Reservoir water-sediment-electricity joint optimization scheduling method based on non-dominated genetic algorithm

Abstract

The invention discloses a reservoir water-sediment-electricity joint optimization scheduling method based on a non-dominated genetic algorithm, and the method comprises the steps: constructing a one-dimensional water-sediment mathematical model, and constructing a reservoir water-sediment-electricity joint scheduling model through the one-dimensional water-sediment mathematical model; optimizing the reservoir water-sediment-electricity combined dispatching model to obtain an optimized reservoir water-sediment-electricity combined dispatching model; iteration is carried out based on the optimized reservoir water-sediment-electricity combined dispatching model, and if the set maximum number of iterations is reached, the Pareto leading edge is output. According to the method, the population quality is improved, the convergence is improved, and the high-dimensional multi-target problem can be better solved.

Description

Technical Field

[0001] The invention relates to the technical field of reservoir sediment dispatching, and in particular to a reservoir water-sediment-electricity joint optimization dispatching method based on a non-dominated genetic algorithm. Background Art

[0002] The construction of a reservoir will break the existing erosion and siltation balance of the river. After the reservoir is built and begins to store water, the river will be channelized, forming a relatively static environment in the reservoir area, fundamentally changing the hydrodynamic conditions of the natural river. The water is deep and the flow is slow, which will inevitably cause the sedimentation in the reservoir area. Many rivers in my country are rich in sediment, and the sediment problem is one of the key factors in the long-term operation and benefit of hydropower stations. The large amount of sedimentation in the reservoir may seriously affect the comprehensive benefits of the reservoir and shorten the service life of the reservoir when there is still a large amount of silt storage capacity. There are many sediment management measures for rivers and reservoirs. Among them, formulating a reasonable reservoir scheduling plan and using floods to discharge sediment are considered to be the most effective non-engineering measures to alleviate reservoir sedimentation.

[0003] However, previous reservoir scheduling rarely considered the scouring and silting of sediment, and only involved runoff scheduling, which is feasible and can achieve relatively good results on rivers with low sediment content and not serious sediment deposition. However, for reservoirs built on rivers with high sediment content, the total amount of sediment deposition in the reservoir and its vertical distribution are related to the long-term effective operation of the reservoir and the realization of flood control goals. First, for reservoirs with flood control tasks, years of sediment deposition will gradually reduce the flood control storage capacity of the reservoir, resulting in a decrease in the flood control capacity of the reservoir, thereby affecting the protection capacity of downstream flood control objects; secondly, when the reservoir is built on a river with a navigation task like the Yangtze River, the changes in the scouring and silting of the reservoir will also affect the shipping benefits. The sediment deposition of the reservoir not only directly affects the benefits of the reservoir, but may also cause some harm. For example, for reservoirs with a large regulation coefficient, the increase of suspended sediment in the water will change the dissolved oxygen content in the water, which will in turn have an adverse effect on the growth of aquatic organisms. Siltation will also aggravate water pollution to a certain extent. In addition, the extension of siltation will increase the scope of flooding upstream and cause land salinization problems.

[0004] Although some studies have been conducted to find effective ways to discharge sediment from reservoirs, such as the double flood water level limit, the basic idea is to use the strong sediment transport capacity of water flow under the combination of large flood flow and low water level in front of the dam to reduce reservoir siltation; others have proposed multiple flood water level limit schemes, early water storage schemes in stages after the flood season, "flood storage and drainage" to extend the water storage period scheme, and flood season open discharge scheme, etc. However, it is impossible to achieve the optimal balance between reservoir sediment discharge and multiple goals such as power generation, flood control, and shipping.

[0005] Therefore, it is necessary to use algorithms to optimize the scheduling of reservoirs. Optimal scheduling is to establish a method with reservoir scouring and siltation and benefits as the objective function. After formulating many constraints that should be met, an optimization algorithm is used to solve the system equations composed of multiple objective functions and constraints so that the goal is achieved. That is, the reservoir control method in which the objective function obtains an extreme value is a reservoir scheduling method that has developed rapidly in recent years.

[0006] Reservoir optimization scheduling methods can be roughly divided into two categories: mathematical programming algorithms and intelligent optimization algorithms. Mathematical programming algorithms mainly include dynamic programming, linear programming and nonlinear programming.

