog optimization algorithm coupled with the gray-green-blue measure of SWMM according to claim 1 is characterized in that: The decision variables of the multi-objective optimization model are: ; in, X A vector representing the decision variables; q represents the flow rate of the drainage pumping station in the study area; p i Representative i Water surface ratio of water body type; x j Representative j The area ratio of green measures, m Represents the number of water body types, n represents the number of green measures; The calculation formula of the objective function is: ;in, f 1 represents the total investment cost, A It is the sum of the present value of investment per unit flow and the present value of operation and management of the drainage pump station; B i For each cubic meter of excavation i the cost of planting water bodies; h i For the i The depth of the water body; F is the area of ​​the study area; P 0,i It is i Current water surface rate; C j For the j The total construction cost and maintenance and operation cost of the LID measures; f 2 It represents the total runoff reduction rate; α A is the runoff coefficient before any measures are implemented; α is the runoff coefficient after the implementation of blue-green measures, △t It is the time for the pump station to drain water; The constraint condition is expressed as: ;in, P i * For the i Minimum water surface ratio of the water body; x i * For the i Minimum deployment area ratio of green measures; P is the rainfall, β is the percentage of discharge flow from the pump station; T Design duration for drainage.

5. The multi-objective segmented optimization method of the wild dog optimization algorithm coupled with the gray-green-blue measure of SWMM according to claim 4 is characterized in that: The formula for calculating the water storage and regulation method for each period is: in, V 1 , V 2 are the river water storage at the beginning and end of the time period respectively; Q 1 , Q 2 are the inflow water volume at the beginning and end of the time period respectively; P 1 , P 2 are the rainfall in each period, φ is the rainfall runoff coefficient for permeable pavement; Δt is the time interval between the beginning and end of the period, F green , F harden The area of ​​green space in the study area and the area of ​​hardened ground in the study area are studied respectively.

6. The multi-objective segmented optimization method of the wild dog optimization algorithm coupled with the gray-green-blue measure of SWMM according to claim 1 is characterized in that: The wild dog optimization algorithm includes: Initialize the population and generate random solutions; Evaluate fitness through non-dominated sorting and select the optimal solution; Simulate wild dog hunting behavior, including siege, pursuit, scavenging and survival rules, and update the solution position; Iterate until the termination condition is met and output the Pareto solution set.

7. The multi-objective segmented optimization method of the wild dog optimization algorithm coupled with the gray-green-blue measure of SWMM according to claim 1 is characterized in that: The SWMM model is constructed based on the basic data and calibrated and verified using historical rainfall data to ensure the accuracy of the SWMM model, including: ArcGIS was used to convert the CAD files of different land use types in the study area into SHP files. The study area was divided into several sub-catchments, and the roads were divided separately, taking into account the elevation changes, land use types, street distribution, and stormwater wells in the study area. ArcGIS was used to split and merge the drainage pipelines in the study area, divide the drainage zones according to the north-south direction of the drainage main pipeline, export the SHP file, and convert the SHP file into an inp file that can be recognized by SWMM through the inpPINS plug-in; The parameters in the SWMM model are divided into deterministic parameters and calibrated parameters; the deterministic parameters include the characteristic width, area, slope, permeability, pipe diameter, burial depth, and depth of the rainwater well of the sub-catchment area; the area of ​​each sub-catchment area is statistically calculated through the "Computational Geometry" function in ArcGIS, and the area proportion of different land use types is calculated; the elevation data is converted and the average slope is calculated through the 3D Analyst and Spatial Analyst functions; the calibrated parameters are determined by referring to the SWMM model manual and the parameter value range of the study area; The model parameters of the SWMM model are preliminarily adjusted using the comprehensive runoff coefficient, the parameter values ​​of different regions are selected according to the empirical range, and the accuracy of the SWMM model is improved through multiple adjustments; The measured data of two historical rainfall events were used for calibration and verification. The maximum overflow of the simulated flood-prone point was evenly distributed over the road area. The maximum simulated surface water depth was calculated and the relative error between the simulated value and the measured water depth was calculated to verify the accuracy of the SWMM model. The formula for the relative error is as follows: ; In the formula, H 1 is the measured value of water accumulation at the monitoring point; H 2 It is the simulated value of water accumulation at the monitoring point.

8. The multi-objective segmented optimization method of the wild dog optimization algorithm coupled with the gray-green-blue measure of SWMM according to claim 1 is characterized in that: The SWMM model is coupled with the Dingo optimization algorithm. Based on the range of decision variables of the first optimization, the objective function and constraints are updated, and the second optimization is performed to generate the Pareto front solution set of the spatial layout of the gray-green-blue measures, including: Based on the range of decision variables under the corresponding scheme of the first optimized Pareto front, the average drainage flow of all schemes and the average water surface rate are used as known variables to update the multi-objective optimization function, and the constraint conditions are updated with the extreme values ​​of the sunken green space rate and the permeable pavement rate; The updated optimization model of multi-objective function is constructed through Python, and the iteration is realized by the wild dog optimization algorithm; The SWMM model was automatically established through Python, and the wild dog optimization algorithm was coupled to achieve automatic optimization of the sunken green space and permeable pavement layout area; Based on the obtained optimization model of multi-objective functions, the wild dog optimization algorithm is adopted for iteration to obtain the Pareto frontier of the second optimization and determine the solution set of spatial layout under green measures.

9. A multi-objective segmented optimization system of gray-green-blue measures coupled with a wild dog optimization algorithm of SWMM, characterized in that: include: The data collection unit is used to determine the current status of waterlogging distribution and simulation results based on the basic data of the study area collected and sorted, and to determine the type, scale, parameters and unit investment cost of the gray-green-blue measures; A measure optimization unit, used for optimizing the type and layout of the gray-green-blue measures based on the current water distribution status; The target construction unit is used to construct a multi-objective optimization model, with the minimization of total investment cost and the maximization of the total runoff reduction rate as the objective function, the pump station drainage flow, water surface ratio, sunken green space ratio and permeable pavement ratio as decision variables, and the constraint conditions are set in combination with the area constraint, effectiveness constraint and the fitting relationship between the decision variables; A multi-objective optimization unit, which is used to analyze the drainage flow of the pump station under different working conditions by using a time-period water storage calculation method based on the gray-green-blue measures and the multi-objective optimization model, and to establish a mathematical relationship expression between decision variables as an equality constraint through multivariate regression analysis; The first optimization unit is used to perform the first iteration optimization on the multi-objective optimization model by using the wild dog optimization algorithm to generate a Pareto solution set and a decision variable range; A model building unit, used to build a SWMM model based on the basic data, and use historical rainfall data for calibration and verification to ensure the accuracy of the SWMM model; The second optimization unit is used to couple the SWMM model with the wild dog optimization algorithm, update the objective function and constraints based on the range of decision variables of the first optimization, perform the second optimization, and generate the Pareto front solution set of the spatial layout of the gray-green-blue measures.

Images