CN114841547B - Renewable energy source correlation scene generation method based on Markov chain and Copula function - Google Patents

Abstract

The invention discloses a renewable energy source correlation scene generation method based on a Markov chain and a Copula function, which comprises the following steps: 1, collecting historical data sequences of a plurality of renewable energy stations; 2, obtaining a scene sequence meeting the time sequence autocorrelation characteristic by using a scene generation method based on a Markov chain model; describing the output correlation of the multi-renewable energy station by using a Copula function theory, and generating a multi-station scene sequence containing cross-correlation characteristics; 4, based on the step 2 and the step 3, generating a random scene which takes the time sequence autocorrelation of the field station and the cross correlation property between the field stations into consideration; the method can realize scene generation considering the time sequence autocorrelation of renewable energy and the cross correlation characteristic of multiple types of energy, and is used for optimizing and dispatching a power system and planning a medium-long-term power grid.

Description

Renewable energy source correlation scene generation method based on Markov chain and Copula function

Technical Field

The invention belongs to the field of renewable energy scene generation methods, and particularly relates to a renewable energy correlation scene generation method based on a Markov chain and a Copula function.

Background

Along with the continuous improvement of the access proportion of renewable energy sources in the power system, the strong randomness and intermittence of the output of the renewable energy sources represented by wind power and photovoltaic power bring great challenges to the safe and stable operation of a power grid and the medium-and-long-term planning of the power system, and how to describe the uncertainty output of the renewable energy sources is an important cut-in point for solving the problems. The scene analysis method is a method for discretizing and describing an uncertainty time sequence, a renewable energy output uncertainty model is constructed and is the basis of renewable resource scene generation technology research, and the existing model is mainly developed from a physical method, a statistical method and a learning method. And for multi-time-section scenes, a Markov chain model, an autoregressive moving average model and other time sequence construction methods are generally needed to be combined to describe the time sequence correlation characteristics of the renewable energy sources. When considering the space-time correlation of the output of a plurality of renewable energy stations, a correlation model of multidimensional random variables is required to be introduced, and the existing method mainly comprises the steps of constructing multi-element joint probability distribution and acquiring correlation information by utilizing Copula functions. The existing scene generation technology can only describe time sequence autocorrelation characteristics of renewable energy sources or cross-correlation characteristics of the output of a plurality of renewable energy sources stations, and can not simultaneously retain the autocorrelation and cross-correlation characteristics.

Disclosure of Invention

The invention aims to solve the defects of the prior art, and provides a renewable energy source correlation scene generation method based on a Markov chain and a Copula function, so as to realize scene generation which takes into consideration the time sequence autocorrelation of renewable energy sources and the cross correlation among multiple types of energy sources, thereby being capable of serving the optimal scheduling of a power system and the medium-long-term power grid planning.

In order to achieve the aim of the invention, the invention adopts the following technical scheme:

The invention relates to a renewable energy source correlation scene generation method based on a Markov chain and a Copula function, which is characterized by comprising the following steps:

Step one, enabling a historical data sequence of a first renewable energy station to be Making the historical data sequence of the second renewable energy station beWherein x _t is the historical output of the first renewable energy station in the t period, and y _t is the historical output of the second renewable energy station in the t period; t _h denotes the total sampling duration of the historical data;

Step two, based on the historical data sequences X and Y, a Markov chain scene generation method is used to obtain a first basic scene sequence X ^m and a second basic scene sequence Y ^m which meet the time sequence autocorrelation characteristic;

Step 2.1, uniformly dividing the output range of the renewable energy source into N state intervals, namely S= { S ₁,s₂,…,s_n,…,s_N }, wherein S _n represents the nth state interval, and And s _n is the upper and lower limits of the nth state interval, respectively;

step 2.2, dividing the historical data sequence X into ΛxΘ groups;

Step 2.2.1, dividing X into lambda groups according to seasonal characteristics, wherein lambda group historical data are recorded as X _λ, lambda epsilon {1,2, …, lambda };

Step 2.2.2, uniformly dividing historical data X _λ of a lambda-th group into theta-day characteristic groups according to different day characteristics, wherein any theta-th day characteristic group is marked as X _λ,θ, theta epsilon {1,2, …, theta };

Step 2.3, defining and initializing a time period tau=1 in the basic scene, and enabling the output of the t' th time period in the basic scene of the first renewable energy station to be as follows And initializeFor x ₁, letThe state interval is S (tau), S (tau) epsilon S; initializing λ=1, θ=1;

