CN112819216B - Wind power sequence scene set-based generation method and system - Google Patents

Abstract

The invention provides a generating method based on a wind power sequence scene set, which comprises the following steps of 1, constructing a prediction box and numbering; 2. generating a wind power static scene set; 3. constructing a wind power state transition matrix; 4. combining the wind power state transition matrix to construct an initial solution, a neighborhood solution and an adaptability function of a tabu search algorithm; 5. and (5) iteratively generating a wind power sequence scene set. The invention further provides a generating system based on the wind power sequence scene set. The invention has the advantages that: the construction process of the initial solution of the tabu search algorithm combines the wind power state transition matrix, so that unreasonable fluctuation of the multi-period transition process of the wind power sequence is effectively filtered, the initial solution quality is improved by improving the time sequence of the initial solution, the iteration times of the tabu search algorithm are reduced, and the timeliness of the algorithm is improved; the construction process of the neighborhood solution of the tabu search algorithm combines the wind power state transition matrix, so that unreasonable fluctuation of the wind power sequence multi-period transition process is effectively filtered, and the time sequence of the sequence scene set is improved.

Description

Wind power sequence scene set-based generation method and system

Technical Field

The invention relates to wind power output uncertainty modeling, in particular to a method for generating a wind power output sequence scene set, which is used for describing various possibilities of a wind power output sequence in a plurality of hours in the future, so as to draw the uncertainty of wind power output at the moment.

Background

With the increasing exhaustion of traditional fossil energy, new energy power generation represented by wind power and photovoltaic power generation occupies an increasingly heavy position in an electric power system. The wind power generation is influenced by weather factors, geographical positions and other factors, the output has randomness, fluctuation and intermittence, the existing wind power prediction output technology only gives one condition of wind power output, and has bottlenecks in prediction accuracy, and the uncertainty of wind power is not considered enough, so that unstable factors are brought to safe and stable operation of a power system. The scene analysis technology describes various possibilities of wind power output through generating a sequence scene set of the wind power output, describes uncertainty information of the wind power output through a deterministic scene, converts an uncertainty problem into a deterministic problem to analyze, and provides data support for planning and scheduling of a power grid.

The existing wind power sequence scene set generation method can be summarized into three types: (1) The historical sequence clustering method is simple in operation, only needs to process historical wind power data, and can reasonably express the change rule of a historical output sequence, but the method is poor in predictability of wind power sequence development trend and poor in effect of describing uncertainty characteristics of future output; (2) The inverse transformation method is used for generating a wind power sequence scene set meeting strong autocorrelation by constructing a multi-element standard normal distribution sequence and inversely transforming the multi-element standard normal distribution sequence, wherein the scene set generated by the method has good effect in describing the time sequence of wind power, but the situation that a small number of scenes are excessively different from an actual output sequence can occur, the change trend of each sequence scene is approximately the same, and the depicting capability of the wind power output randomness is insufficient; (3) Firstly, generating a static scene set in a single-period analysis stage; then, a sequence scene set capable of describing wind power random characteristics is generated in a multi-period analysis stage, and the method has the advantages that the wind power output random characteristics are accurately described, but in the aspect of connection of wind power scenes in adjacent periods, the time sequence and the accuracy of the sequence scene set are difficult to consider. Therefore, the existing wind power sequence scene set generation technology still needs to be improved in the aspect of overall scene set quality.

The patent application with the publication number of CN 111934319A is provided for the team of the inventor, and discloses a generating method and a generating system based on a wind power typical scene set, wherein the method comprises the steps of 1, constructing an adaptive prediction box and numbering; 2. fitting probability distribution of prediction error data in a prediction box; 3. generating a wind power static scene set; 4. generating a wind power dynamic scene set by adopting a tabu search algorithm; 5. carrying out weight assignment on the wind power dynamic scene set; 6. and carrying out clustering reduction on the wind power dynamic scene set to obtain a wind power typical scene set. The invention adopts the self-adaptive prediction box, can describe various possibilities of wind power output, and compensates the problem of insufficient precision of the existing wind power output prediction to a certain extent, but the time sequence and accuracy of the sequence scene set and the overall quality are still to be improved.

Disclosure of Invention

The method aims to solve the technical problems of improving the time sequence and accuracy of a sequence scene set and compensating the problem of insufficient wind power output prediction accuracy to a certain extent by describing the uncertainty of wind power output, so that data support is provided for safe and stable operation of a power grid.