[0007] Dynamic programming is a commonly used mathematical method for solving optimization problems in multi-stage decision-making processes. It has the characteristics of knowability and no aftereffect, and can well solve the nonlinear and random characteristics of water resource systems. However, when the number of reservoirs is large, the "curse of dimensionality" problem will occur, and the amount of calculation is very large. Linear programming is a widely used and relatively simple planning method, which has mature solution methods and applications. However, the reservoir optimization scheduling process has the characteristics of randomness and nonlinearity, so the problem needs to be linearized before the linear programming method is used to solve the reservoir scheduling, and the calculation efficiency is low. Nonlinear programming can effectively deal with inseparable objective functions and nonlinear constraints. However, nonlinear programming is more complicated than linear programming, and there is no universal solution method and procedure, so nonlinear programming is not extensive. Since the computational complexity of mathematical programming algorithms is generally large, it is only suitable for solving small-scale problems and is often not very practical in engineering.

[0008] Intelligent optimization algorithms include: swarm intelligence algorithm, evolutionary algorithm, etc. Swarm intelligence algorithm is an emerging evolutionary computing technology with the advantages of simple operation and good convergence. However, because the initial solution randomly generated during the scheduling process cannot meet the constraints, it leads to slow convergence and the optimization process is prone to fall into local optimality. Evolutionary algorithm can find the global optimal solution to the optimization problem. It has the characteristics that the optimization result is independent of the initial conditions and has strong robustness. It is suitable for solving complex optimization problems, but it has slow convergence speed, poor local search ability, and no definite termination criteria.

[0009] At present, there are also few studies that apply optimization algorithms to reservoirs for joint dispatch of water, sediment and electricity. For example, some researchers use dynamic programming algorithms to construct reservoir optimization dispatch models to optimize the calculation of three goals: reservoir silt reduction, power generation benefits and downstream river regulation needs; or to solve the optimal solution for the two goals of power generation benefits and reservoir silt reduction benefits based on economic benefit quantification. Although the above studies take into account the joint dispatch of silt reduction and power generation, the empirical relationships used in the calculations of sediment transport and siltation cannot accurately reflect the rapid changes in the riverbed in a short time scale. Summary of the invention

[0010] In order to solve the above technical problems, the present invention proposes a reservoir water-sediment-electricity joint optimization scheduling method based on a non-dominated genetic algorithm to solve the problem of weight setting for silt reduction and power generation.

[0011] On the one hand, to achieve the above-mentioned purpose, the present invention provides a reservoir water-sand-electricity joint optimization dispatching method based on a non-dominated genetic algorithm, comprising:

[0012] Constructing a one-dimensional water-sediment mathematical model, and constructing a reservoir water-sediment-electricity joint dispatching model through the one-dimensional water-sediment mathematical model;

[0013] Optimizing the reservoir water-sand-electricity joint dispatching model to obtain an optimized reservoir water-sand-electricity joint dispatching model;

[0014] Iteration is performed based on the optimized reservoir water-sand-electricity joint scheduling model. If the set maximum number of iterations is reached, the Pareto front is output, where the Pareto front is a collection of optimal population individuals, and each individual represents an optional optimization scheduling plan.

[0015] As a preferred solution of the reservoir water-sediment-electricity joint optimization dispatching method based on non-dominated genetic algorithm described in the present invention, the one-dimensional water-sediment mathematical model includes:

[0016] Continuity equation for water flow:

[0017]

[0018] The equation of motion for water flow is:

[0019]

[0020] Suspended load transport equation:

[0021]

[0022] Riverbed deformation equation:

[0023]

[0024] Sand holding force formula:

[0025]

[0026] Where A is the cross-sectional area of ​​the water flow, m 2 ; Q is the average flow rate of the section, m 3 / s; Z S is the water level, m; B is the width of the water surface, m; S is the suspended sediment content, kg / m 3 ; S * is the sand-carrying force of water flow, kg / m 3; A0 is the sediment scouring and deposition area on the cross section, m 2 ; g is the acceleration due to gravity, m / s 2 ; K is the flow modulus, m 2 / s; α is the recovery saturation coefficient, ω is the sediment settling velocity, m / s; γ′ is the dry density of sediment, kg / m 3 , U is flow velocity, h is water depth, t is time, s; x is distance; β is momentum correction coefficient, k is the index of sand-carrying force formula; m is the coefficient of sand-carrying force formula; U is flow velocity, m / s.

[0027] As a preferred solution of the reservoir water-sand-electricity joint optimization dispatching method based on the non-dominated genetic algorithm described in the present invention, the objective function of the reservoir water-sand-electricity joint dispatching model includes:

[0028] Scouring and silting objective function f1:

[0029]

[0030] Among them, ΔV is the sedimentation volume in the reservoir area during the dispatch period, m 3 ; ΔA i is the scouring and silting area of ​​section i, m 2 ; ΔL i is the length of the i-th river section, m; M is the number of sections divided into reservoir areas, △A i+1 is the scouring and silting area of ​​section i+1, m 2 , i is the section number; the scouring and silting area of ​​the section is calculated by the one-dimensional water-sediment mathematical model;

[0031] Power generation objective function f2:

[0032]

[0033] Where E is the total power generation during the dispatch period, kW / h; η l is the turbine efficiency; η g is the generator efficiency; Q t is the power generation reference flow of the unit in period t, m 3 / s;H t is the net water head of the unit in time period t, m; Δt is the time step, h; T is the total number of days in the scheduling period.

