CN113807019A - MCMC wind power simulation method based on improved scene classification and coarse grain removal - Google Patents
MCMC wind power simulation method based on improved scene classification and coarse grain removal Download PDFInfo
- Publication number
- CN113807019A CN113807019A CN202111122162.9A CN202111122162A CN113807019A CN 113807019 A CN113807019 A CN 113807019A CN 202111122162 A CN202111122162 A CN 202111122162A CN 113807019 A CN113807019 A CN 113807019A
- Authority
- CN
- China
- Prior art keywords
- wind power
- output
- algorithm
- improved
- probability density
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 103
- 238000004088 simulation Methods 0.000 title claims abstract description 95
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 146
- 238000005315 distribution function Methods 0.000 claims abstract description 41
- 241000039077 Copula Species 0.000 claims abstract description 35
- 230000008569 process Effects 0.000 claims abstract description 22
- 230000002068 genetic effect Effects 0.000 claims description 37
- 238000000137 annealing Methods 0.000 claims description 29
- 238000009826 distribution Methods 0.000 claims description 24
- 230000000694 effects Effects 0.000 claims description 17
- 230000035772 mutation Effects 0.000 claims description 17
- 238000011156 evaluation Methods 0.000 claims description 13
- 239000011159 matrix material Substances 0.000 claims description 12
- 241000894007 species Species 0.000 claims description 6
- 238000007476 Maximum Likelihood Methods 0.000 claims description 4
- 230000002776 aggregation Effects 0.000 claims description 3
- 238000004220 aggregation Methods 0.000 claims description 3
- 238000004364 calculation method Methods 0.000 claims description 3
- 230000000977 initiatory effect Effects 0.000 claims description 3
- 238000005457 optimization Methods 0.000 claims description 3
- 230000000717 retained effect Effects 0.000 claims description 3
- 238000012546 transfer Methods 0.000 claims description 3
- 230000006872 improvement Effects 0.000 claims description 2
- 230000006870 function Effects 0.000 description 35
- 238000010248 power generation Methods 0.000 description 14
- 230000007704 transition Effects 0.000 description 7
- 238000010586 diagram Methods 0.000 description 6
- 238000005070 sampling Methods 0.000 description 6
- 230000008901 benefit Effects 0.000 description 3
- 230000001186 cumulative effect Effects 0.000 description 3
- 238000003860 storage Methods 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000002922 simulated annealing Methods 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 210000000349 chromosome Anatomy 0.000 description 1
- 238000005314 correlation function Methods 0.000 description 1
- 238000013499 data model Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000003912 environmental pollution Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 230000002441 reversible effect Effects 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 230000001932 seasonal effect Effects 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 238000002948 stochastic simulation Methods 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/12—Computing arrangements based on biological models using genetic models
- G06N3/126—Evolutionary algorithms, e.g. genetic algorithms or genetic programming
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/06—Multi-objective optimisation, e.g. Pareto optimisation using simulated annealing [SA], ant colony algorithms or genetic algorithms [GA]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2113/00—Details relating to the application field
- G06F2113/06—Wind turbines or wind farms
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- General Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Biophysics (AREA)
- Software Systems (AREA)
- Molecular Biology (AREA)
- Mathematical Physics (AREA)
- Biomedical Technology (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Physiology (AREA)
- Computing Systems (AREA)
- Genetics & Genomics (AREA)
- Probability & Statistics with Applications (AREA)
- Medical Informatics (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Wind Motors (AREA)
Abstract
The application provides an MCMC wind power simulation method based on improved scene classification and coarse grain removal, which comprises the following steps: step S100: clustering the historical output data of each day of wind power by using an improved KM clustering algorithm, and dividing the wind power output day into different typical output scenes; step S200: establishing a Markov chain at adjacent moments in a day and a wind power output joint probability density distribution function model at adjacent moments for each classified scene, wherein the wind power output joint probability density distribution function model at adjacent moments is obtained by fitting through a Copula mixed model; step S300: based on the established Markov chain and the wind power output joint probability density distribution function model at the adjacent moment, the existing MCMC flow is improved, and the wind power time sequence output is simulated. According to the method, the historical wind power output data are fully utilized, the wind power output process of a single wind power plant can be effectively simulated, the simulation accuracy is improved, and the method can be applied to the field of wind power dispatching planning.
Description
Technical Field
The application relates to the technical field of new energy scheduling operation of an electric power system, in particular to an MCMC wind power simulation method and device based on improved scene classification and coarse grain removal.
Background
In recent years, with the increasing severity of energy crisis and environmental pollution problems, clean energy power generation such as wind power generation and solar power generation has been vigorously developed. However, although wind energy and solar energy resources are quite abundant in a plurality of provinces in China, the wind energy and solar energy resources have the characteristics of large fluctuation in time distribution, obvious seasonal characteristics and unbalanced distribution in space. Therefore, grid-connected power generation of the power supply has uncontrollable property and can impact safe and stable operation of a power grid. When new energy is connected to the grid on a large scale, the safety, stability and economy of the operation of a power system are seriously affected, and the problems of wind and light abandonment of a power grid are caused. Therefore, the acceptance of the power grid to new energy needs to be evaluated in advance, and the operation scheduling of the power system is planned. The time series simulation of the wind power output can provide operation data which are consistent with the actual operation state, and reference is provided for the scheduling planning work of the power grid. The higher the accuracy, the greater the contribution to the job from the simulation that is more matched to the actual force process.
To realize the simulation of the wind power output, a wind power output model is required to be established. Based on historical wind power output data, the output sequence of the wind power plant can be classified and modeled. The division of the output sequence of the wind power plant mainly comprises a K-means method, an AP clustering method, a fuzzy clustering method and a neural network method. Compared with a K-means method, the AP clustering algorithm does not need to set parameters such as the number of clusters, the clustering center and the like in advance, the fuzzy clustering and neural network clustering method optimizes the clustering precision and the clustering speed, but the application range is small, the algorithm is complex, the K-means algorithm is the most classical algorithm and is one of basic clustering algorithms which are most widely applied, and the problems of poor searching and convergence capabilities, poor clustering precision and the like exist. The modeling of wind power output mainly comprises two methods: indirect modeling based on wind speed and direct modeling based on power. The indirect modeling method based on the wind speed mainly utilizes a wind speed model to generate a wind speed sequence, and then combines actual parameters of a generator and various factors to establish a wind speed-power functional relation, thereby indirectly obtaining a power sequence. However, the function mapping relationship between wind speed and power is complex and is influenced by various factors, and the obtained power hardly meets the actual power distribution characteristics. The direct modeling method utilizes historical wind power output data and utilizes a statistical principle to comprehensively consider output characteristics, and has high reliability. Common methods for constructing a wind power output Model include an Autoregressive Moving Average Model (ARMA), a Copula correlation Model, and a Monte Carlo (MC) stochastic simulation Model. The ARMA algorithm is used for modeling wind power output, parameters of a model need to be estimated, and when the regularity of historical data is poor, the obtained result error is large. The traditional Markov-Monte Carlo (MCMC) algorithm combined with the Markov Chain limits the output interval to a limited number of output states, and using purely random numbers results in coarse grain simulation. Coarse graining has a large influence on the wind power simulation effect, resulting in a large probability distribution deviation.