The invention solves the technical problems by the following technical means: a generation method based on a wind power sequence scene set comprises the following steps:

step 1, constructing a prediction box and numbering:

step 1.1, the predicted output data and the corresponding predicted error data at the same moment in the historical data form data pairs, the data pairs are arranged in ascending order according to the amplitude of the predicted output data, the data pairs are equally divided into H groups according to the amplitude interval, all the data pairs in any group form an initial prediction box, and therefore H initial prediction boxes are obtained and numbered in sequence;

step 1.2, fitting probability distribution of all prediction error data in each initial prediction box, so as to obtain H fitting results;

step 2, generating a wind power static scene set:

step 2.1, predicting future wind power with a known sampling granularity of tForce sequence e= [ E ₁ ,E ₂ ,…,E _g ,…,E _G ]；E _g The predicted wind power output at time t×g is shown, g is [1, G ]]Initializing g=1;

step 2.2, determining a self-adaptive prediction box corresponding to the t×g moment, so as to obtain a fitting result Fg of the probability distribution of the prediction error data at the t×g moment;

step 2.3, fitting the result F to the t×g time error _g Randomly sampling M times to obtain an error sample sequence U at the t×g moment _g ＝[U ¹ _g ,U ² _g ,…,U ^m _g ,…,U ^M _g ]]；U ^m _g The mth error sample at time t×g, m belonging to [1, M]；

Step 2.4, error sample sequence U at time t×g _g Each element is added with wind power predicted force E at t multiplied by g _g Thereby obtaining a static scene set P with the time scale of t multiplied by g being M _g ＝[P ¹ _g ,P ² _g ,…,P ^m _g ,…,P ^M _g ]，P ^m _g The mth wind power static scene at the t multiplied by g moment is represented;

step 2.5, after the value g+1 is assigned to G, if G is less than G+1, executing step 2.2, otherwise, representing that the wind power static scene set at all moments is generated;

step 3, counting the power output data of adjacent moments in the historical wind power output actual measurement data, and constructing a wind power state transition matrix Q;

step 4, constructing an initial solution, a neighborhood solution and a fitness function of a tabu search algorithm:

step 4.1, constructing an initial solution of a tabu search algorithm by combining the wind power state transition matrix Q;

step 4.2, constructing a neighborhood solution of a tabu search algorithm by combining the wind power state transition matrix Q;

step 4.3, constructing an adaptability function of a tabu search algorithm;

and 5, iteratively generating a wind power sequence scene set.

In the invention, the construction process of the initial solution of the tabu search algorithm is combined with the wind power state transition matrix, so that unreasonable fluctuation of the multi-period transition process of the wind power sequence is effectively filtered, the initial solution quality is improved by improving the time sequence of the initial solution, the iteration times of the tabu search algorithm are reduced, and the timeliness of the algorithm is improved;

in the invention, the construction process of the neighborhood solution of the tabu search algorithm is combined with the wind power state transition matrix, so that unreasonable fluctuation of the wind power sequence multi-period transition process is effectively filtered, and the time sequence of the sequence scene set is improved;

according to the method, the adaptability function of the tabu search algorithm gives consideration to the distance and the autocorrelation coefficient of the sequence scene set, the candidate solution with a larger function value is selected as the optimal solution through iteration, so that the dissimilarity between scenes in the optimal solution is ensured, and the effect of the timeliness in the optimizing process is reflected by introducing the autocorrelation coefficient.

Further, the step 4.1 includes:

s411, acquiring the actual wind power output amplitude z in t=0 time period ₀ Calculate the corresponding state b ₀ ，i＝1，t＝1；

S412, randomly extracting a static wind power scene of t time period to be recorded as p _t Calculating p _t Corresponding state b _t ；

S413, record b in state transition matrix Q _t-1 Line b _t The column corresponding elements areIf->Step S412, otherwise, ζ ⁱ _t ,＝p _t ，ξ ⁱ _t Turning to step S414 for the element of the ith row and the tth column;

s414, if t <96, t=t+1, go to step S412, otherwise go to step S415;

s415, if i <100, i=i+1, go to step S412, otherwise end.

Further, the step 4.2 includes:

s422, randomly determining d time periods needing to change the value, and recording as y= [ y ] ₁ ,y ₂ …y _r …y _d ](1≤r≤d,1≤y _r ≤96)，r＝1；

S423, from y _r Randomly extracting a static scene in a static scene set corresponding to a time period as a sequence scene y in the current solution _r If the value of the time interval value is not less than 1 and not more than y _r Less than or equal to 95, go to step S424, otherwise go to step S425;

s424, calculate y _r -1 period, y _r Period of time and y _r State corresponding to wind power generation force in +1 periodAnd->State transition matrix Q +.>Line->Column corresponding element is +.>State transition matrix Q +.>Line->Column corresponding element is +.>If->And->If they are not 0, go to step S426, otherwise go to step S423;

s425, calculate y _r Period-1 and y _r State corresponding to wind power output in time periodIf->If not, go to step S426, otherwise go to step S423;

s426, r=r+1, if r is less than or equal to d, go to step 423, otherwise go to step 427;

s427, judging whether the neighborhood scene is repeated with the sequence scene in the tabu list, if so, turning to step 423, otherwise, finishing the neighborhood scene construction of the sequence scene.

Further, the step 4.2 further includes:

S421，i＝1；

in S427, if the sequence is repeated, r=1, go to step S423, otherwise, the neighbor scene of the ith sequence scene in the current solution is constructed completely, go to step S428;

s428, if i <100, i=i+1, go to step S422, otherwise, the neighborhood solution is constructed.