[0034] As a preferred solution of the reservoir water-sand-electricity joint optimization dispatching method based on the non-dominated genetic algorithm described in the present invention, the constraints of the reservoir water-sand-electricity joint dispatching model include:

[0035] Water balance constraint V t+1 :

[0036] V t+1 =V t +(qt,入库 -q t,出库 )×Δt

[0037] Among them, V t is the storage capacity of the reservoir in the tth period; q t，入库 is the inflow flow in period t, q t，出库 is the outbound flow in period t;

[0038] Water level constraint Z t :

[0039] Z t,min ≤Z t ≤Z t,max

[0040] Among them, Z t，min is the lower limit of the reservoir water level in period t, Z t，max is the upper limit of the reservoir water level in period t;

[0041] Downflow flow constraint q t :

[0042] q t,出库,min ≤q t ≤q t,出库,max

[0043] Among them, q t，出库，min is the lower limit of the reservoir discharge flow in the tth period, q t，出库，max is the upper limit of the discharge flow of the reservoir in period t;

[0044] Generator set overcurrent capacity constraint Q t :

[0045] Q t,min ≤Q t ≤Q t,max

[0046] Among them, Q t，min is the minimum power generation reference flow in period t, Q t，max is the maximum power generation reference flow in period t;

[0047] Generator output capacity constraint 9.81η l η g Q t H t :

[0048] N t,min ≤9.81η l η g Q t H t ≤N t,max

[0049] Among them, N t，minis the minimum output allowed by the generator set in the tth period, N t，max is the maximum output allowed by the generator set in period t;

[0050] End-of-run state constraints:

[0051]

[0052] Among them, Z T is the water storage level at the end of the dispatch period, Z T，目标 The target water storage level to be achieved at the end of the dispatch period;

[0053] Non-negativity constraint: All variables must be non-negative.

[0054] As a preferred solution of the reservoir water-sand-electricity joint optimization dispatching method based on the non-dominated genetic algorithm described in the present invention, the reservoir water-sand-electricity joint dispatching model is optimized, including:

[0055] Optimize the generation of the initial population, determine the crossover method and mutation method to generate new individuals Q and merge them with the parent P, and the individual R is composed of {r1,r2,…,r D}, D is the number of days of discharge; the parent P is composed of {R1, R2, …, R N}, N is the population size;

[0056] Select l-1 layer individuals through fast non-dominated sorting;

[0057] The individuals in the last layer are selected to generate a new population, and the optimized reservoir water-sediment-electricity joint dispatching model is obtained.

[0058] As a preferred solution of the reservoir water-sediment-electricity joint optimization dispatching method based on non-dominated genetic algorithm described in the present invention, the generation of the optimized initial population includes:

[0059] The optimization is performed using real number coded population individuals, where each gene r in the population individual R t Represents the reservoir discharge in one period. The calculation method of the discharge in each period of each individual is:

[0060]

[0061] Among them, r t is the discharge flow in each period, W is the total discharge during the scheduling period, k t is T random numbers uniformly distributed within [0,1] generated, and ΔT is the total discharge time.

[0062] As a preferred solution of the reservoir water-sand-electricity joint optimization dispatching method based on non-dominated genetic algorithm described in the present invention, determining the crossover method and mutation method to generate new individuals and merge them with the parents includes:

[0063] After generating the same number of offspring as the parents through the crossover method and mutation method, the parents and offspring are merged;

[0064] The crossover method is to use uniform crossover, in which two random individuals R a , R b The gene at each corresponding position is exchanged with probability p c Make an exchange;

[0065] Randomly select two individuals R a , R b , and set the exchange probability p c , generate a random number between [0,1], if the random number is greater than p c , then no crossover is performed between the two individuals. If the random number is less than p c , then the two individuals are crossed, and the crossover process is:

[0066] R a The gene r at each position of an individual a，t With R b The gene r at each position of an individual b，t One-to-one correspondence, a random number between [0,1] is generated at each position t of the new individual V for judgment. If the random number is less than or equal to 0.5, the gene v of the individual V at the position t t For r a，t ; If the random number is greater than 0.5, then the gene v of individual V at position t t For r b，t ;