Disclosure of Invention
The present application is directed to solving, at least to some extent, one of the technical problems in the related art.
Therefore, the first purpose of the application is to provide an MCMC wind power simulation method based on improved scene classification and coarse grain removal, the technical problems of poor search capability and low simulation efficiency of the existing method are solved, typical daily clustering is achieved by improving a KM algorithm, a Copula mixed model is adopted to establish wind power output joint probability density at adjacent moments, and coarse grain influence brought by a classification interval is reduced. The wind power output historical data are fully utilized, the wind power output data are classified, the wind power time sequence output process is quickly and accurately simulated under the condition that other data such as meteorological factors, wind motor parameters and the like do not exist, the output characteristics of the wind power historical data are fully explored, an effective wind power output model is established through effective classification and modeling, and support can be provided for operation scheduling planning of wind power.
The second purpose of the present application is to provide an MCMC wind power simulation apparatus based on improved scene classification and coarse grain removal.
In order to achieve the above object, an embodiment of the first aspect of the present application provides an MCMC wind power simulation method based on improved scene classification and coarse grain removal, including: step S100: clustering the historical output data of each day of wind power by using an improved KM clustering algorithm, and dividing the wind power output day into different typical output scenes; step S200: establishing a Markov chain at adjacent moments in a day and a wind power output joint probability density distribution function model at adjacent moments for each classified scene, wherein the wind power output joint probability density distribution function model at adjacent moments is obtained by fitting through a Copula mixed model; step S300: based on the established Markov chain and the wind power output joint probability density distribution function model at the adjacent moment, the existing MCMC flow is improved, and the wind power time sequence output is simulated.
Optionally, in an embodiment of the present application, the improved KM clustering algorithm includes the following steps:
step S110: designing a clustering effect evaluation index, and determining the number of clustering populations according to the size of the clustering effect evaluation index;
step S120: and improving the KM algorithm by using a genetic algorithm and an annealing algorithm, and selecting, crossing and mutating the population.
Optionally, in an embodiment of the present application, the cluster effect evaluation index is expressed as:
wherein D isinIs the average value of the distances within the species group, DoutIs the minimum out-of-class spacing of the population, C is the tuning constant, d (x)i,vj) Is a sample point xiAnd a clustering center vjD (v) ofi,vj) Is the center of the cluster viAnd vjThe distance of (c).
Optionally, in an embodiment of the present application, the KM algorithm is modified by a genetic algorithm and an annealing algorithm, and the selection, crossing and mutation of the population are performed, including the following steps:
step S121: determining the scale of a genetic population, a genetic algebra, an annealing initial temperature and an annealing speed, and normalizing the wind power historical data;
step S122: each population individual completes one-time complete KM clustering operation, calculates individual fitness, performs selection operation according to the fitness, and the eliminated individuals are selectively retained by using a wheel disc selection algorithm;
step S123: carrying out cross and mutation operations on the selected new individuals;
step S124: recalculating the fitness value of the individual, and selecting the population individual according to the fitness;
step S125: the steps S122, S123 and S124 are repeated until the maximum number of generations of inheritance is reached.
Optionally, in an embodiment of the present application, the individual fitness value is calculated using an individual fitness function, which is expressed as:
wherein k is the number of species clusters, n is the total number of samples, viIs the cluster center of the ith population,
is the average of all the samples and is,
is the average of the i-th sample, Tr(SB) And Tr (S)w) Respectively used for measuring the difference degree between classes and the aggregation degree of samples in the classes.
Optionally, in an embodiment of the present application, the wheel disc selection algorithm is improved by an annealing algorithm, and the probability of the wheel disc selection algorithm accepting rejected individuals is:
wherein T is the current annealing temperature, a is the annealing speed, T0As annealing initiation temperature, fbestThe fitness value of the best population individual is K which is a constant, K is the global optimization times, namely the genetic algebra, and p is the reselected probability of the eliminated individual.
Optionally, in an embodiment of the present application, the probability of occurrence of performing the crossover and mutation operations is:
wherein p iscTo cross probability, pmAs the mutation probability, pcmaxAnd pcminRespectively preset maximum and minimum cross probability, pmmaxAnd pmminRespectively preset maximum and minimum mutation probabilities, fmeanIs a population mean fitness value, f1And f2Fitness values of two intersecting individuals, respectively, fmaxIs the maximum fitness value of the population.
Optionally, in an embodiment of the present application, for each classified scene, the establishing of the wind power output joint probability density distribution function model at the adjacent time includes the following steps:
step S210: according to the historical wind power output data of the inspected classification scene, a distribution function of the wind power output under the classification scene and a probability density distribution histogram between adjacent moments are obtained through statistical calculation;
step S220: selecting a mixed Copula function model according to the shape of the probability density distribution histogram between adjacent moments for modeling to generate a mixed Copula function model;
step S230: and estimating parameters of the hybrid Copula function model by using a maximum likelihood estimation method according to the observation data to obtain a wind power output joint probability density distribution function model at the adjacent moment.
Optionally, in an embodiment of the present application, based on the established markov chain and the wind power output joint probability density model at the adjacent time, the existing MCMC process is improved, and the wind power time series output simulation is performed, including the following steps:
step S310: obtaining classified typical output days by using an improved KM algorithm, and establishing a Markov chain, a wind power output probability density function and a fluctuation quantity probability density function of different typical days, wherein the Markov chain divides the standard wind power output into a plurality of output intervals;
step S320: generating power generation power of a first day and a first hour by initialization according to output probability density distribution, generating output at the next moment by using a wind power output joint probability density model at adjacent moments, selecting an output interval at the next moment where the output at the next moment is located according to a Markov chain, generating accurate output by using fluctuation quantity probability density, and generating an output day type of the next day by using a day transfer matrix after the output length at the day meets the requirement;
step S330: and repeating the step S320 until the length of the simulated output sequence meets the requirement.
In order to achieve the above object, an embodiment of a second aspect of the present application provides an MCMC wind power simulation apparatus based on improved scene classification and coarse grain removal, including: a dividing module, a model establishing module and a simulation module, wherein,
the dividing module is used for clustering the historical output data of each day of wind power by using an improved KM clustering algorithm and dividing the wind power output day into different typical output scenes;
the model establishing module is used for establishing a Markov chain at adjacent moments in a day and a wind power output joint probability density distribution function model at adjacent moments for each classified scene, wherein the wind power output joint probability density model at the adjacent moments is obtained by fitting a Copula mixed model;
and the simulation module is used for improving the existing MCMC flow and simulating the wind power time sequence output based on the established Markov chain and the wind power output joint probability density distribution function model at the adjacent moment.
