Water resource optimization regulation and control method for long-vine melon-bearing water network irrigation system

Abstract

The invention discloses a water resource optimization regulation and control method for a long-vine melon-bearing water network irrigation system. The method comprises the following steps: (1) establishing a generalized and network topological relation between a water supply end, a water taking end and a water transmission and distribution end of the long-vine melon-bearing water network system; (2) designing a three-layer hierarchical control structure of'total adjustment, sub-adjustment and self-adjustment 'according to the concept of'multi-level hierarchical control structure'; wherein the total regulation layer is used for dispatching a backbone water source, and water is lifted to a key node of each backbone branch canal through a canal head pump station; the water quantity is dispatched to a water distribution point of each branch canal by the'branch dispatching layer '; the'self-adjusting layer 'forms a point-to-point relationship among a plurality of water distribution points of the main and branch canals, local ponds, namely low watermelons, and water users, so that the redistribution of water is carried out; (3) constructing a long-vine melon-bearing reservoir-canal combined multi-target water resource optimization regulation and control model, and performing solution by using NSGA-II, the model comprising: determining decision variables, establishing a target function, and constructing constraint conditions; scientific basis is provided for water resource optimization regulation and control of the long-vine melon-bearing water network irrigation system.

Description

Technical Field

[0001] The invention relates to the technical field of water resource management and regulation, and in particular to a water resource optimization regulation method for a long-vine melon water network irrigation system. Background Art

[0002] The "long vine and melon" irrigation system is a common water resource management model in the southern hilly areas. It has the advantages of widely opening up water sources, uniformly allocating resources, expanding irrigation benefits, and reducing the operating costs of backbone channels. Through the canal head backbone water supply project, the problem of uneven spatial distribution of water resources can be effectively solved, while the small and medium-sized water source projects in the irrigation area use the "filling water in idle time and irrigating fields in busy time" model to adjust the imbalance in time distribution through the storage mechanism of reservoirs and ponds. The precise joint dispatch of reservoir-channel water resources is of great significance in ensuring urban and rural water use safety, coordinating water and soil resources, and promoting high-quality development of irrigation areas.

[0003] Different from general water resource dispatching research, the optimization and allocation of multiple water sources in reservoir-canal connected irrigation areas faces its own unique complexity, uncertainty and particularity. The complexity is reflected in the fact that the problem spans multiple fields and tasks, involving multiple links such as runoff forecasting, irrigation demand forecasting, reservoir dispatching, water resource allocation and water transmission dispatching; the system needs to take into account irrigation, water supply, power generation and ecological needs, and the "water lifting-regulation-water storage-transmission and distribution" process in each link makes the problem more difficult. Uncertainty comes from the diverse forms of reservoirs and ponds, the complex topological structure of the water transmission and distribution system, and the scattered and diverse water supply objects. At the same time, climate change and social and economic development make the prediction of water resource supply and demand full of challenges. The particularity is that this problem is a staged, dynamically linked, ultra-high-dimensional nonlinear optimization problem, which contains many dispatching control nodes. If it is processed in blocks, it is difficult to solve the overall problem. If a global solution is performed, it is easy to cause the curse of dimensionality or local optimal problems. Summary of the invention

[0004] Purpose of the invention: The purpose of the present invention is to provide a method for optimizing the regulation of water resources in a long-vine melon water network irrigation system, so as to solve the problems of complex water resource regulation and low utilization efficiency in the existing long-vine melon water network irrigation system.

[0005] Technical solution: The water resource optimization and control method of the long-vine melon water network irrigation system described in the present invention comprises the following steps:

[0006] (1) Establish the generalization and network topology relationship between the water supply end, water intake end and water transmission and distribution end of the Changtengjiegua water network system;

[0007] (2) Based on the concept of "multi-level hierarchical control structure", a three-layer hierarchical control structure of "general regulation-divided regulation-self-regulation" is designed; the "general regulation layer" is the backbone water source regulation, which pumps water to the key nodes of each backbone branch canal through the headwater pumping station; the "divided regulation layer" dispatches water to the water diversion points of each branch canal; the "self-regulation layer" establishes a point-to-point relationship between the many water diversion points of the trunk and branch canals, local reservoirs and ponds, and water users, thereby redistributing water;

[0008] (3) A multi-objective water resources optimization and regulation model for the Changtengjieguakuqu was constructed and solved using NSGA-II. The model includes: determining decision variables, establishing objective functions, and constructing constraints.

[0009] Furthermore, step (1) includes the following steps:

[0010] (11) The water supply end is generalized into several nodes, each node represents a different water source type and location; for the river water supply system, the generalized nodes represent the main river estuary locations; for the reservoir water supply system, the nodes represent the water intake locations of the reservoir; the nodes of the groundwater supply system represent the main water wells; each water supply node is classified and connected according to its geographical location, water supply capacity and the relationship between them.

[0011] (12) The water intake end is generalized according to functional requirements. Different water intake end nodes represent different user groups or areas. Each water intake node is numbered according to the actual water demand, describing its hydraulic connection relationship with the water supply end, and setting a transmission channel with the water supply end.

[0012] (13) The water transmission and distribution ends are connected through canals and pipelines, simplifying the complex water transmission and distribution system into several main water transmission trunks and branch pipelines; the flow, capacity and hydraulic conditions of each water transmission route are modeled according to the terrain and engineering conditions, and hierarchical management is carried out;

[0013] (14) Generalization of water replenishment logic First, the water source volume at the water supply node and the water demand at the water intake node are monitored in real time to generate a supply-demand balance state. At the same time, the regional control priority is set according to the supply and demand situation, and the transmission path is adjusted in real time through a dynamic feedback mechanism to optimize the efficiency of water resource allocation.

[0014] (15) By generalizing the water supply end, water intake end, and water transmission and distribution end, each node is connected to form a complete water network system; the network topology is constructed using graph theory methods, each node represents the water supply, water intake, or water transmission and distribution end, and the edge represents the path of water flow; each edge is attached with flow and pressure loss parameters.

[0015] Furthermore, step (2) includes the following steps:

[0016] (21) The overall dispatching layer is the backbone water source dispatching: At this level, the water source dispatching of the headwater pumping station is carried out by controlling the water volume Q 总调 (t) The formula for key nodes supplying trunk and branch canals is:

[0017] Q 总调 (t) = f 1 (t,Q 输入 ,X 渠首泵站 ) (1.1)

[0018] Where t is the time variable, representing the time-varying nature of the system; Q 输入 is the water source input flow, X 渠首泵站 It is the status of the headworks pumping station (pumping station switch status, pumping station capacity, etc.).

[0019] (22) Branch canal dispatching: At this level, the water volume of branch canals is dispatched and controlled to each water diversion point Q 分调 (t), the water volume is provided by the output of the total adjustment layer, the formula is as follows:

[0020] Q 分调 (t) = f 2 (t,Q 总调 (t),X 支渠 ) (1.2)

[0021] In the formula, Q 总调 (t) is the amount of water input from the upper layer; X 支渠 It is the state variable of the branch canal (including the flow capacity, leakage, flow distribution, etc. of the canal).