Step 2.4, calculating an accumulated transition probability matrix Q _λ,θ of a theta day characteristic group X _λ,θ in a lambda th group history number; wherein any ith row and jth column element Q _λ,θ (i, j) in the cumulative transition probability matrix Q _λ,θ represents the cumulative transition probability from state i to state j, i, j e S;

Step 2.5, randomly generating a random number epsilon obeying uniform distribution, wherein epsilon [0,1]; calculating the renewable energy output of tau+1 time period by using the formula (1) State interval s (τ+1):

Step 2.6, generating another [0,1] uniformly distributed random number epsilon', and calculating the renewable energy output in tau+1 time period by using the formula (2)

Step 2.7, updating lambda and theta:

step 2.7.1 if And mod (λ+1, Λ) +.Λ, then mod (λ+1, Λ) is assigned to λ; if it isAnd mod (λ+1, Λ) =Λ, then Λ is assigned to λ, where mod (a, b) represents the remainder operation; otherwise, sequentially executing the steps 2.7.2;

Step 2.7.2, if mod [ mod (τ, 24), 24/Θ ] =0 and mod (θ+1, Θ) +.Θ, assigning mod (θ+1, Θ) to θ; if mod [ mod (τ, 24), 24/Θ ] =0 and mod (θ+1, Θ) =Θ, then Θ is assigned to θ; otherwise, sequentially executing the step 2.8;

step 2.8, assigning τ+1 to τ, if τ > T, returning to step 2.4, otherwise, obtaining a first base scene sequence with time sequence autocorrelation characteristics Wherein T is the total duration of a Markov chain generated scene;

step 2.9, processing the historical data sequence Y according to the process from step 2.2 to step 2.8, thereby generating a second basic scene sequence with time sequence autocorrelation characteristics Wherein, Indicating the output of the t time period in Y ^m;

step three, generating a scene sequence containing variable cross-correlation characteristics by using a Copula function model based on the historical data sequences X and Y;

Step 3.1, constructing a binary group (X, Y), calculating an empirical Copula function (X, Y), selecting a theoretical Copula function with the smallest distance error by calculating the Euclidean distance between the theoretical Copula function and the empirical Copula function, and fitting the shape parameter of the theoretical Copula function with the smallest distance error to obtain a Copula function C (u, v) describing the output correlation of two renewable energy stations;

step 3.2, generating a binary random number sequence (U, V) with a value of [0,1] which has consistent correlation with a binary group (X, Y) by using a Copula function C (U, V);

step 3.3, performing inverse probability integral transformation on the binary random number sequence (U, V) with the value of [0,1] to obtain a first scene search sequence And a second scene search sequenceWherein, AndThe output of the kth period in the first scene search sequence and the output of the kth period in the second scene search sequence are respectively, K is the total duration of a Copula function generating scene, and K is more than T;

step four, obtaining a renewable energy station scene generation sequence (X ', Y') based on the sequences (X ^m,Y^m) and (X ^c,Y^c);

Step 4.1, taking a basic scene sequence (X ^m,Y^m) as a basic sequence set, taking a scene search sequence (X ^c,Y^c) as a search sequence set, and initializing a period t' =1; initializing a cycle counter Assign sequence (X ^c,Y^c) to the firstSub-cyclic scene search sequence

Step 4.2, establishing an equal step length correlation scene search optimization model:

Step 4.2.1, establishing an objective function of the equal step-length correlation scene search optimization model by utilizing the step (3)

In the formula (3), t' is the current firstA starting period of scene generation in the sub-cycle, t' +Γ -1 is a termination period, Γ is a step length of scene searching; Is the first The first renewable energy scene in the sub-cycle generates the output of the ith period in sequence X',Is the firstThe output of the ith period in the second renewable energy station scene generation sequence Y' in the secondary cycle; And The output average values of the basic scene sequences X ^m and Y ^m are respectively; Pearson correlation coefficients for sequences X ^m and Y ^m;

Step 4.2.2, establishing constraint conditions of an equal step correlation scene search optimization model by using the formula (4) and the formula (5):

in the formulas (4) and (5), Representation ofThe value of (2) isThe (k _i) th element of the (c) tree,Representation ofThe value of (2) isThe k _i th element, k _i and k _j represent the kSearching decision variables of an optimization model by using a sub-loop medium step size correlation scene, wherein i, j is epsilon { t ', t ' +1, …, t ' +Γ -1};