The MCMC wind power simulation method and device based on improved scene classification and coarsening removal solve the technical problems of poor searching capability and low simulation efficiency of the existing method, realize typical daily clustering by improving a KM algorithm, establish wind power output joint probability density at adjacent moments by adopting a Copula mixed model, fully utilize historical wind power output data, classify the wind power output data, realize quick and accurate simulation of a wind power time series output process under the condition that other data such as meteorological factors, wind motor parameters and the like do not exist, fully develop the output characteristics of the wind power historical data, establish an effective wind power output model by effective classification and modeling, and provide support for operation scheduling planning of wind power.
Additional aspects and advantages of the present application will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of the present application.
Drawings
The foregoing and/or additional aspects and advantages of the present application will become apparent and readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings of which:
fig. 1 is a flowchart of an MCMC wind power simulation method based on improved scene classification and coarse grain removal according to an embodiment of the present application;
fig. 2 is a flowchart of an algorithm of each genetic individual in an improved KM algorithm of an MCMC wind power simulation method based on improved scene classification and de-coarsening according to an embodiment of the present application;
fig. 3 is a diagram of improved individual progeny in an improved KM algorithm of an MCMC wind power simulation method based on improved scene classification and coarse grain removal according to an embodiment of the present application;
fig. 4 is an overall flowchart of an improved KM algorithm using a genetic algorithm and an annealing algorithm of the MCMC wind power simulation method based on improved scene classification and coarsening removal according to the embodiment of the present application;
fig. 5 is an overall flowchart of an actual wind power output process simulation performed by using wind power historical output data in the MCMC wind power simulation method based on improved scene classification and coarsening removal according to the embodiment of the present application;
fig. 6 is an evaluation index graph of different scene classification numbers of the MCMC wind power simulation method based on improved scene classification and coarse grain removal according to the embodiment of the present application;
fig. 7 is a comparison graph of results of an improved KM algorithm and an unmodified KM algorithm of the MCMC wind power simulation method based on improved scene classification and coarse grain removal according to the embodiment of the present application;
fig. 8 is a diagram illustrating a classification result of a historical wind power generation day obtained by using an improved KM algorithm according to an MCMC wind power simulation method based on improved scene classification and coarse grain removal according to an embodiment of the present application;
fig. 9 is a graph of a wind power generation sequence adjacent time output frequency distribution histogram and Copula fitting result based on the MCMC wind power simulation method for improving scene classification and coarse grain removal according to the embodiment of the present application;
fig. 10 is a wind power historical data of an MCMC wind power simulation method based on improved scene classification and coarse grain removal and a probability density distribution graph simulated by an MCMC algorithm, an improved MCMC algorithm not classified by a KM algorithm, and an improved MCMC algorithm classified by a KM algorithm, respectively according to an embodiment of the present application;
fig. 11 is a wind power historical data of the MCMC wind power simulation method based on improved scene classification and coarse grain removal and an autocorrelation coefficient curve graph of an output sequence obtained by simulation using the MCMC algorithm, the improved MCMC algorithm not classified by the KM algorithm, and the improved MCMC algorithm classified by the KM algorithm, respectively, according to the embodiment of the present application;
FIG. 12 is a comparison graph of simulated output and historical data of various output scenes 24h after classification based on the MCMC wind power simulation method for improving scene classification and de-coarsening according to the embodiment of the application;
fig. 13 is a wind power historical data of the MCMC wind power simulation method based on improved scene classification and coarse grain removal according to the embodiment of the present application, and a 150-hour simulation result comparison graph of a simulation performed by the MCMC algorithm, the improved MCMC algorithm not classified by the KM algorithm, and the improved MCMC algorithm classified by the KM algorithm, respectively;
fig. 14 is a schematic structural diagram of an MCMC wind power simulation apparatus based on improved scene classification and coarse grain removal according to the second embodiment of the present application.
Detailed Description
Reference will now be made in detail to the embodiments of the present application, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to the same or similar elements or elements having the same or similar function throughout. The embodiments described below with reference to the drawings are exemplary and intended to be used for explaining the present application and should not be construed as limiting the present application.
The MCMC wind power simulation method and apparatus based on improved scene classification and de-coarsening according to the embodiment of the present application are described below with reference to the accompanying drawings.
Fig. 1 is a flowchart of an MCMC wind power simulation method based on improved scene classification and coarse grain removal according to an embodiment of the present application.
As shown in fig. 1, the MCMC wind power simulation method based on improved scene classification and de-coarsening includes the following steps:
step S100: clustering the historical output data of each day of wind power by using an improved KM clustering algorithm, and dividing the wind power output day into different typical output scenes;
step S200: establishing a Markov chain at adjacent moments in a day and a wind power output joint probability density distribution function model at adjacent moments for each classified scene, wherein the wind power output joint probability density distribution function model at adjacent moments is obtained by fitting through a Copula mixed model;
step S300: based on the established Markov chain and the wind power output joint probability density distribution function model at the adjacent moment, the existing MCMC flow is improved, and the wind power time sequence output is simulated.
The MCMC wind power simulation method based on improved scene classification and coarse grain removal in the embodiment of the application comprises the following steps of S100: clustering the historical output data of each day of wind power by using an improved KM clustering algorithm, and dividing the wind power output day into different typical output scenes; step S200: establishing a Markov chain at adjacent moments in a day and a wind power output joint probability density distribution function model at adjacent moments for each classified scene, wherein the wind power output joint probability density distribution function model at adjacent moments is obtained by fitting through a Copula mixed model; step S300: based on the established Markov chain and the wind power output joint probability density distribution function model at the adjacent moment, the existing MCMC flow is improved, and the wind power time sequence output is simulated. Therefore, the technical problems of poor searching capability and low simulation efficiency of the existing method can be solved, typical daily clustering is realized by improving a KM algorithm, wind power output joint probability density at adjacent moments is established by adopting a Copula mixed model, wind power output historical data is fully utilized and classified, the wind power output process is quickly and accurately simulated under the condition that other data such as meteorological factors, wind motor parameters and the like do not exist, the output characteristics of the wind power historical data are fully developed, an effective wind power output model is established through effective classification and modeling, and support can be provided for operation scheduling planning of wind power.
The application provides an improved wind power output simulation method, which comprises the following steps: s1, processing the original data, and determining the clustering number of the improved KM clustering algorithm; s2, initializing a genetic algebra and a fitness function of a genetic algorithm, and then completing one-time KM clustering by all individuals to generate an initial generation population; s3, selecting, crossing and mutating the initial generation population; s4, continuously circulating the previous steps until the maximum genetic algebra is reached; s5, carrying out statistics on the classified data of each typical output type to obtain a Markov transformation matrix, an output power probability density distribution function and an output fluctuation amount probability density distribution function; s6, establishing an accumulative state transition matrix among different output scenes and a joint probability density distribution function of adjacent output; s7, generating wind power output initial output day state d based on output probability density distribution0And initial time state t0(ii) a And S8, correcting the random state number v of the traditional MCMC algorithm by using the joint probability density distribution function and the fluctuation amount probability density distribution function to perform output simulation.