[0022] (23) Self-regulating layer, i.e. point-to-point dispatching: The self-regulating layer is the redistribution of water between water diversion outlets, local reservoirs or ponds and water users. The formula is as follows:

[0023] Q 自调 (t) = f 3 (t,Q 分调 (t),X 库塘 ,X 用水户 ) (1.3)

[0024] In the formula, Q 分调 (t) is the amount of water input from the sub-adjustment layer; X 库塘 is the state variable of the local reservoir or pond; X 用水户 It is the needs and status of water users (including user water consumption, water intake points, etc.).

[0025] (24) Construct a multi-level hierarchical control structure for the entire system: The water dispatching of the entire system can be expressed as a multi-layer feedback control system through hierarchical control theory: The overall mathematical expression of the system is described by a nested control equation:

[0026] Q(t)=f 3 (t,f 2 (t,f 1 (t,Q 输入 ,X 渠首泵站 ),X 支渠 ),X 库塘 ,X 用水户 ) (1.4)

[0027] Furthermore, in step (3), the decision variables are as follows: the hydrological year is used as the scheduling period, and the scheduling period is T;

[0028] Taking the optimized reservoir water supply W, optimized reservoir discharge flow Q, channel filling volume C to the Digua Reservoir and channel direct irrigation water volume B to farmland in each period as decision variables, after introducing the regulation period, the total number of decision variables of this multi-objective optimization problem is (W+Q+C+B)×T.

[0029] Furthermore, in step (3), the objective function is as follows: the minimum sum of squares of water shortage rates in the regional period is taken as the water supply target to coordinate the competitive relationship between the water supply target and other targets. The mathematical expression is:

[0030]

[0031] Where: D k,t is the water demand of the kth water user in the tth period; Q i,k,t is the water supply from the ith water supply source in the region to the kth water user in the tth period; α j,k The importance coefficient of the kth water user to get water supply priority relative to other water users.

[0032] The mathematical expression of the economic target is:

[0033]

[0034] Where: T 1 is the specific water inflow peak period 1, T 2 is the specific peak water supply period2; β is the weight of the pumping station water extraction cost in time allocation, β 1 (Replenish water during period 1) <β 2 (Water replenishment in period 2). M is the number of pumping stations, m = 1, 2, ..., M; N is the number of units, n = 1, 2, ... N; ρ is the density of water; q t,m,n is the pumping flow of the nth unit of the mth pumping station in the tth period; H t,m is the head of the mth pumping station in the tth period; η t,m,n The efficiency of the nth unit of the mth pumping station in the tth period; Δt is the time interval; c t is the electricity price in period t.

[0035] Furthermore, in step (3), the constraints include:

[0036] Reservoir water balance constraints

[0037] V t -V t-1 =(WR t +B t -W t -Q t -C t )Δt (1.7)

[0038] Where: V t 、V t-1 is the reservoir storage at the end and beginning of the tth period; WR t is the average inflow in the tth period. For the Quanmutang Reservoir, the water volume lifted from the dam needs to be subtracted from the natural runoff. t is the water volume of the channel filling the Digua Reservoir in the tth period (if any); C t Q is the amount of water released from Gaogua Reservoir to the channel during the tth period (if any); t is the discharge flow of the reservoir in the tth period; W t is the water supply of the reservoir in the tth period;

[0039] Reservoir water supply constraints

[0040] W t ≤W kg (1.8)

[0041] Where: W kg It is the water supply of the reservoir in the tth period in the design level year.

[0042] Reservoir outflow restrictions

[0043] Q t,min ≤Q t ≤Q t,max (1.9)

[0044] Where: Q t,min is the minimum allowable outflow during period t, which is set as the ecological base flow under the reservoir dam; Q t,max is the maximum allowable outflow during period t, i.e. the maximum discharge of the reservoir;

[0045] Reservoir water level restrictions

[0046] Z t,min ≤Z t ≤Z t,max (1.10)

[0047]

[0048] Downstream flow rate fluctuation constraint

[0049] |Q t -Q t-1 |≤Δ (1.12)

[0050] Where: Δ is the maximum amplitude of reservoir outflow;

[0051] Water balance constraints of water diversion section

[0052] O t =I t -W t (1.13)

[0053] Where: O t ,I t , W t are the outflow, inflow and water diversion of the river section in the tth period respectively.

[0054] Section water balance constraints

[0055]

[0056] Where: Q i,t is the average flow rate of the i-th section in time period t; Q i-1,t-1 is the average flow rate of the i-1th section in the t-1 period; Q i取,t , Q i入,t , Q i退,t , Q i损,t are the water intake flow, interval inflow, water withdrawal flow and loss flow of the river section from i-1 to i in the period t-1 respectively; τ i,t is the propagation time of the flow from the i-1th to the i-th section in the t period;

[0057] Section ecological flow constraints

[0058] O t ≥Q t,e (1.15)

[0059] Where: O t is the discharge flow of the river section in period t; Q t,e It is the minimum or suitable ecological flow of the river section;

[0060] Channel flow balance constraints

[0061] Q j,t =Q j-1,t +Q l,t (1.16)

[0062] Q l,t =βALQ j-1,t1-m / 100 (1.17)

[0063] Where: Q j,t is the gross flow of the first section of the j-th water conveyance channel during period t; Q j-1,t is the net flow at the end of the j-th water channel (or the beginning of the j-1-th water channel) in period t; Q l,t is the water loss of the j-th channel in period t; it is estimated according to the empirical formula, where L is the channel length, km; A is the water permeability coefficient of the channel bed soil, light clay, A = 2.65; m is the water permeability index of the channel bed soil, medium permeability, m = 0.45; β is the reduction coefficient of the channel bed leakage after anti-seepage measures are taken, the channel adopts concrete surface protection, β = 0.15;

[0064] Channel water distribution flow constraints

[0065] q min ≤q t ≤q max (1.18)

[0066] Where: q t is the channel water distribution flow in period t; q max ,q min The upper and lower limits of the water flow rate at each water intake of the channel;

[0067] Channel water capacity constraints

[0068] Q j,t ≤Q j,cmax (1.19)

[0069] Where: Q j,t is the flow rate of the jth channel section in the tth period; Q j,cmax is the corresponding channel flow capacity.

[0070] Pump station head constraints

[0071] H m,min ≤H m,t ≤H m,max (1.20)

[0072] Where: H m,min , H m,max is the minimum and maximum head of the mth pump station; H m,t is the head of the mth pumping station during period t;

[0073] Pumping station overcurrent capacity constraints and power constraints

[0074] q n,min ≤q n,t ≤q n,max (1.21)

[0075] ξ n,t ≤ξ n,max (1.22)

[0076] Where: q n,min ,q n,max are the minimum and maximum pumping flow rates of the nth unit respectively; q n,t is the pumping flow of the nth unit in period t; ξ n,max is the maximum power of the nth unit, ξ n,t is the pumping power of the nth unit in period t;

[0077] Water demand constraints

[0078]

[0079] Water supply guarantee rate constraint

[0080]

[0081] Where: p k is the water supply guarantee rate of the kth water user; is the design guarantee rate of the kth water user.