Step 4.3, solving the equal step length correlation scene search optimization model to obtain the first step First scene generation sequence of sub-loopAnd (d)Second scene generation sequence of sub-loopWherein, Represent the firstThe magnitude of the output of the first renewable energy farm of the secondary cycle during period t',Represent the firstThe output of the second renewable energy station of the secondary cycle in the period t';

Step 4.4, delete the first In the first scene generation sequence of the sub-loopCorresponding toElements of (a) And will be deletedPost sequenceAssignment toDelete the firstIn the second scene generation sequence of the sub-loopCorresponding toMiddle element And will be deletedPost sequenceAssignment to

Step 4.5, if T '+Γ -1 is less than or equal to T', then T '+Γ is assigned to T',Assignment toAnd returning to the step 4.2 for sequential execution, otherwise, the sequence (X ', Y ') of renewable energy generation scenes taking the dual correlation characteristics into consideration is obtained, wherein the first renewable energy scene generation sequence X ' = { X ' (1), X ' (2), …, X ' (T ') …, X ' (T ') } and the second renewable energy scene generation sequence Y ' = { Y ' (1), Y ' (2), …, Y ' (T '), …, Y ' (T ') } are represented, X ' (T ') represents the output of the first renewable energy scene in the period T ', Y ' (T ') represents the output of the second renewable energy scene in the period T ', and T ' represents the total duration of the generation scenes (X ', Y ').

Compared with the prior art, the invention has the beneficial effects that:

1. The invention combines the advantages of the Markov chain method and the Copula function model in the aspect of correlation description, avoids the unilaterality of the traditional single method applied to scene generation, and takes account of the scene generation of the time sequence autocorrelation of renewable energy and the cross correlation among multiple types of energy sources.

2. The invention provides an equal step length correlation scene search optimization model, which can realize the periodical generation of a renewable energy output scene by solving the model, and reserves the time sequence autocorrelation characteristic of renewable energy historical data and the cross correlation characteristic among a plurality of renewable energy stations.

3. According to the method, the seasonal characteristic and the intra-day characteristic of the renewable energy are considered, the historical data are grouped based on the seasonal characteristic and the intra-day characteristic, the change rule of the wind speed is counted independently, a corresponding Markov transition probability matrix is established, and the accuracy of scene generation is improved.

Drawings

FIG. 1 is a flow chart of a basic scene sequence calculation based on a Markov chain method in the invention;

FIG. 2 is a flow chart of a method for generating renewable energy source correlation scene based on Markov chain and Copula function.

Detailed Description

In this embodiment, a method for generating a renewable energy source correlation scene based on a markov chain and a Copula function is performed according to the following steps:

Step one, enabling a historical data sequence of a first renewable energy station to be Making the historical data sequence of the second renewable energy station beWherein x _t is the historical output of the first renewable energy station in the t period, and y _t is the historical output of the second renewable energy station in the t period; t _h denotes the total sampling duration of the historical data;

Step two, based on the historical data sequences X and Y, a Markov chain scene generation method is used to obtain a first basic scene sequence X ^m and a second basic scene sequence Y ^m which meet the time sequence autocorrelation characteristic, as shown in figure 1;

Step 2.1, uniformly dividing the output range of the renewable energy source into N state intervals, namely S= { S ₁,s₂,…,s_n,…,s_N }, wherein S _n represents the nth state interval, and And s _n are the upper and lower limits of the nth state interval, respectively;

step 2.2, dividing the historical data sequence X into ΛxΘ groups;

Step 2.2.1, dividing X into lambda groups according to seasonal characteristics, wherein lambda group historical data are recorded as X _λ, lambda epsilon {1,2, … and lambda }, lambda is a factor of 12, and generally 4,6 or 12 are taken;

step 2.2.2, uniformly dividing historical data X _λ of a lambda-th group into theta-day characteristic groups according to different day characteristics, wherein any theta-th day characteristic group is marked as X _λ,θ, theta epsilon {1,2, …, theta } is a factor of 24, and generally 3,4 or 6 is taken, so that inapplicability of a Markov chain method caused by too small value of grouping theta or reduction of day characteristic modeling accuracy caused by too large value of theta is avoided;

Step 2.3, defining and initializing a time period tau=1 in the basic scene, and enabling the output of the t' th time period in the basic scene of the first renewable energy station to be as follows And initializeFor x ₁, letThe state interval is S (tau), S (tau) epsilon S, wherein x ₁ is a given value, and can be any value in the upper limit and the lower limit of the output of the renewable energy source; initializing λ=1, θ=1;