According to the method and the device, under the single condition that only wind power historical output data exist, the historical output days are firstly classified, and the simulation precision is improved through independent modeling of different output days. Meanwhile, the clustering speed and the clustering effect are improved by using an improved KM clustering algorithm. And then, on the basis of clustering, the MCMC algorithm improved by the statistical function is further utilized to simulate the output sequence, so that the simulation precision can be further improved, and the simulation process is more consistent with the actual situation.
The method for constructing the wind power plant power generation time sequence simulation scene considering time correlation and volatility has the advantages that on one hand, under the condition that only wind power historical output data exist, the modeling accuracy can be improved by fully extracting data features; on the other hand, the operation data which accords with the actual situation can be provided for the operation scheduling of the power grid by constructing the wind power generation simulation model, so that the contribution to the stable and safe operation of the power system is made. Firstly, a genetic algorithm and an annealing algorithm are utilized to improve a KM algorithm, so that the classification speed and precision are improved, the wind power historical output date is accurately classified, different data models are respectively established by analyzing different types of output data, and the modeling error is reduced to the maximum extent; and then, correcting the random sampling process in the MCMC algorithm by utilizing a mixed Copula model of adjacent output in time and a probability density distribution function of fluctuation quantity, reducing the coarse graining influence caused by a power interval in a pure random process, further improving the simulation efficiency, and enabling the data generated by the established wind power output model to better accord with the actual situation.
The effect of the classification of the output day can directly influence the establishment of the wind power model, and the genetic algorithm is used for improving the defects of poor searching capability and low convergence rate of the traditional KM algorithm. Meanwhile, the simulated annealing algorithm is combined with the genetic algorithm, so that the population diversity of the genetic algorithm is increased, the population evolution speed is increased, and the speed and the precision of the whole classification process are improved.
Analyzing by taking the output data of a certain wind power plant in Gansu province as a sample, wherein the data sampling interval is 15min, and performing data normalization treatment:
wherein, PiFor input sampling power, PMRated capacity, P, of wind farmsoIs the normalized output power.
Further, in the embodiment of the present application, the improved KM clustering algorithm includes the following steps:
step S110: designing a clustering effect evaluation index, and determining the number of clustering populations according to the size of the clustering effect evaluation index;
step S120: and improving the KM algorithm by using a genetic algorithm and an annealing algorithm, and selecting, crossing and mutating the population.
Genetic individuals are not initialized in a coding manner, but a KM clustering process is regarded as an individual.
Further, in the embodiment of the present application, the clustering effect evaluation index is expressed as:
wherein D isinIs the average value of the distances within the species group, DoutIs the minimum out-of-class spacing of the population, C is the tuning constant, d (x)i,vj) Is a sample point xiAnd a clustering center vjD (v) ofi,vj) Is the center of the cluster viAnd vjThe distance of (c).
Further, in the embodiment of the present application, the KM algorithm is improved by using a genetic algorithm and an annealing algorithm, and population selection, crossing and mutation are performed, including the following steps:
step S121: determining the scale of a genetic population, a genetic algebra, an annealing initial temperature and an annealing speed, and normalizing the wind power historical data;
step S122: each population individual completes one-time complete KM clustering operation, calculates individual fitness, performs selection operation according to the fitness, and the eliminated individuals are selectively retained by using a wheel disc selection algorithm;
step S123: carrying out cross and mutation operations on the selected new individuals;
step S124: recalculating the fitness value of the individual, and selecting the population individual according to the fitness;
step S125: the steps S122, S123 and S124 are repeated until the maximum number of generations of inheritance is reached.
The selection process of the population parent individuals comprises the following steps: calculating the fitness values of all individuals in the population, and directly reserving the individuals with the optimal fitness to filial generations; selecting the rest individuals by adopting a wheel disc selection algorithm until the individuals with the population size of 50% are selected, and then transmitting the individuals to filial generations; and screening out the individuals with the population size of 30% to be transmitted to offspring according to the probability obtained by the simulated annealing algorithm.
The chromosome crossing and the mutation are performed on the clustering center of each individual, the crossing is performed by randomly exchanging the clustering centers of two populations, the mutation is that the clustering center of a certain population is changed immediately, the probability of the two operations is changed along with the evolution of the population and the change of the fitness value of the individual, the population is continuously updated until the maximum genetic algebra is reached, and the classification result of the wind power generation day can be obtained.
Further, in the embodiment of the present application, the individual fitness value is calculated using an individual fitness function, which is expressed as:
wherein k is the number of species clusters, n is the total number of samples, viIs the cluster center of the ith population,
is the average of all the samples and is,
is the average of the i-th sample, Tr(SB) And Tr (S)w) Respectively used for measuring the difference degree between classes and the aggregation degree of samples in the classes.
Further, in the embodiment of the present application, the wheel disc selection algorithm is improved by the annealing algorithm, and the probability of the wheel disc selection algorithm receiving rejected individuals is:
wherein T is the current annealing temperature, a is the annealing speed, T0As annealing initiation temperature, fbestThe fitness value of the best population individual is K which is a constant, K is the global optimization times, namely the genetic algebra, and p is the reselected probability of the eliminated individual.
Further, in the embodiment of the present application, the probability of occurrence of performing the crossover and mutation operations is:
wherein p iscTo cross probability, pmAs the mutation probability, pcmaxAnd pcminRespectively preset maximum and minimum cross probability, pmmaxAnd pmminRespectively preset maximum and minimum mutation probabilities, fmeanIs a population mean fitness value, f1And f2Fitness values of two intersecting individuals, respectively, fmaxIs the maximum fitness value of the population.
Further, in the embodiment of the application, for each classified scene, the establishment of the wind power output joint probability density distribution function model at the adjacent time comprises the following steps:
step S210: according to the historical wind power output data of the inspected classification scene, a distribution function of the wind power output under the classification scene and a probability density distribution histogram between adjacent moments are obtained through statistical calculation;
step S220: selecting a mixed Copula function model according to the shape of the probability density distribution histogram between adjacent moments for modeling to generate a mixed Copula function model;
step S230: and estimating parameters of the hybrid Copula function model by using a maximum likelihood estimation method according to the observation data to obtain a wind power output joint probability density distribution function model at the adjacent moment.
The distribution function F (x) of the wind power output is calculated by the formula
Wherein the content of the first and second substances,
for the wind power output probability density estimation function, phi (·) is a standard normal distribution function, h is a window width, and an empirical rule is adopted for solving:
wherein
Is the standard deviation of the sample, xiAnd i is 1, …, and n is wind power output historical data.
The hybrid Copula function model is:
wherein λ iskFor each Copula function Ck(u,v；θk) Weight coefficient in the mixing function, θkIs the parameter of the kth Copula function.