[0067] Crossover is performed until the same number of N offspring individuals {V1, V2, ..., V N}until;

[0068] The mutation method is to use polynomials to perform mutation, and set the mutation probability p v , select an individual V after the crossover operation, generate a random number between [0,1], if the random number is greater than p v , the individual does not mutate, if the random number is less than p v , then the individual mutates, and the mutation form is:

[0069] v′ k =v k +δ×(u k -l k )

[0070] Among them, v′ k is the gene at position k of the new individual after update, v k is the gene at position k of the individual after the crossover operation, u k is the maximum leakage flow, l k is the minimum leakage flow;

[0071] The calculation method of δ is:

[0072]

[0073] in, u is a random number in [0,1], θ m is the distribution index selected by the user;

[0074] {v′1,v′2,…,v′ D} constitutes the mutated offspring individual V′, that is, a new D-day outbound flow process is obtained after one optimization iteration. After mutation, the offspring population Q is generated, which consists of {V′1, V′2, …, V′ N}constitute.

[0075] As a preferred solution of the reservoir water-sand-electricity joint optimization dispatching method based on non-dominated genetic algorithm described in the present invention, wherein the l-1 layer individuals are selected by the fast non-dominated sorting, including:

[0076] Arrange the individuals in the population according to the function value of the scouring and silting objective function f1. If the objective function values ​​of the scouring and silting objective function f1 are equal, arrange them according to the function value of the power generation objective function f2. If the function values ​​of the scouring and silting objective function f1 and the power generation objective function f2 are equal, arrange them arbitrarily.

[0077] Create a sequential search strategy based on the permutation results to rank and rank each individual.

[0078] As a preferred solution of the reservoir water-sand-electricity joint optimization dispatching method based on the non-dominated genetic algorithm described in the present invention, the last layer of individuals are selected to generate a new population, including:

[0079] S1. Create reference points, the coordinate value of each reference point is:

[0080]

[0081] In the formula, H is the number of divisions for each optimization target, and i is the reference point number;

[0082] S2. Adaptive normalization of the objective function:

[0083]

[0084] Among them, f i (x) represents the objective function value of the i-th target, represents the minimum objective function of the i-th target, a i Represents the intercept of the hyperplane formed by the population extreme points on the coordinate axis, f i n (x) is the standardized objective function value;

[0085] S3, associating the population with the reference point: connecting the origin and the reference point, drawing a reference line, and then calculating the distance between the individual and each reference point. The reference point closest to the individual means that the individual belongs to the reference point;

[0086] S4, select the reference point with the least number of associated individuals in the l-1 layer, select the individual closest to the reference point in the individuals in the lth layer and add it to the new population, and the number of individuals associated with the reference point is increased by 1;

[0087] S5. Repeat S1-S4 until the maximum population size is reached.

[0088] On the other hand, the present invention also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of a reservoir water-sand-electricity joint optimization scheduling method based on a non-dominated genetic algorithm.

[0089] Compared with the prior art, the present invention has the following advantages and technical effects:

[0090] The present invention introduces widely distributed reference points to maintain population diversity, and by storing all individuals in layers, the best individuals will not be lost, which rapidly improves the population quality and convergence, thereby being able to better handle high-dimensional multi-objective problems; the present method also couples the reservoir scheduling model with the one-dimensional water-sediment model, so that more comprehensive river channel scouring and silting change information can be obtained during the scheduling process, including not only the total amount of siltation, but also the riverbed deformation of each section, providing water-sediment and scouring and silting prediction data with higher temporal and spatial resolution for the valuation of the objective function. BRIEF DESCRIPTION OF THE DRAWINGS

[0091] The drawings constituting a part of the present application are used to provide a further understanding of the present application. The illustrative embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0092] Figure 1 It is a flow chart of a reservoir water-sand-electricity joint optimization dispatching method based on a non-dominated genetic algorithm according to an embodiment of the present invention;

[0093] Figure 2This is a diagram of optimization results of the reservoir water-sand-electricity joint optimization scheduling algorithm based on the non-dominated genetic algorithm in an embodiment of the present invention. DETAILED DESCRIPTION

[0094] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0095] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0096] The present invention proposes a reservoir water-sediment-electricity joint optimization dispatching method based on a non-dominated genetic algorithm. Figure 1 ,include:

[0097] Constructing a one-dimensional water-sediment mathematical model, and constructing the reservoir water-sediment-electricity joint dispatching model through the one-dimensional water-sediment mathematical model;

[0098] Optimize the reservoir water-sand-electricity joint dispatching model to obtain the optimized reservoir water-sand-electricity joint dispatching model;

[0099] Iteration is performed based on the optimized reservoir water-sand-electricity joint scheduling model. If the set maximum number of iterations is reached, the Pareto front is output.