[0082] Furthermore, in step (3), the model solution process is as follows: The optimization of water resources in the long-vine melon irrigation water network system involves a multi-objective nonlinear problem. A global search method is used to avoid falling into the "dead loop" problem of the local minimum. The overall process is expressed as:

[0083] P t+1 =Selection(P t )+Crossover(P t )+Mutation(P t ) (1.25)

[0084] Furthermore, a multi-objective genetic algorithm based on non-dominated sorting (NSGA-II) is selected to solve the model, as follows: First, the fitness function of non-dominated sorting is constructed:

[0085]

[0086] Where R(x) represents the non-dominated level of individual x; the goal of NSGA-II is to find an approximate non-dominated solution set S through the fitness function nd , and keeps approaching the true non-inferior solution set S.

[0087] Then, the projection pursuit optimization model is used to solve the solution set S nd Evaluate and sort the various solutions in the high-dimensional space; suppose the data in the high-dimensional space is X=(x 1 ,x2 ,…,x d ), the projection direction is ω, then the projection value in the low-dimensional space is:

[0088]

[0089] In the formula, ω T is the transpose of the projection vector, ω i is the weight in the projection direction, and x i is the eigenvalue in high-dimensional data;

[0090] The optimal projection vector ω is found in the iterative process. The process is expressed as:

[0091] max w Q(ω,X)(1.28) where Q(ω,X) represents the objective function of projection pursuit, reflecting the quality of the projection result in the low-dimensional space; by maximizing Q(ω,X), we can find the optimal projection vector ω * , thereby ensuring that the data structure in the low-dimensional space can most effectively reflect the characteristics of the high-dimensional data; the optimal projection direction and the corresponding projection value obtained can be used for subsequent analysis, visualization, and decision support, thereby achieving the solution set S nd Effective evaluation and ranking of solutions.

[0092] An electronic device described in the present invention includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program, when loaded into the processor, implements any one of the methods for optimizing and regulating water resources in a long-vine melon water network irrigation system.

[0093] A storage medium described in the present invention stores a computer program, characterized in that when the computer program is executed by a processor, it implements any one of the methods for optimizing and controlling water resources in a long-vine melon water network irrigation system.

[0094] Compared with the prior art, the present invention has the following beneficial effects: The present invention simplifies the complex water resource system structure and clarifies the relationship between the various components by generalizing and designing the network topology of the water supply end, water intake end and water transmission and distribution end of the long vine melon water network irrigation system. The multi-level canal hierarchical control structure is adopted to decompose the complex water conservancy system into multiple levels, realize hierarchical optimization control, and improve the efficiency and accuracy of water resource scheduling; the constructed reservoir-channel joint multi-objective water resource optimization and control model comprehensively considers multiple objectives such as water supply and economy, takes into account the needs of various water users and the operating costs of the pumping station, and realizes the optimal allocation of water resources. The application of advanced multi-objective genetic algorithms and multi-attribute decision-making methods improves the efficiency of model solving and the reliability of results, providing a scientific basis for the optimization and control of water resources in the long vine melon water network irrigation system. BRIEF DESCRIPTION OF THE DRAWINGS

[0095] Figure 1 It is a schematic diagram of the process of the present invention;

[0096] Figure 2 It is a regional schematic diagram of the present invention;

[0097] Figure 3-1 to Figure 3-2 The water resource system overview diagram and canal water distribution diagram of the present invention;

[0098] Figures 4-1 to 4-4 It is a multi-objective Pareto front diagram under various water inflow frequencies of the present invention;

[0099] Figure 5 is the target projection vector at each water inflow frequency of the present invention;

[0100] Figures 6-1 to 6-3 It is a typical reservoir dispatching process line of the present invention. DETAILED DESCRIPTION

[0101] The technical solution of the present invention is further described below in conjunction with the accompanying drawings.

[0102] like Figure 1 As shown, the embodiment of the present invention provides a method for optimizing and controlling water resources in a long-vine melon water network irrigation system, comprising the following steps:

[0103] S10, generalization of the water supply end, water intake end and water transmission and distribution end of the Changtengjiegua water network system and network topology; including the following steps: S11, generalization of the water supply end

[0104] The water supply end includes multiple water sources such as rivers, reservoirs, and groundwater sources. In order to simplify the complexity of the water network, the water supply end is generalized into several nodes, each of which represents a different type and location of water sources. For the river water supply system, the generalized nodes can represent the main river estuary locations; for the reservoir water supply system, the nodes represent the water intake locations of the reservoir; and the nodes of the groundwater supply system represent the main water wells. Each water supply node is classified and connected according to its geographical location, water supply capacity, and relationship with each other.

[0105] S12, generalization of water intake end

[0106] The water intake is a key node for water resource use, which usually includes various demand types such as farmland irrigation, urban water supply, and industrial water use. For the convenience of management and scheduling, the water intake is generalized according to functional requirements, and different water intake nodes represent different user groups or regions. Each water intake node will be numbered according to the actual water demand, describe its hydraulic connection relationship with the water supply end, and set up a transmission channel with the water supply end.

[0107] S13, generalization of water transmission and distribution

[0108] The water transmission and distribution end is the intermediate link from the water supply end to the water intake end, mainly including various water transmission channels, storage tanks, distribution pumping stations, etc. The water transmission and distribution end is connected through channels and pipelines, simplifying the complex water transmission and distribution system into several main water transmission trunks and branch pipelines. The flow, capacity and hydraulic conditions of each water transmission path are modeled according to the terrain and engineering conditions, and hierarchical management is carried out to ensure that the water transmission channel system at all levels can effectively transfer water resources.

[0109] S14. Generalization of water filling and replenishment logic

[0110] The logic of water replenishment and replenishment is summarized by dynamically monitoring and regulating the water supply end, water intake end and water transmission and distribution end to ensure the balance of water resource allocation at each node in the system. First, the water source volume of the water supply node and the water demand of the water intake node are monitored in real time to generate a supply and demand balance state. When a water intake node is short of water resources, the system identifies the water supply node and generates the optimal water resource allocation path to efficiently transport water resources to the node that needs water replenishment. At the same time, the regional regulation priority is set according to the supply and demand situation to ensure that the emergency area gets water resource replenishment first, and the transmission path is adjusted in real time through the dynamic feedback mechanism to optimize the efficiency of water resource allocation and ensure the rational use and continuous balance of water resources in the system.

[0111] S15. Network topology construction

[0112] By generalizing the water supply end, water intake end and water distribution end, each node is connected to form a complete water network system. The topological structure of the network can reflect the path, flow rate and mutual influence between nodes of water resources flowing from the water supply node to the water intake node. The network topology is constructed using graph theory. Each node represents the water supply, water intake or water distribution end, and the edge represents the path of water flow. Each edge is attached with parameters such as flow rate and pressure loss for analysis and calculation in the subsequent water resource optimization model.