Estimating parameters of the hybrid Copula function model by using a maximum likelihood estimation method according to observation data:
and performing statistical analysis by using the wind power historical data to respectively obtain the probability density distribution of the output power and the probability density distribution of the fluctuation quantity.
Initializing a power day type S with a first day d equal to 00Output data P with m and initial time t 00And state z0. The current moment output is PtIn the state ztGenerating a random number v, if Q (z) using a hybrid Copula functiont,j-1)<v≤Q(ztJ), then the state at time t +1 is zt+1＝j；
Further, in the embodiment of the present application, based on the established markov chain and the wind power output joint probability density model at the adjacent time, the existing MCMC process is improved, and the wind power time series output simulation is performed, including the following steps:
step S310: obtaining classified typical output days by using an improved KM algorithm, and establishing a Markov chain, a wind power output probability density function and a fluctuation quantity probability density function of different typical days, wherein the Markov chain divides the standard wind power output into a plurality of output intervals;
step S320: generating power generation power of a first day and a first hour by initialization according to output probability density distribution, generating output at the next moment by using a wind power output joint probability density model at adjacent moments, selecting an output interval at the next moment where the output at the next moment is located according to a Markov chain, generating accurate output by using fluctuation quantity probability density, and generating an output day type of the next day by using a day transfer matrix after the output length at the day meets the requirement;
step S330: and repeating the step S320 until the length of the simulated output sequence meets the requirement.
The output probability density is obtained by counting the output of the wind power at all single moments, obtaining a probability density distribution function by utilizing kernel density estimation, and fitting the fluctuation quantity probability density function by utilizing the kernel density function by counting the difference value of the wind power output at adjacent moments.
And generating random fluctuation amount based on the joint probability density function of the fluctuation amount to further correct the output data at the next moment.
The state transition matrix is defined as:
wherein n isijThe number of transitions from scene i to scene j.
The cumulative state transition matrix is:
establishing a cumulative transition probability matrix Q of all the contributing daysdIn each output scene, establishing a frequency distribution histogram of output at adjacent time, and obtaining a difference value through output at adjacent time to obtain fluctuation quantity probability density:
ΔP＝Pt-Pt-1
selecting a proper binary Coipla function C (u, v; theta) for modeling according to the frequency distribution histogram, and performing modeling according to wind power output historical data at adjacent moments
And the EM algorithm to estimate the parameter θ of the mixed function model.
Generating a random fluctuation amount beta based on the previously obtained fluctuation amount probability density functiontAnd the predicted generated power at the next moment is Pt+1＝Pt+βtIf P ist+1∈{P|z＝zt+1Fifthly, continuing; otherwise, the fluctuation amount is extracted again. And after the output sequence of one day is generated, the cumulative probability matrix of the output day is utilized, and the typical day type of the next day is generated by random sampling.
Fig. 2 is a flowchart of an algorithm of each genetic individual in an improved KM algorithm based on an MCMC wind power simulation method for improved scene classification and de-coarsening according to an embodiment of the present application.
As shown in fig. 2, the algorithm of each genetic individual in the improved KM algorithm is to input data X and a parameter k, and select k samples as initial clustering centers; distributing all samples to different classes according to a nearest distance principle; recalculating the average value of each cluster class; judging whether the clustering center of each cluster changes or not, and if not, ending; and if so, re-distributing all the samples to different classes according to the nearest principle until the clustering center of each cluster is unchanged. Each genetic individual was subjected to one KM clustering process.
Fig. 3 is a diagram of improved individual progeny in an improved KM algorithm of an MCMC wind power simulation method based on improved scene classification and coarse grain removal according to an embodiment of the present application.
As shown in fig. 3, in the improved KM algorithm of the MCMC wind power simulation method based on improved scene classification and coarse grain removal, each time the population updating process, the offspring population is composed of the optimal individuals, the wheel disc selection individuals, and the culled individuals.
Fig. 4 is an overall flowchart of an improved KM algorithm using a genetic algorithm and an annealing algorithm of the MCMC wind power simulation method based on improved scene classification and coarse grain removal according to the embodiment of the present application.
As shown in fig. 4, the step of improving the KM algorithm by using the genetic algorithm and the annealing algorithm is to input data X and determine the value range of the sample cluster number k; initializing a genetic population; selecting the genetic population; performing cross variation on the genetic population; recalculating the fitness value of the individuals in the population; judging whether the maximum inheritance times is reached, if so, ending; and if not, carrying out selection operation and cross variation on the genetic population again until the maximum genetic times are reached.
Fig. 5 is an overall flowchart of an actual wind power output process simulation performed by using wind power historical output data in the MCMC wind power simulation method based on improved scene classification and coarsening removal according to the embodiment of the present application.
As shown in fig. 5, in the MCMC wind power simulation method based on improved scene classification and coarse grain removal, historical output data of wind power is processed, abnormal output power is deleted, and output power is standardized; clustering the data to complete typical day division; calculating a state transition matrix and a Copula time sequence correlation function C (U, V) of each scene; calculating an accumulative transition probability matrix Q; sampling to generate a state quantity z at the next moment, and determining the output range at the next moment; establishing a fluctuation amount probability density distribution model to generate random fluctuation amount betatAnd the predicted generated power at the next moment is Pt+1＝Pt+βtIf P ist+1∈{P|z＝zt+1Continuing, otherwise, extracting fluctuation quantity again; after generating a power output sequence of one day, generating a typical day type of the next day by random sampling by using an accumulated probability matrix of the power output day; and repeating the two steps continuously until a simulation running sequence with the length meeting the requirement is generated.
Fig. 6 is an evaluation index graph of different scene classification numbers of the MCMC wind power simulation method based on improved scene classification and coarse grain removal according to the embodiment of the present application.
As shown in fig. 6, according to the MCMC wind power simulation method based on improved scene classification and coarse grain removal, the clustering effect is better as the index value is smaller as can be seen from the formula of the clustering evaluation index. When the classification number is 6, the classification effect is good, and when the classification number exceeds 6, the operation burden is increased while the improvement on the clustering effect is not obvious, so that the typical scene classification number k is selected to be 6.
Fig. 7 is a comparison graph of results of the improved KM algorithm and the unmodified KM algorithm of the MCMC wind power simulation method based on improved scene classification and coarse grain removal according to the embodiment of the present application.
As shown in fig. 7, the improved KM algorithm based on the MCMC wind power simulation method with improved scene classification and de-coarsening has faster convergence speed and better classification effect than the non-improved KM algorithm.
Fig. 8 is a diagram illustrating a classification result of a historical wind power generation day obtained by using an improved KM algorithm according to an MCMC wind power simulation method based on improved scene classification and coarse grain removal in the embodiment of the present application.
As shown in fig. 8, in the MCMC wind power simulation method based on improved scene classification and coarse grain removal, the evaluation index is used to determine the clustering number of the KM algorithm for the outgoing day classification, the fitness function of the genetic algorithm in the improved algorithm is selected, and the improved KM algorithm is used to classify the historical outgoing day to obtain the classification result.