[0100] It is worth noting that this embodiment introduces widely distributed reference points to maintain population diversity, and by storing all individuals in layers, the best individuals will not be lost, which rapidly improves the population quality and convergence, thereby being able to better handle high-dimensional multi-objective problems; this method also couples the reservoir scheduling model with the one-dimensional water-sediment model, so that more comprehensive river channel scouring and silting change information can be obtained during the scheduling process, including not only the total amount of siltation, but also the riverbed deformation of each section, providing water-sediment and scouring and silting prediction data with higher temporal and spatial resolution for the valuation of the objective function.

[0101] Furthermore, the one-dimensional water-sand mathematical model includes:

[0102] Continuity equation for water flow:

[0103]

[0104] The equation of motion for water flow is:

[0105]

[0106] Suspended load transport equation:

[0107]

[0108] Riverbed deformation equation:

[0109]

[0110] Sand holding force formula:

[0111]

[0112] Where A is the cross-sectional area of ​​the water flow, m 2 ; Q is the average flow rate of the section, m 3 / s; Z S is the water level, m; B is the width of the water surface, m; S is the suspended sediment content, kg / m 3 ; S * is the sand-carrying force of water flow, kg / m 3 ; A0 is the sediment scouring and deposition area on the cross section, m 2 ; g is the acceleration due to gravity, m / s 2 ; K is the flow modulus, m 2 / s; α is the recovery saturation coefficient, ω is the sediment settling velocity, m / s; γ′ is the dry density of sediment, kg / m 3 , U is flow velocity, h is water depth, t is time, s; x is distance, m; β is momentum correction coefficient; k is coefficient of sand entraining force formula; m is index of sand entraining force formula; U is flow velocity, m / s.

[0113] The objective function of the reservoir water-sediment-electricity joint dispatching model includes:

[0114] Scouring and silting objective function f1:

[0115]

[0116] Among them, ΔV is the sedimentation volume in the reservoir area during the dispatch period, m 3 ; ΔA i is the scouring and silting area of ​​section i, m 2 ; ΔL i is the length of the i-th river section, m; M is the number of sections divided into reservoir areas, △A i+1 is the scouring and silting area of ​​section i+1, m 2 ; i is the section number; the scouring and silting area of ​​the section is calculated by the one-dimensional water-sediment mathematical model;

[0117] Power generation objective function f2:

[0118]

[0119] Where E is the total power generation during the dispatch period, kW / h; η l is the turbine efficiency; η g is the generator efficiency; Q tis the power generation reference flow of the unit in period t, m 3 / s;H t is the net water head of the unit in time period t, m; Δt is the time step, h; T is the total number of days in the scheduling period.

[0120] Furthermore, the constraints of the reservoir water-sand-electricity joint dispatch model include:

[0121] Water balance constraint V t+1 :

[0122] V t+1 =V t +(q t,入库 -q t,出库 )×Δt

[0123] Among them, V t is the storage capacity of the reservoir in the tth period; q t，入库 is the inflow flow in period t, q t，出库 is the outbound flow in period t;

[0124] Water level constraint Z t :

[0125] Z t,min ≤Z t ≤Z t,max

[0126] Among them, Z t，min is the lower limit of the reservoir water level in period t, Z t，max is the upper limit of the reservoir water level in period t;

[0127] Downflow flow constraint q t :

[0128] q t,出库,min ≤q t ≤q t,出库,max

[0129] Among them, q t，出库，min is the lower limit of the reservoir discharge flow in the tth period, q t，出库，max is the upper limit of the discharge flow of the reservoir in period t;

[0130] Generator set overcurrent capacity constraint Q t :

[0131] Q t,min ≤Q t ≤Q t,max

[0132] Among them, N t，min is the minimum power generation reference flow in period t, Q t，max is the maximum power generation reference flow in period t;

[0133] Generator output capacity constraint 9.81η l η g Q t H t :

[0134] N t,min ≤9.81η l η g Q t H t ≤N t,max

[0135] Among them, N t，min is the minimum output allowed by the generator set in the tth period, N t，max is the maximum output allowed by the generator set in period t;

[0136] End-of-run state constraints:

[0137] Z T =Z T,目标

[0138] Among them, Z T is the water storage level at the end of the dispatch period, Z T，目标 The target water storage level to be achieved at the end of the dispatch period;

[0139] Non-negativity constraint: All variables must be non-negative.