[0113] S20. Multi-level canal hierarchical control structure design of long vine melon water network irrigation system; including the following steps: S21. General adjustment layer (backbone water source adjustment)

[0114] In this embodiment, the master regulation layer of the Changtengjiegua water network irrigation system realizes the water source dispatching of the key nodes of the backbone and branch canals through the canal head pump station. 输入 ) is obtained from natural water sources (such as rivers or reservoirs) and lifted to the trunk and branch canals through the headwater pumping station. The switch status and flow control of the pumping station are dynamically adjusted according to the real-time water resource demand and system status.

[0115] The status parameters of the pump station include the switch status of the pump station (operating or closed), the capacity of the pump station (maximum water volume that can be lifted), and the operating time. The water volume scheduling formula for this part is:

[0116] Q 总调 (t) = f 1 (t,Q 输入 ,X 渠首泵站 ) (2.1)

[0117] During implementation, the system will use sensors to monitor the water level and flow changes at each key node in real time, adjust the working status of the headwater pump station based on the regional water demand and river flow fluctuations, and ensure that the trunk and branch canals receive sufficient water supply.

[0118] S22, sub-dispatching layer (branch canal dispatching)

[0119] In the sub-regulation layer of this embodiment, the water volume dispatching of the branch canal system is adjusted by the output water volume of the upper general regulation layer.

[0120] The water volume of the branch canal flows into the main regulation point and is distributed to the lower-level ditches and farmland irrigation areas through the diversion outlets of the branch canal. The state variables of the branch canal include the water delivery capacity, leakage and flow distribution ratio of the branch canal.

[0121] Each branch canal in the system is equipped with an automated control device (such as a gate or flow meter) to monitor and regulate the water distribution at the diversion outlet. The control formula for this process is:

[0122] Q 分调 (t) = f 2 (t,Q 总调 (t),X 支渠 )(2.2)

[0123] The water allocation in the branch canal is dynamically adjusted according to the irrigation needs of the farmland, and real-time monitoring and feedback are carried out using an information management platform to ensure that each water outlet can receive a reasonable water supply.

[0124] S23, self-regulating layer (branch canal dispatching)

[0125] The self-regulation layer is the lowest water resource scheduling level in this embodiment, responsible for further distributing water from the diversion outlets of the branch canals and local reservoirs (or ponds) to water users. This level adopts a point-to-point water resource allocation method to ensure that each water user obtains the appropriate amount of water according to its actual needs.

[0126] At the self-regulation level, the system monitors the water level of local reservoirs and ponds and the needs of water users in real time to achieve accurate water scheduling. The water volume control formula at this level is:

[0127] Q 自调 (t) = f 3 (t,Q 分调 (t),X 库塘 ,X 用水户 ) (2.3)

[0128] Among them, the state variables of reservoirs include current water level, storage capacity, evaporation and other factors; the status of water users includes the user's water consumption, crop type and irrigation cycle.

[0129] S24, system overall multi-level hierarchical control

[0130] The entire system adopts a hierarchical control structure, which gradually distributes water from the general adjustment layer to the self-adjustment layer. The general adjustment layer controls the water volume of the headwater pump station according to the input of the backbone water source; the sub-adjustment layer then transports the water volume to the water outlets of each branch canal; and the self-adjustment layer transports water resources from the branch canal to farmland and water users.

[0131] The system uses hierarchical control and feedback mechanisms to ensure that each level of the system can regulate water volume according to actual needs, and coordinate the lower subsystems through the upper system to achieve overall optimization of the entire water network system. The overall mathematical model of the system is:

[0132] Q(t)=f 3 (t,f 2 (t,f 1 (t,Q 输入 ,X 渠首泵站 ),X 支渠 ),X 库塘 ,X 用水户 ) (2.4)

[0133] This hierarchical control structure realizes the dynamic optimization allocation of water resources in time and space, effectively improving the efficiency and water resource utilization rate of the long-vine melon water network irrigation system.

[0134] Through the multi-level hierarchical control structure of the present invention, the long-vine melon water network irrigation system can effectively respond to changes in regional water resource demand, reasonably allocate water supply, reduce water resource waste, and ensure coordinated optimization of systems at all levels to improve overall water resource management efficiency.

[0135] S30, construct a multi-objective water resources optimization and control model for Changtengjieguakuqu. Based on the multi-objective optimization model, efficient control of water resources in the system is achieved. The scheduling cycle of this method is based on the hydrological year, and the scheduling period is T. By optimizing the reservoir water supply (W t ), reservoir discharge (Q t ), the amount of water discharged from the channel to the Digua Reservoir (C t ) and the amount of direct irrigation water from the channel to the farmland (B t ) to conduct multi-objective water resource scheduling. It includes the following steps:

[0136] S31. Setting of decision variables

[0137] In the present invention, the decision variables include:

[0138] Reservoir water supply W t : Refers to the water supply of each reservoir in the system during each time period to ensure that the needs of water users are met.

[0139] Reservoir discharge flow Q t : The amount of water discharged from the reservoir to the downstream river or other water bodies to ensure the ecological flow requirements.

[0140] Water filling amount C t : The amount of water that the channel fills into the Digua Reservoir meets the region's water resource allocation needs.

[0141] Direct irrigation water volume B t : The amount of irrigation water directly supplied from the channel to farmland to meet agricultural water needs.

[0142] The total number of decision variables is (W+Q+C+B)×T, and is adjusted dynamically as the regulation period changes.

[0143] S32. Objective function

[0144] In order to achieve multi-objective water resources optimization, this paper introduces two main objective functions:

[0145] (1) Water supply target

[0146] The optimization of water supply target aims to ensure that the water demand in the region is met to the greatest extent by minimizing the sum of the squares of the regional water shortage rate in each period. The mathematical expression of the water supply target is:

[0147]

[0148] Where: D k,t is the water demand of the kth water user in the tth period; Q i,k,t is the water supply from the ith water supply source in the region to the kth water user in the tth period; α j,k The importance coefficient of the kth water user to get water supply priority relative to other water users.

[0149] (2) Economic objectives

[0150] In order to reduce the operating cost of the pump station, the present invention optimizes the operating time and electricity cost of the pump station during different water source peak hours. The mathematical expression of the economic goal is:

[0151]

[0152] Where: T 1 is the specific water inflow peak period 1, T 2 is the specific peak water supply period2; β is the weight of the pumping station water extraction cost in time allocation, β 1 (Replenish water during period 1) <β2 (Water replenishment in period 2). M is the number of pumping stations, m = 1, 2, ..., M; N is the number of units, n = 1, 2, ... N; ρ is the density of water; q t,m,n is the pumping flow rate; H t,m is the pump station head; η t,m,n is the unit efficiency; Δt is the time interval; c t is the electricity price in period t.

[0153] S33. Constraints

[0154] The multi-objective optimization scheduling model in the present invention also needs to meet a series of constraints to ensure that the actual operation of the system meets the requirements. The water volume of the reservoir needs to maintain a dynamic balance, and the inflow, water supply, discharge flow and filling volume should be reasonably matched to ensure the stability of the reservoir water storage. The water supply of the reservoir needs to be less than the designed water supply. At the same time, the outflow of the reservoir should vary between the minimum ecological flow and the maximum allowable discharge to avoid affecting the ecosystem and flood control safety. The water level of the reservoir should be kept between the upper and lower limits specified, and the change range of the outflow flow cannot exceed the set value. The river section needs to maintain water balance, the outflow flow should meet the ecological flow requirements, and the flow during the process of water inflow, water intake, and water withdrawal needs to be balanced. The flow balance in the channel system is equally important. The flow of each section should be operated within the design capacity, the loss should be minimized during the water delivery process, and the water intake flow must meet the upper and lower limit requirements. The head, flow and power of the pump station need to operate within the design capacity of the equipment to ensure safe and efficient operation of the system. Finally, it is necessary to ensure that the water needs of water users are met, and the water supply guarantee rate should meet the design standards.

[0155] S34, multi-objective intelligent optimization solution

[0156] The present invention aims at solving the multi-objective water resource optimization problem in the long-vine melon water network irrigation system by using an intelligent optimization algorithm to ensure that the system can operate efficiently under complex conditions. The goal of multi-objective intelligent optimization is to consider multiple factors such as water supply demand, economic cost and ecological protection at the same time to find an optimal or suboptimal solution.

[0157] S35. Multi-attribute decision analysis

[0158] In the multi-objective intelligent optimization solution of the present invention, multiple Pareto front solution sets are obtained, which reflect different water resource regulation schemes. In order to select the best scheme from them, the present invention adopts a multi-attribute decision analysis method based on the projection pursuit model, and evaluates and ranks each scheme in the solution set in combination with multiple attributes, so as to determine the regulation scheme that best meets the actual needs.

[0159] In practical applications, multi-attribute decision analysis based on projection pursuit models can not only help decision makers choose the optimal water resources regulation scheme, but also dynamically respond to changes in the system. When water supply demand, environmental conditions or other factors change, the projection pursuit model can quickly re-evaluate the Pareto solution set and adjust the projection direction to ensure that the system always chooses the best regulation scheme.

[0160] Example 2

[0161] On the basis of Example 1, this embodiment uses a specific example to illustrate the water resource optimization and control method of the long-vine melon water network irrigation system of Example 1.

[0162] This embodiment selects the Quanmutang irrigation area as an implementation case. Figure 2 As shown. The irrigation area of ​​the Quanmutang Reservoir Project is located in the central area of ​​the "Hengshao Drought Corridor" in Hunan Province, spanning the Zishui River Basin and the Xiangshui River Basin. The irrigation area covers four cities and eight counties (districts and cities), namely Shaoyang County, Shaodong City, Daxiang District, Shuangqing District of Shaoyang City, Lengshuitan District and Qiyang County of Yongzhou City, Qidong County of Hengyang City, and Shuangfeng County of Loudi City, that is, 111°10′~112°10′ east longitude and 26°30′~27°30′ north latitude. The irrigation area belongs to the subtropical monsoon climate zone, with warm, humid climate, abundant rainfall, distinct four seasons, short severe cold period, and long frost-free period. May to June is the plum rain season, with dull weather and high humidity. From July to August, under the control of the subtropical high pressure in the western Pacific, the extreme maximum temperature reaches about 40.0℃. The average temperature for many years from 1979 to 2018 is 17.43℃. The flood season of the basin is from April to October. The most rainfall in the basin is in spring and summer. The rainfall from April to June accounts for about 42% of the total annual rainfall. The rainfall in winter is the smallest. The rainfall from December to February accounts for only 14% of the total annual rainfall. The average annual precipitation is 1354.56mm. There are 23 rivers and streams in the irrigation area and its surrounding areas, with a total length of 996.02km. The basin area is 200km 2 The above rivers include Tanjiang, Shaoshui, Chajiang, Qishui, Shuangjiangkou, Guiyanghe, Meishui and Si'anbu, with a drainage area of ​​200km. 2 The following rivers include 15 ditches, including Daba, Leijia Temple, Louduodi, Buyunqiao, Songjiapai, Chebitang, Huihekou, Gaoqiao, Huangtupu, Yangjiazhou, Epotang, Dongxiashanshui, Shizijiang, Gantangpu, and Hongqihe.

[0163] Step S10: The water supply end, water intake end and water transmission and distribution end of the Changtengjugua water network system are generalized and the network topology method is specifically as follows:

[0164] S11. Generalization elements of water supply

[0165] The water supply sources in the study area are shown in the following table. The water supply sources involved in the optimization and regulation model in this invention include medium-sized reservoirs and small type I reservoirs, which are marked with * in the table. Other water supply bodies follow the principle of "use water when it is available" for runoff regulation calculation.

[0166]

[0167]

[0168] S12, generalized elements of water intake

[0169] The Quanmutang irrigation area involves four urban districts, namely, Shaodong City, Qidong County, Daxiang District, and Shuangqing District, as well as 26 towns. Urban and rural domestic water is mainly taken from local rivers and medium and small reservoirs. Some towns take water from local rivers and wells, which are not generalized due to their small scale. When the local water conservancy facilities are insufficient in water supply, consider pumping water from Quanmutang to the reservoir and then supplying water to the towns. The water withdrawal situation of each urban district and town is shown in the following table. D represents the generalized urban and rural area number, totaling 17; W1~W25 are urban and rural water supply and demand nodes, totaling 25, of which the mark * is the optimization decision object.

[0170]

[0171]

[0172] According to the positional relationship between the reservoir in the design level year and the Changtengjiegua irrigation channel, medium-sized reservoirs, small reservoirs, mountain ponds, and river dams supply water to local agriculture according to the location of their irrigation areas. According to the layout of the Quanmutang irrigation area, water replenishment to the Changlinchong Reservoir and the Jinjitang Reservoir through the canal system can meet the water replenishment requirements to the Hongqi River. The water replenishment volume is the minimum ecological flow of the Hongqi River estuary section. In this study, two ecological water replenishment nodes were set, namely the discharge flow of the Jinjitang Reservoir and the Changlinchong Reservoir, both of which are optimization decision variables.

[0173] S13. Generalization elements of water transmission and distribution

[0174] The irrigation area canal system includes 1 main canal and 10 backbone branch canals, with a total length of 223.3km, including the main canal Q0, Jiugongqiao branch canal Q1, Daxiang branch canal Q2, Xiakoushan branch canal Q3, Huangjiaba main canal Q4, Jiangkou branch canal Q5, Qiyang main canal Q6, Yangjiatai branch canal Q7, Qidong main canal Q8, Sandu main canal Q9, Shaofeng main canal Q10, channel water diversion nodes q1~q78, and 5 generalized pump station nodes.

[0175] S14. Generalization of water filling and replenishment logic

[0176] Water is replenished according to the coupling relationship between the main and branch canals and the logical order of the water diversion nodes of each subordinate canal. B represents the water replenishment node of the generalized canal to the irrigation area, totaling 13, and C represents the water filling node to the connected low melon reservoir, totaling 30; B and C are both the optimization decision objects of this study; q is the generalized water diversion node proposed according to the distribution of canal field acres and the canal water distribution map.

[0177] S15. System overview diagram

[0178] A generalized diagram of the water resources system of the Quanmutang Irrigation Project is shown in Figure 3-1 As shown in the canal water distribution diagram Figure 3-2 shown.

[0179] Step S20 is to design a multi-level canal hierarchical control structure for the long-vine melon water network irrigation system, specifically:

[0180] S21, general dispatching layer (backbone water source dispatching)

[0181] The general regulation layer of the Quanmutang Changtengjiegua water network irrigation system realizes the water source dispatching of key nodes of the backbone and branch canals through the canal head pump station. 输入 ) is obtained from the Quanmutang Reservoir and lifted to the trunk branch canal through the headworks pumping station.

[0182] The design flow and increased flow results of the main nodes of the irrigation area’s main canal are shown in the table below.

[0183]

[0184] S22, sub-dispatching layer (branch canal dispatching)

[0185] In the branch regulation layer of this embodiment, the water dispatch of the branch canal system is adjusted by the output water of the upper general regulation layer. The water volume of the branch canal flows in from the general regulation point and is distributed to the lower-level ditches and farmland irrigation areas through the branch canal's water diversion outlet. The state variables of the branch canal include the branch canal's water delivery capacity, leakage, and flow distribution ratio. The distribution of some branch canals in the irrigation area is shown in the table below.

[0186]

[0187]

[0188] S23, self-regulating layer (branch canal dispatching)

[0189] The self-regulation layer is the lowest water resource scheduling level in this embodiment, responsible for further distributing water from the diversion outlets of the branch canals and local reservoirs (or ponds) to water users. This level adopts a point-to-point water resource allocation method to ensure that each water user obtains the appropriate amount of water according to its actual needs.

[0190] Basic information of each irrigation section of the self-regulating small type I reservoir.

[0191]

[0192]

[0193]

[0194] The basic information of small type II reservoir bundling is as follows.

[0195]

[0196] Step S30 constructs a multi-objective water resources optimization and regulation model for long vine melons, which specifically includes:

[0197] S31. Decision variables

[0198] The hydrological year is used as the dispatching cycle (starting from early April), and the dispatching period T is set to ten days. The optimized reservoir water supply W (W1~W59), optimized reservoir discharge flow Q (Q1~Q43), channel filling volume C (C1~C30) to the low melon reservoir, and channel direct irrigation water volume B (B1~B13) to farmland are used as decision variables in each period. After introducing the regulation period, the total number of decision variables of this multi-objective optimization problem is 145×36=5220.

[0199] S32, multi-objective optimization solution

[0200] Based on the above model and solution algorithm, taking the hydrological year as the allocation cycle and the ten-day period as the regulation period, combined with the decision variables, the multi-objective allocation model of water resources in the Quanmutang irrigation area is solved based on the NSGA-Ⅱ algorithm, and the competitive game relationship between multiple objectives such as water supply and pump station operation in the Quanmutang irrigation area is comprehensively analyzed.

[0201] The model optimization population is set to 1000, the number of iterations is set to 2000, and the multi-objective optimization model is solved to obtain the multi-objective non-inferior solution set of the Changtengjieguakuqu joint multi-objective optimization regulation model. The Pareto frontier lines under different water inflow frequencies (25%, 50%, 80%, 95%) are as follows: Figure 4-1 to Figure 4-4 shown.

[0202] It can be seen from the figure that the non-inferior solution set target value ranges of the sum of squares of water shortage rate under each water inflow frequency (25%, 50%, 80%, 95%) are 2.04-4.79 (average value 2.32), 2.86-9.21 (average value 3.99), 3.4-9.66 (average value 7.82), 5.81-28.6 (average value 22.24), and the non-inferior solution set target value ranges of pump station operation cost are 33.1244-38.0112 million yuan (average value 37.2244 million yuan), 32.2428-40.8036 million yuan (average value 38.1268 million yuan), 31.8048-40.0856 million yuan (average value 33.1992 million yuan), and 33.9916-53.0728 million yuan (average value 37.298 million yuan), as shown in the following table.

[0203]

[0204] The projection of the non-inferior frontier solution on pump station operation-water supply is relatively concentrated. By analyzing the competitive relationship between the objectives, the following conclusions can be drawn: with the increase of the pump station operation cost, the sum of the squares of the water shortage rate in the Quanmutang irrigation area is decreasing. There is a relatively obvious competitive relationship between the pump station operation and the water supply function attributes in the region. The competitive relationship between the pump station operation and water supply functions is mainly because the water supply goal pursues the minimum sum of the squares of the water shortage rate in the basin, and the water volume is evenly distributed according to the size of the water demand. The pump station operation cost goal pursues the minimum regional pump station operation cost. The water volume prefers water users with smaller pump station operation costs generated by water supply. There is bound to be a conflict between the two, resulting in an obvious competitive relationship.

[0205] S33. Optimal Decision Making

[0206] Based on the projection pursuit method, a water resource multi-objective regulation scheme optimization model for the Quanmutang irrigation area was constructed, and the genetic algorithm was combined to solve it, and the projection eigenvalues ​​of the population in a typical year were obtained, and the best regulation scheme was selected. The values ​​of each objective function in the water resource multi-objective optimization regulation model were selected as multi-objective evaluation indicators, namely, the minimum sum of squares of water shortage rate and the minimum operating cost of the pump station, both of which are negative indicators. Combined with the population size of 100 in the NSGA-Ⅱ algorithm of the multi-objective optimization regulation model, the Pareto set of water supply and economic objectives of 100 populations was selected as the evaluation index and scheme set.

[0207] The multi-objective optimization control scheme for the Quanmutang Irrigation District in typical years is shown in the following table. Scheme I is the optimal water supply target, Scheme II is the optimal system operation cost target, and Scheme III is the optimal comprehensive benefit of all targets.

[0208]

[0209] The results show that lower water supply targets match higher pumping station cost targets, and vice versa. The unit projection vectors of water supply targets and pumping station cost targets in wet years, normal years, dry years, and extremely dry years are 0.674 / 0.739, 0.705 / 0.709, 0.721 / 0.693, and 0.742 / 0.670, respectively, indicating that the economic target in wet years is a relatively important evaluation indicator, and the water supply target is a more important evaluation indicator under other water inflow frequencies. With the increase of water inflow frequency, the unit projection vector of the water supply target increases, and the importance of the water supply target increases accordingly. Thus, the optimal scheme for the joint multi-objective regulation of reservoirs and channels in wet years, normal years, dry years, and extremely dry years in the Quanmutang Irrigation District is obtained, as follows: Figure 5 .

[0210] S34. Optimal water resource regulation plan for typical years (normal water year)

[0211] The summary table of urban and rural water resources allocation, irrigation water resources allocation and pumping station water extraction capacity for each plan in a normal water year is as follows, and the detailed allocation plans (Plan III) are as follows.

[0212] Summary of urban and rural water resources allocation in each plan in a normal water year (10,000 m 3 ) as shown in the following table.

[0213]

[0214]

[0215] Summary of irrigation water resource allocation for each scheme in a normal water year (10,000 m 3 ) as shown in the following table.

[0216]

[0217] S35. Typical annual pond filling and reservoir maintenance dispatching rules

[0218] Taking the water resources optimization and regulation scheme III with the best comprehensive benefits in each typical year as an example, the corresponding filling and reservoir maintenance dispatching process of each optimized reservoir (medium-sized reservoir, small type I bundled reservoir) in the Changtengjiegua irrigation water network can be obtained, including the reservoir inflow process line, the Quanmutang reservoir filling line (the net water volume at the target reservoir section), the water supply process line, the discharge process line, and the reservoir storage change process. The statistical results are shown in the following table. Taking the Huangjiaba Reservoir, Longjiangqiao Reservoir, and Qunli Reservoir as typical reservoirs, their reservoir dispatching process lines are as follows: Figures 6-1 to 6-3 .

[0219]

[0220]

[0221] An embodiment of the present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program, when loaded into the processor, implements any one of the methods for optimizing and regulating water resources in a long-vine melon water network irrigation system.

[0222] An embodiment of the present invention further provides a storage medium storing a computer program, wherein the computer program, when executed by a processor, implements any one of the methods for optimizing and controlling water resources in a long-vine melon water network irrigation system.

Claims

1. A method for optimizing and controlling water resources in a water network irrigation system for long-vine melons, characterized in that: The following steps are involved: (1) Establish the generalization and network topology relationship between the water supply end, water intake end and water transmission and distribution end of the Changtengjiegua water network system; (2) Based on the concept of "multi-level hierarchical control structure", a three-layer hierarchical control structure of "general regulation-divided regulation-self-regulation" was designed; the "general regulation layer" refers to the dispatching of the main water source, which pumps water to the key nodes of each main branch canal through the headwater pumping station; the "divided regulation layer" dispatches water to the water diversion points of each branch canal; and the "self-regulation layer" establishes a point-to-point relationship between the many water diversion points of the main and branch canals, local reservoirs and ponds, and water users, thereby redistributing water; (3) A multi-objective water resources optimization and regulation model for the Changtengjieguakuqu was constructed and solved using NSGA-II. The model includes: determining decision variables, establishing objective functions, and constructing constraints.

2. The method for optimizing and controlling water resources in a water network irrigation system for long-vine melons according to claim 1, characterized in that: Step (1) comprises the following steps: (11) The water supply end is generalized into several nodes, each node represents a different water source type and location; for the river water supply system, the generalized nodes represent the main river estuary locations; for the reservoir water supply system, the nodes represent the water intake locations of the reservoir; the nodes of the groundwater supply system represent the main water wells; each water supply node is classified and connected according to its geographical location, water supply capacity and the relationship between them. (12) The water intake end is generalized according to functional requirements. Different water intake end nodes represent different user groups or areas. Each water intake node is numbered according to the actual water demand, describing its hydraulic connection relationship with the water supply end, and setting a transmission channel with the water supply end. (13) The water transmission and distribution ends are connected through canals and pipelines, simplifying the complex water transmission and distribution system into several main water transmission trunks and branch pipelines; the flow, capacity and hydraulic conditions of each water transmission route are modeled according to the terrain and engineering conditions, and hierarchical management is carried out; (14) Generalization of water replenishment logic First, the water source volume at the water supply node and the water demand at the water intake node are monitored in real time to generate a supply-demand balance state. At the same time, the regional control priority is set according to the supply and demand situation, and the transmission path is adjusted in real time through a dynamic feedback mechanism to optimize the efficiency of water resource allocation. (15) By generalizing the water supply end, water intake end, and water transmission and distribution end, each node is connected to form a complete water network system; the network topology is constructed using graph theory methods, each node represents the water supply, water intake, or water transmission and distribution end, and the edge represents the path of water flow; each edge is attached with flow and pressure loss parameters.

3. The method for optimizing and controlling water resources in a water network irrigation system for long-vine melons according to claim 1 is characterized in that: Step (2) comprises the following steps: (21) The overall dispatching layer is the backbone water source dispatching: At this level, the water source dispatching of the headwater pumping station is carried out by controlling the water volume Q 总调 (t) The formula for key nodes supplying trunk and branch canals is: Q 总调 (t)=f1(t,Q 输入 ,X 渠首泵站 ) (1.1) Where t is the time variable, representing the time-varying nature of the system; Q 输入 is the water source input flow, X 渠首泵站 It is the status of the headworks pumping station (pumping station switch status, pumping station capacity, etc.). (22) Branch canal dispatching: At this level, the water volume of branch canals is dispatched and controlled to each water diversion point Q 分调 (t), the water volume is provided by the output of the total adjustment layer, the formula is as follows: Q 分调 (t)=f2(t,Q 总调 (t),X 支渠 ) (1.2) In the formula, Q 总调 (t) is the amount of water input from the upper layer; X 支渠 It is the state variable of the branch canal (including the flow capacity, leakage, flow distribution, etc. of the canal). (23) Self-regulating layer, i.e. point-to-point dispatching: The self-regulating layer is the redistribution of water between water diversion outlets, local reservoirs or ponds and water users. The formula is as follows: Q 自调 (t)=f3(t,Q 分调 (t),X 库塘 ,X 用水户 ) (1.3) In the formula, Q 分调 (t) is the amount of water input from the sub-adjustment layer; X 库塘 is the state variable of the local reservoir or pond; X 用水户 It is the needs and status of water users (including user water consumption, water intake points, etc.). (24) Construct a multi-level hierarchical control structure for the entire system: The water dispatching of the entire system can be expressed as a multi-layer feedback control system through hierarchical control theory: The overall mathematical expression of the system is described by a nested control equation: Q(t)=f3(t,f2(t,f1(t,Q 输入 ,X 渠首泵站 ),X 支渠 ),X 库塘 ,X 用水户 ) (1.4) 4. The method for optimizing and controlling water resources in a water network irrigation system for long-vine melons according to claim 1 is characterized in that: In step (3), the decision variables are as follows: the hydrological year is used as the scheduling cycle and the scheduling period is T; the optimized reservoir water supply W, the optimized reservoir discharge flow Q, the channel filling volume C to the low melon reservoir and the channel direct irrigation water volume B to the farmland are used as decision variables in each period. After introducing the regulation period, the total number of decision variables of this multi-objective optimization problem is (W+Q+C+B)×T.

5. The method for optimizing and controlling water resources in a water network irrigation system for long-vine melons according to claim 1, characterized in that: In step (3), the objective function is as follows: the minimum sum of squares of water shortage rate in the regional period is taken as the water supply target to coordinate the competitive relationship between the water supply target and other targets. The mathematical expression is: Where: D k,t is the water demand of the kth water user in the tth period; Q i,k,t is the water supply from the ith water supply source in the region to the kth water user in the tth period; α j,k The importance coefficient of the kth water user to get water supply priority relative to other water users. The mathematical expression of the economic target is: Where: T1 is the specific water peak period 1, T2 is the specific water peak period 2; β is the weight of the pumping station water extraction cost in time allocation, β1 (water replenishment in period 1) < β2 (water replenishment in period 2). M is the number of pumping stations, m = 1, 2, ..., M; N is the number of units, n = 1, 2, ... N; ρ is the density of water; q t,m,n is the pumping flow of the nth unit of the mth pumping station in the tth period; H t,m is the head of the mth pumping station in the tth period; η t,m,n The efficiency of the nth unit of the mth pumping station in the tth period; Δt is the time interval; c t is the electricity price in period t.

6. The method for optimizing and controlling water resources in a water network irrigation system for long-vine melons according to claim 1, characterized in that: In step (3), the constraints include: Reservoir water balance constraints V t -V t-1 =(WR t +B t -W t -Q t -C t )Δt (1.7) Where: V t 、V t-1 is the reservoir storage at the end and beginning of the tth period; WR t is the average inflow in the tth period. For the Quanmutang Reservoir, the water volume lifted from the dam needs to be subtracted from the natural runoff. t is the water volume of the channel filling the Digua Reservoir in the tth period (if any); C t Q is the amount of water released from Gaogua Reservoir to the channel during the tth period (if any); t is the discharge flow of the reservoir in the tth period; W t is the water supply of the reservoir in the tth period; Reservoir water supply constraints IN t ≤In kg (1.8) Where: W kg It is the water supply of the reservoir in the tth period in the design level year. Reservoir outflow restrictions Q t,min ≤Q t ≤Q t,max (1.9) Where: Q t,min is the minimum allowable outflow during period t, which is set as the ecological base flow under the reservoir dam; Q t,max is the maximum allowable outflow during period t, i.e. the maximum discharge of the reservoir; Reservoir water level restrictions WITH t,min ≤Z t ≤Z t,max (1.10) Downstream flow rate fluctuation constraint |Q t -Q t-1 |≤Δ (1.12) Where: Δ is the maximum amplitude of reservoir outflow; Water balance constraints of water diversion section O t =I t -W t (1.13) Where: O t ,I t , W t are the outflow, inflow and water diversion of the river section in the tth period respectively. Section water balance constraints Where: Q i,t is the average flow rate of the i-th section in time period t; Q i-1,t-1 is the average flow rate of the i-1th section in the t-1 period; Q i取,t , Q i入,t , Q i退,t , Q i损,t are the water intake flow, interval inflow, water withdrawal flow and loss flow of the river section from i-1 to i in the period t-1 respectively; τ i,t is the propagation time of the flow from the i-1th to the i-th section in the t period; Section ecological flow constraints THE t ≥Q t,e (1.15) Where: O t is the discharge flow of the river section in period t; Q t,e It is the minimum or suitable ecological flow of the river section; Channel flow balance constraints Q j,t =Q j-1,t +Q l,t (1.16) Q l,t =βALQ j-1,t 1-m / 100 (1.17) Where: Q j,t is the gross flow of the first section of the j-th water diversion channel during period t; Q j-1,t is the net flow at the end of the j-th water channel (or the beginning of the j-1-th water channel) in period t; Q l,t is the water loss of the j-th channel in period t; it is estimated according to the empirical formula, where L is the channel length, km; A is the water permeability coefficient of the channel bed soil, light clay, A = 2.65; m is the water permeability index of the channel bed soil, medium permeability, m = 0.45; β is the reduction coefficient of the channel bed leakage after anti-seepage measures are taken, the channel adopts concrete surface protection, β = 0.15; Channel water distribution flow constraints q min ≤q t ≤q max (1.18) Where: q t is the channel water distribution flow in period t; q max ,q min The upper and lower limits of the water flow rate at each water intake of the channel; Channel water capacity constraints Q j,t ≤Q j,cmax (1.19) Where: Q j,t is the flow rate of the jth channel section in the tth period; Q j,cmax is the corresponding channel flow capacity. Pump station head constraints H m,min ≤H m,t ≤H m,max (1.20) Where: H m,min , H m,max is the minimum and maximum head of the mth pump station; H m,t is the head of the mth pumping station during period t; Pumping station overcurrent capacity constraints and power constraints q n,min ≤q n,t ≤q n,max (1.21) x n,t ≤ξ n,max (1.22) Where: q n,min ,q n,max are the minimum and maximum pumping flow rates of the nth unit respectively; q n,t is the pumping flow of the nth unit in period t; ξ n,max is the maximum power of the nth unit, ξn ,t is the pumping power of the nth unit in period t; Water demand constraints Water supply guarantee rate constraint Where: p k is the water supply guarantee rate of the kth water user; is the design guarantee rate of the kth water user.

7. The method for optimizing and controlling water resources in a water network irrigation system for long-vine melons according to claim 1, characterized in that: In step (3), the model solution process is as follows: the optimization of water resources in the long-vine melon irrigation water network system involves a multi-objective nonlinear problem. A global search method is used to avoid falling into the "dead loop" problem of local minimum values. The overall process is expressed as: P t+1 =Selection(P t )+Crossover(P t )+Mutation(P t ) (1.25) The multi-objective genetic algorithm based on non-dominated sorting (NSGA-II) is selected to solve the model, as follows: First, the fitness function of non-dominated sorting is constructed: Where R(x) represents the non-dominated rank of individual x; The goal of NSGA-II is to find an approximate non-dominated solution set S through the fitness function nd , and keeps approaching the true non-inferior solution set S. Then, the projection pursuit optimization model is used to solve the solution set S nd Evaluate and sort the solutions in the high-dimensional space; suppose the data in the high-dimensional space is X=(x1,x2,…,x d ), the projection direction is ω, then the projection value in the low-dimensional space is: In the formula, ω T is the transpose of the projection vector, ω i is the weight in the projection direction, and x i is the eigenvalue in high-dimensional data; The optimal projection vector ω is found in the iterative process. The process is expressed as: max w Q(ω,X) (1.28) Among them, Q(ω,X) represents the objective function of projection pursuit, which reflects the quality of the projection result in the low-dimensional space; by maximizing Q(ω,X), we can find the optimal projection vector ω * , thereby ensuring that the data structure in the low-dimensional space can most effectively reflect the characteristics of the high-dimensional data; the optimal projection direction and the corresponding projection value obtained can be used for subsequent analysis, visualization, and decision support, thereby achieving the solution set S nd Effective evaluation and ranking of solutions.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the computer program is loaded into the processor, it implements the water resource optimization and control method of the long-vine melon water network irrigation system according to any one of claims 1-7.

9. A storage medium storing a computer program, characterized in that: When the computer program is executed by the processor, a method for optimizing and controlling water resources in a long-vine melon water network irrigation system according to any one of claims 1 to 7 is implemented.

Images