Long-term random cooperative scheduling method for hydropower-wind power-photovoltaic complementary system

Abstract

The invention provides a hydropower-wind power-photovoltaic complementary system long-term random collaborative scheduling method, which comprises the following steps: acquiring attribute information of cascade power stations in a drainage basin, including a resource variable historical value time sequence of each resource variable; based on a Markov chain and a Copula function, time correlation of each resource variable of the drainage basin cascade power station and correlation among the resource variables are quantified, an initial simulation scene set is generated, and scene reduction is carried out; and constructing and solving a watershed cascade hydropower-wind power-photovoltaic complementary system long-term random collaborative scheduling model, obtaining the ex-warehouse flow of the hydropower station in the scheduling period by time period suitable for all the comprehensive scenes, and obtaining the planned water level process and the planned output process of the hydropower station through the ex-warehouse flow. According to the method, long-term random collaborative scheduling of the hydropower-wind power-photovoltaic complementary system is carried out under the condition of considering the uncertainty wind and light output scene, and a reliable scheduling process can be obtained by comprehensively considering multiple actual operation constraints of the power station.

Description

Technical Field

[0001] The present invention belongs to the technical field of multi-energy coordinated complementary scheduling, and specifically relates to a long-term random coordinated scheduling method for a hydropower-wind power-photovoltaic complementary system. Background Art

[0002] With the continuous growth of economy and population, energy demand is increasing day by day. The contradiction between energy demand and the limited and non-renewable nature of traditional fossil fuel energy is becoming more prominent. It is crucial to increase the penetration rate of clean renewable energy such as solar energy, wind energy, geothermal energy, hydropower, wave energy and tidal energy in conventional power systems. However, wind, light and other energies are affected by meteorological factors and have intermittent, random and volatile characteristics, posing a threat to the safe and stable operation of the power grid. Making full use of the flexibility of hydropower and the natural complementarity between different energy sources for complementary power generation is an effective way to promote the consumption of new energy, and has become a research focus in the power system and energy field in recent years.

[0003] At present, common multi-energy complementary systems include hydro-solar complementary systems, hydro-wind complementary systems, and hydro-wind-solar complementary systems. Domestic and foreign research on multi-energy complementary systems is mainly divided into three levels: unit capacity configuration, hydropower and wind-solar energy complementarity characteristics, and multi-energy complementary system operation and scheduling. However, based on current research results, how to consider the correlation between different resources of hydropower, wind and solar in the long-term multi-energy complementary scheduling process to more reasonably describe the uncertainty in the complementary system, and how to solve the robustness of multi-objective optimization methods in different scheduling scenarios have not been well solved. At the same time, due to the significant regional differences and industry characteristics of the long-term scheduling problem of multi-energy complementarity, most of the research in various countries is carried out for a specific case in combination with their own needs and multi-energy structure and distribution characteristics. The long-term multi-energy complementary random scheduling method used to guide cascade hydropower to absorb wind and solar is still an important scientific problem that needs to be solved urgently. Summary of the invention

[0004] In view of the defects of the prior art, the present invention provides a long-term random coordinated scheduling method for a hydropower-wind power-photovoltaic complementary system to overcome the defects of the effective solution to the long-term random coordinated scheduling problem of a hydropower-wind power-photovoltaic complementary system in the background technology, which can effectively cope with the uncertainty brought about by the large-scale grid connection of new energy and improve the reliability and robustness of hydropower plans.

[0005] The technical solution adopted by the present invention is as follows:

[0006] The present invention provides a long-term random coordinated scheduling method for a hydropower-wind power-photovoltaic complementary system, comprising the following steps:

[0007] Step S1, determine resource variables They are runoff inflow, wind power output and photovoltaic output respectively; obtain the attribute information of the cascade power stations in the basin, which includes the constraint boundaries of each hydropower station and the resource variables of each resource in historical years. The resource variable historical value time series, where ;

[0008] Step S2, scene generation:

[0009] Step S2.1: quantify each resource variable of the cascade hydropower stations in the basin based on Markov chain and Copula function The time correlation and the correlation between resource variables generate each resource variable The initial set of simulation scenarios;

[0010] Step S2.2: Based on the probability distance, each resource variable The initial simulation scenario set is reduced to obtain each resource variable The reduced A simulation scenario set of simulation scenarios;

[0011] Therefore, for the runoff, wind power output and photovoltaic output resource variables, we obtain The simulation scenario set of each resource variable is randomly arranged to obtain the final runoff simulation scenario set. , Wind power output simulation scenario set and photovoltaic output simulation scenario set ;

[0012] Set the runoff simulation scenario , Wind power output simulation scenario set and photovoltaic output simulation scenario set Combine to form a comprehensive scene set ; Among them, for the comprehensive scene set Comprehensive scene in , , with the following relationship: At the same time, the comprehensive scene Probability ; Runoff simulation scenarios , Wind power output simulation scenario And photovoltaic output simulation scene The probability of

[0013] Step S3, constructing a long-term random coordinated dispatch model for the basin cascade hydropower-wind power-photovoltaic complementary system;

[0014] Step S4, with comprehensive scene set As input, the long-term random coordinated dispatch model of the cascade hydropower-wind power-photovoltaic complementary system in the basin is solved to obtain the hydropower station applicable to all comprehensive scenarios. In each period of the scheduling cycle Outbound flow , to achieve long-term random coordinated scheduling of water, wind and solar power.

[0015] Preferably, the constraint boundaries of the hydropower station include the maximum transmission capacity of each transmission section, the comprehensive output coefficient of each hydropower station, the upper and lower limits of the total outflow flow and power generation flow of each reservoir, the upper and lower limits of the output and storage capacity of each reservoir, the maximum and minimum water level limits and initial and final water levels of each reservoir, the water level-storage capacity characteristic curve and the tail water level-outflow flow characteristic curve data and installed capacity.

[0016] Preferably, step S2.1 is specifically:

[0017] Step S2.1.1, resource variables is a random variable; the resource variables are fitted using the nonparametric kernel density estimation method. The marginal distribution of

[0018] Step S2.1.2, determine the central variable of the first layer of the vine structure:

[0019] Calculate resource variables separately and The Kendall coefficient between , resource variables and The Kendall coefficient between , and resource variables and The Kendall coefficient between ;

[0020] Compute and Resource Variables The combined coefficient of association = ; Computation and resource variables The combined coefficient of association = ; Computation and resource variables The combined coefficient of association = ;

[0021] Comparison of combination coefficients , and , determine the maximum combination coefficient, the resource variable corresponding to the maximum combination coefficient is the central variable of the first layer of the vine structure;

[0022] Step S2.1.3, build three-dimensional Rattan structure:

[0023] The central variable of the first layer of the vine structure is expressed as , and the other two resource variables are represented as and , thus the first layer of vine structure is established as and , the second layer of vine structure is Three-dimensional Rattan structure; among them, and Respectively and The joint distribution function of the determined copula, and and The joint distribution function of the determined copula;

[0024] Step S2.1.4, based on the resource variables of step S2.1.1 The marginal distribution of the three-dimensional The maximum likelihood method is used to estimate the parameters of each Coupla function, and the AIC criterion is used to determine the Coupla function of each edge. The Coupla function with the smallest AIC value is selected as the final Coupla function of each edge, thereby obtaining the Coupla function connection rattan model;

[0025] Step S2.1.5, generate the central variable of the first layer of the vine structure Considering the time correlation scenario, and determining the cumulative probability of the time correlation scenario as a uniform variable ,and ;

[0026] Step S2.1.6, let are three sets of solution variables to be solved, so , Treated as solution variable The sample points;

[0027] Step S2.1.7, generate two random numbers that follow a uniform distribution, defined as uniform variables , ;

[0028] Use the binary method to solve the formula , the solution is the variable to be solved of sample points; among them, Indicates known Under the conditions The distribution function of is the conditional distribution function; express and Coupla probability distribution function between ;

[0029] Step S2.1.8, according to the formula , and the solution is ;in: Indicates known Under the conditions The distribution function of Indicates known Under the conditions The distribution function of Indicates known Under the conditions The distribution function of Indicates known Under the conditions, and Coupla probability distribution function between ;

[0030] According to the formula , solve to get the variable to be solved of sample points; among them, express and Coupla probability distribution function between ;

[0031] Step S2.1.9, respectively Perform inverse transform and sample multiple times to obtain The corresponding random sample set, The corresponding random sample set, The corresponding random sample set; each random sample set is an initial simulation scenario set of a resource variable that comprehensively considers the time correlation and the correlation between heterogeneous energy sources; thus, each resource variable is obtained The initial set of simulation scenarios.

[0032] Preferably, step S2.1.5 generates the central variable of the first layer of the vine structure Considering the time correlation scenario, and determining the cumulative probability of the time correlation scenario as a uniform variable , specifically:

[0033] Step A1: Set the central variable of the first layer of the vine structure The corresponding resource variable is the resource variable , ; Determine resource variables The corresponding state space; then, apply the Markov chain to the resource variable The historical value time series is analyzed for time correlation to obtain the resource variable The cumulative probability state transfer matrix ;

[0034] Determine the research time scale as a period Length ; Set the number of simulation scenarios B and resource variables that need to be initially generated The initial value of the resource variable The cumulative probability state transfer matrix , using random numbers to generate the The resource variable The simulation value time series forms a simulation scenario, that is, a scenario considering time correlation; the loop is executed B times to obtain the resource variable of A simulation scenario is formed to form the resource variable The initial set of simulation scenarios;

[0035] Step A2: Analyze the cumulative probability of the initial simulation scenario set in step A1 and determine it as a uniform variable .

[0036] Preferably, step A1 specifically comprises:

[0037] Step A11, for resource variables , determine the corresponding state space; the state space has states, namely state 1, state 2, ..., state ;

[0038] Step A12, applying the Markov chain to the resource variable The historical value time series is analyzed for time correlation to obtain the resource variable The state transition probability matrix :

[0039]

[0040] in: Represents a resource variable From the status To status The probability of transition between states; ; ;

[0041] Step A13, based on resource variables The state transition probability matrix , get the resource variable The cumulative probability state transfer matrix :

[0042]

[0043] in: Represents a resource variable From the status To status The cumulative probability of state transition; ;

[0044] Step A14, generate resource variables A simulation scenario:

[0045] Step A141, set resource variables Initial value of , according to the initial value , get the initial state ;in, ;

[0046] Step A142, set the time period =1;

[0047] Step A143, generate a random number , ; In the cumulative probability state transfer matrix In the state The corresponding row vector is: Row vector of rows: ;

[0048] Step A144, traverse the row vector from front to back When the element in is greater than or equal to the random number Elements hour, , the element The corresponding state after state transfer , which is the time period The state of , re-expressed as state ;state Corresponding resource variables The value range is: ;

[0049] Step A145, generate a random number , ; Through the formula , get the resource variable In the period The simulated value of ;in, For resource variables In the period The simulated value of

[0050] Step A146, determine the time period Whether the total number of preset scheduling periods has been reached If it is reached, execute step A147; if it is not reached, set , return to step A143;

[0051] Step A147, output the resource variable obtained at this time The simulated values ​​in time period 1, time period 2, ..., time period T form a simulated value time series , forming a simulation scenario;

[0052] Step A148, determine whether the current If there is no simulation scenario, the process returns to step A142 and forms another simulation scenario by executing steps A142 to A147. This process is repeated until a simulation scenario is obtained. A simulation scenario is formed to form the resource variable The initial set of simulation scenarios.

[0053] Preferably, step S2.2, based on the probability distance, each resource variable The initial simulation scenario set is reduced to obtain each resource variable The reduced The simulation scenario set of simulation scenarios is as follows:

[0054] Step S2.2.1, resource variables The initial set of simulation scenarios includes simulation scenarios, expressed as: simulation scenario set ; Respectively Simulation scenarios;

[0055] Step S2.2.2, set each simulation scenario The initial probability Equal, both ;in, ; Create a collection , used to store scenes that need to be eliminated;

[0056] Step S2.2.3, set up the collection is empty;

[0057] Step S2.2.4, traverse the simulation scene set Each simulation scenario in ;

[0058] For each simulation scenario traversed , determine the distance simulation scene Recent simulation scenarios , that is, it satisfies the relationship:

[0059]

[0060] in: Represents a simulation scenario and simulation scenarios The distance between Represents a simulation scenario and recent simulation scenarios The distance between

[0061] Using formula , calculate the simulation scenario The corresponding minimum distance probability ;

[0062] Step S2.2.5, compare simulation scenario sets Various simulation scenarios The minimum distance probability , using the formula , get the simulation scene set The minimum distance probability in The minimum value of ;Should Corresponding simulation scenario , distance simulation scene The most recent simulation scenario is ;

[0063] Step S2.2.6, let the simulation scenario set = -{ }, = +{ }, and let ; Determine the simulation scenario set Is the number of simulation scenarios reduced to If yes, execute step S2.2.7; if no, return to step S2.2.4;

[0064] Step S2.2.7, obtain the reduced A simulation scenario set of simulation scenarios, and each simulation scenario in the simulation scenario set has a probability.

[0065] Preferably, the constructed long-term random coordinated dispatch model of the basin cascade hydropower-wind power-photovoltaic complementary system is based on the hydropower station In the period Outbound flow As the decision variable of the model; the objective function includes the objective function And the objective function ;

[0066] Objective Function :Maximize the total power generation expectation of the water-wind-solar complementary system

[0067]

[0068] in: Indicates hydropower station In the comprehensive scene Below the period contribution; ; is the total number of long-term scheduling periods; ; Indicates the number of hydropower stations; ; Represents a comprehensive scene set The number of comprehensive scenarios in the Represents a comprehensive scenario The probability of Indicates the length of the long-term scheduling period; Indicates that wind turbines are in comprehensive scenarios Below the period contribution; Indicates that the photovoltaic unit is in a comprehensive scenario Below the period contribution; In the comprehensive scene Under the constraints The degree of constraint violation when insufficient; In the comprehensive scene Under the constraints the degree of constraint violation when exceeded; Representing constraints The penalty factor of ; Indicates the set of constraints that require adding penalty factor terms;

[0069] Objective Function :Maximize the minimum output expectation

[0070]

[0071] The constraints include:

[0072] (1) Water balance equation:

[0073] in: , Hydropower Station In the period and time period Storage capacity; For hydropower station In the period In comprehensive scenes Runoff inflow under For hydropower station In the period of runoff inflow for all composite scenarios; For hydropower station The upstream hydropower station In the period of runoff inflow; Hydropower Station a collection of upstream hydroelectric stations;

[0074] (2) Outbound flow equation:

[0075] in: , Hydropower Station In the period The power generation flow and abandoned water flow;

[0076] (3) Outbound flow restrictions:

[0077]

[0078] in: , Hydropower Station In the period The lower and upper bounds of the total outbound flow; It represents the degree of constraint violation when constraint 1 is insufficient in all comprehensive scenarios; Indicates constraints in all comprehensive scenarios The degree of constraint violation when exceeded; constraint 1 is the outbound flow restriction constraint;

[0079] (4) Power generation flow limit:

[0080]

[0081] in: , Hydropower Station In the period Lower and upper bounds on the generation flow; It represents the constraint violation degree when constraint 2 is insufficient in all comprehensive scenarios; Indicates constraints in all comprehensive scenarios The constraint violation degree when exceeded; Constraint 2 is the power generation flow limit constraint;

[0082] (5) Output limit:

[0083]

[0084] in: , Hydropower Station In the period Lower and upper bounds on output; Indicated in comprehensive

[0085] Scenario The constraint violation degree when the lower constraint condition 3 is insufficient; In the comprehensive scene Under the constraints The constraint violation degree when exceeded; constraint 3 is the output limit constraint;

[0086] (6) Storage capacity limit:

[0087] in: , Hydropower Station In the period The lower and upper limits of the reservoir capacity; In the comprehensive scene The constraint violation degree when the lower constraint condition 4 is insufficient; In the comprehensive scene Under the constraints The degree of constraint violation when exceeded; constraint 4 is the storage capacity limit constraint;

[0088] (7) Water level constraints at the beginning and end of the dispatch period:

[0089]

[0090]

[0091] in: , Hydropower Station The initial and final water levels; , Hydropower Station Water level in period 1 and period T; In the comprehensive scene The constraint violation degree when the lower constraint condition 5 is insufficient; In the comprehensive scene Under the constraints The constraint violation degree when exceeded; Constraint 5 is the water level constraint at the beginning and end of the dispatch period;

[0092] (8) Water level-reservoir capacity constraints:

[0093]

[0094] in: It is a nonlinear function between reservoir water level and water storage; For hydropower station In the period In comprehensive scenes The water level below;

[0095] (9) Tailwater level-discharge constraint:

[0096] in: For hydropower station Non-linear function between tailwater level and outflow rate; For hydropower station In the period In comprehensive scenes The tailwater level below;

[0097] (10) Hydropower output function:

[0098]

[0099] in: For hydropower station In the period In comprehensive scenes The power generation head under For hydropower station Output coefficient; For hydropower station In the period In comprehensive scenes The water level below;

[0100] (11) Transmission capacity limitation constraints:

[0101]

[0102] in: Represents hydroelectric power station A collection of associated wind farms; Represents hydroelectric power station Related wind farms contribution; Represents hydroelectric power station A collection of associated photovoltaic power plants; Represents hydroelectric power station The output of the associated photovoltaic power station sol; Indicates hydropower station Total output of For hydropower station The maximum value of integrated water, wind and solar transmission capacity; It represents the constraint violation degree when constraint 6 is insufficient in all comprehensive scenarios; Indicates constraints in all comprehensive scenarios The degree of constraint violation when exceeded; constraint 6 is the transmission capacity limit constraint.

[0103] Preferably, step S4 is specifically:

[0104] Step S4.1, generate individual code :

[0105] ;in, Meaning: Hydropower Station In the period The runoff inflow under all comprehensive scenarios is the decision variable; according to the individual coding , according to the set population size, generate the corresponding number of individuals with initialization solutions to form an initialized population;

[0106] Step S4.2, determine the population size NI and ratio ; Among them, the ratio = The number of individuals in the dominant population / the number of individuals in the basic population;

[0107] Step S4.3, using the inter-individual dominance algorithm to determine the dominant population and the basic population;

[0108] Step S4.4, adopt the multi-recombination operator fusion strategy to perform crossover operation on the dominant population and the basic population to generate a new population; then return to step S4.3, re-adopt the inter-individual domination algorithm to divide the new population into the dominant population and the basic population, and repeat this cycle until the set number of cycles is reached, and then execute step S4.5;

[0109] Step S4.5, for each individual in the new population, a heuristic constraint processing method is used to perform heuristic constraint processing on the solution value of each individual, so as to obtain each individual with an updated solution value;

[0110] Step S4.6, based on the updated solution value of each individual obtained in step S4.5, calculate the objective function value of the objective function F1 and the objective function F2 corresponding to each individual;

[0111] Step S4.7, based on the objective function values ​​corresponding to each individual, determine whether the iteration termination condition is met through the restart mechanism. If so, execute step S4.8; if not, determine whether a restart mechanism is needed. If so, start the restart mechanism, update the basic population, and return to step S4.4; if not, return to step S4.3;

[0112] Step S4.8, output the solution value of each individual in the dominant population. The solution value of each individual is a scheduling plan, thereby obtaining a set of multi-objective optimization scheduling plans, thereby completing the long-term random coordinated scheduling of the hydropower station group.

[0113] Preferably, step S4.3 is specifically:

[0114] Step S4.3.1, setting parameters The value of

[0115] Step S4.3.2, with the objective function The function value of the horizontal axis is the objective function The function value of the vertical coordinate is For the grid side length, establish a gridded objective function space;

[0116] Step S4.3.3, for each current individual X c , c=1,2,…, NI, according to its objective function The function value and objective function The function value of , determines the grid where it is located in the objective function space;

[0117] Step S4.3.4, traverse each individual and calculate each individual X c The dominance number NC is calculated as:

[0118] For individual X c , the initial value of its dominance number NC is 0; traverse the other NI-1 individuals in turn, and for the currently traversed individual X d , d = 1, 2, ..., NI, and d ≠ c, if individual X d and individual X c The function values ​​of X are located in the same grid in the objective function space, then Algorithm 1 is used to determine the individual X c Whether it dominates individual X d , if it dominates, then the domination number NC increases by 1; if individual X d and individual X c If the function value of is located in different grids in the objective function space, then Algorithm 2 is used to determine the individual X c Whether it dominates individual X dIf it dominates, the domination number NC increases by 1; thus, the traversal of the other NI-1 individuals is completed, and the value of the domination number NC is obtained, which is the value of individual X c The domination number NC represents the domination number of individual X. c The number of individuals controlled;

[0119] Among them: Algorithm 1 and Algorithm 2 are:

[0120] Individual X c , individual X c The objective function The function value is ; Individual X c The objective function The function value is , individual X c The objective function value vector is expressed as: = ;

[0121] Individual X d The objective function The function value is ; Individual X d The objective function The function value is , individual X d The objective function value vector is expressed as: = ;

[0122] Algorithm 1 is: When conditions 1 and 2 are met, individual X c Dominant individual X d ;

[0123] Condition 1: ≥ ,and ≥ ;

[0124] Condition 2: Existence , x= , satisfying the relationship: ,and, ;

[0125] Algorithm 2 is: if and only if one of conditions 3 and 4 is met, individual X c Dominant individual X d :

[0126] Condition 3:

[0127] in: represent and The relationship between conditions 1 and 2 is satisfied. represents round down;

[0128] Condition 4:

[0129] in: is the modulo operation; is logical AND;

[0130] Step S4.3.5, based on population size NI and ratio , and get the number of individuals in the dominant population and the number of individuals in the basic population ;

[0131] Sort the individuals in descending order of dominance numbers. If two individuals have the same dominance numbers, sort them randomly. Among the sorted NI individuals, select the one with the highest dominance number. A certain number of individuals form the dominant population, and the other individuals form the basic population.

[0132] Preferably, step S4.4 is specifically:

[0133] Step S4.4.1, setting a crossover operator pool; the crossover operator pool has N_CO crossover operators;

[0134] Step S4.4.2, initially, each crossover operator CO is set to have the same selection probability PC;

[0135] Step S4.4.3, select a crossover operator to perform a crossover operation, and update the selection probability PC of the crossover operator:

[0136] ① Based on the selection probability PC of each current crossover operator, select the crossover operator A from the crossover operator pool; ② According to the number of parents N required by the crossover operator A A , randomly select a parent individual X from the dominant population A1 , select N from the basic population through a tournament A -1 parent individual, expressed as: parent individual X AB ,B=2,3,…,N A ;

[0137] ③Use the crossover operator A to perform crossover on the parent individual X A1 and parent individual X AB Perform crossover operation to generate N A offspring individuals, and N A The original parent individuals are deleted from the population;

[0138] ④According to N AThe function value of the offspring individuals is used to determine the grid in the objective function space, and the offspring individuals within the dominant population range are added to the dominant population, and the other offspring individuals are added to the basic population; where: the number of offspring individuals added to the dominant population generated by the crossover operator A this time is ;

[0139] ⑤ Update the selection probability PC of the crossover operator A using the following formula: A :

[0140]

[0141] in: Represents the total number of offspring individuals added to the dominant population generated by the crossover operator V from the first cycle to the current cycle, V=1,2,…,N_CO; is a constant, ;

[0142] ⑥When this cycle ends, return to step ① and continue the next cycle until the set number of cycles is reached;

[0143] In step S4.7, it is determined whether a restart mechanism is needed. If necessary, the restart mechanism is started to update the basic population. Specifically, when the algorithm search is stagnant, the restart mechanism is started: according to the current ratio and the current number of dominant population individuals , using the formula / , get the number of individuals in the basic population; clear the current basic population, and fill the corresponding number of individuals with the highest NC ranking in the current dominant population into the basic population, and the insufficient part is obtained through the mutation of individuals in the dominant population, so as to get the updated basic population.

[0144] The long-term random coordinated dispatching method of a hydropower-wind power-photovoltaic complementary system provided by the present invention has the following advantages:

[0145] Based on historical information and statistical laws, the present invention quantifies the spatiotemporal relationship between hydropower, wind power and photovoltaic power generation based on the Markov chain and Copula joint distribution function, and generates corresponding runoff, wind and photovoltaic output scenarios to simulate uncertainty. On this basis, a multi-scenario long-term random collaborative scheduling optimization model for cascade power stations in a river basin is constructed considering the uncertainty of wind and solar power. Compared with the scenario generation method that regards different resources as independent individuals without considering the correlation between resources, the water, wind and solar scenario generation method formulated by the present invention is more reliable. Combined with the construction of a long-term random collaborative scheduling optimization model for cascade power stations in a river basin, it can effectively reduce scheduling risks and provide a new technical approach for the preparation of hydropower plans for hydropower-dominated high-proportion renewable energy power systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0146] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0147] Figure 1 It is a flow chart of a long-term random coordinated dispatching method for a hydropower-wind power-photovoltaic complementary system provided by the present invention;

[0148] Figure 2 This is an example diagram of a cascade hydropower-wind power-photovoltaic complementary system provided by an embodiment of the present invention;

[0149] Figure 3 It is a wind power-photovoltaic Frank-Copula probability density function and joint distribution function diagram provided by an embodiment of the present invention;

[0150] Figure 4 It is a runoff-photovoltaic Clayton-Copula probability density function and a joint distribution function diagram provided by an embodiment of the present invention;

[0151] Figure 5 It is a graph of the probability density function and joint distribution function of the water-wind-light Gumbel Copula provided by an embodiment of the present invention;

[0152] Figure 6 This is a scene diagram after water, wind and light reduction provided by an embodiment of the present invention;

[0153] Figure 7 It is a process diagram of water level process optimization calculation results provided by an embodiment of the present invention;

[0154] Figure 8 It is a process diagram of the optimization calculation results of the hydropower output process provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0155] In order to make the technical problems, technical solutions and beneficial effects solved by the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0156] Those skilled in the art will appreciate that the following examples are only used to illustrate the present invention and should not be considered to limit the scope of the present invention. If no specific techniques or conditions are specified in the examples, the techniques or conditions described in the literature in the art or the product specifications are used. If the manufacturer of the materials or equipment used is not specified, they are all conventional products that can be purchased.

[0157] See also Figure 1 The present invention provides a long-term random coordinated scheduling method for a hydropower-wind power-photovoltaic complementary system, comprising the following steps:

[0158] Step S1, determine resource variables They are runoff inflow, wind power output and photovoltaic output respectively; obtain the attribute information of the cascade power stations in the basin, which includes the constraint boundaries of each hydropower station and the resource variables of each resource in historical years. The resource variable historical value time series, where The constraint boundaries of the hydropower station include the maximum transmission capacity of each transmission section, the comprehensive output coefficient of each hydropower station, the upper and lower limits of the total outflow and power generation flow of each reservoir, the upper and lower limits of the output and storage capacity of each reservoir, the maximum and minimum water level limits and initial and final water levels of each reservoir, the water level-storage capacity characteristic curve and the tail water level-outflow characteristic curve data and installed capacity.

[0159] Step S2, scene generation:

[0160] Step S2.1: quantify each resource variable of the cascade hydropower stations in the basin based on Markov chain and Copula function The time correlation and the correlation between resource variables generate each resource variable The initial set of simulation scenarios;

[0161] Step S2.1 is specifically as follows:

[0162] Step S2.1.1, resource variables is a random variable; the resource variables are fitted using the nonparametric kernel density estimation method. The marginal distribution of

[0163] Step S2.1.2, determine the central variable of the first layer of the vine structure:

[0164] Calculate resource variables separately and The Kendall coefficient between , resource variables and The Kendall coefficient between , and resource variables and The Kendall coefficient between ;

[0165] Compute and Resource Variables The combined coefficient of association = ; Computation and resource variables The combined coefficient of association = ; Computation and resource variables The combined coefficient of association = ;

[0166] Comparison of combination coefficients , and , determine the maximum combination coefficient, the resource variable corresponding to the maximum combination coefficient is the central variable of the first layer of the vine structure;

[0167] Step S2.1.3, build three-dimensional Rattan structure:

[0168] The central variable of the first layer of the vine structure is expressed as , and the other two resource variables are represented as and , thus the first layer of vine structure is established as and , the second layer of vine structure is Three-dimensional Rattan structure; among them, and Respectively and The joint distribution function of the determined copula, and and The joint distribution function of the determined copula;

[0169] Step S2.1.4, based on the resource variables of step S2.1.1 The marginal distribution of the three-dimensional The maximum likelihood method is used to estimate the parameters of each Coupla function, and the AIC criterion is used to determine the Coupla function of each edge. The Coupla function with the smallest AIC value is selected as the final Coupla function of each edge, thereby obtaining the Coupla function connection rattan model;

[0170] Step S2.1.5, generate the central variable of the first layer of the vine structure Considering the time correlation scenario, and determining the cumulative probability of the time correlation scenario as a uniform variable ,and ;

[0171] The specific steps of this step are:

[0172] Step A1: Set the central variable of the first layer of the vine structure The corresponding resource variable is the resource variable , ; Determine resource variables The corresponding state space; then, apply the Markov chain to the resource variable The historical value time series is analyzed for time correlation to obtain the resource variable The cumulative probability state transfer matrix ;

[0173] Determine the research time scale as a period Length ; Set the number of simulation scenarios B and resource variables that need to be initially generated The initial value of the resource variable The cumulative probability state transfer matrix , using random numbers to generate the The resource variable The simulation value time series forms a simulation scenario, that is, a scenario considering time correlation; the loop is executed B times to obtain the resource variable of A simulation scenario is formed to form the resource variable The initial set of simulation scenarios;

[0174] Step A1 is specifically as follows:

[0175] Step A11, for resource variables , determine the corresponding state space; the state space has states, namely state 1, state 2, ..., state ;

[0176] Step A12, applying the Markov chain to the resource variable The historical value time series is analyzed for time correlation to obtain the resource variable The state transition probability matrix :

[0177]

[0178] in: Represents a resource variable From the status To status The probability of transition between states; ; ; State transition probability matrix In , the sum of the transition probabilities between states in each row is 1.

[0179] The expression is as follows:

[0180]

[0181] in: and It means and Time period resource variables status, For resource variables From the status Transfer to state The frequency of , , g is the conditional transition probability; in the present invention, the time period It can be a different time scale such as month, week, day, etc.

[0182] Step A13, based on resource variables The state transition probability matrix , get the resource variable Tired

[0183] Calculate the probability state transfer matrix :

[0184]

[0185] in: Represents a resource variable From the status To status The cumulative probability of state transition; ;

[0186] Step A14, generate resource variables A simulation scenario:

[0187] Step A141, set resource variables Initial value of , according to the initial value , get the initial state ;in, ;

[0188] Step A142, set the time period =1;

[0189] Step A143, generate a random number , ; In the cumulative probability state transfer matrix In the state The corresponding row vector is: Row vector of rows: ;

[0190] Step A144, traverse the row vector from front to back When the element in is greater than or equal to the random number Elements hour, , the element The corresponding state after state transfer , which is the time period The state of , re-expressed as state ;state Corresponding resource variables The value range is: ;

[0191] Step A145, generate a random number , ; Through the formula , get the resource variable In the period The simulated value of ;in, For resource variables In the period The simulated value of

[0192] Step A146, determine the time period Whether the total number of preset scheduling periods has been reached If it is reached, execute step A147; if it is not reached, set , return to step A143;

[0193] Step A147, output the resource variable obtained at this time The simulated values ​​in time period 1, time period 2, ..., time period T form a simulated value time series , forming a simulation scenario;

[0194] Step A148, determine whether the current If there is no simulation scenario, the process returns to step A142 and forms another simulation scenario by executing steps A142 to A147. This process is repeated until a simulation scenario is obtained. A simulation scenario is formed to form the resource variable The initial set of simulation scenarios.

[0195] Step A2: Analyze the cumulative probability of the initial simulation scenario set in step A1 and determine it as a uniform variable .

[0196] Step S2.1.6, let are three sets of solution variables to be solved, so , Treated as solution variable The sample points;

[0197] Step S2.1.7, generate two random numbers that follow a uniform distribution, defined as uniform variables , ;

[0198] because and All are known quantities, and the formula is solved by bisection method , the solution is the variable to be solved of sample points; among them, Indicates known Under the conditions The distribution function of is the conditional distribution function; express and Coupla probability distribution function between ;

[0199] Step S2.1.8, similarly, due to , It has been defined, according to the formula , and the solution is ;in: Indicates known Under the conditions The distribution function of Indicates known Under the conditions The distribution function of Indicates known Under the conditions The distribution function of Indicates known Under the conditions, and Coupla probability distribution function between ;

[0200] According to the formula , solve to get the variable to be solved of sample points; among them, express and Coupla probability distribution function between ;

[0201] Step S2.1.9, respectively Perform inverse transform and sample multiple times to obtain The corresponding random sample set, The corresponding random sample set, The corresponding random sample set; each random sample set is an initial simulation scenario set of a resource variable that comprehensively considers the time correlation and the correlation between heterogeneous energy sources; thus, each resource variable is obtained The initial set of simulation scenarios.

[0202] Step S2.2: Based on the probability distance, each resource variable The initial simulation scenario set is reduced to obtain each resource variable The reduced A simulation scenario set of simulation scenarios;

[0203] Therefore, for the runoff, wind power output and photovoltaic output resource variables, we obtain The simulation scenario set of each resource variable is randomly arranged to obtain the final runoff simulation scenario set. , Wind power output simulation scenario set and photovoltaic output simulation scenario set ;

[0204] Set the runoff simulation scenario , Wind power output simulation scenario set and photovoltaic output simulation scenario set Combine to form a comprehensive scene set ; Among them, for the comprehensive scene set Comprehensive scene in , , with the following relationship: At the same time, the comprehensive scene Probability ; Runoff simulation scenarios , Wind power output simulation scenario And photovoltaic output simulation scene The probability of

[0205] In this step, each resource variable is classified based on the probability distance. The initial simulation scenario set is reduced to obtain each resource variable The reduced The simulation scenario set of simulation scenarios is as follows:

[0206] Step S2.2.1, resource variables The initial set of simulation scenarios includes simulation scenarios, expressed as: simulation scenario set ; Respectively Simulation scenarios;

[0207] Step S2.2.2, set each simulation scenario The initial probability Equal, both ;in, ; Create a collection , used to store scenes that need to be eliminated;

[0208] Step S2.2.3, set up the collection is empty;

[0209] Step S2.2.4, traverse the simulation scene set Each simulation scenario in ;

[0210] For each simulation scenario traversed , determine the distance simulation scene Recent simulation scenarios , that is, it satisfies the relationship:

[0211]

[0212] in: Represents a simulation scenario and simulation scenarios The distance between Represents a simulation scenario and recent simulation scenarios The distance between

[0213] Using formula , calculate the simulation scenario The corresponding minimum distance probability ;

[0214] Step S2.2.5, compare simulation scenario sets Various simulation scenarios The minimum distance probability , using the formula , get the simulation scene set The minimum distance probability in The minimum value of ;Should Corresponding simulation scenario , distance simulation scene The most recent simulation scenario is ;

[0215] Step S2.2.6, let the simulation scenario set = -{ }, = +{ }, and let ; Determine the simulation scenario set Is the number of simulation scenarios reduced to If yes, execute step S2.2.7; if no, return to step S2.2.4;

[0216] Step S2.2.7, obtain the reduced A simulation scenario set of simulation scenarios, and each simulation scenario in the simulation scenario set has a probability.

[0217] The steps of scene reduction can be described as:

[0218] 1. Determine the original scene probability .

[0219] 2. Calculate the probability distance between each scene. The scene with the smallest probability distance is the scene that needs to be eliminated.

[0220] 3. Change the total number of scenarios and the corresponding probability of the corresponding scenarios: Total number of scenarios , add the probability of the removed scene to the scene closest to it to ensure that the sum of the probabilities of the retained scenes is 1.

[0221] 4. If the total number of remaining scenes If the number of scenes is greater than the specified number of scenes to be retained, the process returns to step 2 until the number of scenes to be retained is reduced to the specified number of scenes to be retained.

[0222] In the above steps: A1 is the time correlation analysis of the water-wind-solar hybrid system; Step S2.1.1-Step S2.1.5

[0223] It is the correlation analysis between resources; Step S2.1.6-Step S2.1.9 is the uncertainty scenario generation; Step S2.2 is the uncertainty scenario reduction.

[0224] Step S3, constructing a long-term random coordinated dispatch model for the basin cascade hydropower-wind power-photovoltaic complementary system;

[0225] The long-term random coordinated dispatch model of the basin cascade hydropower-wind power-photovoltaic complementary system is constructed, with hydropower stations as the In the period Outbound flow As the decision variable of the model; the objective function includes the objective function And the objective function ;

[0226] The present invention is used to construct a set of solutions that are suitable for processing in multiple scenarios. However, due to the existence of some extreme scenarios, there are situations in the solution that do not meet the constraints. In order to deal with this situation, indicators are used to measure the quality of the solution, and a model with a penalty term is established for this purpose.

[0227] Objective Function :Maximize the total power generation expectation of the water-wind-solar complementary system

[0228]

[0229] in: Indicates hydropower station In the comprehensive scene Below the period contribution; ; is the total number of long-term scheduling periods; ; Indicates the number of hydropower stations; ; Represents a comprehensive scene set The number of comprehensive scenarios in the Represents a comprehensive scenario The probability of Indicates the length of the long-term scheduling period; Indicates that wind turbines are in comprehensive scenarios Below the period contribution; Indicates that the photovoltaic unit is in a comprehensive scenario Below the period contribution; In the comprehensive scene Under the constraints The degree of constraint violation when insufficient; In the comprehensive scene Under the constraints the degree of constraint violation when exceeded; Representing constraints The penalty factor of ; Indicates the set of constraints that require adding penalty factor terms;

[0230] Objective Function :Maximize the minimum output expectation

[0231]

[0232] The constraints include:

[0233] (1) Water balance equation:

[0234] in: , Hydropower Station In the period and time period Storage capacity; For hydropower station In the period In comprehensive scenes Runoff inflow under For hydropower station In the period of runoff inflow for all composite scenarios; For hydropower station The upstream hydropower station In the period of runoff inflow; Hydropower Station a collection of upstream hydroelectric stations;

[0235] (2) Outbound flow equation:

[0236] in: , Hydropower Station In the period The power generation flow and abandoned water flow;

[0237] (3) Outbound flow restrictions:

[0238]

[0239] in: , Hydropower Station In the period The lower and upper bounds of the total outbound flow; It represents the degree of constraint violation when constraint 1 is insufficient in all comprehensive scenarios; Indicates constraints in all comprehensive scenarios The degree of constraint violation when exceeded; constraint 1 is the outbound flow restriction constraint;

[0240] (4) Power generation flow limit:

[0241]

[0242] in: , Hydropower Station In the period Lower and upper bounds on the generation flow; It represents the constraint violation degree when constraint 2 is insufficient in all comprehensive scenarios; Indicates constraints in all comprehensive scenarios The constraint violation degree when exceeded; Constraint 2 is the power generation flow limit constraint;

[0243] (5) Output limit:

[0244]

[0245] in: , Hydropower Station In the period Lower and upper bounds on output; Indicated in comprehensive

[0246] Scenario The constraint violation degree when the lower constraint condition 3 is insufficient; In the comprehensive scene Under the constraints The constraint violation degree when exceeded; constraint 3 is the output limit constraint;

[0247] (6) Storage capacity limit:

[0248] in: , Hydropower Station In the period The lower and upper limits of the reservoir capacity; In the comprehensive scene The constraint violation degree when the lower constraint condition 4 is insufficient; In the comprehensive scene Under the constraints The degree of constraint violation when exceeded; constraint 4 is the storage capacity limit constraint;

[0249] (7) Water level constraints at the beginning and end of the dispatch period:

[0250]

[0251]

[0252] in: , Hydropower Station The initial and final water levels; , Hydropower Station Water level in period 1 and period T; In the comprehensive scene The constraint violation degree when the lower constraint condition 5 is insufficient; In the comprehensive scene Under the constraints The constraint violation degree when exceeded; Constraint 5 is the water level constraint at the beginning and end of the dispatch period;

[0253] (8) Water level-reservoir capacity constraints:

[0254]

[0255] in: It is a nonlinear function between reservoir water level and water storage; For hydropower station In the period In comprehensive scenes The water level below;

[0256] (9) Tailwater level-discharge constraint:

[0257] in: For hydropower station Non-linear function between tailwater level and outflow rate; For hydropower station In the period In comprehensive scenes The tailwater level below;

[0258] (10) Hydropower output function:

[0259]

[0260] in: For hydropower station In the period In comprehensive scenes The power generation head under For hydropower station Output coefficient; For hydropower station In the period In comprehensive scenes The water level below;

[0261] (11) Transmission capacity limitation constraints:

[0262]

[0263] in: Represents hydroelectric power station A collection of associated wind farms; Represents hydroelectric power station Related wind farms contribution; Represents hydroelectric power station A collection of associated photovoltaic power plants; Represents hydroelectric power station The output of the associated photovoltaic power station sol; Indicates hydropower station Total output of For hydropower station The maximum value of integrated water, wind and solar transmission capacity; It represents the constraint violation degree when constraint 6 is insufficient in all comprehensive scenarios; Indicates constraints in all comprehensive scenarios The degree of constraint violation when exceeded; constraint 6 is the transmission capacity limit constraint.

[0264] Step S4, with comprehensive scene set As input, the long-term random coordinated dispatch model of the cascade hydropower-wind power-photovoltaic complementary system in the basin is solved to obtain the hydropower station applicable to all comprehensive scenarios. In each period of the scheduling cycle Outbound flow , to achieve long-term random coordinated scheduling of water, wind and solar power.

[0265] Step S4 is specifically as follows:

[0266] Step S4.1, generate individual code :

[0267] ;in, Meaning: Hydropower Station In the period The runoff inflow under all comprehensive scenarios is the decision variable; according to the individual coding , according to the set population size, generate the corresponding number of individuals with initialization solutions to form an initialized population;

[0268] Step S4.2, determine the population size NI and ratio ; Among them, the ratio = The number of individuals in the dominant population / the number of individuals in the basic population;

[0269] Step S4.3, using the inter-individual dominance algorithm to determine the dominant population and the basic population;

[0270] Step S4.3 is specifically as follows:

[0271] Step S4.3.1, setting parameters The value of

[0272] Step S4.3.2, with the objective function The function value of the horizontal axis is the objective function The function value of the vertical coordinate is For the grid side length, establish a gridded objective function space;

[0273] Step S4.3.3, for each current individual X c , c=1,2,…, NI, according to its objective function The function value and objective function The function value of , determines the grid where it is located in the objective function space;

[0274] Step S4.3.4, traverse each individual and calculate each individual X c The dominance number NC is calculated as:

[0275] For individual X c , the initial value of its dominance number NC is 0; traverse the other NI-1 individuals in turn, and for the currently traversed individual X d , d = 1, 2, ..., NI, and d ≠ c, if individual X d and individual X c The function values ​​of X are located in the same grid in the objective function space, then Algorithm 1 is used to determine the individual X c Whether it dominates individual X d , if it dominates, then the domination number NC increases by 1; if individual X d and individual X c If the function value of is located in different grids in the objective function space, then Algorithm 2 is used to determine the individual X c Whether it dominates individual X d If it dominates, the domination number NC increases by 1; thus, the traversal of the other NI-1 individuals is completed, and the value of the domination number NC is obtained, which is the value of individual X c The domination number NC represents the domination number of individual X. c The number of individuals controlled;

[0276] Among them: Algorithm 1 and Algorithm 2 are:

[0277] Individual X c , individual X c The objective function The function value is ; Individual X c The objective function The function value is , individual X c The objective function value vector is expressed as: = ;

[0278] Individual X d The objective function The function value is ; Individual X d The objective function The function value is , individual X d The objective function value vector is expressed as: = ;

[0279] Algorithm 1 is: When conditions 1 and 2 are met, individual X c Dominant individual X d ;

[0280] Condition 1: ≥ ,and ≥ ;

[0281] Condition 2: Existence , x= , satisfying the relationship: ,and, ;

[0282] Algorithm 2 is: if and only if one of conditions 3 and 4 is met, individual X c Dominant individual X d :

[0283] Condition 3:

[0284] in: represent and The relationship between conditions 1 and 2 is satisfied. represents round down;

[0285] Condition 4:

[0286] in: is the modulo operation; is logical AND;

[0287] Step S4.3.5, based on population size NI and ratio , and get the number of individuals in the dominant population and the number of individuals in the basic population ;

[0288] Sort the individuals in descending order of dominance numbers. If two individuals have the same dominance numbers, sort them randomly. Among the sorted NI individuals, select the one with the highest dominance number. A certain number of individuals form the dominant population, and the other individuals form the basic population.

[0289] Step S4.4, adopt the multi-recombination operator fusion strategy to perform crossover operation on the dominant population and the basic population to generate a new population; then return to step S4.3, re-adopt the inter-individual domination algorithm to divide the new population into the dominant population and the basic population, and repeat this cycle until the set number of cycles is reached, and then execute step S4.5;

[0290] Step S4.4 is specifically as follows:

[0291] Step S4.4.1, setting a crossover operator pool; the crossover operator pool has N_CO crossover operators;

[0292] Step S4.4.2, initially, each crossover operator CO is set to have the same selection probability PC;

[0293] Step S4.4.3, select a crossover operator to perform a crossover operation, and update the selection probability PC of the crossover operator:

[0294] ① Based on the selection probability PC of each current crossover operator, select the crossover operator A from the crossover operator pool; ② According to the number of parents N required by the crossover operator A A , randomly select a parent individual X from the dominant population A1 , select N from the basic population through a tournament A -1 parent individual, expressed as: parent individual X AB ,B=2,3,…,N A ;

[0295] ③Use the crossover operator A to perform crossover on the parent individual X A1 and parent individual X AB Perform crossover operation to generate N A offspring individuals, and N A The original parent individuals are deleted from the population;

[0296] ④According to N A The function value of the offspring individuals is used to determine the grid in the objective function space, and the offspring individuals within the dominant population range are added to the dominant population, and the other offspring individuals are added to the basic population; where: the number of offspring individuals added to the dominant population generated by the crossover operator A this time is ;

[0297] ⑤ Update the selection probability PC of the crossover operator A using the following formula: A :

[0298]

[0299] in: Represents the total number of offspring individuals added to the dominant population generated by the crossover operator V from the first cycle to the current cycle, V=1,2,…,N_CO; is a constant, ;

[0300] The constant Prevents operator probabilities from reaching 0, thus ensuring that no operators are "lost" during algorithm execution.

[0301] ⑥When this cycle ends, return to step ① and continue the next cycle until the set number of cycles is reached;

[0302] The multi-group operator fusion strategy of the present invention uses a pool including multiple crossover operators and establishes a feedback mechanism to compare the performance of operators during the search process. Operators with excellent performance will receive higher selection weights. Specifically, the selection weights of all operators are equal at the beginning, and their probabilities in the next selection are updated through the feedback mechanism, and the adjustment strategy is based on the number of solutions in the dominant population generated by the operator.

[0303] Step S4.5, for each individual in the new population, a heuristic constraint processing method is used to perform heuristic constraint processing on the solution value of each individual, so as to obtain each individual with an updated solution value;

[0304] Step S4.6, based on the updated solution value of each individual obtained in step S4.5, calculate the objective function value of the objective function F1 and the objective function F2 corresponding to each individual;

[0305] Step S4.7, based on the objective function values ​​corresponding to each individual, determine whether the iteration termination condition is met through the restart mechanism. If so, execute step S4.8; if not, determine whether a restart mechanism is needed. If so, start the restart mechanism, update the basic population, and return to step S4.4; if not, return to step S4.3;

[0306] Determine whether a restart mechanism is needed. If necessary, start the restart mechanism and update the basic population. Specifically, when the algorithm search is stagnant, start the restart mechanism: according to the current ratio and the current number of dominant population individuals , using the formula / , get the number of individuals in the basic population; clear the current basic population, and fill the corresponding number of individuals with the highest NC ranking in the current dominant population into the basic population, and the insufficient part is obtained through the mutation of individuals in the dominant population, so as to get the updated basic population.

[0307] Therefore, in the present invention, in order to prevent the algorithm from falling into a local optimum, a population restart mechanism is designed to detect the stagnation of the algorithm search. If the algorithm search is stagnant, the restart mechanism is used to regenerate the basic population. This mechanism updates the basic population individuals so that the algorithm can search again when it finds stagnation or falls into a local optimal solution to find a better solution.

[0308] Step S4.8, output the solution value of each individual in the dominant population. The solution value of each individual is a scheduling plan, thereby obtaining a set of multi-objective optimization scheduling plans, thereby completing the long-term random coordinated scheduling of the hydropower station group.

[0309] In practical applications, when solving the decision variables After that, it can be further transformed into the planned water level process and planned output process of the hydropower station. Specifically, after obtaining the decision variables After that, the specific outflow conditions of each hydropower station in different time periods can be determined. At this time, the water balance equation can be used to calculate the outflow conditions of each hydropower station in different time periods. Get storage capacity ; Get the storage capacity of the corresponding period Afterwards, the water level is calculated according to the water level-reservoir capacity constraint formula in the model, and the output of each power station is calculated according to the hydropower output generation function formula.

[0310] In order to make the technical problems, technical solutions and advantages to be solved by the present invention more clear, a detailed description will be given below with reference to the accompanying drawings and specific embodiments.

[0311] The model and method of the present invention are tested by taking the Hongshui River Basin power grid as an actual engineering background. Figure 1 A schematic flow chart of a long-term random coordinated dispatching method for a hydropower-wind-photovoltaic complementary system is shown. Figure 2 provides a hydropower-wind-photovoltaic complementary system structure, with Tianshengqiao I, Longtan and Yantan in the Hongshui River Basin as cascade power stations in the basin, of which Tianshengqiao I is located in the upstream, Longtan is located in the middle, and Yantan is located in the downstream. The wind and solar output of each power station is bundled with the hydropower station. The historical runoff data from 1951 to 2010 and the historical wind and solar output of Guangxi in 2020 were collected as samples.

[0312] A long-term random coordinated dispatching method for a hydropower-wind power-photovoltaic complementary system comprises the following steps:

[0313] S1: Obtain the attribute information of cascade power stations in the basin, including historical data of runoff and wind and solar power output, the maximum transmission capacity of each transmission section, the comprehensive output coefficient of each hydropower station, the upper and lower limits of the total outflow and power generation flow of each reservoir, the upper and lower limits of the output and storage capacity of each reservoir, the maximum and minimum water level limits and initial and final water levels of each reservoir, the water level-storage capacity characteristic curve and the tailwater level-outflow characteristic curve data and installed capacity;

[0314] S2: Scene Generation:

[0315] Based on Markov chain and Copula function, the time correlation of hydropower, wind power and photovoltaic power output and the correlation between resources of cascade power stations in the basin are quantified to obtain the initial scenario set; based on the initial scenario set, the initial scenarios are reduced based on the probability distance;

[0316] S3: Model building:

[0317] Construct a long-term random coordinated dispatch model for the basin cascade hydropower-wind power-photovoltaic complementary system;

[0318] According to the water, wind and light data of the study area, the present invention divides each variable into four states, establishes the corresponding Markov chain state transfer probability matrix and cumulative probability state transfer matrix to prove that the single variable of water, wind and light has time correlation.

[0319] The correlation between water, wind and photovoltaic resources is analyzed according to the Copula function. First, based on the runoff data, wind power output and photovoltaic output data of the cascade reservoirs, their respective empirical distribution functions are calculated, and the cumulative probability distribution of the three variables is determined by the non-parametric estimation method. By comparison, it is found that the two curves of every two variables basically coincide, which proves that the non-parametric kernel density estimation method is suitable for fitting the marginal distribution of each variable.

[0320] Then, the Kendall coefficients between each variable were calculated according to the Kendall correlation coefficient calculation formula. The Kendall coefficients of photovoltaic output, runoff and wind power output were 0.455 and 0.576 respectively, which were moderately correlated and met the maximum sum of the absolute values ​​of the Kendall coefficients. Therefore, the photovoltaic variable was used as the central variable of the first layer of the vine structure. Let runoff, wind power and photovoltaic output be variables respectively. , the first layer of vine structure is and , the second layer of vine structure is Three-dimensional Vine structure.

[0321] Then let the marginal distributions of X and Y be and , the maximum likelihood method is used to estimate the linear correlation parameters of five commonly used Copula functions, namely Gaussian Copula, t Copula, Gumbel Copula, Clayton Copula and Frank Copula functions and the degrees of freedom of t-copula, and then the AIC criterion is used to determine the most suitable copula function for each side. The specific results are shown in Table 1. (Take Longtan Power Station as an example)

[0322] Table 1 AIC value calculation results of the copula function on each side of the vine structure

[0323] According to the calculation results in the above table, according to the principle that the lower the AIC value, the better the model, In the first layer of the vine model, Frank Copula is used for wind power output and photovoltaic output, and its parameter value is -1.7789; Clayton Copula is used for runoff and photovoltaic output, and its parameter value is 0.2511; for the second layer, Gumbel Copula is used to describe the correlation between water, wind and photovoltaic, and its parameter value is 2.6015. The probability density function and joint distribution function of each Copula function are shown in the following figure: Figure 3-5 shown.

[0324] Then, according to the scenario generation method of the present invention, since the hydropower station of the present invention belongs to a cascade reservoir, the runoff flow rules of the upstream and downstream reservoirs need to be considered when generating the runoff scenario, that is, the flow data of the downstream reservoir must be greater than both the midstream runoff flow and the upstream runoff flow, so the upstream data is used as the initial input data, and the midstream and downstream data are obtained by combining the obtained upstream sample data with the Copula function. First, according to the calculation results of the Kendall coefficient, the photovoltaic variable is selected as the central variable, the photovoltaic variable time correlation scenario is generated, and the corresponding cumulative probability is recorded as Secondly, according to the scenario generation steps S2.1.6-S2.1.7, the PV variable sample points are obtained. , wind power variable sample points Since the Tianshengqiao Hydropower Station is located upstream, its relevant data is input into step S2.1.8 to obtain the upstream sample point of runoff variable Then, according to the Copula function principle, the data of Tianshengqiao Level 1 and Longtan were taken as a group to perform Copula function fitting, and the data of Longtan and Yantan were taken as a group to perform Copula function fitting, following the same steps as in step S2.1.1 to step S2.1.3. The specific results are shown in Table 2.

[0325] Table 2 Calculation results of AIC values ​​of copula functions of three hydropower stations

[0326] According to the data in the table, the t-Copula function is selected for fitting between Tianshengqiao Level 1 and Longtan, with a parameter of 0.9213 and a degree of freedom of 5.8306; the Frank Copula function is selected for fitting between Longtan and Yantan, with a parameter of 61.6272. Using the fitted Copula function, the Tianshengqiao Level 1 sample points obtained in step S2.1.8 are used as runoff input data to obtain the Longtan runoff sample points and the Yantan runoff sample points in turn. Then, the sample points of the three hydropower stations and the wind power and photovoltaic sample points are sampled by inverse transformation to obtain an uncertainty scenario set for water, wind and light that takes into account time correlation and correlation between heterogeneous resources. Finally, the initial scenario is reduced according to the scenario reduction method in step S2.2 to obtain the final water, wind and light scenario set. The specific results are as follows. Figure 6 (The picture shows the water scenery scene taking the first level of Tiansheng Bridge as an example).

[0327] according to Figure 1 Principle The obtained scene (the reduced scene is set to 5) is brought into the model constructed by the present invention for simulation testing. The test environment is as follows: operating system: Windows 10; CPU: Inter i7-8086K; memory: 32 GB; Gurobi version: Gurobi 10.0.3 Education Edition.

[0328] The simulation test results show that the constructed model is reasonable. Figure 7 It is the process diagram of the optimization calculation results of the water level process of each power station obtained by solving the solution. Figure 8 It is a process diagram of the optimization calculation results of the output process of each power station. It can be seen that the dispatch results in all scenarios are within the feasible range.

[0329] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be considered as the scope of protection of the present invention.

Claims

1. A long-term random coordinated dispatching method for a hydropower-wind power-photovoltaic complementary system, characterized in that: The steps include: Step S1, determine resource variables They are runoff inflow, wind power output and photovoltaic output respectively; obtain the attribute information of the cascade power stations in the basin, which includes the constraint boundaries of each hydropower station and the resource variables of each resource in historical years. The resource variable historical value time series, where ; Step S2, scene generation: Step S2.1: quantify each resource variable of the cascade hydropower stations in the basin based on Markov chain and Copula function The time correlation and the correlation between resource variables generate each resource variable The initial set of simulation scenarios; Step S2.2: Based on the probability distance, The initial simulation scenario set is reduced to obtain each resource variable The reduced A simulation scenario set of simulation scenarios; Therefore, for the runoff, wind power output and photovoltaic output resource variables, we obtain The simulation scenario set of each resource variable is randomly arranged to obtain the final runoff simulation scenario set. , Wind power output simulation scenario set and photovoltaic output simulation scenario set ; Set the runoff simulation scenario , Wind power output simulation scenario set and photovoltaic output simulation scenario set Combine to form a comprehensive scene set ; Among them, for the comprehensive scene set Comprehensive scene in , , with the following relationship: At the same time, the comprehensive scene Probability ; Runoff simulation scenarios , Wind power output simulation scenario And photovoltaic output simulation scene probability; Step S3, constructing a long-term random coordinated dispatch model for the basin cascade hydropower-wind power-photovoltaic complementary system; Step S4, with comprehensive scene set As input, the long-term random coordinated dispatch model of the cascade hydropower-wind power-photovoltaic complementary system in the basin is solved to obtain the hydropower station applicable to all comprehensive scenarios. In each period of the scheduling cycle Outbound flow , to achieve long-term random coordinated scheduling of water, wind and solar power.

2. A long-term random coordinated scheduling method for a hydropower-wind power-photovoltaic complementary system according to claim 1, characterized in that: The constraint boundaries of the hydropower station include the maximum transmission capacity of each transmission section, the comprehensive output coefficient of each hydropower station, the upper and lower limits of the total outflow and power generation flow of each reservoir, the upper and lower limits of the output and storage capacity of each reservoir, the maximum and minimum water level limits and initial and final water levels of each reservoir, the water level-storage capacity characteristic curve and the tail water level-outflow characteristic curve data and installed capacity.

3. The long-term random coordinated scheduling method of a hydropower-wind power-photovoltaic complementary system according to claim 1 is characterized in that: Step S2.1 is specifically as follows: Step S2.1.1, resource variables is a random variable; the resource variables are fitted using the nonparametric kernel density estimation method. The marginal distribution of Step S2.1.2, determine the central variable of the first layer of the vine structure: Calculate resource variables separately and The Kendall coefficient between , resource variables and The Kendall coefficient between , and resource variables and The Kendall coefficient between ; Compute and Resource Variables The combined coefficient of association = ; Computation and resource variables The combined coefficient of association = ; Computation and resource variables The combined coefficient of association = ; Comparison of combination coefficients , and , determine the maximum combination coefficient, the resource variable corresponding to the maximum combination coefficient is the central variable of the first layer of the vine structure; Step S2.1.3, build three-dimensional Rattan structure: The central variable of the first layer of the vine structure is expressed as , and the other two resource variables are represented as and , thus the first layer of vine structure is established as and , the second layer of vine structure is Three-dimensional Rattan structure; among them, and Respectively and The joint distribution function of the determined copula, and and The joint distribution function of the determined copula; Step S2.1.4, based on the resource variables of step S2.1.1 The marginal distribution of the three-dimensional The maximum likelihood method is used to estimate the parameters of each Coupla function, and the AIC criterion is used to determine the Coupla function of each edge. The Coupla function with the smallest AIC value is selected as the final Coupla function of each edge, thereby obtaining the Coupla function connection rattan model; Step S2.1.5, generate the central variable of the first layer of the vine structure Considering the time correlation scenario, and determining the cumulative probability of the time correlation scenario as a uniform variable ,and ; Step S2.1.6, let are three sets of solution variables to be solved, so , Treated as solution variable The sample points; Step S2.1.7, generate two random numbers that follow a uniform distribution, defined as uniform variables , ; Use the binary method to solve the formula , the solution is the variable to be solved of sample points; among them, Indicates known Under the conditions The distribution function of is the conditional distribution function; express and Coupla probability distribution function between ; Step S2.1.8, according to the formula , and we can solve ;in: Indicates known Under the conditions The distribution function of Indicates known Under the conditions The distribution function of Indicates known Under the conditions The distribution function of Indicates known Under the conditions, and Coupla probability distribution function between ; According to the formula , solve to get the variable to be solved of sample points; among them, express and Coupla probability distribution function between ; Step S2.1.9, respectively Perform inverse transform and sample multiple times to obtain The corresponding random sample set, The corresponding random sample set, The corresponding random sample set; each random sample set is an initial simulation scenario set of a resource variable that comprehensively considers the time correlation and the correlation between heterogeneous energy sources; thus, each resource variable is obtained The initial set of simulation scenarios.

4. A long-term random coordinated dispatching method for a hydropower-wind power-photovoltaic complementary system according to claim 3, characterized in that: Step S2.1.5, generate the central variable of the first layer of the vine structure Considering the time correlation scenario, and determining the cumulative probability of the time correlation scenario as a uniform variable , specifically: Step A1: Set the central variable of the first layer of the vine structure The corresponding resource variable is the resource variable , ; Determine resource variables The corresponding state space; then, apply the Markov chain to the resource variable The historical value time series is analyzed for time correlation to obtain the resource variable The cumulative probability state transfer matrix ; Determine the research time scale as a period Length ; Set the number of simulation scenarios B and resource variables that need to be initially generated The initial value of the resource variable The cumulative probability state transfer matrix , using random numbers to generate the The resource variable The simulation value time series forms a simulation scenario, that is, a scenario considering time correlation; the loop is executed B times to obtain the resource variable of A simulation scenario is formed to form the resource variable The initial set of simulation scenarios; Step A2: Analyze the cumulative probability of the initial simulation scenario set in step A1 and determine it as a uniform variable .

5. A long-term random coordinated dispatching method for a hydropower-wind power-photovoltaic complementary system according to claim 4, characterized in that: Step A1 is specifically as follows: Step A11, for resource variables , determine the corresponding state space; the state space has states, namely state 1, state 2, ..., state ; Step A12, applying the Markov chain to the resource variable The historical value time series is analyzed for time correlation to obtain the resource variable The state transition probability matrix : in: Represents a resource variable From the status To status The probability of transition between states; ; ; Step A13, based on resource variables The state transition probability matrix , get the resource variable The cumulative probability state transfer matrix : in: Represents a resource variable From the status To status The cumulative probability of state transition; ； Step A14, generate resource variables A simulation scenario: Step A141, set resource variables Initial value of , according to the initial value , get the initial state ;in, ; Step A142, set the time period =1; Step A143, generate a random number , ; In the cumulative probability state transfer matrix In the state The corresponding row vector is: Row vector of rows: ; Step A144, traverse the row vector from front to back When the element in is greater than or equal to the random number Elements hour, , the element The corresponding state after state transfer , which is the time period The state of , re-expressed as state ;state Corresponding resource variables The value range is: ; Step A145, generate a random number , ; Through the formula , get the resource variable In the period The simulated value of ;in, For resource variables In the period The simulated value of Step A146, determine the time period Whether the total number of preset scheduling periods has been reached If it is reached, execute step A147; if it is not reached, set , return to step A143; Step A147, output the resource variable obtained at this time The simulated values ​​in time period 1, time period 2, ..., time period T form a simulated value time series , forming a simulation scenario; Step A148, determine whether the current If there is no simulation scenario, return to step A142, and form another simulation scenario by executing steps A142 to A147; and repeat this cycle until a simulation scenario is obtained. A simulation scenario is formed to form the resource variable The initial set of simulation scenarios.

6. A long-term random coordinated dispatching method for a hydropower-wind power-photovoltaic complementary system according to claim 5, characterized in that: Step S2.2: Based on the probability distance, The initial simulation scenario set is reduced to obtain each resource variable The reduced The simulation scenario set of simulation scenarios is as follows: Step S2.2.1, resource variables The initial set of simulation scenarios includes simulation scenarios, expressed as: simulation scenario set ; Respectively Simulation scenarios; Step S2.2.2, set each simulation scenario The initial probability Equal, both ;in, ; Create a collection , used to store scenes that need to be eliminated; Step S2.2.3, set up the collection is empty; Step S2.2.4, traverse the simulation scene set Each simulation scenario in ; For each simulation scenario traversed , determine the distance simulation scene Recent simulation scenarios , that is, it satisfies the relationship: in: Represents a simulation scenario and simulation scenarios The distance between Represents a simulation scenario And the most recent simulation scenario The distance between Using formula , calculate the simulation scenario The corresponding minimum distance probability ; Step S2.2.5, compare simulation scenario sets Various simulation scenarios The minimum distance probability , using the formula , get the simulation scene set The minimum distance probability in The minimum value of ;Should Corresponding simulation scenario , distance simulation scene The most recent simulation scenario is ; Step S2.2.6, let the simulation scenario set = -{ }, = +{ }, and let ; Determine the simulation scenario set Is the number of simulation scenarios reduced to If yes, execute step S2.2.7; if no, return to step S2.2.4; Step S2.2.7, obtain the reduced A simulation scenario set of simulation scenarios, and each simulation scenario in the simulation scenario set has a probability.

7. A long-term random coordinated dispatching method for a hydropower-wind power-photovoltaic complementary system according to claim 1, characterized in that: The long-term random coordinated dispatch model of the basin cascade hydropower-wind power-photovoltaic complementary system is constructed, with hydropower stations as the In the period Outbound flow As decision variables of the model; The objective function includes the objective function And the objective function ; Objective Function :Maximize the total power generation expectation of the water-wind-solar complementary system in: Indicates hydropower station In the comprehensive scene Below the period contribution; ; is the total number of long-term scheduling periods; ; Indicates the number of hydropower stations; ; Represents a comprehensive scene set The number of comprehensive scenarios in the Represents a comprehensive scenario probability; Indicates the length of the long-term scheduling period; Indicates that wind turbines are in comprehensive scenarios Below the period contribution; Indicates that the photovoltaic unit is in a comprehensive scenario Below the period contribution; In the comprehensive scene Under the constraints The degree of constraint violation when insufficient; In the comprehensive scene Under the constraints the degree of constraint violation when exceeded; Representing constraints The penalty factor of ; Indicates the set of constraints that require adding penalty factor terms; Objective Function :Maximize the minimum output expectation The constraints include: (1) Water balance equation: in: , Hydropower Station In the period and time period Storage capacity; For hydropower station In the period In comprehensive scenes Runoff inflow under For hydropower station In the period of runoff inflow for all composite scenarios; For hydropower station The upstream hydropower station In the period of runoff inflow; Hydropower Station a collection of upstream hydroelectric power stations; (2) Outbound flow equation: in: , Hydropower Station In the period The power generation flow and abandoned water flow; (3) Outbound flow restrictions: in: , Hydropower Station In the period The lower and upper bounds of the total outbound flow; It represents the degree of constraint violation when constraint 1 is insufficient in all comprehensive scenarios; Indicates constraints in all comprehensive scenarios The degree of constraint violation when exceeded; constraint 1 is the outbound flow restriction constraint; (4) Power generation flow limit: in: , Hydropower Station In the period Lower and upper bounds on the generation flow; It represents the constraint violation degree when constraint 2 is insufficient in all comprehensive scenarios; Indicates constraints in all comprehensive scenarios The constraint violation degree when exceeded; Constraint 2 is the power generation flow limit constraint; (5) Output limit: in: , Hydropower Station In the period Lower and upper bounds on output; Indicated in comprehensive Scenario The constraint violation degree when the lower constraint condition 3 is insufficient; In the comprehensive scene Under the constraints The constraint violation degree when exceeded; constraint 3 is the output limit constraint; (6) Storage capacity limit: in: , Hydropower Station In the period The lower and upper limits of the reservoir capacity; In the comprehensive scene The constraint violation degree when the lower constraint condition 4 is insufficient; In the comprehensive scene Under the constraints The degree of constraint violation when exceeded; constraint 4 is the storage capacity limit constraint; (7) Water level constraints at the beginning and end of the dispatch period: in: , Hydropower Station The initial and final water levels; , Hydropower Station Water level in period 1 and period T; In the comprehensive scene The constraint violation degree when the lower constraint condition 5 is insufficient; In the comprehensive scene Under the constraints The constraint violation degree when exceeded; Constraint 5 is the water level constraint at the beginning and end of the dispatch period; (8) Water level-reservoir capacity constraints: in: It is a nonlinear function between reservoir water level and water storage; For hydropower station In the period In comprehensive scenes The water level below; (9) Tailwater level-discharge constraint: in: For hydropower station Non-linear function between tailwater level and outflow rate; For hydropower station In the period In comprehensive scenes The tailwater level below; (10) Hydropower output function: in: For hydropower station In the period In comprehensive scenes The power generation head under For hydropower station Output coefficient; For hydropower station In the period In comprehensive scenes The water level below; (11) Transmission capacity limitation constraints: in: Represents hydroelectric power station A collection of associated wind farms; Represents hydroelectric power station Related wind farms contribution; Represents hydroelectric power station A collection of associated photovoltaic power plants; Represents hydroelectric power station The output of the associated photovoltaic power station sol; Indicates hydropower station Total output of For hydropower station The maximum value of integrated water, wind and solar transmission capacity; It represents the constraint violation degree when constraint 6 is insufficient in all comprehensive scenarios; Indicates constraints in all comprehensive scenarios The degree of constraint violation when exceeded; constraint 6 is the transmission capacity limit constraint.

8. The long-term random coordinated dispatching method of a hydropower-wind power-photovoltaic complementary system according to claim 1, characterized in that: Step S4 is specifically as follows: Step S4.1, generate individual code : ;in, Meaning: Hydropower Station In the period The runoff inflow under all comprehensive scenarios is the decision variable; according to the individual coding , according to the set population size, generate the corresponding number of individuals with initialization solutions to form an initialized population; Step S4.2, determine the population size NI and ratio ; Among them, the ratio = The number of individuals in the dominant population / the number of individuals in the basic population; Step S4.3, using the inter-individual dominance algorithm to determine the dominant population and the basic population; Step S4.4, adopt the multi-recombination operator fusion strategy to perform crossover operation on the dominant population and the basic population to generate a new population; then return to step S4.3, re-adopt the inter-individual domination algorithm to divide the new population into the dominant population and the basic population, and repeat this cycle until the set number of cycles is reached, and then execute step S4.5; Step S4.5, for each individual in the new population, a heuristic constraint processing method is used to perform heuristic constraint processing on the solution value of each individual, so as to obtain each individual with an updated solution value; Step S4.6, based on the updated solution value of each individual obtained in step S4.5, calculate the objective function value of the objective function F1 and the objective function F2 corresponding to each individual; Step S4.7, based on the objective function values ​​corresponding to each individual, determine whether the iteration termination condition is met through the restart mechanism. If so, execute step S4.8; if not, determine whether a restart mechanism is needed. If so, start the restart mechanism, update the basic population, and return to step S4.4; if not, return to step S4.3; Step S4.8, output the solution value of each individual in the dominant population. The solution value of each individual is a scheduling plan, thereby obtaining a set of multi-objective optimization scheduling plans, thereby completing the long-term random coordinated scheduling of the hydropower station group.

9. A long-term random coordinated dispatching method for a hydropower-wind power-photovoltaic complementary system according to claim 8, characterized in that: Step S4.3 is specifically as follows: Step S4.3.1, setting parameters The value of Step S4.3.2, with the objective function The function value is the horizontal axis, with the objective function The function value of the vertical coordinate is For the grid side length, establish a gridded objective function space; Step S4.3.3, for each current individual X c , c=1,2,…, NI, according to its objective function The function value and objective function The function value of , determines the grid where it is located in the objective function space; Step S4.3.4, traverse each individual and calculate each individual X c The dominance number NC is calculated as: For individual X c , the initial value of its dominance number NC is 0; traverse the other NI-1 individuals in turn, and for the currently traversed individual X d , d = 1, 2, ..., NI, and d ≠ c, if individual X d and individual X c The function values ​​of X are located in the same grid in the objective function space, then Algorithm 1 is used to determine the individual X c Whether it dominates individual X d , if it dominates, then the domination number NC increases by 1; if individual X d and individual X c If the function value of is located in different grids in the objective function space, then Algorithm 2 is used to determine the individual X c Whether it dominates individual X d If it dominates, the domination number NC increases by 1; thus, the traversal of the other NI-1 individuals is completed, and the value of the domination number NC is obtained, which is the value of individual X c The domination number NC represents the domination number of individual X. c The number of individuals controlled; Among them: Algorithm 1 and Algorithm 2 are: Individual X c , individual X c The objective function The function value is ; Individual X c The objective function The function value is , individual X c The objective function value vector is expressed as: = ; Individual X d The objective function The function value is ; Individual X d The objective function The function value is , individual X d The objective function value vector is expressed as: = ; Algorithm 1 is: When conditions 1 and 2 are met, individual X c Dominant individual X d ; Condition 1: ≥ ,and ≥ ; Condition 2: Existence , x= , satisfying the relationship: ,and, ; Algorithm 2 is: if and only if one of conditions 3 and 4 is met, individual X c Dominant individual X d : Condition 3: in: represent and The relationship between conditions 1 and 2 is satisfied. represents round down; Condition 4: in: is the modulo operation; is logical AND; Step S4.3.5, based on population size NI and ratio , and get the number of individuals in the dominant population and the number of individuals in the basic population ; Sort the individuals in descending order of dominance numbers. If two individuals have the same dominance numbers, sort them randomly. Among the sorted NI individuals, select the one with the highest dominance number. A certain number of individuals form the dominant population, and the other individuals form the basic population.

10. A long-term random coordinated dispatching method for a hydropower-wind power-photovoltaic complementary system according to claim 9, characterized in that: Step S4.4 is specifically as follows: Step S4.4.1, setting a crossover operator pool; the crossover operator pool has N_CO crossover operators; Step S4.4.2, initially, each crossover operator CO is set to have the same selection probability PC; Step S4.4.3, select a crossover operator to perform a crossover operation, and update the selection probability PC of the crossover operator: ① Based on the selection probability PC of each crossover operator, select the crossover operator A from the crossover operator pool; ② According to the number of parents N required by the crossover operator A A , randomly select a parent individual X from the dominant population A1 , select N from the basic population through a tournament A -1 parent individual, expressed as: parent individual X AB ,B=2,3,…,N A ; ③Use the crossover operator A to perform crossover on the parent individual X A1 and parent individual X AB Perform crossover operation to generate N A offspring individuals, and N A The original parent individuals are deleted from the population; ④According to N A The function value of the offspring individuals is used to determine the grid in the objective function space, and the offspring individuals within the dominant population are added to the dominant population, and the other offspring individuals are added to the basic population; where: the number of offspring individuals added to the dominant population generated by the crossover operator A this time is ; ⑤ Update the selection probability PC of the crossover operator A using the following formula: A : in: Represents the total number of offspring individuals added to the dominant population generated by the crossover operator V from the first cycle to the current cycle, V=1,2,…,N_CO; is a constant, ; ⑥When this cycle ends, return to step ① and continue the next cycle until the set number of cycles is reached; In step S4.7, it is determined whether a restart mechanism is needed. If necessary, the restart mechanism is started to update the basic population. Specifically, when the algorithm search is stagnant, the restart mechanism is started: according to the current ratio and the current number of dominant population individuals , using the formula / , get the number of individuals in the basic population; clear the current basic population, and fill the corresponding number of individuals with the highest NC ranking in the current dominant population into the basic population, and the insufficient part is obtained through the mutation of individuals in the dominant population, so as to get the updated basic population.

Images