[0140] Optimize the reservoir water-sand-electricity joint dispatch model, including:

[0141] Optimize the generation of the initial population, determine the crossover method and mutation method to generate new individuals and merge them with the parents;

[0142] Select l-1 layer individuals through fast non-dominated sorting;

[0143] The individuals in the last layer are selected to generate a new population, and the optimized reservoir water-sediment-electricity joint dispatching model is obtained.

[0144] Furthermore, the generation of the initial population is optimized, including:

[0145] The optimization is performed using real number coded population individuals, where each gene r in the population individual R t Represents the reservoir discharge in one period. The calculation method of the discharge in each period of each individual is:

[0146]

[0147] Among them, r t is the discharge flow in each period, W is the total discharge during the scheduling period, k tis T random numbers uniformly distributed within [0,1] generated, and ΔT is the total discharge time.

[0148] Parent P t By {R1,R2,…,R N}, N is the population size, and t is the number of iterations.

[0149] Determine the crossover method and mutation method to generate new individuals and merge them with the parents, including:

[0150] After generating the same number of offspring as the parents through the crossover method and mutation method, the parents and offspring are merged;

[0151] The crossover method is to use uniform crossover, where the gene at each corresponding position is exchanged between two random individuals with an exchange probability P;

[0152] The mutation method is to use polynomials for mutation, and the mutation form is:

[0153] v′ k =v k +δ×(u k -l k )

[0154] Among them, v′ k is the gene at position k of the new individual after update, v k is the gene at position k of the previous generation individual, u k is the maximum leakage flow, l k is the minimum leakage flow;

[0155] The calculation method of δ is:

[0156]

[0157] in, u is a random number in [0,1], θ m is the distribution index selected by the user.

[0158] The offspring Q generated after crossover mutation t With the parent P t After the merger, individuals of the new population are selected.

[0159] Furthermore, the individuals in the l-1 layer are selected by fast non-dominated sorting, including:

[0160] Arrange the individuals in the population according to the function value of the scouring and silting objective function f1. If the objective function values ​​of the scouring and silting objective function f1 are equal, arrange them according to the function value of the power generation objective function f2. If the function values ​​of the scouring and silting objective function f1 and the power generation objective function f2 are equal, arrange them arbitrarily.

[0161] Create a sequential search strategy based on the permutation results to rank and rank each individual.

[0162] Select the last layer of individuals to generate a new population, including:

[0163] S1. Create reference points: In this embodiment, a total of reference points; among them, is the number of combinations, M is the number of targets, which is 2 in this embodiment, and H is the number of segments of the target function. The coordinate value of each reference point is:

[0164]

[0165] In the formula, H is the number of segments of the objective function, and i is the reference point number;

[0166] S2. Adaptive normalization of the objective function:

[0167]

[0168] Among them, f i (x) represents the objective function value of the i-th target, represents the minimum objective function of the i-th target, a i Represents the intercept of the hyperplane formed by the population extreme points on the coordinate axis, f i n (x) is the standardized objective function value;

[0169] S3. Associating the population with the reference point: connecting the origin and the reference point, drawing a reference line, and then calculating the distance between the individual and each reference point. The reference point to which the individual is closest belongs to the reference point.

[0170] S4, microhabitat preservation operation: select the reference point with the least number of associated individuals in the l-1 layer, select an individual in the lth layer that is closest to the reference point and add it to the new population, and the number of individuals associated with the reference point is increased by 1, and repeat the above operation until the maximum population size is reached.

[0171] S5. Repeat S1-S4 until the maximum population size is reached.

[0172] Output the result. If the maximum number of iterations is reached, the Pareto front is output.

[0173] Furthermore, a stored computer program is used to solve the problem, and when a set maximum number of iterations is reached, the iteration is stopped and the Pareto front is output.

[0174] This embodiment also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above method when executing the computer program.

[0175] A computer-readable storage medium stores a computer program, which implements the steps of the above method when executed by a processor.

[0176] It should be noted that the computer device may be a server. The computer device includes a processor, a memory, an input / output interface (I / O for short) and a communication interface. The processor, the memory and the input / output interface are connected via a system bus, and the communication interface is connected to the system bus via the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store data cluster data of the power monitoring system. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, a reservoir water-sand-electricity joint optimization scheduling method based on a non-dominated genetic algorithm (NSGA-Ⅲ) is implemented.

[0177] In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through experiments:

[0178] The water and sediment data of Xiaolangdi Reservoir from April 20 to October 20, 2012 were input into the optimization operation model for optimization calculation, and the following results were obtained: Figure 2 The Pareto front result of the proposed scheme is obviously better than that of the original scheme.

