A combined dispatching system and method for a drainage pipe network pump station group based on a SWMM model

Abstract

This invention discloses a joint scheduling system and method for drainage network pumping station groups based on the SWMM model, belonging to the field of drainage network scheduling technology. The system includes a hydraulic modeling module, an optimal scheduling model construction module, an objective function and constraint setting module, an initial condition input module, an SWMM model engine driving module, a simulation-optimization loop module, and a scheduling scheme output and feedback module. The method constructs a hydraulic model of the drainage network and pumping station groups, establishes an optimal scheduling model based on artificial intelligence optimization algorithms using a simulation-optimization, feedforward-feedback mode, and outputs the optimal scheduling scheme through multiple rounds of simulation-optimization loop iterations, while also providing feedback adjustments. This invention overcomes the shortcomings of existing technologies, such as low optimization efficiency, poor real-time performance, and difficulty in handling multi-objective and multi-constraint problems. It achieves globally optimal or near-optimal real-time control of complex drainage systems, improving the operating efficiency, reliability, and adaptability of drainage systems, and reducing energy consumption and overflow risks.

Description

Technical Field

[0001] This invention relates to the field of joint scheduling of drainage network pumping station groups, specifically a joint scheduling system and method for drainage network pumping station groups based on the SWMM model. Background Technology

[0002] With the acceleration of urbanization, the drainage pressure on urban drainage networks is increasing, especially under extreme weather conditions such as rainstorms. The effectiveness of the coordinated operation of pumping station groups directly affects urban drainage safety and residents' quality of life. Traditional drainage network scheduling methods rely heavily on manual experience rules or static control strategies, lacking the ability to adapt to dynamic changes and uncertainties in the system.

[0003] In existing technologies, some scheduling methods use mathematical models for simulation and optimization, but these methods have many drawbacks:

[0004] 1. The optimization efficiency is low, making it difficult to quickly respond to dynamic changes such as uneven spatial and temporal distribution of rainfall and pipeline siltation;

[0005] 2. Poor real-time performance; unable to effectively combine real-time monitoring data and rainfall forecast data for dynamic adjustments;

[0006] 3. Insufficient multi-objective processing capability; optimization methods based on a single objective function are difficult to fully address multiple needs such as energy consumption control, drainage efficiency improvement, and overflow risk reduction.

[0007] 4. The constraints are not fully considered, and the ability to coordinate and handle constraints such as pump station capacity, water level safety, and water balance is lacking.

[0008] Therefore, there is an urgent need to develop a joint scheduling system and method for drainage network pumping stations that can achieve global optimization, dynamic adjustment, and high efficiency and reliability. Summary of the Invention

[0009] The purpose of this invention is to provide a joint scheduling system and method for drainage network pumping station groups based on the SWMM model, so as to solve the defects of the prior art such as low optimization efficiency, poor real-time performance, and difficulty in handling multi-objective and multi-constraint problems.

[0010] A joint scheduling system for drainage network pumping station groups based on the SWMM model includes:

[0011] The hydraulic modeling module is used to construct hydraulic models of drainage pipe networks and pumping station groups. The hydraulic models include pipe network topology, pumping station characteristics, hydrological data, and sub-catchment parameters.

[0012] The optimized scheduling model construction module is based on artificial intelligence optimization algorithms and adopts a simulation-optimization and feedforward-feedback composite control mode to establish an optimized scheduling model. The simulation-optimization mode optimizes the scheduling rules through iterative optimization, while the feedforward-feedback mode dynamically adjusts the scheduling strategy by combining rainfall forecast data and real-time monitoring data.

[0013] The objective function and constraint setting module is used to determine the multi-objective function and constraints of the optimization scheduling model;

[0014] The initial condition input module is used to input the initial conditions for the operation of the hydraulic model and the initial scheduling rules for the pump station group; the initial conditions include initial water level, flow data, historical rainfall data and rainfall forecast data, and the initial scheduling rules include pump start-up and shutdown rules based on historical experience;

[0015] The SWMM model engine driver module is used to drive the SWMM model engine, using scheduling rules as control parameters to perform hydraulic simulation calculations and generate simulation results of the hydraulic state of the pipeline network.

[0016] The simulation-optimization loop module is used to evaluate the simulation results based on the objective function, select the best-performing scheduling scheme through the fitness function, and generate new scheduling rules by combining the selection, crossover, and mutation operations of the intelligent optimization algorithm. The simulation-optimization process is repeated until the stopping condition is met; the stopping condition includes an iteration number threshold or an objective function convergence threshold.

[0017] The scheduling scheme output and feedback module is used to output the optimal solution set and the corresponding optimal scheduling scheme results, issue the scheduling scheme for execution, and provide feedback on the effectiveness through comparison and analysis of real-time monitoring data and simulation data, and dynamically adjust the model parameters.

[0018] As a preferred technical solution of the present invention, the artificial intelligence optimization algorithm includes at least one of genetic algorithm, particle swarm optimization algorithm, and neural network algorithm; when a genetic algorithm is used, the scheduling rules are encoded as chromosome vectors, and the chromosome vectors represent the start-stop state combination of each pump station pump within a preset time period.

[0019] As a preferred technical solution of the present invention, the multi-objective function includes at least one of minimizing system energy consumption, minimizing overflow, and maximizing drainage efficiency, and the constraints include water balance constraints, pump station capacity constraints, and water level constraints.

[0020] As a preferred technical solution of the present invention, in the feedforward-feedback composite control mode, the feedforward control generates a pre-scheduling scheme based on future rainfall forecast data, and the feedback control compares the data with the simulated values ​​in real time through the level gauges and flow meters of key nodes. When the deviation exceeds the threshold, it triggers the fine-tuning or re-optimization of the model parameters.

[0021] This invention also provides a joint scheduling method for drainage network pumping station groups based on the SWMM model, comprising the following steps:

[0022] Step S1: Hydraulic modeling is carried out for drainage pipe network and pump station group. Geographic information data, pipe network structure data, hydrological and meteorological data and operation data are collected. A SWMM model including sub-catchment area, pipeline, inspection well, pump station and storage tank is constructed. The model is calibrated and verified by historical operation data to ensure that the Nash efficiency coefficient NSE of the simulation data and the actual monitoring data is ≥0.75.

[0023] Step S2: Based on artificial intelligence optimization algorithms, an optimized scheduling model is established using a simulation-optimization and feedforward-feedback composite control mode. The simulation-optimization mode optimizes the scheduling rules through iterative optimization, while the feedforward-feedback mode dynamically adjusts the scheduling strategy by combining rainfall forecast data and real-time monitoring data.

[0024] Step S3: Determine the multi-objective function and constraints of the optimization scheduling model; the multi-objective function includes minimizing system energy consumption and minimizing overflow, and the constraints include water balance constraints, pump station capacity constraints, and water level safety range;

[0025] Step S4: Input the initial conditions for the operation of the hydraulic model and the initial scheduling rules for the pump station group; the initial conditions include the initial water level of the inspection well at the current time, the initial flow rate of the pipeline, the initial water volume of the storage tank, and the rainfall forecast data for the preset duration in the future; the initial scheduling rules include pump start-up and shutdown rules based on historical experience.

[0026] Step S5: Drive the SWMM model engine to perform simulation calculations, evaluate the calculation results based on the objective function, select the best-performing scheduling scheme through the fitness function, and generate new scheduling rules by combining the selection, crossover, and mutation operations of the intelligent optimization algorithm.

[0027] Step S6: Repeat the simulation-optimization process of step S5 until the stopping condition is met;

[0028] Step S7: Output the optimal solution set and the corresponding optimal scheduling scheme results, and issue the scheduling scheme for execution; conduct performance feedback analysis by comparing the real-time monitoring data of key nodes with the simulation data; when the deviation exceeds the preset threshold, start short-term domain re-optimization; after the rainfall event ends, update the historical database and periodically perform automatic calibration of SWMM model parameters.

[0029] As a preferred technical solution of the present invention, the SWMM model constructed by the hydraulic model modeling module, after being calibrated with historical operating data, has a Nash efficiency coefficient (NSE) of ≥0.75 between the simulated data and the actual monitoring data.

[0030] As a preferred technical solution of the present invention, the objective function in step S3 is a multi-objective optimization function, and the constraints include water balance constraints, pump station capacity constraints, and water level safety range.

[0031] The objective function expression for minimizing system energy consumption is:

[0032]

[0033] In the formula: f is the operating cost function of the drainage system; T is the number of time periods; I is the number of pumps; P(i,t) is the power of pump station i in time period t, which is calculated from its head, flow rate and efficiency according to the performance curve; Δt is the scheduling period; R_i(t) is the unit electricity cost in time period t.

[0034] The objective function expression for minimizing the overflow is:

[0035]

[0036] In the formula: V_i——the overflow amount of water accumulation node i, and m is the total number of water accumulation nodes.

[0037] The expression for the water balance constraint is:

[0038]

[0039] In the formula: These represent the total water diversion flow of the water system at times t+1 and t, respectively. Δt represents the total outflow of the water system at times t+1 and t, respectively; Δt represents the unit time; and ΔV represents the change in water volume of the water system per unit time.

[0040] The expression for the pump station capacity constraint is:

[0041]

[0042] In the formula: The pumping capacity of the pumping station is (m³ / h). This is the minimum pumping capacity of the pumping station (m³ / h). This is the maximum pumping capacity of the pumping station (m³ / h).

[0043] The expression for the water level constraint is:

[0044]

[0045] In the formula: H(j,t) represents the actual water level of the node or pump station forebay during the scheduling period t (unit: meters); H_min(j) represents the minimum water level allowed for node or pump station j (usually the minimum water level to prevent silt deposition or meet the pump station's suction depth, unit: meters); H_max(j) represents the maximum water level allowed for node or pump station j (usually the maximum warning water level or the elevation of the top of the pump station's inlet pipe, unit: meters).

[0046] As a preferred technical solution of the present invention, the intelligent optimization algorithm in step S5 is a genetic algorithm, which evaluates the scheduling scheme through the fitness function and generates new scheduling rules through crossover and mutation operations.

[0047] As a preferred embodiment of the present invention, the stopping condition in step S6 includes an iteration number threshold or an objective function convergence threshold.

[0048] The beneficial effects of this invention are as follows:

[0049] By adopting a simulation-optimization and feedforward-feedback composite control mode, and combining rainfall forecast data and real-time monitoring data, the scheduling strategy can be planned and dynamically corrected in a forward-looking manner, which significantly improves the real-time performance and adaptability of the scheduling and can effectively cope with uncertainties such as uneven spatial and temporal distribution of rainfall and pipeline siltation.

[0050] The optimization design is based on a multi-objective function, taking into account multiple objectives such as system energy consumption, overflow, and drainage efficiency. It provides a variety of scheduling schemes through the optimal solution set to meet the management needs of different scenarios and overcomes the limitations of traditional single-objective optimization methods.

[0051] By integrating the advantages of artificial intelligence optimization algorithms and SWMM models, it achieves global optimal or near-optimal control of complex drainage systems through iterative optimization, improving optimization efficiency and the scientific nature of scheduling schemes, and reducing system energy consumption and overflow risks.

[0052] A complete feedback and adjustment mechanism has been established. Through real-time data comparison and correction and historical data model calibration, the model has achieved self-evolution, ensuring the long-term effectiveness and accuracy of the scheduling scheme. It is applicable to multiple fields such as urban drainage management and flood control and disaster reduction. Attached Figure Description

[0053] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0054] Figure 1This is a flowchart of a joint scheduling system and method for drainage network pumping station groups based on the SWMM model according to the present invention;

[0055] Figure 2 This is a detailed flowchart of the simulation-optimization loop process in the embodiment. Detailed Implementation

[0056] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings and embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0057] Example 1:

[0058] This paper takes the drainage system of the central urban area of ​​City A, a coastal city in East China, as an example. The area covers approximately 50 square kilometers, with a total drainage network length of about 430 kilometers, including 18 key drainage pumping stations (P1-P18), a large regulating reservoir (T1), and the outlets of three sewage treatment plants. This region faces problems such as frequent summer downpours and high tides, and traditional scheduling methods often lead to localized flooding and energy waste. This invention utilizes the joint scheduling of the pumping station group. The implementation process of this embodiment can be found in [reference needed]. Figure 1-2 The method flowchart shown includes the following steps in its execution process:

[0059] S1. Perform hydraulic modeling for drainage pipe network and pump station group;

[0060] Specifically, in this embodiment, the implementation process of S1 is as follows:

[0061] S101. Data collection and processing, including the following:

[0062] Geographic information data: Obtain geographic information data such as DEM digital elevation model and land use map of the area from municipal departments;

[0063] Pipeline structure data: Extract all pipeline structure data (diameter, length, slope, material), inspection wells (location, bottom elevation), pumping stations (performance curves of P1-P18, start and stop water levels, design flow), and regulating reservoirs (volume of T1, effluent rules) from the drainage pipeline network GIS data coordinated by the housing and construction department.

[0064] Hydrometeorological data: Collect historical rainfall data from the past 5 years, as well as typical design rainfall curves (such as 24-hour rainfall process lines with 2-year, 5-year, and 10-year return periods). Connect to the meteorological bureau's real-time rainfall forecast API to obtain high spatiotemporal resolution rainfall forecast data for the next 2 hours.

[0065] Operational data: Obtain historical operation records and energy consumption data for each pumping station over the past year;

[0066] S102, SWMM Model Construction and Calibration:

[0067] Using the above data, a hydraulic model of the region was established in the SWMM environment. The sub-catchments were divided into smaller sub-basins based on land use type, and parameters such as Manning coefficient, sinkhole capacity, and permeability were set. The connection relationships and attribute parameters of elements such as pipelines, nodes, pumping stations, and regulating reservoirs were accurately input. In particular, for the 18 pumping stations, their performance curves were defined in SWMM using Type 3 pump curves.

[0068] After the model was built, it was calibrated and verified using historical operating data. By adjusting parameters such as impermeability and pipe roughness, the Nash efficiency coefficient (NSE) of both the simulated water level and flow rate data and the actual monitoring data was made to be greater than 0.75, thus ensuring the reliability of the model.

[0069] S2. Based on artificial intelligence optimization algorithms, an optimized scheduling model is established using simulation-optimization and feedforward-feedback modes.

[0070] Specifically, in this embodiment, the implementation process of S2 is as follows:

[0071] S201. First, the optimization algorithm is selected. In this embodiment, a multi-objective genetic algorithm is selected as the core optimization algorithm because it can effectively handle multi-objective and nonlinear problems and output a set of optimal solutions.

[0072] S202. Secondly, the scheduling mode is designed, adopting a feedforward-feedback composite control mode;

[0073] Feedforward control: Based on rainfall forecast data for the next 2 hours, the model is driven to perform forward simulation optimization to generate a pre-scheduling scheme.

[0074] Feedback control: Real-time monitoring data is obtained by deploying level gauges and flow meters at key nodes (such as pump station forebays and flood-prone areas) and compared with simulated values. If the deviation exceeds the threshold (such as 10%), the model parameters are fine-tuned or re-optimized to achieve dynamic correction.

[0075] S3. Determine the objective function and constraints of the optimization scheduling model;

[0076] Specifically, in this embodiment, the implementation process of S3 is as follows:

[0077] S301. Determine the objective function;

[0078] This embodiment sets two core objectives, constituting a bi-objective optimization problem:

[0079] Objective 1 (f1): Minimize the total system overflow.

[0080]

[0081] In the formula: — Overflow amount of water accumulation node i.

[0082] Objective 2 (f2): Minimize the total energy consumption of the system;

[0083]

[0084] In the formula: I represents the number of time periods; I represents the number of water pumps. Δt is the power of pump station i during time period t, which is calculated from its head, flow rate and efficiency based on the performance curve; Δt is the scheduling period (set to 5 minutes in this example). The unit electricity cost for time period t;

[0085] S302. Determine the constraints, which include:

[0086] Water balance equation:

[0087]

[0088] In the formula: These represent the total water diversion flow of the water system at times t+1 and t, respectively. These represent the total outflow rate of the water system at times t+1 and t, respectively. Unit of time; This represents the change in water volume within a unit of time.

[0089] Pump station capacity constraints:

[0090]

[0091] In the formula: The pumping capacity of the pumping station is (m³ / h). This is the minimum pumping capacity of the pumping station (m³ / h). This is the maximum pumping capacity of the pumping station (m³ / h).

[0092] Water level constraints:

[0093]

[0094] In the formula: H(j,t) represents the actual water level of the forebay of the node or pump station during the scheduling period t (unit: meters); H_min(j) represents the minimum water level allowed for the node or pump station j (unit: meters); H_max(j) represents the maximum water level allowed for the node or pump station j (unit: meters).

[0095] S4. Input the initial conditions for the operation of the hydraulic model and the initial scheduling rules for the pump station group;

[0096] Specifically, in this embodiment, the implementation process of S4 is as follows:

[0097] S401. Set initial conditions: Input the initial water level of all inspection wells, the initial flow rate of the pipeline, and the initial water volume of the storage tank at the current moment. Obtain the rainfall process line data for the next 2 hours from the weather forecast API.

[0098] S402. Set the initial scheduling rules for the pump station group: Input a scheduling scheme based on the original empirical rules, for example:

[0099] If the water level in the forebay of the pumping station is > 2.0m, then start pump No. 1.

[0100] If the water level in the forebay of the pumping station is > 2.5m, then start pump No. 2.

[0101] This rule will serve as one of the initial population individuals for the genetic algorithm.

[0102] S5 drives the SWMM model engine and performs a simulation-optimization loop;

[0103] Specifically, in this embodiment, the implementation process of S5 is as follows:

[0104] S501, Encoding and Initialization: The scheduling rules are encoded into chromosomes of a genetic algorithm. Each chromosome is a vector representing the start / stop status (0 or 1) of each pump in 18 pumping stations within the next 2 hours (24 5-minute intervals). The initial population size is 100, containing randomly generated individuals and individuals based on the aforementioned empirical rules.

[0105] S502. Simulation and Evaluation, the steps are as follows:

[0106] Loop begins: The genetic algorithm generates 100 scheduling schemes (chromosomes);

[0107] Calling the SWMM engine: The program automatically writes each scheduling scheme as a control rule into the SWMM parameter setting module and drives the SWMM engine to perform hydraulic simulation for the next 2 hours.

[0108] Results Extraction and Evaluation: After the simulation, the time series data of each node and pipeline is extracted from the output files (.rpt and .out) of SWMM, and the two objective function values ​​(f1, f2) corresponding to each scheduling scheme are calculated.

[0109] S503, genetic manipulation, the steps are as follows:

[0110] Selection: Select superior individuals from the current population based on Pareto order and crowding calculation;

[0111] Crossover and mutation: Selected individuals are subjected to crossover (crossover rate 0.8) and mutation (mutation rate 0.02) operations to generate a new generation of population;

[0112] S6. Repeat the simulation-optimization process of step S5 until the stopping condition is met;

[0113] Repeat the above simulation-optimization process until the number of iterations reaches the preset 200 generations, or the improvement of the Pareto front is less than 1% for 20 consecutive generations, at which point the loop stops.

[0114] S7. Output the optimal solution and analyze the feedback of results;

[0115] Specifically, in this embodiment, the implementation process of S7 is as follows:

[0116] S701. Solution Output and Execution: After optimization, the system outputs the Pareto optimal solution set. Dispatchers can select the solution with the smallest f1 based on current management preferences, such as prioritizing flood control during severe rainstorms and selecting the solution with the smallest f2 during normal rainfall to consider energy conservation. The final dispatching solution is then selected and sent to the PLCs of each pumping station for execution via the SCADA system.

[0117] S702, Real-time Feedback and Model Updates:

[0118] Real-time correction: During system operation, real-time monitoring data is continuously compared with simulated and predicted values. Once the water level error at a critical node continues to exceed the threshold, the system automatically initiates a short-term re-optimization, using the latest measured data as initial conditions to generate corrected scheduling instructions.

[0119] Effectiveness Analysis and Model Self-Learning: After each rainfall event, the system automatically generates an effectiveness analysis report, comparing the improvements in overflow and energy consumption between the optimized scheduling and traditional rule-based scheduling. Simultaneously, the complete rainfall and operational dataset is stored in a historical database for periodic automatic calibration of the SWMM model parameters, enabling the model to adapt to long-term changes such as pipeline siltation and urban underlying surface variations, thus achieving self-evolution.

[0120] This invention is not limited to the above embodiments, and other intelligent optimization algorithms, such as particle swarm optimization and simulated annealing, can also be used. The objective function and constraints can be adjusted according to actual needs.

[0121] Finally, it should be noted that the above description represents a preferred embodiment of the present invention. It should be pointed out that although preferred embodiments have been described, those skilled in the art, once they understand the basic inventive concept of the present invention, can make various improvements and modifications without departing from the principles described herein. These improvements and modifications should also be considered within the scope of protection of the present invention. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the embodiments of the present invention.

Claims

1. A SWMM model-based combined drainage pipe network pump station group dispatching system, characterized in that, include: The hydraulic modeling module is used to construct hydraulic models of drainage pipe networks and pumping station groups. The hydraulic models include pipe network topology, pumping station characteristics, hydrological data, and sub-catchment parameters. The optimized scheduling model construction module is based on artificial intelligence optimization algorithms and adopts a simulation-optimization and feedforward-feedback composite control mode to establish an optimized scheduling model. The simulation-optimization mode optimizes the scheduling rules through iterative optimization, while the feedforward-feedback mode dynamically adjusts the scheduling strategy by combining rainfall forecast data and real-time monitoring data. The objective function and constraint setting module is used to determine the multi-objective function and constraints of the optimization scheduling model; The initial condition input module is used to input the initial conditions for the operation of the hydraulic model and the initial scheduling rules for the pump station group; the initial conditions include initial water level, flow data, historical rainfall data and rainfall forecast data, and the initial scheduling rules include pump start-up and shutdown rules based on historical experience; The SWMM model engine driver module is used to drive the SWMM model engine, using scheduling rules as control parameters to perform hydraulic simulation calculations and generate simulation results of the hydraulic state of the pipeline network. The simulation-optimization loop module is used to evaluate the simulation results based on the objective function, select the best-performing scheduling scheme through the fitness function, and generate new scheduling rules by combining the selection, crossover, and mutation operations of the intelligent optimization algorithm. The simulation-optimization process is repeated until the stopping condition is met; the stopping condition includes an iteration number threshold or an objective function convergence threshold. The scheduling scheme output and feedback module is used to output the optimal solution set and the corresponding optimal scheduling scheme results, issue the scheduling scheme for execution, and provide feedback on the effectiveness through comparison and analysis of real-time monitoring data and simulation data, and dynamically adjust the model parameters.

2. The SWMM model-based drainage network pump station group joint scheduling system according to claim 1, characterized in that, The artificial intelligence optimization algorithm includes at least one of genetic algorithm, particle swarm optimization algorithm, and neural network algorithm; when a genetic algorithm is used, the scheduling rules are encoded as chromosome vectors, which represent the start-stop state combination of pumps in each pumping station within a preset time period.

3. The SWMM model-based drainage network pump station group joint scheduling system according to claim 1, characterized in that, The multi-objective function includes at least one of minimizing system energy consumption, minimizing overflow, and maximizing drainage efficiency, and the constraints include water balance constraints, pump station capacity constraints, and water level constraints.

4. The SWMM model-based drainage network pump station group joint scheduling system according to claim 1, characterized in that, In the feedforward-feedback composite control mode, the feedforward control generates a pre-scheduling scheme based on future rainfall forecast data, and the feedback control compares the data with the simulated values ​​in real time through the level gauges and flow meters of key nodes. When the deviation exceeds the threshold, it triggers the fine-tuning or re-optimization of the model parameters.

5. A joint scheduling method for drainage network pumping station groups based on the SWMM model, characterized in that, Includes the following steps: Step S1: Perform hydraulic modeling for drainage pipe network and pumping station group, collect geographic information data, pipe network structure data, hydrological and meteorological data and operation data, and construct SWMM model including sub-catchment areas, pipes, inspection wells, pumping stations and regulating reservoirs; Step S2: Based on artificial intelligence optimization algorithms, an optimized scheduling model is established using a simulation-optimization and feedforward-feedback composite control mode. The simulation-optimization mode optimizes the scheduling rules through iterative optimization, while the feedforward-feedback mode dynamically adjusts the scheduling strategy by combining rainfall forecast data and real-time monitoring data. Step S3: Determine the multi-objective function and constraints of the optimization scheduling model; the multi-objective function includes minimizing system energy consumption and minimizing overflow, and the constraints include water balance constraints, pump station capacity constraints, and water level safety range; Step S4: Input the initial conditions for the operation of the hydraulic model and the initial scheduling rules for the pump station group; the initial conditions include the initial water level of the inspection well at the current time, the initial flow rate of the pipeline, the initial water volume of the storage tank, and the rainfall forecast data for the preset duration in the future; the initial scheduling rules include pump start-up and shutdown rules based on historical experience. Step S5: Drive the SWMM model engine to perform simulation calculations, evaluate the calculation results based on the objective function, select the best-performing scheduling scheme through the fitness function, and generate new scheduling rules by combining the selection, crossover, and mutation operations of the intelligent optimization algorithm. Step S6: Repeat the simulation-optimization process of step S5 until the stopping condition is met; Step S7: Output the optimal solution set and the corresponding optimal scheduling scheme results, and issue the scheduling scheme for execution; conduct performance feedback analysis by comparing the real-time monitoring data of key nodes with the simulation data; when the deviation exceeds the preset threshold, start short-term domain re-optimization; after the rainfall event ends, update the historical database and periodically perform automatic calibration of SWMM model parameters.

6. The method for joint scheduling of drainage network pumping station groups based on the SWMM model according to claim 1, characterized in that, The SWMM model constructed by the hydraulic model modeling module, after being calibrated using historical operating data, has a Nash efficiency coefficient (NSE) of ≥0.75 between the simulated data and the actual monitoring data.

7. The method for joint scheduling of drainage network pumping station groups based on the SWMM model according to claim 5, characterized in that, The simulation-optimization mode in step S2 includes iterative optimization of scheduling rules, and the feedforward-feedback mode includes adjusting scheduling strategies based on real-time data.

8. The method for joint scheduling of drainage network pumping station groups based on the SWMM model according to claim 5, characterized in that, The objective function in step S3 is a multi-objective optimization function, and the constraints include water balance constraints, pump station capacity constraints, and water level safety range. The objective function expression for minimizing system energy consumption is: In the formula: The function is the operating cost of the drainage system; I represents the number of time periods; I represents the number of water pumps. It is the power of pump station i in time period t, which is calculated from its head, flow rate and efficiency based on the performance curve; Δt is the scheduling period; The unit electricity cost for time period t; The objective function expression for minimizing the overflow is: In the formula: —The overflow amount of water accumulation node i, where m is the total number of water accumulation nodes; The expression for the water balance constraint is: In the formula: These represent the total water diversion flow of the water system at times t+1 and t, respectively. These represent the total outflow rate of the water system at times t+1 and t, respectively. Unit of time; This refers to the change in water volume within a unit of time. The expression for the pump station capacity constraint is: In the formula: The pumping capacity of the pumping station is (m³ / h). This is the minimum pumping capacity of the pumping station (m³ / h). This is the maximum pumping capacity of the pumping station (m³ / h). The expression for the water level constraint is: In the formula: H(j,t) represents the actual water level of the node or pump station forebay during the scheduling period t (unit: meters); H_min(j) represents the minimum water level allowed for node or pump station j (usually the minimum water level to prevent silt deposition or meet the pump station's suction depth, unit: meters); H_max(j) represents the maximum water level allowed for node or pump station j (usually the maximum warning water level or the elevation of the top of the pump station's inlet pipe, unit: meters).

9. The method for joint scheduling of drainage network pumping station groups based on the SWMM model according to claim 5, characterized in that, The intelligent optimization algorithm in step S5 is a genetic algorithm, which evaluates the scheduling scheme through the fitness function and generates new scheduling rules through crossover and mutation operations.

10. The method for joint scheduling of drainage network pumping station groups based on the SWMM model according to claim 5, characterized in that, The stopping conditions in step S6 include an iteration count threshold or an objective function convergence threshold.

Images