Step 2.4, calculating an accumulated transition probability matrix Q _λ,θ of a theta day characteristic group X _λ,θ in a lambda th group history number; wherein any ith row and jth column element Q _λ,θ (i, j) in the cumulative transition probability matrix Q _λ,θ represents the cumulative transition probability from state i to state j, i, j e S;

Step 2.5, randomly generating a random number epsilon obeying uniform distribution, wherein epsilon [0,1]; calculating the renewable energy output of tau+1 time period by using the formula (1) State interval s (τ+1):

Step 2.6, generating another [0,1] uniformly distributed random number epsilon', and calculating the renewable energy output in tau+1 time period by using the formula (2)

Step 2.7, updating lambda and theta:

step 2.7.1 if And mod (λ+1, Λ) +.Λ, then mod (λ+1, Λ) is assigned to λ; if it isAnd mod (λ+1, Λ) =Λ, then Λ is assigned to λ, where mod (a, b) represents the remainder operation; otherwise, sequentially executing the steps 2.7.2;

Step 2.7.2, if mod [ mod (τ, 24), 24/Θ ] =0 and mod (θ+1, Θ) +.Θ, assigning mod (θ+1, Θ) to θ; if mod [ mod (τ, 24), 24/Θ ] =0 and mod (θ+1, Θ) =Θ, then Θ is assigned to θ; otherwise, sequentially executing the step 2.8;

step 2.8, assigning τ+1 to τ, if τ > T, returning to step 2.4, otherwise, obtaining a first base scene sequence with time sequence autocorrelation characteristics Wherein T is the total duration of a Markov chain generated scene;

step 2.9, processing the historical data sequence Y according to the process from step 2.2 to step 2.8, thereby generating a second basic scene sequence with time sequence autocorrelation characteristics Wherein, Indicating the output of the t time period in Y ^m;

step three, generating a scene sequence containing variable cross-correlation characteristics by using a Copula function model based on the historical data sequences X and Y;

Step 3.1, constructing a binary group (X, Y), calculating an empirical Copula function (X, Y), selecting a theoretical Copula function with the smallest distance error by calculating the Euclidean distance between the theoretical Copula function and the empirical Copula function, and fitting the shape parameter of the theoretical Copula function with the smallest distance error to obtain a Copula function C (u, v) describing the output correlation of two renewable energy stations;

step 3.2, generating a binary random number sequence (U, V) with a value of [0,1] which has consistent correlation with a binary group (X, Y) by using a Copula function C (U, V);

step 3.3, performing inverse probability integral transformation on the binary random number sequence (U, V) with the value of [0,1] to obtain a first scene search sequence And a second scene search sequenceWherein, AndThe output of the kth period in the first scene search sequence and the output of the kth period in the second scene search sequence are respectively, K is the total duration of a Copula function generating scene, and K is more than T;

Step four, obtaining a renewable energy station scene generation sequence (X ', Y') based on the sequences (X ^m,Y^m) and (X ^c,Y^c), as shown in figure 2;

Step 4.1, taking a basic scene sequence (X ^m,Y^m) as a basic sequence set, taking a scene search sequence (X ^c,Y^c) as a search sequence set, and initializing a period t' =1; initializing a cycle counter Assign sequence (X ^c,Y^c) to the firstSub-cyclic scene search sequence

Step 4.2, establishing an equal step length correlation scene search optimization model:

Step 4.2.1, establishing an objective function of the equal step-length correlation scene search optimization model by utilizing the step (3)

In the formula (3), the objective functionMinimizing the difference between the correlation index representing the sequence of the generated scene and the correlation index of the sequence (X ^m,Y^m), t' being the current firstA starting period of scene generation in the sub-cycle, t' +Γ -1 is a termination period, Γ is a step length of scene searching; Is the first The first renewable energy scene in the sub-cycle generates the output of the ith period in sequence X',Is the firstThe output of the ith period in the second renewable energy station scene generation sequence Y' in the secondary cycle; And The output average values of the basic scene sequences X ^m and Y ^m are respectively; Pearson correlation coefficients for sequences X ^m and Y ^m;