Further, in the step 4.3, constructing an fitness function of the tabu search algorithm specifically includes:

recording any one candidate solution as lambda, lambda being a matrix of 100 x 96, each behavior being a sequence scene,

using the formula

Calculating the distance between any two sequence scenes in the candidate solution, wherein,

λ ⁱ and lambda (lambda) ^j (1 is less than or equal to i, j is less than or equal to S) is respectively the ith and jth sequence scenes in the candidate solution;

using the formula

Calculating an autocorrelation coefficient of any one sequence scene in the candidate solution, wherein,

A ^k for the autocorrelation coefficient of the kth sequence scene in the candidate solution, k is more than or equal to 1 and less than or equal to 100, c and v are covariance and variance respectively, lambda ^k _e (1.ltoreq.e.ltoreq.96) is the element of the kth row and the e column of the candidate solution lambda;

using the formula

And calculating a fitness function f and S of the candidate solution, wherein S is the scale of the sequence scene set in the candidate solution.

Further, step 5, iteratively generating a wind power sequence scene set includes the following steps:

s51, the scale of the scene set of the given wind power sequence is 100, and the termination criterion θ=10 ^-4 The number of neighborhood solutions in each iteration is 1000, the number of time periods for changing the value of the current solution is 24, and an initial solution H is constructed according to the step 4.1 ₀ Step 4.3 calculating the fitness function value f of the initial solution ₀ The tabu table is set to null and the iteration number ip=1;

s52, repeating the method of step 4.2 for 1000 times, constructing 1000 neighborhood solutions of the IP iteration, calculating the fitness function value of each neighborhood solution, and recording the maximum value as f ^IP The corresponding neighborhood solution is H ^IP ，

S53, recording the maximum fitness function value of the candidate solution as f _IP Taking f _IP ＝max{f _IP-1 ，f ^IP As the fitness function value of the current solution of the second iteration, the corresponding current neighborhood solution is marked as H _IP ；

S54, if |f _IP -f _IP-1 |/f _IP >θ, if ip=ip+1, adding sequence scenes in all candidate solutions to the tabu table, turning to step S52, otherwise outputting the optimal solution H _best ＝H _IP The iteration ends.

The invention also discloses a generating system based on the wind power sequence scene set, which comprises the following modules:

the prediction box construction module is used for constructing the prediction box and numbering, and comprises the following units:

the initial prediction box unit is used for forming each data pair by the prediction output data and the prediction error data corresponding to the prediction output data at the same moment in the historical data, dividing the data pairs into H groups according to the amplitude interval after the data pairs are arranged in an ascending order according to the amplitude of the prediction output data, and forming an initial prediction box by all the data pairs in any group, so that H initial prediction boxes are obtained and numbered in sequence;

the fitting unit is used for fitting probability distribution of all prediction error data in each initial prediction box so as to obtain H fitting results;

the wind power static scene set generation module is used for generating a wind power static scene set and comprises the following units:

a processing sequence prediction unit for predicting a force sequence E= [ E ] according to future wind power with a known sampling granularity of t ₁ ,E ₂ ,…,E _g ,…,E _G ]；E _g The predicted wind power output at time t×g is shown, g is [1, G ]]Initializing g=1;

the fitting unit is used for determining a self-adaptive prediction box corresponding to the t×g moment so as to obtain a fitting result Fg of the probability distribution of the prediction error data at the t×g moment;

sampling unit for fitting result F to time error of t×g _g Randomly sampling M times to obtain an error sample sequence U at the t×g moment _g ＝[U ¹ _g ,U ² _g ,…,U ^m _g ,…,U ^M _g ]]；U ^m _g The mth error sample at time t×g, m belonging to [1, M]；

A static scene set unit for collecting error sample sequence U at time t×g _g Each element is added with wind power predicted force E at t multiplied by g _g Thereby obtainingStatic scene set P with scale M up to t×g time _g ＝[P ¹ _g ,P ² _g ,…,P ^m _g ,…,P ^M _g ]，P ^m _g The mth wind power static scene at the t multiplied by g moment is represented;

the assignment unit is used for assigning g+1 to G, if G is less than g+1, executing step 2.2, otherwise, representing generating a wind power static scene set at all moments;

the wind power state transition matrix construction module is used for counting the power output data of adjacent moments in the historical wind power output actual measurement data to construct a wind power state transition matrix Q;

the tabu search algorithm module is used for constructing an initial solution, a neighborhood solution and a fitness function of the tabu search algorithm and comprises the following units:

the initial solution construction unit is used for constructing an initial solution of a tabu search algorithm by combining the wind power state transition matrix Q;

the neighborhood solution construction unit is used for constructing a neighborhood solution of a tabu search algorithm by combining the wind power state transition matrix Q;

the fitness function construction unit is used for constructing a fitness function of a tabu search algorithm;

and the generating wind power sequence scene set iteration module is used for iteratively generating a wind power sequence scene set.

