Rain flood resource water shortage-peak clipping double-target cooperative regulation and control method based on distributed virtual reservoir

Abstract

The invention discloses a rainfall flood resource water shortage-peak clipping double-target cooperative regulation and control method based on a distributed virtual reservoir, and the method comprises the following steps: constructing a virtual reservoir according to the division of an administrative region in a drainage basin, and calculating and determining the storage capacity of the virtual reservoir based on the storage capacity of the reservoir in the drainage basin according to an area weight method; collecting monthly water demand data, rainfall data and natural runoff data in the drainage basin, and determining a decision variable as a virtual reservoir to regulate and control the discharged flow; the initial water quantity of the virtual reservoir is calculated, the initial water quantity is determined based on the drainage basin area, the initial precipitation quantity and the natural runoff data, and the surface water accounting for the total water use and demand coefficient is calculated according to the water resource announcement; constructing a multi-objective optimization model, solving the multi-objective optimization model by adopting an NSGA-II algorithm, and generating a Pareto optimal solution set, so as to obtain a monthly optimization regulation and control scheme of the virtual reservoir; the method can be widely applied to rainfall flood resource optimization in the Gangwan river basin and other arid and semi-arid areas.

Description

Technical Field

[0001] This invention relates to the field of water resource utilization engineering technology, specifically to a dual-objective coordinated regulation method for rainwater and flood resources based on a distributed virtual reservoir, addressing both water shortage and peak shaving. Background Technology

[0002] Virtual reservoirs, as a non-physical water resource regulation carrier, simulate the interception, retention, and regulation processes of rainwater and floodwater through hydrological models. They not only overcome the geographical and engineering scale limitations of traditional concrete reservoirs but also flexibly integrate the simulation of complex watershed elements such as silt-retention dams. This effectively solves the generalization problems and insufficient adaptability issues faced by traditional regulation methods in complex watersheds (such as arid and semi-arid regions), becoming a key technological means to achieve efficient utilization of rainwater and floodwater resources. Currently, the formulation of rainwater and floodwater regulation schemes for virtual reservoirs largely relies on empirical formulas or single-objective optimization methods. The former, based on fitting regulation rules to historical regulation data, is difficult to adapt to the uncertainty of water inflow under climate change and does not consider the trade-offs between multiple objectives. The latter, using "maximizing rainwater and floodwater utilization" or "minimizing flood control risk" as a single objective, simplifies calculations but ignores the synergistic needs of multiple objectives such as rainwater and flood control safety and ecological protection. This can easily lead to unbalanced regulation schemes, resulting in an excessive pursuit of rainwater and floodwater utilization, thereby increasing the watershed's flood control risk, or strict control of flood control risk, leading to a decrease in rainwater and floodwater resource utilization. Existing multi-objective water resource optimization methods based on traditional optimization algorithms (such as genetic algorithms) have two limitations: First, the algorithms have low convergence efficiency. In scenarios with multiple decision variables (such as monthly water release, water storage threshold, and water replenishment priority) and multiple constraints (such as reservoir flood control capacity, downstream ecological base flow, and water use guarantee rate) in virtual reservoirs, they are prone to getting trapped in local optima and it is difficult to generate a globally optimal Pareto solution set. Second, the solution set lacks diversity and cannot provide decision-makers with a rich set of objective trade-offs. Especially when climate change leads to volatile water inflow scenarios, the adaptability of a single optimization result is poor. In addition, the formulation of existing virtual reservoir stormwater utilization schemes relies heavily on human intervention, such as adjusting control parameters through expert experience. This is not only highly subjective and inefficient, but also difficult to achieve batch processing and dynamic updates, failing to meet the needs of real-time control of stormwater resources at the watershed scale. Summary of the Invention

[0003] Purpose of the invention: The purpose of this invention is to provide a dual-objective coordinated regulation method for rainwater and flood resources based on distributed virtual reservoirs, which solves the problems existing in the background technology by optimizing the NSGA-II algorithm and improving the regulation logic.

[0004] Technical solution: The present invention provides a method for coordinated regulation of rainwater and flood resources based on a distributed virtual reservoir, targeting both water shortage and peak shaving, comprising the following steps:

[0005] (1) A virtual reservoir is constructed based on the administrative division within the basin, wherein the capacity of the virtual reservoir is determined by the area weighting method based on the capacity of the reservoirs within the basin;

[0006] (2) Collect monthly water demand data, rainfall data and natural runoff data within the watershed, and determine the decision variable as the flow rate of the virtual reservoir for regulation and discharge;

[0007] (3) Calculate the initial water volume of the virtual reservoir, where the initial water volume is determined based on the watershed area, initial precipitation and natural runoff data, and calculate the surface water as a percentage of total water demand based on the water resources bulletin;

[0008] (4) Construct a multi-objective optimization model. The multi-objective optimization model takes minimizing water shortage and maximizing potential peak shaving capacity as objective functions, and uses virtual reservoir carrying capacity constraints, rainwater collection and utilization constraints, ecological flow constraints and non-negativity constraints as constraints.