Fig. 9 is a graph of a wind power generation sequence adjacent time output frequency distribution histogram and Copula fitting result based on the MCMC wind power simulation method for improving scene classification and coarse grain removal according to the embodiment of the present application.
As shown in fig. 9, the Copula function is used for measuring the time correlation of the output sequence, and establishing an accurate and effective Copula function model can greatly improve the simulation precision and efficiency, and the MCMC wind power simulation method based on improved scene classification and coarse grain removal selects a binary joint probability density model by constructing a frequency histogram of the output of wind power adjacent time to obtain the observation data at the time t and the time t +1
It can be seen that the constructed Copula function model can fit the output data well.
Fig. 10 is a wind power history data of the MCMC wind power simulation method based on improved scene classification and coarse grain removal and probability density distribution maps of simulations performed by the MCMC algorithm, the improved MCMC algorithm not classified by the KM algorithm, and the improved MCMC algorithm classified by the KM algorithm, respectively, according to the embodiment of the present application.
Fig. 11 is a wind power history data of the MCMC wind power simulation method based on improved scene classification and coarse grain removal and an autocorrelation coefficient curve graph of a power output sequence obtained by simulation using the MCMC algorithm, the improved MCMC algorithm not classified by the KM algorithm, and the improved MCMC algorithm classified by the KM algorithm, respectively.
Fig. 12 is a comparison graph of simulated output and historical data of various output scenes 24h after classification based on the MCMC wind power simulation method with improved scene classification and de-coarsening according to the embodiment of the present application.
Fig. 13 is a wind power historical data of the MCMC wind power simulation method based on improved scene classification and coarse grain removal according to the embodiment of the present application, and a 150-hour simulation result comparison graph of a simulation performed by using the MCMC algorithm, an improved MCMC algorithm not using the KM algorithm, and an improved MCMC algorithm using the KM algorithm.
Fig. 10, fig. 11, fig. 12, and fig. 13 respectively simulate comparison graphs of data and historical data on probability density distribution, autocorrelation coefficient change trend, 24h time scale, and 150h time scale, and it can be seen from the results that the modeling method provided by the application has higher simulation accuracy, the operating state better conforms to the actual power generation condition of wind power, and meanwhile, the simulation effect classified by the improved KM algorithm is better.
The table is a comparison of mean and standard deviation of the wind power history data and data obtained by simulation using the MCMC algorithm, the modified MCMC algorithm not classified using the KM algorithm, and the modified MCMC algorithm classified using the KM algorithm, respectively, in the present application. And the second table is the comparison of the standard deviation of the wind power history data and the data obtained by six classification scenes obtained by respectively simulating by using the MCMC algorithm and the improved MCMC algorithm. It can be seen that the classified simulation data better conforms to the actual situation, and the result obtained by the improved MCMC algorithm is more accurate.
The output day data is simulated through a traditional MCMC algorithm, an improved MCMC algorithm which does not classify the output day and the classified improved MCMC algorithm, and the obtained output data pair is shown in a table I and a table II.
Watch two
Fig. 13 is a schematic structural diagram of an MCMC wind power simulation apparatus based on improved scene classification and coarse grain removal according to the second embodiment of the present application.
As shown in fig. 13, the MCMC wind power simulation apparatus based on improved scene classification and de-coarsening includes: a dividing module, a model establishing module and a simulation module, wherein,
the dividing module 10 is configured to cluster the wind power daily historical output data by using an improved KM clustering algorithm, and divide the wind power output day into different typical output scenes;
the model establishing module 20 is used for establishing a Markov chain at adjacent time in a day and a wind power output joint probability density distribution function model at adjacent time for each classified scene, wherein the wind power output joint probability density model at adjacent time is obtained by fitting a Copula mixed model;
and the simulation module 30 is configured to improve the existing MCMC process and simulate the wind power time series output based on the established markov chain and the wind power output joint probability density distribution function model at the adjacent time.
The MCMC wind power simulation device based on improved scene classification and coarse grain removal comprises: the device comprises a dividing module, a model establishing module and a simulation module, wherein the dividing module is used for clustering the daily historical output data of the wind power by using an improved KM clustering algorithm and dividing the wind power output day into different typical output scenes; the model establishing module is used for establishing a Markov chain at adjacent moments in a day and a wind power output joint probability density distribution function model at adjacent moments for each classified scene, wherein the wind power output joint probability density model at the adjacent moments is obtained by fitting a Copula mixed model; and the simulation module is used for improving the existing MCMC flow and simulating the wind power time sequence output based on the established Markov chain and the wind power output joint probability density distribution function model at the adjacent moment. Therefore, the technical problems of poor searching capability and low simulation efficiency of the existing method can be solved, typical daily clustering is realized by improving a KM algorithm, wind power output joint probability density at adjacent moments is established by adopting a Copula mixed model, wind power output historical data is fully utilized and classified, the wind power output process is quickly and accurately simulated under the condition that other data such as meteorological factors, wind motor parameters and the like do not exist, the output characteristics of the wind power historical data are fully developed, an effective wind power output model is established through effective classification and modeling, and support can be provided for operation scheduling planning of wind power.
In the description herein, reference to the description of the term "one embodiment," "some embodiments," "an example," "a specific example," or "some examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the application. In this specification, the schematic representations of the terms used above are not necessarily intended to refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Furthermore, various embodiments or examples and features of different embodiments or examples described in this specification can be combined and combined by one skilled in the art without contradiction.
Furthermore, the terms "first", "second" and "first" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one such feature. In the description of the present application, "plurality" means at least two, e.g., two, three, etc., unless specifically limited otherwise.
Any process or method descriptions in flow charts or otherwise described herein may be understood as representing modules, segments, or portions of code which include one or more executable instructions for implementing steps of a custom logic function or process, and alternate implementations are included within the scope of the preferred embodiment of the present application in which functions may be executed out of order from that shown or discussed, including substantially concurrently or in reverse order, depending on the functionality involved, as would be understood by those reasonably skilled in the art of the present application.
The logic and/or steps represented in the flowcharts or otherwise described herein, e.g., an ordered listing of executable instructions that can be considered to implement logical functions, can be embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus, or device, such as a computer-based system, processor-containing system, or other system that can fetch the instructions from the instruction execution system, apparatus, or device and execute the instructions. For the purposes of this description, a "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of the computer-readable medium would include the following: an electrical connection (electronic device) having one or more wires, a portable computer diskette (magnetic device), a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium could even be paper or another suitable medium upon which the program is printed, as the program can be electronically captured, via for instance optical scanning of the paper or other medium, then compiled, interpreted or otherwise processed in a suitable manner if necessary, and then stored in a computer memory.
It should be understood that portions of the present application may be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, the various steps or methods may be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. If implemented in hardware, as in another embodiment, any one or combination of the following techniques, which are known in the art, may be used: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application specific integrated circuit having an appropriate combinational logic gate circuit, a Programmable Gate Array (PGA), a Field Programmable Gate Array (FPGA), or the like.