Further, the working process of the initial solution construction unit includes:

s411, acquiring the actual wind power output amplitude z in t=0 time period ₀ Calculate the corresponding state b ₀ ，i＝1，t＝1；

S412, randomly extracting a static wind power scene of t time period to be recorded as p _t Calculating p _t Corresponding state b _t ；

S413, record b in state transition matrix Q _t-1 Line b _t The column corresponding elements areIf->Go to step S412, otherwise->ξ ⁱ _t Turning to step S414 for the element of the ith row and the tth column;

s414, if t <96, t=t+1, go to step S412, otherwise go to step S415;

s415, if i <100, i=i+1, go to step S412, otherwise end.

Further, the working process of the neighborhood solution building unit includes:

s422, randomly determining d time periods needing to change the value, and recording as y= [ y ] ₁ ,y ₂ …y _r …y _d ](1≤r≤d,1≤y _r ≤96)，r＝1；

S423, from y _r Randomly extracting a static scene in a static scene set corresponding to a time period as a sequence scene y in the current solution _r If the value of the time interval value is not less than 1 and not more than y _r Less than or equal to 95, go to step S424, otherwise go to step S425;

s424, calculate y _r -1 period, y _r Period of time and y _r State corresponding to wind power generation force in +1 periodAnd->State transition matrix Q +.>Line->Column corresponding element is +.>State transition matrix Q +.>Line->Column corresponding element is +.>If->And->If they are not 0, go to step S426, otherwise go to step S423;

s425, calculate y _r Period-1 and y _r State corresponding to wind power output in time periodIf->If not, go to step S426, otherwise go to step S423;

s426, r=r+1, if r is less than or equal to d, go to step 423, otherwise go to step 427;

s427, judging whether the neighborhood scene is repeated with the sequence scene in the tabu list, if so, turning to step 423, otherwise, finishing the neighborhood scene construction of the sequence scene.

Further, the working process of the fitness function construction unit includes:

using the formula

Calculating the distance between any two sequence scenes in the candidate solution, wherein,

λ ⁱ and lambda (lambda) ^j (1 is less than or equal to i, j is less than or equal to S) is respectively the ith and jth sequence scenes in the candidate solution;

using the formula

Calculating an autocorrelation coefficient of any one sequence scene in the candidate solution, wherein,

A ^k for the autocorrelation coefficients of the kth sequence scene in the candidate solution, c and v are covariance and variance, respectively, lambda ^k _e (1.ltoreq.e.ltoreq.96) is the element of the kth row and the e column of the candidate solution lambda;

using the formula

And calculating a fitness function f and S of the candidate solution, wherein S is the scale of the sequence scene set in the candidate solution.

The invention has the advantages that:

(1) In the invention, the construction process of the initial solution of the tabu search algorithm is combined with the wind power state transition matrix, so that unreasonable fluctuation of the multi-period transition process of the wind power sequence is effectively filtered, the initial solution quality is improved by improving the time sequence of the initial solution, the iteration times of the tabu search algorithm are reduced, and the timeliness of the algorithm is improved;

(2) In the invention, the construction process of the neighborhood solution of the tabu search algorithm is combined with the wind power state transition matrix, so that unreasonable fluctuation of the wind power sequence multi-period transition process is effectively filtered, and the time sequence of the sequence scene set is improved;

(3) According to the method, the adaptability function of the tabu search algorithm gives consideration to the distance and the autocorrelation coefficient of the sequence scene set, the candidate solution with a larger function value is selected as the optimal solution through iteration, so that the dissimilarity between scenes in the optimal solution is ensured, and the effect of the timeliness in the optimizing process is reflected by introducing the autocorrelation coefficient.

Drawings

FIG. 1 is a flow chart of a method for generating a wind power sequence scene set according to an embodiment of the invention;

fig. 2 is a flow chart of the tabu search algorithm of the present invention.

Detailed Description

For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the technical solutions in the embodiments of the present invention will be clearly and completely described in the following in conjunction with the embodiments of the present invention, and it is apparent that the described embodiments are some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.

In this embodiment, as shown in fig. 1, a method for generating a scene set based on a wind power sequence includes the following steps:

step 1, constructing a prediction box and numbering;

the wind power prediction error has better statistical characteristics than wind power prediction output, the prediction error is related to wind power prediction output amplitude, the prediction box technology can describe probability distribution conditions of the prediction error under different prediction output amplitudes, and the condition correlation of the prediction error and the wind power prediction output amplitude is established.

Step 1.1, the predicted output data at each moment in the historical data and the corresponding predicted error data form each data pair, the data pairs are arranged in ascending order according to the amplitude of the predicted output data, and then divided into 50 groups according to the amplitude interval, all the data pairs in any group form an initial prediction box, so that 50 initial prediction boxes are obtained and numbered in sequence;

step 1.2, fitting probability distribution of all prediction error data in each prediction box by adopting a ksdensity function in MATLAB, so as to obtain 50 fitting results;

the ksdensity function in MATLAB belongs to a non-parameter kernel density estimation function, and in view of different probability distribution of prediction errors in different output prediction intervals, the unified fitting effect is poor by using a certain function, and the applicability of fitting the wind power prediction errors can be improved by adopting the non-parameter kernel density estimation method.

Step 2, generating a wind power static scene set;

step 2.1, predicting a force sequence E= [ E ] according to the future wind power with the known sampling granularity of 15min ₁ ,E ₂ ,…,E _g ,…,E ₉₆ ]；E _g The predicted wind power output at 15 Xg min is shown, g is [1,96 ]]The method comprises the steps of carrying out a first treatment on the surface of the Initializing g=1; (24 hours)

Step 2.2, determining a self-adaptive prediction box corresponding to the 15 Xg min moment, so as to obtain a fitting result Fg of the probability distribution of the prediction error data at the moment;

step 2.3, fitting the result F to the 15 Xg min time error _g Performing 100 random samplings to obtain an error sample sequence U at the moment _g ＝[U ¹ _g ,U ² _g ,…,U ^m _g ,…,U ¹⁰⁰ _g ]]；U ^m _g The mth error sample representing the 15 Xg min time, m belonging to [1,100 ]]；

Step 2.4, error sample sequence U at 15 Xg min time _g Respectively adding the wind power predicted output E at the moment to each element of the wind power generator _g Thereby obtaining a static scene set P with the 15 Xg min moment scale of 100 _g ＝[P ¹ _g ,P ² _g ,…,P ^m _g ,…,P ¹⁰⁰ _g ]，P ^m _g The mth wind power static scene at the 15 Xg min moment is represented;

step 2.5, after the value of g+1 is assigned to g, if g is smaller than 101, executing step 2.2, otherwise, generating a wind power static scene set at all moments;

step 3, constructing a wind power state transition matrix;

counting the adjacent moment output data in the historical wind power output actual measurement data, dividing the data into 4 states, and constructing a state transition matrix of the wind power output at the adjacent moment as follows:

in the formula (4), q _i,j (1.ltoreq.i, j.ltoreq.4) is that the wind power is transferred from the state i to the state j after 15minProbability.

The wind power state transition matrix counts the probability of the wind power output from one state to another state after 15min, and if the probability is 0, the transition is not existed historically and is not allowed to appear in the sequence scene.

Step 4, constructing an initial solution, a neighborhood solution and an adaptability function of a tabu search algorithm;

step 4.1, constructing an initial solution of a tabu search algorithm;

matrix with initial solution of 100 x 96, xi ⁱ _t Is i (1 is less than or equal to i)<100 Line t (1 is less than or equal to t)<96 Column element).

S411, acquiring the actual wind power output amplitude z in t=0 time period ₀ Calculate the corresponding state b ₀ ，i＝1，t＝1；

S412, randomly extracting a static wind power scene of t time period to be recorded as p _t Calculating p _t Corresponding state b _t ；

S413, record b in state transition matrix Q _t-1 Line b _t The column corresponding elements areIf->Step S412, otherwise, ζ ⁱ _t ,＝p _t Step S414;

s414, if t <96, t=t+1, go to step S412, otherwise go to step S415;

s415, if i <100, i=i+1, go to step S412, otherwise end;

the initial solution is used as the current solution for the first iteration of the tabu search algorithm, and the quality of the initial solution often affects the time cost of the algorithm. Wind power fluctuation in the initial solution is limited through the state transition matrix, so that the quality of the initial solution is effectively improved, and the time of an algorithm is reduced.

Step 4.2, constructing a neighborhood solution of a tabu search algorithm;

S421，i＝1；

s422, randomly confirmThe 24 time periods requiring the change of the value are recorded as y= [ y ] ₁ ,y ₂ …y _r …y ₂₄ ](1≤r≤24,1≤y _r ≤96)，r＝1；

S423, from y _r Randomly extracting a static scene in a static scene set corresponding to a period of time as a current solution y _r If the value of the time interval value is not less than 1 and not more than y _r Less than or equal to 95, go to step S424, otherwise go to step S425;

s424, calculate y _r -1 period, y _r Period of time and y _r State corresponding to wind power generation force in +1 periodAnd->State transition matrix Q +.>Line->Column corresponding element is +.>State transition matrix Q +.>Line->Column corresponding element is +.>If->And->All are not 0, the steps are changedS426, otherwise go to step S423;

s425, calculate y _r Period-1 and y _r State corresponding to wind power output in time periodIf->If not, go to step S426, otherwise go to step S423;

s426, r=r+1, if r is less than or equal to 24, go to step 42S3, otherwise go to step S427;

s427, judging whether the neighborhood scene is repeated with the sequence scene in the tabu list, if so, r=1, turning to step S423, otherwise, finishing the neighborhood scene construction of the ith sequence scene in the current solution, and turning to step S428;

s428, if i <100, i=i+1, go to step S422, otherwise, the neighborhood solution is constructed;

the neighborhood solution and the current solution are collectively called as candidate solutions of the current iteration, the construction process of the neighborhood solution meets the fluctuation limit of wind power, the time sequence of the neighborhood solution is effectively improved, the quality of the candidate solutions is greatly improved, and the effects of improving the iteration speed and enhancing the time sequence of a sequence scene set are obvious.

Step 4.3, constructing an adaptability function of a tabu search algorithm;

any one candidate solution is recorded as lambda, lambda is a matrix of 100 x 96, and each row is a sequence scene.

Using the formula

Calculating the distance between any two sequence scenes in the candidate solution, wherein,

λ ⁱ and lambda (lambda) ^j (1 is less than or equal to i, j is less than or equal to 100) is the ith and jth sequence scenes in the candidate solution respectively;

using the formula

Calculating an autocorrelation coefficient of any one sequence scene in the candidate solution, wherein,

A ^k (1.ltoreq.k.ltoreq.100) is the autocorrelation coefficients of the kth sequence scene in the candidate solution, c, v are covariance and variance, respectively, λ ^k _e (1.ltoreq.e.ltoreq.96) is the element of the kth row and the e column of the candidate solution lambda;

using the formula

An fitness function of the candidate solution is calculated, wherein,

f is the fitness function value, S is the scale of the sequence scene set in the candidate solution;

the fitness function gives consideration to the distance of the sequence scene set and the autocorrelation coefficient, and the candidate solution with a larger function value is selected as the optimal solution through iteration, so that the dissimilarity between scenes in the optimal solution is ensured, and the effect of the timeliness in the optimizing process is reflected by introducing the autocorrelation coefficient.

Step 5, iteratively generating a wind power sequence scene set;

s51, the scale of the scene set of the given wind power sequence is 100, and the termination criterion θ=10 ^-4 The number of neighborhood solutions in each iteration is 1000, the number of time periods for changing the value of the current solution is 24, and an initial solution H is constructed according to the step 4.1 ₀ Step 4.3 calculating the fitness function value f of the initial solution ₀ The tabu table is set to null and the iteration number ip=1;

s52, repeating the method of step 4.2 for 1000 times, constructing 1000 neighborhood solutions of the IP iteration, calculating the fitness function value of each neighborhood solution, and recording the maximum value as f ^IP The corresponding neighborhood solution is H ^IP ，

S53, recording the maximum fitness function value of the candidate solution as f _IP Taking f _IP ＝max{f _IP-1 ，f ^IP As fitness function value of the current solution of the IP-th iterationThe corresponding current neighborhood solution is denoted as H _IP ；

S54, if |f _IP -f _IP-1 |/f _IP >θ, if ip=ip+1, adding sequence scenes in all candidate solutions to the tabu table, turning to step S52, otherwise outputting the optimal solution H _best ＝H _IP The iteration ends.

Compared with the meta-heuristic algorithms such as the particle swarm algorithm, the tabu search algorithm can effectively jump out the local optimal solution through the tabu table, so that the search of a larger area is realized, and a wind power sequence scene set with more comprehensive time sequence performance is obtained.

The above embodiments are only for illustrating the technical solution of the present invention, and are not limiting; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims (10)

1. The generation method based on the wind power sequence scene set is characterized by comprising the following steps of:

step 1, constructing a prediction box and numbering:

step 1.1, the predicted output data and the corresponding predicted error data at the same moment in the historical data form data pairs, the data pairs are arranged in ascending order according to the amplitude of the predicted output data, the data pairs are equally divided into H groups according to the amplitude interval, all the data pairs in any group form an initial prediction box, and therefore H initial prediction boxes are obtained and numbered in sequence;

step 1.2, fitting probability distribution of all prediction error data in each initial prediction box, so as to obtain H fitting results;

step 2, generating a wind power static scene set:

step 2.1, predicting a force sequence E= [ E ] according to future wind power with known sampling granularity t ₁ ,E ₂ ,…,E _g ,…,E _G ]；E _g The predicted wind power output at time t×g is shown, g is [1, G ]]Initializing g=1;

step 2.2, determining a self-adaptive prediction box corresponding to the t×g moment, so as to obtain a fitting result Fg of the probability distribution of the prediction error data at the t×g moment;

step 2.3, fitting the result F to the t×g time error _g Randomly sampling M times to obtain an error sample sequence U at the t×g moment _g ＝[U ¹ _g ,U ² _g ,…,U ^m _g ,…,U ^M _g ]]；U ^m _g The mth error sample at time t×g, m belonging to [1, M]；

Step 2.4, error sample sequence U at time t×g _g Each element is added with wind power predicted force E at t multiplied by g _g Thereby obtaining a static scene set P with the time scale of t multiplied by g being M _g ＝[P ¹ _g ,P ² _g ,…,P ^m _g ,…,P ^M _g ]，P ^m _g The mth wind power static scene at the t multiplied by g moment is represented;

step 2.5, after the value g+1 is assigned to G, if G is less than G+1, executing step 2.2, otherwise, representing that the wind power static scene set at all moments is generated;

step 3, counting the power output data of adjacent moments in the historical wind power output actual measurement data, and constructing a wind power state transition matrix Q;

step 4, constructing an initial solution, a neighborhood solution and a fitness function of a tabu search algorithm:

step 4.1, constructing an initial solution of a tabu search algorithm by combining the wind power state transition matrix Q;

step 4.2, constructing a neighborhood solution of a tabu search algorithm by combining the wind power state transition matrix Q;

step 4.3, constructing an adaptability function of a tabu search algorithm;

and 5, iteratively generating a wind power sequence scene set.

2. The method according to claim 1, wherein the step 4.1 comprises:

s411, acquiring the actual wind power output amplitude z in t=0 time period ₀ Calculate the corresponding state b ₀ ，i＝1，t＝1；

S412, randomly extracting a static wind power scene of t time period to be recorded as p _t Calculating p _t Corresponding state b _t ；

S413, record b in state transition matrix Q _t-1 Line b _t The column corresponding elements areIf->Step S412, otherwise, ζ ⁱ _t ,＝p _t ，ξ ⁱ _t Turning to step S414 for the element of the ith row and the tth column;

s414, if t <96, t=t+1, go to step S412, otherwise go to step S415;

s415, if i <100, i=i+1, go to step S412, otherwise end.

3. The method according to claim 1, wherein the step 4.2 comprises:

s422, randomly determining d time periods needing to change the value, and recording as y= [ y ] ₁ ,y ₂ …y _r …y _d ](1≤r≤d,1≤y _r ≤96)；

S423, from y _r Randomly extracting a static scene in a static scene set corresponding to a time period as a sequence scene y in the current solution _r If the value of the time interval value is not less than 1 and not more than y _r Less than or equal to 95, go to step S424, otherwise go to step S425;

s424, calculate y _r -1 period, y _r Period of time and y _r State corresponding to wind power generation force in +1 periodAnd->State transition matrix Q +.>Line->Column corresponding element is +.>State transition matrix Q +.>Line->Column corresponding element is +.>If->And->If they are not 0, go to step S426, otherwise go to step S423;

s425, calculate y _r Period-1 and y _r State corresponding to wind power output in time periodIf->If not, go to step S426, otherwise go to step S423;

s426, r=r+1, if r is less than or equal to d, go to step 423, otherwise go to step 427;

s427, judging whether the neighborhood scene is repeated with the sequence scene in the tabu list, if so, turning to step 423, otherwise, finishing the neighborhood scene construction of the sequence scene.

4. A method of generating as claimed in claim 3, wherein said step 4.2 further comprises:

S421，i＝1；

in S427, if the sequence is repeated, r=1, go to step S423, otherwise, the neighbor scene of the ith sequence scene in the current solution is constructed completely, go to step S428;

s428, if i <100, i=i+1, go to step S422, otherwise, the neighborhood solution is constructed.

5. The generating method according to claim 1, wherein the step 4.3 of constructing the fitness function of the tabu search algorithm specifically includes:

recording any one candidate solution as lambda, lambda being a matrix of 100 x 96, each behavior being a sequence scene,

using the formula

Calculating the distance between any two sequence scenes in the candidate solution, wherein,

λ ⁱ and lambda (lambda) ^j (1 is less than or equal to i, j is less than or equal to S) is the ith and jth sequence scenes in the candidate solution respectively;

using the formula

Calculating an autocorrelation coefficient of any one sequence scene in the candidate solution, wherein,

A ^k the k is more than or equal to 1 and less than or equal to 100, c is the autocorrelation coefficient of the kth sequence scene in the candidate solution,v is covariance and variance, respectively, lambda ^k _e E is more than or equal to 1 and less than or equal to 96, and is the element of the kth line and the e column of the candidate solution lambda;

using the formula

And calculating a fitness function f and S of the candidate solution, wherein S is the scale of the sequence scene set in the candidate solution.

6. The method of generating according to claim 1, wherein step 5 of iteratively generating a wind sequence scene set comprises the steps of:

s51, the scale of the scene set of the given wind power sequence is 100, and the termination criterion θ=10 ^-4 The number of neighborhood solutions in each iteration is 1000, the number of time periods for changing the value of the current solution is 24, and an initial solution H is constructed according to the step 4.1 ₀ Step 4.3 calculating the fitness function value f of the initial solution ₀ The tabu table is set to null and the iteration number ip=1;

s52, repeating the method of step 4.2 for 1000 times, constructing 1000 neighborhood solutions of the IP iteration, calculating the fitness function value of each neighborhood solution, and recording the maximum value as f ^IP The corresponding neighborhood solution is H ^IP ，

S53, recording the maximum fitness function value of the candidate solution as f _IP Taking f _IP ＝max{f _IP-1 ，f ^IP As the fitness function value of the current solution of the second iteration, the corresponding current neighborhood solution is marked as H _IP ；

S54, if |f _IP -f _IP-1 |/f _IP >θ, if ip=ip+1, adding sequence scenes in all candidate solutions to the tabu table, turning to step S52, otherwise outputting the optimal solution H _best ＝H _IP The iteration ends.

7. The generation system based on the wind power sequence scene set is characterized by comprising the following modules:

the prediction box construction module is used for constructing the prediction box and numbering, and comprises the following units:

the initial prediction box unit is used for forming each data pair by the prediction output data and the prediction error data corresponding to the prediction output data at the same moment in the historical data, dividing the data pairs into H groups according to the amplitude interval after the data pairs are arranged in an ascending order according to the amplitude of the prediction output data, and forming an initial prediction box by all the data pairs in any group, so that H initial prediction boxes are obtained and numbered in sequence;

the fitting unit is used for fitting probability distribution of all prediction error data in each initial prediction box so as to obtain H fitting results;

the wind power static scene set generation module is used for generating a wind power static scene set and comprises the following units:

a processing sequence prediction unit for predicting a force sequence E= [ E ] according to future wind power with a known sampling granularity of t ₁ ,E ₂ ,…,E _g ,…,E _G ]；E _g The predicted wind power output at time t×g is shown, g is [1, G ]]Initializing g=1;

the fitting unit is used for determining a self-adaptive prediction box corresponding to the t×g moment so as to obtain a fitting result Fg of the probability distribution of the prediction error data at the t×g moment;

sampling unit for fitting result F to time error of t×g _g Randomly sampling M times to obtain an error sample sequence U at the t×g moment _g ＝[U ¹ _g ,U ² _g ,…,U ^m _g ,…,U ^M _g ]]；U ^m _g The mth error sample at time t×g, m belonging to [1, M]；

A static scene set unit for collecting error sample sequence U at time t×g _g Each element is added with wind power predicted force E at t multiplied by g _g Thereby obtaining a static scene set P with the time scale of t multiplied by g being M _g ＝[P ¹ _g ,P ² _g ,…,P ^m _g ,…,P ^M _g ]，P ^m _g The mth wind power static scene at the t multiplied by g moment is represented;

the assignment unit is used for assigning g+1 to G, if G is less than g+1, executing step 2.2, otherwise, representing generating a wind power static scene set at all moments;

the wind power state transition matrix construction module is used for counting the power output data of adjacent moments in the historical wind power output actual measurement data to construct a wind power state transition matrix Q;

the tabu search algorithm module is used for constructing an initial solution, a neighborhood solution and a fitness function of the tabu search algorithm and comprises the following units:

the initial solution construction unit is used for constructing an initial solution of a tabu search algorithm by combining the wind power state transition matrix Q;

the neighborhood solution construction unit is used for constructing a neighborhood solution of a tabu search algorithm by combining the wind power state transition matrix Q;

the fitness function construction unit is used for constructing a fitness function of a tabu search algorithm;

and the generating wind power sequence scene set iteration module is used for iteratively generating a wind power sequence scene set.

8. The generating system of claim 7, wherein the initial solution building unit operation comprises:

s411, acquiring the actual wind power output amplitude z in t=0 time period ₀ Calculate the corresponding state b ₀ ，i＝1，t＝1；

S412, randomly extracting a static wind power scene of t time period to be recorded as p _t Calculating p _t Corresponding state b _t ；

S413, record b in state transition matrix Q _t-1 Line b _t The column corresponding elements areIf->Step S412, otherwise, ζ ⁱ _t ,＝p _t ，ξ ⁱ _t For the elements of row i and column t,turning to step S414;

s414, if t <96, t=t+1, go to step S412, otherwise go to step S415;

s415, if i <100, i=i+1, go to step S412, otherwise end.

9. The generation system of claim 7, wherein the neighborhood solution building element operation comprises:

s422, randomly determining d time periods needing to change the value, and recording as y= [ y ] ₁ ,y ₂ …y _r …y _d ](1≤r≤d,1≤y _r ≤96)；

S423, from y _r Randomly extracting a static scene in a static scene set corresponding to a time period as a sequence scene y in the current solution _r If the value of the time interval value is not less than 1 and not more than y _r Less than or equal to 95, go to step S424, otherwise go to step S425;

s424, calculate y _r -1 period, y _r Period of time and y _r State corresponding to wind power generation force in +1 periodAnd->State transition matrix Q +.>Line->Column corresponding element is +.>State transition matrix Q +.>Line->Column corresponding element is +.>If->And->If they are not 0, go to step S426, otherwise go to step S423;

s425, calculate y _r Period-1 and y _r State corresponding to wind power output in time periodIf->If not, go to step S426, otherwise go to step S423;

s426, r=r+1, if r is less than or equal to d, go to step 423, otherwise go to step 427;

s427, judging whether the neighborhood scene is repeated with the sequence scene in the tabu list, if so, turning to step 423, otherwise, finishing the neighborhood scene construction of the sequence scene.

10. The generating system of claim 7, wherein the fitness function building unit operation comprises:

using the formula

Calculating the distance between any two sequence scenes in the candidate solution, wherein,

λ ⁱ and lambda (lambda) ^j (1.ltoreq.i, j.ltoreq.S) are respectively within the candidate solutionsIth and jth sequence scenes;

using the formula

Calculating an autocorrelation coefficient of any one sequence scene in the candidate solution, wherein,

A ^k for the autocorrelation coefficients of the kth sequence scene in the candidate solution, c and v are covariance and variance, respectively, lambda ^k _e (1.ltoreq.e.ltoreq.96) is the element of the kth row and the e column of the candidate solution lambda;

using the formula

And calculating a fitness function f and S of the candidate solution, wherein S is the scale of the sequence scene set in the candidate solution.