Step 4.2.2, establishing constraint conditions of an equal step correlation scene search optimization model by using the formula (4) and the formula (5):

in the formulas (4) and (5), The value of x' (i) isThe (k _i) th element of the (c) tree,The value of y' (i) isThe k _i th element, k _i and k _j represent the kThe sub-loop mid-step correlation scene searches for decision variables of the optimization model and i, j e { t ', t ' +1, …, t ' +Γ -1}, equation (5) represents the thAny two decision variables k _i in the secondary cycle are different in value;

step 4.3, solving an equal step length correlation scene search optimization model to obtain the first step First scene generation sequence of sub-loopAnd (d)Second scene generation sequence of sub-loopWherein, Represent the firstThe magnitude of the output of the first renewable energy farm of the secondary cycle during period t',Represent the firstThe output of the second renewable energy station of the secondary cycle in the period t';

Step 4.4, delete the first In the first scene generation sequence of the sub-loopCorresponding toElements of (a) Will deletePost sequenceAssignment toDelete the firstIn the second scene generation sequence of the sub-loopCorresponding toMiddle element Will deletePost sequenceAssignment to

Step 4.5, if T '+Γ -1 is less than or equal to T', then T '+Γ is assigned to T',Assignment toAnd returning to the step 4.2 for sequential execution, otherwise, the sequence (X ', Y ') of renewable energy generation scenes taking the dual correlation characteristics into consideration is obtained, wherein the first renewable energy scene generation sequence X ' = { X ' (1), X ' (2), …, X ' (T ') …, X ' (T ') } and the second renewable energy scene generation sequence Y ' = { Y ' (1), Y ' (2), …, Y ' (T '), …, Y ' (T ') } are represented, X ' (T ') represents the output of the first renewable energy scene in the period T ', Y ' (T ') represents the output of the second renewable energy scene in the period T ', and T ' represents the total duration of the generation scenes (X ', Y ').

Claims (1)

1. A renewable energy source correlation scene generation method based on a Markov chain and a Copula function is characterized by comprising the following steps:

Step one, enabling a historical data sequence of a first renewable energy station to be Making the historical data sequence of the second renewable energy station beWherein x _t is the historical output of the first renewable energy station in the t period, and y _t is the historical output of the second renewable energy station in the t period; t _h denotes the total sampling duration of the historical data;

Step two, based on the historical data sequences X and Y, a Markov chain scene generation method is used to obtain a first basic scene sequence X ^m and a second basic scene sequence Y ^m which meet the time sequence autocorrelation characteristic;

Step 2.1, uniformly dividing the output range of the renewable energy source into N state intervals, namely S= { S ₁,s₂,…,s_n,…,s_N }, wherein S _n represents the nth state interval, and And s _n is the upper and lower limits of the nth state interval, respectively;

step 2.2, dividing the historical data sequence X into ΛxΘ groups;

Step 2.2.1, dividing X into lambda groups according to seasonal characteristics, wherein lambda group historical data are recorded as X _λ, lambda epsilon {1,2, …, lambda };

Step 2.2.2, uniformly dividing historical data X _λ of a lambda-th group into theta-day characteristic groups according to different day characteristics, wherein any theta-th day characteristic group is marked as X _λ,θ, theta epsilon {1,2, …, theta };

Step 2.3, defining and initializing a time period tau=1 in the basic scene, and enabling the output of the t' th time period in the basic scene of the first renewable energy station to be as follows And initializeFor x ₁, letThe state interval is S (tau), S (tau) epsilon S; initializing λ=1, θ=1;

Step 2.4, calculating an accumulated transition probability matrix Q _λ,θ of a theta day characteristic group X _λ,θ in a lambda th group history number; wherein any ith row and jth column element Q _λ,θ (i, j) in the cumulative transition probability matrix Q _λ,θ represents the cumulative transition probability from state i to state j, i, j e S;

Step 2.5, randomly generating a random number epsilon obeying uniform distribution, wherein epsilon [0,1]; calculating the renewable energy output of tau+1 time period by using the formula (1) State interval s (τ+1):

Step 2.6, generating another [0,1] uniformly distributed random number epsilon', and calculating the renewable energy output in tau+1 time period by using the formula (2)

Step 2.7, updating lambda and theta:

step 2.7.1 if And mod (λ+1, Λ) +.Λ, then mod (λ+1, Λ) is assigned to λ; if it isAnd mod (λ+1, Λ) =Λ, then Λ is assigned to λ, where mod (a, b) represents the remainder operation; otherwise, sequentially executing the steps 2.7.2;

Step 2.7.2, if mod [ mod (τ, 24), 24/Θ ] =0 and mod (θ+1, Θ) +.Θ, assigning mod (θ+1, Θ) to θ; if mod [ mod (τ, 24), 24/Θ ] =0 and mod (θ+1, Θ) =Θ, then Θ is assigned to θ; otherwise, sequentially executing the step 2.8;

step 2.8, assigning τ+1 to τ, if τ > T, returning to step 2.4, otherwise, obtaining a first base scene sequence with time sequence autocorrelation characteristics Wherein T is the total duration of a Markov chain generated scene;

step 2.9, processing the historical data sequence Y according to the process from step 2.2 to step 2.8, thereby generating a second basic scene sequence with time sequence autocorrelation characteristics Wherein, Indicating the output of the t time period in Y ^m;

step three, generating a scene sequence containing variable cross-correlation characteristics by using a Copula function model based on the historical data sequences X and Y;

Step 3.1, constructing a binary group (X, Y), calculating an empirical Copula function (X, Y), selecting a theoretical Copula function with the smallest distance error by calculating the Euclidean distance between the theoretical Copula function and the empirical Copula function, and fitting the shape parameter of the theoretical Copula function with the smallest distance error to obtain a Copula function C (u, v) describing the output correlation of two renewable energy stations;

step 3.2, generating a binary random number sequence (U, V) with a value of [0,1] which has consistent correlation with a binary group (X, Y) by using a Copula function C (U, V);

step 3.3, performing inverse probability integral transformation on the binary random number sequence (U, V) with the value of [0,1] to obtain a first scene search sequence And a second scene search sequenceWherein, AndThe output of the kth period in the first scene search sequence and the output of the kth period in the second scene search sequence are respectively, K is the total duration of a Copula function generating scene, and K is more than T;

step four, obtaining a renewable energy station scene generation sequence (X ', Y') based on the sequences (X ^m,Y^m) and (X ^c,Y^c);

Step 4.1, taking a basic scene sequence (X ^m,Y^m) as a basic sequence set, taking a scene search sequence (X ^c,Y^c) as a search sequence set, and initializing a period t' =1; initializing a cycle counter Assign sequence (X ^c,Y^c) to the firstSub-cyclic scene search sequence

Step 4.2, establishing an equal step length correlation scene search optimization model:

Step 4.2.1, establishing an objective function of the equal step-length correlation scene search optimization model by utilizing the step (3)

In the formula (3), t' is the current firstA starting period of scene generation in the sub-cycle, t' +Γ -1 is a termination period, Γ is a step length of scene searching; Is the first The first renewable energy scene in the sub-cycle generates the output of the ith period in sequence X',Is the firstThe output of the ith period in the second renewable energy station scene generation sequence Y' in the secondary cycle; And The output average values of the basic scene sequences X ^m and Y ^m are respectively; Pearson correlation coefficients for sequences X ^m and Y ^m;

Step 4.2.2, establishing constraint conditions of an equal step correlation scene search optimization model by using the formula (4) and the formula (5):

in the formulas (4) and (5), Representation ofThe value of (2) isThe (k _i) th element of the (c) tree,Representation ofThe value of (2) isThe k _i th element, k _i and k _j represent the kSearching decision variables of an optimization model by using a sub-loop medium step size correlation scene, wherein i, j is epsilon { t ', t ' +1, …, t ' +Γ -1};

Step 4.3, solving the equal step length correlation scene search optimization model to obtain the first step First scene generation sequence of sub-loopAnd (d)Second scene generation sequence of sub-loopWherein, Represent the firstThe magnitude of the output of the first renewable energy farm of the secondary cycle during period t',Represent the firstThe output of the second renewable energy station of the secondary cycle in the period t';

Step 4.4, delete the first In the first scene generation sequence of the sub-loopCorresponding toElements of (a) And will be deletedPost sequenceAssignment toDelete the firstIn the second scene generation sequence of the sub-loopCorresponding toMiddle elementAnd will be deletedPost sequenceAssignment to

Step 4.5, if T '+Γ -1 is less than or equal to T', then T '+Γ is assigned to T',Assignment toAnd returning to the step 4.2 for sequential execution, otherwise, the sequence (X ', Y ') of renewable energy generation scenes taking the dual correlation characteristics into consideration is obtained, wherein the first renewable energy scene generation sequence X ' = { X ' (1), X ' (2), …, X ' (T ') …, X ' (T ') } and the second renewable energy scene generation sequence Y ' = { Y ' (1), Y ' (2), …, Y ' (T '), …, Y ' (T ') } are represented, X ' (T ') represents the output of the first renewable energy scene in the period T ', Y ' (T ') represents the output of the second renewable energy scene in the period T ', and T ' represents the total duration of the generation scenes (X ', Y ').