[0009] (5) The NSGA-II algorithm is used to solve the multi-objective optimization model and generate the Pareto optimal solution set, thereby obtaining the monthly optimization control scheme of the virtual reservoir.

[0010] Further, step (1) is as follows: the basin is divided into upstream parallel virtual reservoirs and downstream virtual reservoirs according to administrative regions; the total area of ​​each administrative region and the total reservoir capacity within the region are calculated, and the capacity of each virtual reservoir is determined according to the area ratio.

[0011] Further, step (2) is as follows: obtain monthly water demand data using historical water use datasets; simulate natural runoff data using hydrological models and verify rainfall runoff data.

[0012] Further, step (3) is as follows: using the time-history method, select the water inflow model for normal water year and moderate dry year, combine the monthly water demand scheme to perform water balance calculation, and determine the initial water storage of the virtual reservoir; based on the water resources bulletin, obtain the water use data and surface water supply of each administrative region, and calculate the surface water proportion coefficient.

[0013] Furthermore, in step (4), the objective function is as follows: Minimize the water shortage:

[0014] ;

[0015] Where T is the total number of calculation periods; This represents the total number of virtual reservoirs within the basin. This is the surface water ratio coefficient; For the first The virtual reservoir basin in the first Water demand during a given time period for Within the virtual reservoir basin, the first Available rainfall during a given time period;

[0016] Maximize potential peak-shaving capability:

[0017] ;

[0018] in, For the first Virtual reservoir capacity over a given time period This refers to precipitation under natural conditions. for The drainage area of ​​the watershed; For the first Outflow rate after reservoir regulation during the specified time period: For characteristic functions: when > If the area is under flood warning, the mitigation benefits need to be calculated; otherwise... The sum reflects the virtual reservoir's ability to reduce flood events.

[0019] Furthermore, in step (4), the constraint formula is as follows:

[0020] Virtual reservoir carrying capacity constraints:

[0021] ;

[0022] in, This represents the maximum carrying capacity of virtual reservoir i.

[0023] Constraints on rainwater harvesting and utilization:

[0024] ;

[0025] in, The actual amount of rainwater collected within the k-basin during time period t. Let be the amount of rainwater that can be collected within watershed k during time period t. The rainfall and runoff discharge within the k-basin during time period t;

[0026] Ecological flow constraints:

[0027] ;

[0028] Nonnegativity constraint:

[0029] ; .

[0030] Further, step (5) is as follows: In the initial population generation stage, a multi-strategy collaborative mechanism is adopted to ensure that the population covers the solution space and improves the quality of the initial solution; in combination with the water carrying capacity threshold of the virtual reservoir and the rain and flood pattern, an initial population with balanced water supply load is constructed; based on the regional water demand distribution characteristics, an initial solution that meets the basic needs of key water use nodes is generated first; with the goal of dynamic balance of water storage in the reservoir group, an initial solution is generated to maintain the water level of each virtual reservoir within the safe operating range.

[0031] Furthermore, step (5) also includes: constructing a multi-criteria evaluation system using the analytic hierarchy process, comprehensively evaluating the Pareto optimal solution set based on the index weights and normalized index values; calculating the comprehensive evaluation score using the weighted summation method, and selecting the unique optimal solution from the monthly Pareto frontier.

[0032] Beneficial Effects: Compared with existing technologies, this invention has the following significant advantages: This invention integrates the dual objectives of water shortage and potential peak-shaving capacity into the water resource optimization model, significantly improving water supply and flood control capabilities. This invention employs the NSGA-II multi-objective evolutionary algorithm to address problems such as uneven Pareto solution set distribution and susceptibility to local optima under multi-objective conditions, improving solution set diversity and global applicability. This invention is adaptable to scenarios such as increased frequency of extreme rainstorms and uneven spatiotemporal distribution of precipitation under climate change, exhibiting stronger robustness. It can adaptively adjust control parameters under different inflow scenarios, generating highly adaptable optimization schemes. This invention improves algorithm convergence speed and the proportion of feasible solutions through multi-source initialization and intelligent repair, adapting to complex real-world constraints. This invention can be widely applied to rainwater and flood resource optimization in the Zhuanlongwan watershed and other arid and semi-arid regions, and is also suitable for other water resource system optimization scenarios with multiple reservoirs, multiple objectives, and complex constraints. Attached Figure Description