It will be understood by those skilled in the art that all or part of the steps carried by the method for implementing the above embodiments may be implemented by hardware related to instructions of a program, which may be stored in a computer readable storage medium, and when the program is executed, the program includes one or a combination of the steps of the method embodiments.
In addition, functional units in the embodiments of the present application may be integrated into one processing module, or each unit may exist alone physically, or two or more units are integrated into one module. The integrated module can be realized in a hardware mode, and can also be realized in a software functional module mode. The integrated module, if implemented in the form of a software functional module and sold or used as a stand-alone product, may also be stored in a computer readable storage medium.
The storage medium mentioned above may be a read-only memory, a magnetic or optical disk, etc. Although embodiments of the present application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present application, and that variations, modifications, substitutions and alterations may be made to the above embodiments by those of ordinary skill in the art within the scope of the present application.
Claims (10)
1. An MCMC wind power simulation method based on improved scene classification and coarse grain removal is characterized by comprising the following steps:
step S100: clustering the historical output data of each day of wind power by using an improved KM clustering algorithm, and dividing the wind power output day into different typical output scenes;
step S200: establishing a Markov chain at adjacent moments in a day and a wind power output joint probability density distribution function model at the adjacent moments for each classified scene, wherein the wind power output joint probability density distribution function model at the adjacent moments is obtained by fitting through a Copula mixed model;
step S300: based on the established Markov chain and the wind power output joint probability density distribution function model at the adjacent moment, the existing MCMC flow is improved, and the wind power time sequence output is simulated.
2. The method of claim 1, wherein the improved KM clustering algorithm comprises the steps of:
step S110: designing a clustering effect evaluation index, and determining the number of clustering populations according to the size of the clustering effect evaluation index;
step S120: and improving the KM algorithm by using a genetic algorithm and an annealing algorithm, and selecting, crossing and mutating the population.
3. The method according to claim 2, wherein the clustering effect evaluation index is expressed as:
wherein D isinIs the average value of the distances within the species group, DoutIs the minimum out-of-class spacing of the population, C is the tuning constant, d (x)i,vj) Is a sample point xiAnd a clustering center vjD (v) ofi,vj) Is the center of the cluster viAnd vjThe distance of (c).
4. The method of claim 2, wherein the improvement of the KM algorithm by genetic algorithm and annealing algorithm for population selection, crossing and mutation comprises the steps of:
step S121: determining the scale of a genetic population, a genetic algebra, an annealing initial temperature and an annealing speed, and normalizing the wind power historical data;
step S122: each population individual completes one-time complete KM clustering operation, calculates individual fitness, performs selection operation according to the fitness, and the eliminated individuals are selectively retained by using a wheel disc selection algorithm improved by an annealing algorithm;
step S123: carrying out cross and mutation operations on the selected new individuals;
step S124: recalculating the fitness value of the individual, and selecting the population individual according to the fitness;
step S125: the steps S122, S123 and S124 are repeated until the maximum number of generations of inheritance is reached.
5. The method of claim 4, wherein the individual fitness value is calculated using an individual fitness function represented as:
wherein k is the number of species clusters, n is the total number of samples, viIs the cluster center of the ith population,
is the average of all the samples and is,
is the average of the i-th sample, Tr(SB) And Tr (S)w) Respectively used for measuring the difference degree between classes and the aggregation degree of samples in the classes.
6. The method of claim 4, wherein the wheel selection algorithm is refined by an annealing algorithm, the wheel selection algorithm accepting the culled individuals with a probability of:
wherein T is the current annealing temperature, a is the annealing speed, T0As annealing initiation temperature, fbestThe fitness value of the best population individual is K which is a constant, K is the global optimization times, namely the genetic algebra, and p is the reselected probability of the eliminated individual.
7. The method of claim 4, wherein the probability of occurrence of performing the crossover and mutation operations is:
wherein p iscTo cross probability, pmAs the mutation probability, pcmaxAnd pcminRespectively preset maximum and minimum cross probability, pmmaxAnd pmminRespectively preset maximum and minimum mutation probabilities, fmeanIs a population mean fitness value, f1And f2Fitness values of two intersecting individuals, respectively, fmaxIs the maximum fitness value of the population.
8. The method of claim 1, wherein the building of the wind power output joint probability density distribution function model at adjacent moments for each classified scene comprises the following steps:
step S210: according to the historical wind power output data of the inspected classification scene, a distribution function of the wind power output under the classification scene and a probability density distribution histogram between adjacent moments are obtained through statistical calculation;
step S220: selecting a mixed Copula function model according to the shape of the probability density distribution histogram between the adjacent moments for modeling to generate a mixed Copula function model;
step S230: and estimating parameters of the hybrid Copula function model by using a maximum likelihood estimation method according to observation data to obtain the wind power output joint probability density distribution function model at the adjacent moment.
9. The method of claim 1, wherein the existing MCMC process is improved to perform the simulation of the wind power time series output based on the established Markov chain and adjacent moment wind power output joint probability density model, and the method comprises the following steps:
step S310: obtaining classified typical output days by using an improved KM algorithm, and establishing a Markov chain, a wind power output probability density function and a fluctuation quantity probability density function of different typical days, wherein the Markov chain divides the standard wind power output into a plurality of output intervals;
step S320: generating the generated power of the first day and the first hour by initialization according to output probability density distribution, generating output at the next moment by using the wind power output joint probability density model at the adjacent moment, then selecting the output interval at the next moment where the output at the next moment is located according to the Markov chain, generating accurate output by using fluctuation quantity probability density, and generating the output day type of the next day by using a day transfer matrix after the output length of the day meets the requirement;
step S330: and repeating the step S320 until the length of the simulated output sequence meets the requirement.