[0179] The above are only preferred specific implementations of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A reservoir water-sediment-electricity joint optimization dispatching method based on a non-dominated genetic algorithm, characterized in that: include: Constructing a one-dimensional water-sediment mathematical model, and constructing a reservoir water-sediment-electricity joint dispatching model through the one-dimensional water-sediment mathematical model; Optimizing the reservoir water-sand-electricity joint dispatching model to obtain an optimized reservoir water-sand-electricity joint dispatching model; Iteration is performed based on the optimized reservoir water-sand-electricity joint scheduling model. If the set maximum number of iterations is reached, the Pareto front is output, where the Pareto front is a collection of optimal population individuals, and each individual represents an optional optimization scheduling plan.

2. The reservoir water-sediment-electricity joint optimization dispatching method based on non-dominated genetic algorithm according to claim 1 is characterized in that: The one-dimensional water-sediment mathematical model comprises: Continuity equation for water flow: The equation of motion for water flow is: Suspended load transport equation: Riverbed deformation equation: Sand holding force formula: Where A is the cross-sectional area of ​​the water flow, m 2 ; Q is the average flow rate of the section, m 3 / s; Z S is the water level, m; B is the width of the water surface, m; S is the suspended sediment content, kg / m 3 ; S * is the sand-carrying force of water flow, kg / m 3 ; A0 is the sediment scouring and deposition area on the cross section, m 2 ; g is the acceleration due to gravity, m / s 2 ; K is the flow modulus, m 2 / s; α is the recovery saturation coefficient, ω is the sediment settling velocity, m / s; γ′ is the dry density of sediment, kg / m 3 , U is flow velocity, h is water depth, t is time, s; x is distance; β is momentum correction coefficient, k is the index of sand-carrying force formula; m is the coefficient of sand-carrying force formula; U is flow velocity, m / s.

3. The reservoir water-sand-electricity joint optimization dispatching method based on non-dominated genetic algorithm according to claim 2 is characterized in that: The objective function of the reservoir water-sand-electricity joint dispatching model includes: The scouring and silting objective function is: Among them, ΔV is the sedimentation volume in the reservoir area during the dispatch period, m 3 ; ΔA i is the scouring and silting area of ​​section i, m 2 ; ΔL i is the length of the i-th river section, m; M is the number of sections divided into reservoir areas, ΔA i+1 is the scouring and silting area of ​​section i+1, m 2 , i is the section number; the scouring and silting area of ​​the section is calculated by the one-dimensional water-sediment mathematical model; Power generation objective function f2: Where E is the total power generation during the dispatch period, kW / h; η l is the turbine efficiency; η g is the generator efficiency; Q t is the power generation reference flow of the unit in period t, m 3 / s;H t is the net water head of the unit in time period t, m; Δt is the time step, h; T is the total number of days in the scheduling period.

4. The reservoir water-sand-electricity joint optimization dispatching method based on non-dominated genetic algorithm according to claim 3 is characterized in that: The constraints of the reservoir water-sand-electricity joint dispatching model include: Water balance constraint V t+1 : V t+1 =V t +(q t,入库 -q t，出库 )×Δt Among them, V t is the storage capacity of the reservoir in the tth period; q t，入库 is the inflow flow in period t, q t，出库 is the outbound flow in period t; Water level constraint Z t : WITH t，min ≤Z t ≤Z t，max Among them, Z t，min is the lower limit of the reservoir water level in period t, Z tmax is the upper limit of the reservoir water level in period t; Downflow flow constraint q t : q t，出库，min ≤q t ≤q t，出库，max Among them, q t，出库，min is the lower limit of the reservoir discharge flow in the tth period, q t，出库，max is the upper limit of the discharge flow of the reservoir in period t; Generator set overcurrent capacity constraint Q t : Q t，min ≤Q t ≤Q t，max Among them, Q t，min is the minimum power generation reference flow in period t, Q t，max is the maximum power generation reference flow in period t; Generator output capacity constraint 9.81η l η g Q t H t : N t，min ≤9.81th l or g Q t H t ≤N t，max Among them, N t，min is the minimum output allowed by the generator set in the tth period, N t，max is the maximum output allowed by the generator set in period t; End-of-run state constraints: WITH T =Z T，目标 Among them, Z T is the water storage level at the end of the dispatch period, Z T，目标 The target water storage level to be achieved at the end of the dispatch period; Non-negativity constraint: All variables must be non-negative.

5. The reservoir water-sediment-electricity joint optimization dispatching method based on non-dominated genetic algorithm according to claim 1 is characterized in that: The reservoir water-sand-electricity joint dispatching model is optimized, including: Optimize the generation of the initial population, determine the crossover method and mutation method to generate new individuals Q and merge them with the parent P. Individual R consists of {r1, r2, ..., r D }, D is the number of days of discharge; the parent P is composed of {R1, R2, ..., R N }, N is the population size; Select l-1 layer individuals through fast non-dominated sorting; The individuals in the last layer are selected to generate a new population, and the optimized reservoir water-sediment-electricity joint dispatching model is obtained.

6. The reservoir water-sediment-electricity joint optimization dispatching method based on non-dominated genetic algorithm according to claim 5 is characterized in that: The generation of the optimized initial population includes: The optimization is performed using real number coded population individuals, where each gene r in the population individual R t Represents the reservoir discharge in one period. The calculation method of the discharge in each period of each individual is: Among them, r t is the discharge flow in each period, W is the total discharge during the scheduling period, k t is T random numbers uniformly distributed within [0, 1] generated, and ΔT is the total discharge time.

7. The reservoir water-sediment-electricity joint optimization dispatching method based on non-dominated genetic algorithm according to claim 5 is characterized in that: Determining the crossover method and mutation method to generate new individuals and merging them with the parents includes: After generating the same number of offspring as the parents through the crossover method and mutation method, the parents and offspring are merged; The crossover method is to use uniform crossover, in which two random individuals R a , R b The gene at each corresponding position is exchanged with probability p c Make an exchange; Randomly select two individuals R a , R b , and set the exchange probability p c , generate a random number between [0, 1], if the random number is greater than p c , then no crossover is performed between the two individuals. If the random number is less than p c , then the two individuals are crossed, and the crossover process is: R a The gene r at each position of an individual a，t With R b The gene r at each position of an individual b，t One-to-one correspondence, a random number between [0, 1] is generated at each position t of the new individual V for judgment. If the random number is less than or equal to 0.5, the gene v of the individual V at the position t t For r a，t ; If the random number is greater than 0.5, then the gene v of individual V at position t t For r b，t ; Crossover is performed until the same number of N offspring individuals {V1, V2, ..., V N }until; The mutation method is to use polynomials to perform mutation, and set the mutation probability p v , select an individual V after the crossover operation, generate a random number between [0, 1], if the random number is greater than p v , the individual does not mutate, if the random number is less than p v , then the individual mutates, and the mutation form is: v′ k =v k +δ×(u k -l k ) Among them, v′ k is the gene at position k of the new individual after update, v k is the gene at position k of the individual after the crossover operation, u k is the maximum leakage flow, l k is the minimum leakage flow; The calculation method of δ is: in, u is a random number in [0, 1], θ m is the distribution index selected by the user; {v′1, v′2, ..., v′ D } constitutes the mutated offspring individual V′, that is, a new D-day outbound flow process is obtained after one optimization iteration. After mutation, the offspring population Q is generated, which consists of {V′1, V′2, ..., V′ N }constitute.

8. The reservoir water, sand and electricity joint optimization dispatching method based on non-dominated genetic algorithm according to claim 5 is characterized in that: The 1-1 layer individuals are selected by the fast non-dominated sorting, including: Arrange the individuals in the population according to the function value of the scouring and silting objective function f1. If the objective function values ​​of the scouring and silting objective function f1 are equal, arrange them according to the function value of the power generation objective function f2. If the function values ​​of the scouring and silting objective function f1 and the power generation objective function f2 are equal, arrange them arbitrarily. Create a sequential search strategy based on the permutation results to rank and rank each individual.

9. The reservoir water, sand and electricity joint optimization dispatching method based on non-dominated genetic algorithm according to claim 5 is characterized in that: Select the last layer of individuals to generate a new population, including: S1. Create reference points, the coordinate value of each reference point is: In the formula, H is the number of divisions for each optimization target, and i is the reference point number; S2. Adaptive normalization of the objective function: Among them, f i (x) represents the objective function value of the i-th target, represents the minimum objective function of the i-th target, a i Represents the intercept of the hyperplane formed by the population extreme points on the coordinate axis, f i n (x) is the standardized objective function value; S3, associating the population with the reference point: connecting the origin and the reference point, drawing a reference line, and then calculating the distance between the individual and each reference point. The reference point closest to the individual means that the individual belongs to the reference point; S4, select the reference point with the least number of associated individuals in the l-1 layer, select the individual closest to the reference point in the first layer and add it to the new population, and the number of individuals associated with the reference point is increased by 1; S5. Repeat S1-S4 until the maximum population size is reached.

10. A computer device, characterized in that: It comprises a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the reservoir water-sand-electricity joint optimization scheduling method based on a non-dominated genetic algorithm as described in any one of claims 1 to 9.

Images