[0033] Figure 1 This is a flowchart of the present invention;

[0034] Figure 2 This is a line graph of the virtual reservoir regulation and discharge flow from 1996 to 2023, based on the present invention; wherein, Figure 2 (a) represents the outflow from the virtual reservoir in Dongsheng District from 1996 to 2023. Figure 2 (b) in the figure represents the discharge flow regulated by the virtual reservoir in Ejin Horo Banner; Figure 2 (c) represents the discharge flow regulated by the Kangbashi Virtual Reservoir. Detailed Implementation

[0035] The technical solution of the present invention will be further described below with reference to the accompanying drawings.

[0036] like Figure 1As shown, this invention provides a dual-objective coordinated regulation method for rainwater and flood resources based on a distributed virtual reservoir, addressing both water shortage and peak shaving. The method uses the NSGA-II algorithm to solve a multi-objective optimization regulation model for the water supply system. Taking the efficient utilization of rainwater and flood resources in the Zhuanlongwan watershed from 1996 to 2023 as an application scenario, the method includes the following steps:

[0037] Step S1: Construct a virtual reservoir based on the administrative divisions within the basin. The reservoir capacity is calculated using the area-weighted method based on the reservoir capacity within the basin.

[0038] Step S2: Collect monthly water demand (HSWUD dataset), rainfall data, and natural runoff data to determine the decision variable as the flow rate of the virtual reservoir's regulated discharge.

[0039] ;

[0040] Where R = {Dongsheng Reservoir, Yijinhuoluo Banner Reservoir, Kangbashi Reservoir}, and T is the set of control periods;

[0041] Step S3: Calculate the initial water volume of the virtual reservoir and calculate the surface water as a percentage of total water demand based on the water resources bulletin;

[0042] ;

[0043] Where T is the total number of calculation periods; This represents the total number of virtual reservoirs within the basin. for The drainage area of ​​the watershed; For the first The outflow of natural runoff during a given period. For the first The precipitation in the month preceding the initial date within a virtual reservoir basin.

[0044] The initial water storage capacity of the virtual reservoir was calculated, with the Dongsheng Reservoir having a capacity of 2.16 million cubic meters. 3 The Yijinhuoluo Banner Reservoir has a capacity of 3.19 million cubic meters. 3 The Kangbashi Reservoir has a capacity of 51.43 million cubic meters. 3 According to the water resources bulletin, the multi-year average surface water as a percentage of total water demand is calculated as follows: Dongsheng Reservoir is 0.029, Yijinhuoluo Banner Reservoir is 0.065, and Kangbashi Reservoir is 0.123.

[0045] Step S4: Construct a multi-objective function that includes minimizing water shortage and maximizing potential peak shaving capacity, and apply constraints such as water supply capacity, water balance, demand satisfaction, and non-negativity. Solve the function using the NSGA-II algorithm. Specifically, this includes:

[0046] Step S4.1: The objective function of the multi-objective stormwater resource efficiency utilization model is:

[0047] Objective 1: Minimize water shortage

[0048] ;

[0049] Where T is the total number of calculation periods; This represents the total number of virtual reservoirs within the basin. This is the surface water ratio coefficient; For the first The virtual reservoir basin in the first Water demand during a given time period for Within the virtual reservoir basin, the first Available rainfall during a given period.

[0050] Objective 2: Maximize potential peak shaving capability

[0051] ;

[0052] in, For the first Virtual reservoir capacity over a given time period This refers to precipitation under natural conditions. for The drainage area of ​​the watershed; For the first Outflow rate after reservoir regulation during the specified time period: For characteristic functions: when > , (Under flood warning conditions, reduction benefits need to be calculated); otherwise The sum reflects the virtual reservoir's ability to reduce flood events.

[0053] Step S4.2: Considering constraints such as the effective utilization of water resources and supply-demand balance, the following conditions apply:

[0054] ① Virtual reservoir carrying capacity constraints

[0055] ;

[0056] In the formula, This represents the maximum carrying capacity of virtual reservoir i.

[0057] ②Constraints on rainwater harvesting and utilization

[0058] ;

[0059] In the formula, The actual amount of rainwater collected within the k-basin during time period t. Let be the amount of rainwater that can be collected within watershed k during time period t. Let be the rainfall and runoff discharge within the k-basin during time period t.

[0060] ③ Ecological flow constraints

[0061] ;

[0062] ④ Non-negativity constraint

[0063] ; .

[0064] Step S4.3: The solution process uses the NSGA-II algorithm, specifically including: encoding decision variables into multi-dimensional vector form; using diversified strategies to generate the initial population to ensure the coverage of the solution space; introducing a penalty function mechanism in the fitness evaluation stage to handle constraint violation problems and ensure the feasibility of the solution; designing adaptive crossover and mutation operators, and combining them with feasibility repair operators to optimize the search process and improve the convergence efficiency and optimization ability of the algorithm; dynamically evaluating the convergence and distribution of the solution set through the analytic hierarchy process (AHP); and finally outputting the Pareto optimal solution set that satisfies the optimization objective, as shown in Table 1.

[0065] Table 1. Simulation results of monthly optimized rainwater utilization model schemes from 1996 to 2023 ;

[0066] Depend on Figure 2 It is evident that natural runoff exhibits typical seasonal characteristics: "weak baseflow during dry seasons and abrupt flood peaks during rainy seasons." Regulated discharge significantly mitigates flood peaks through the "peak shaving and valley filling" effect of the virtual reservoir, while maintaining a relatively stable discharge during the dry season. In terms of regional differentiation in the magnitude of regulation, Kangbashi Banner had the highest extreme natural runoff, which decreased to approximately 35 cubic meters per second after regulation, representing a regulation magnitude of 41.7%. Its natural runoff coefficient of variation was 1.52, decreasing to 0.91 after regulation. In Ejin Horo Banner, the extreme natural runoff was approximately 35 cubic meters per second, decreasing to approximately 30 cubic meters per second after regulation, representing a regulation magnitude of 14.3%. The natural runoff coefficient of variation was 1.18, decreasing to 0.75 after regulation. In Dongsheng District, the extreme natural runoff was approximately 30 cubic meters per second, decreasing to approximately 25 cubic meters per second after regulation, representing a regulation magnitude of 16.7%. The natural runoff coefficient of variation was 1.25, decreasing to 0.82 after regulation. This shows that Kangbashi Banner, with its more drastic fluctuations in natural runoff, has the largest range of regulation; the virtual reservoir has effectively reduced the runoff variation coefficient and improved the availability of water resources in all three regions.

[0067] From the long-term series of 1996 to 2023, the regulatory effect of the virtual reservoir exhibits interannual stability: it can continuously achieve "peak shaving and valley filling" regardless of whether it is a high-water year or a low-water year. For example, in years with concentrated flood peaks such as 2005 and 2017, the peak values ​​of the regulated discharge in the three locations were significantly lower than the natural runoff; while in low-water years such as 2011, the regulated discharge was still able to maintain a basic discharge, avoiding the risk of interruption of natural runoff.

[0068] The results of multi-objective optimization from 1996 to 2023 show that, in terms of the spatiotemporal distribution characteristics of water shortage and potential peak shaving capacity, water shortage exhibits significant interannual and intra-annual differences in time dimension. During the period from 1996 to 2023, the peak water shortage in some years (such as 2005 and 2006) exceeded 35 million cubic meters, and the distribution of water shortage in each month of the year was uneven. The water shortage in summer (June-August) was generally at a low level, while the water shortage in spring and winter (January-March and November-December) was often in a higher range. Potential peak shaving capacity also showed obvious spatiotemporal differentiation, with a value range of 0-50 cubic meters per second. Interannually, the potential peak shaving capacity was relatively strong in some years (such as 2017 and 2018), reaching more than 40 cubic meters per second. In terms of intra-annual distribution, the potential peak shaving capacity was relatively prominent in summer (June-August) and relatively weak in winter (December-February). In summary, water shortage and potential peak shaving capacity are complementary in their spatiotemporal distribution. That is, during periods or years with high water shortage, the potential peak shaving capacity often shows a corresponding trend. This provides data support for the regulation and management of regional water resources, helps to further optimize the scheduling strategy of virtual reservoirs, and achieve the coordinated promotion of efficient water resource utilization and flood control and disaster reduction.

Claims

1. A method for coordinated regulation of rainwater and flood resources based on a distributed virtual reservoir, characterized in that: Includes the following steps: (1) A virtual reservoir is constructed based on the administrative division within the basin, wherein the capacity of the virtual reservoir is determined by the area weighting method based on the capacity of the reservoirs within the basin; (2) Collect monthly water demand data, rainfall data and natural runoff data within the watershed, and determine the decision variable as the flow rate of the virtual reservoir for regulation and discharge; (3) Calculate the initial water volume of the virtual reservoir, where the initial water volume is determined based on the watershed area, initial precipitation and natural runoff data, and calculate the surface water as a percentage of total water demand based on the water resources bulletin; (4) Construct a multi-objective optimization model. The multi-objective optimization model takes minimizing water shortage and maximizing potential peak shaving capacity as objective functions, and uses virtual reservoir carrying capacity constraints, rainwater collection and utilization constraints, ecological flow constraints and non-negativity constraints as constraints. (5) The NSGA-II algorithm is used to solve the multi-objective optimization model and generate the Pareto optimal solution set, thereby obtaining the monthly optimization control scheme of the virtual reservoir.

2. The method for coordinated regulation of rainwater and flood resources based on a distributed virtual reservoir, with dual objectives of water shortage and peak shaving, as described in claim 1, is characterized in that... Step (1) is as follows: Divide the basin into upstream parallel virtual reservoirs and downstream virtual reservoirs according to administrative regions; calculate the total area of ​​each administrative region and the total reservoir capacity within the region, and determine the capacity of each virtual reservoir based on the area ratio.

3. The method for coordinated regulation of rainwater and flood resources based on a distributed virtual reservoir, with dual objectives of water shortage and peak shaving, as described in claim 1, is characterized in that... Step (2) is as follows: Use historical water use datasets to obtain monthly water demand data; use hydrological models to simulate natural runoff data and verify rainfall runoff data.

4. The method for coordinated regulation of rainwater and flood resources based on a distributed virtual reservoir, with dual objectives of water shortage and peak shaving, as described in claim 1, is characterized in that... Step (3) is as follows: Using the time-history method, select the water inflow model for normal water year and moderate dry year, combine the monthly water demand scheme to perform water balance calculation, and determine the initial water storage of the virtual reservoir; obtain water use data and surface water supply of each administrative region based on the water resources bulletin, and calculate the surface water proportion coefficient.

5. The method for coordinated regulation of rainwater and flood resources based on a distributed virtual reservoir, with dual objectives of water shortage and peak shaving, as described in claim 1, is characterized in that... In step (4), the objective function is as follows: Minimize the water shortage: ； Where T is the total number of calculation periods; This represents the total number of virtual reservoirs within the basin. This is the surface water ratio coefficient; For the first The virtual reservoir basin in the first Water demand during a given time period for Within the virtual reservoir basin, the first Available rainfall during a given time period; Maximize potential peak-shaving capability: ； in, For the first Virtual reservoir capacity over a given time period This refers to precipitation under natural conditions. for The drainage area of ​​the watershed; For the first Outflow rate after reservoir regulation during the specified time period: For characteristic functions: when > If the area is under flood warning, the mitigation benefits need to be calculated; otherwise... The sum reflects the virtual reservoir's ability to reduce flood events.

6. The method for coordinated regulation of rainwater and flood resources based on a distributed virtual reservoir, with dual objectives of water shortage and peak shaving, as described in claim 1, is characterized in that... In step (4), the constraint formula is as follows: Virtual reservoir carrying capacity constraints: ； in, This represents the maximum carrying capacity of virtual reservoir i. Constraints on rainwater harvesting and utilization: ； in, The actual amount of rainwater collected within the k-basin during time period t. Let be the amount of rainwater that can be collected within watershed k during time period t. The rainfall and runoff discharge within the k-basin during time period t; Ecological flow constraints: ； Nonnegativity constraint: ； 。 7. The method for coordinated regulation of rainwater and flood resources based on a distributed virtual reservoir, with dual objectives of water shortage and peak shaving, as described in claim 1, is characterized in that... Step (5) is as follows: In the initial population generation stage, a multi-strategy collaborative mechanism is adopted to ensure that the population covers the solution space and improves the quality of the initial solution; in combination with the water carrying capacity threshold of the virtual reservoir and the rain and flood pattern, an initial population with balanced water supply load is constructed; based on the regional water demand distribution characteristics, an initial solution that meets the basic needs of key water use nodes is generated first; with the goal of dynamic balance of water storage in the reservoir group, an initial solution is generated to maintain the water level of each virtual reservoir within the safe operating range.

8. A method for coordinated regulation of rainwater and flood resources based on a distributed virtual reservoir, with dual objectives of water shortage and peak shaving, as described in claim 1, is characterized in that... Step (5) also includes: constructing a multi-criteria evaluation system using the analytic hierarchy process, comprehensively evaluating the Pareto optimal solution set based on index weights and normalized index values; calculating the comprehensive evaluation score using the weighted summation method, and selecting the unique optimal solution from the monthly Pareto frontier.

Images