10. An MCMC wind power simulation device based on improved scene classification and coarse grain removal is characterized by comprising: a dividing module, a model establishing module and a simulation module, wherein,
the dividing module is used for clustering the historical output data of each day of wind power by using an improved KM clustering algorithm and dividing the wind power output day into different typical output scenes;
the model establishing module is used for establishing a Markov chain at adjacent time in a day and a wind power output joint probability density distribution function model at adjacent time for each classified scene, wherein the wind power output joint probability density model at adjacent time is obtained by fitting a Copula mixed model;
and the simulation module is used for improving the existing MCMC flow and simulating the wind power time sequence output based on the established Markov chain and the wind power output joint probability density distribution function model at the adjacent moment.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111122162.9A CN113807019A (en) | 2021-09-24 | 2021-09-24 | MCMC wind power simulation method based on improved scene classification and coarse grain removal |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111122162.9A CN113807019A (en) | 2021-09-24 | 2021-09-24 | MCMC wind power simulation method based on improved scene classification and coarse grain removal |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113807019A true CN113807019A (en) | 2021-12-17 |
Family
ID=78940354
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111122162.9A Pending CN113807019A (en) | 2021-09-24 | 2021-09-24 | MCMC wind power simulation method based on improved scene classification and coarse grain removal |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113807019A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114896765A (en) * | 2022-04-19 | 2022-08-12 | 国网甘肃省电力公司电力科学研究院 | Multi-scene switching wind power sequence simulation method and device based on flexible time boundary |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060023723A1 (en) * | 2004-07-27 | 2006-02-02 | Michele Morara | Object oriented library for markov chain monte carlo simulation |
| CN107464007A (en) * | 2016-06-02 | 2017-12-12 | 南京理工大学 | Continuous time Probabilistic Load Flow Forecasting Methodology based on Markov theory and pro rate principle |
| CN109524993A (en) * | 2018-12-19 | 2019-03-26 | 中国农业大学 | The typical week power output scene generating method of wind-powered electricity generation photovoltaic for Mid-long Term Optimized Scheduling |
| CN109783841A (en) * | 2018-11-28 | 2019-05-21 | 河海大学 | A kind of photovoltaic power output time series analogy method based on more scene state transfer matrixes and conditional probability sampling |
| CN109904878A (en) * | 2019-02-28 | 2019-06-18 | 西安交通大学 | A kind of windy electric field electricity-generating timing simulation scenario building method |
| CN110570122A (en) * | 2019-09-10 | 2019-12-13 | 重庆大学 | Offshore wind power plant reliability assessment method considering wind speed seasonal characteristics and current collection system element faults |
| CN110909911A (en) * | 2019-09-29 | 2020-03-24 | 中国农业大学 | Method for aggregating multidimensional time series data considering space-time correlation |
| CN111639808A (en) * | 2020-05-29 | 2020-09-08 | 国网山东省电力公司经济技术研究院 | Multi-wind-farm output scene generation method and system considering time-space correlation |
| CN111859812A (en) * | 2020-07-30 | 2020-10-30 | 南方电网科学研究院有限责任公司 | Reliability evaluation method for offshore wind power plant and flexible direct grid-connected system under weather influence |
| CN112307590A (en) * | 2019-10-16 | 2021-02-02 | 国网福建省电力有限公司 | Wind power plant output time sequence curve simulation method considering correlation |
| CN112884270A (en) * | 2020-12-31 | 2021-06-01 | 国网山东省电力公司聊城供电公司 | Multi-scene power distribution network planning method and system considering uncertainty factors |
-
2021
- 2021-09-24 CN CN202111122162.9A patent/CN113807019A/en active Pending
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060023723A1 (en) * | 2004-07-27 | 2006-02-02 | Michele Morara | Object oriented library for markov chain monte carlo simulation |
| CN107464007A (en) * | 2016-06-02 | 2017-12-12 | 南京理工大学 | Continuous time Probabilistic Load Flow Forecasting Methodology based on Markov theory and pro rate principle |
| CN109783841A (en) * | 2018-11-28 | 2019-05-21 | 河海大学 | A kind of photovoltaic power output time series analogy method based on more scene state transfer matrixes and conditional probability sampling |
| CN109524993A (en) * | 2018-12-19 | 2019-03-26 | 中国农业大学 | The typical week power output scene generating method of wind-powered electricity generation photovoltaic for Mid-long Term Optimized Scheduling |
| CN109904878A (en) * | 2019-02-28 | 2019-06-18 | 西安交通大学 | A kind of windy electric field electricity-generating timing simulation scenario building method |
| CN110570122A (en) * | 2019-09-10 | 2019-12-13 | 重庆大学 | Offshore wind power plant reliability assessment method considering wind speed seasonal characteristics and current collection system element faults |
| CN110909911A (en) * | 2019-09-29 | 2020-03-24 | 中国农业大学 | Method for aggregating multidimensional time series data considering space-time correlation |
| CN112307590A (en) * | 2019-10-16 | 2021-02-02 | 国网福建省电力有限公司 | Wind power plant output time sequence curve simulation method considering correlation |
| CN111639808A (en) * | 2020-05-29 | 2020-09-08 | 国网山东省电力公司经济技术研究院 | Multi-wind-farm output scene generation method and system considering time-space correlation |
| CN111859812A (en) * | 2020-07-30 | 2020-10-30 | 南方电网科学研究院有限责任公司 | Reliability evaluation method for offshore wind power plant and flexible direct grid-connected system under weather influence |
| CN112884270A (en) * | 2020-12-31 | 2021-06-01 | 国网山东省电力公司聊城供电公司 | Multi-scene power distribution network planning method and system considering uncertainty factors |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114896765A (en) * | 2022-04-19 | 2022-08-12 | 国网甘肃省电力公司电力科学研究院 | Multi-scene switching wind power sequence simulation method and device based on flexible time boundary |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111369070B (en) | Multimode fusion photovoltaic power prediction method based on envelope clustering | |
| CN103488869A (en) | Wind power generation short-term load forecast method of least squares support vector machine | |
| CN112288164B (en) | Wind power combined prediction method considering spatial correlation and correcting numerical weather forecast | |
| CN107292502B (en) | Power distribution network reliability assessment method | |
| Zhang et al. | Wind speed prediction research considering wind speed ramp and residual distribution | |
| CN106649479A (en) | Probability graph-based transformer state association rule mining method | |
| CN116449218B (en) | Lithium battery health state estimation method | |
| CN114792156A (en) | Photovoltaic output power prediction method and system based on curve characteristic index clustering | |
| CN113313139A (en) | Wind power prediction uncertainty quantification method based on dynamic characteristics of unit | |
| CN115714420A (en) | Combined power station operation optimization method and system based on high-precision wind and light output prediction | |
| CN111461404A (en) | Short-term load and hydropower prediction method based on neural network prediction interval | |
| CN112819238A (en) | Short-term wind power prediction method based on chaotic chicken flock optimization algorithm | |
| CN115600729A (en) | Grid load prediction method considering multiple attributes | |
| CN117744501B (en) | Water network system regulation node optimal scheduling and decision-making method considering ecological flow | |
| CN113807019A (en) | MCMC wind power simulation method based on improved scene classification and coarse grain removal | |
| CN110807508A (en) | Bus peak load prediction method considering complex meteorological influence | |
| CN108694475B (en) | Short-time-scale photovoltaic cell power generation capacity prediction method based on hybrid model | |
| CN117060408B (en) | New energy power generation prediction method and system | |
| Zhu et al. | Wind power interval and point prediction model using neural network based multi-objective optimization | |
| CN113762591A (en) | Short-term electric quantity prediction method and system based on GRU and multi-core SVM counterstudy | |
| CN113033898A (en) | Electrical load prediction method and system based on K-means clustering and BI-LSTM neural network | |
| CN114266186B (en) | Air conditioner energy consumption data missing value filling method, equipment and medium | |
| CN115409317A (en) | Transformer area line loss detection method and device based on feature selection and machine learning | |
| CN115660425A (en) | Windage yaw flashover risk evaluation method, system, equipment and readable storage medium | |
| CN113902492A (en) | Time-sharing electricity price prediction method and system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |