CN111191815A - Ultra-short-term output prediction method and system for wind power cluster - Google Patents

Ultra-short-term output prediction method and system for wind power cluster Download PDF

Info

Publication number
CN111191815A
CN111191815A CN201911168558.XA CN201911168558A CN111191815A CN 111191815 A CN111191815 A CN 111191815A CN 201911168558 A CN201911168558 A CN 201911168558A CN 111191815 A CN111191815 A CN 111191815A
Authority
CN
China
Prior art keywords
sub
time period
power
cluster
historical time
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.)
Granted
Application number
CN201911168558.XA
Other languages
Chinese (zh)
Other versions
CN111191815B (en
Inventor
梅生伟
刘宸宇
张雪敏
黄少伟
杨滢璇
刘锋
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Tsinghua University
State Grid Corp of China SGCC
China Electric Power Research Institute Co Ltd CEPRI
Original Assignee
Tsinghua University
State Grid Corp of China SGCC
China Electric Power Research Institute Co Ltd CEPRI
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Tsinghua University, State Grid Corp of China SGCC, China Electric Power Research Institute Co Ltd CEPRI filed Critical Tsinghua University
Priority to CN201911168558.XA priority Critical patent/CN111191815B/en
Publication of CN111191815A publication Critical patent/CN111191815A/en
Application granted granted Critical
Publication of CN111191815B publication Critical patent/CN111191815B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/29—Geographical information databases
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • G06N20/10—Machine learning using kernel methods, e.g. support vector machines [SVM]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/06—Energy or water supply
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
    • Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
    • Y04S10/00—Systems supporting electrical power generation, transmission or distribution
    • Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Business, Economics & Management (AREA)
  • Data Mining & Analysis (AREA)
  • Economics (AREA)
  • Software Systems (AREA)
  • Human Resources & Organizations (AREA)
  • General Engineering & Computer Science (AREA)
  • Strategic Management (AREA)
  • Mathematical Physics (AREA)
  • Databases & Information Systems (AREA)
  • Mathematical Optimization (AREA)
  • Pure & Applied Mathematics (AREA)
  • Health & Medical Sciences (AREA)
  • Mathematical Analysis (AREA)
  • Computational Mathematics (AREA)
  • General Business, Economics & Management (AREA)
  • Tourism & Hospitality (AREA)
  • Marketing (AREA)
  • Operations Research (AREA)
  • Game Theory and Decision Science (AREA)
  • Public Health (AREA)
  • Quality & Reliability (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Remote Sensing (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Medical Informatics (AREA)
  • Probability & Statistics with Applications (AREA)
  • Evolutionary Computation (AREA)
  • Artificial Intelligence (AREA)
  • Evolutionary Biology (AREA)
  • Development Economics (AREA)
  • Algebra (AREA)
  • Computing Systems (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Water Supply & Treatment (AREA)

Abstract

The embodiment of the invention provides a method and a system for ultra-short term output prediction of a wind power cluster, wherein the method comprises the following steps: performing sub-region division on a wind power plant of the wind power cluster to be predicted to obtain a sub-region division combination of the wind power cluster to be predicted; acquiring a power real value of each subarea in the subarea division combination in a first preset historical time period and a power real value of each subarea in a second preset historical time period; predicting the power of each sub-region of the first preset historical time period according to the real power value of the second preset historical time period to obtain a cluster power predicted value of the first preset historical time period in each division combination mode; and acquiring errors between the real power value of the first preset historical time period and the predicted cluster power value in each partition combination form, and taking the sub-region partition combination corresponding to the minimum error as the optimal sub-region partition combination to obtain the ultra-short term output prediction of the wind power cluster. According to the embodiment of the invention, the prediction accuracy of the ultra-short term output of the wind power cluster is improved.

Description

Ultra-short-term output prediction method and system for wind power cluster
Technical Field
The invention relates to the technical field of wind power data processing, in particular to an ultra-short-term output prediction method and system for a wind power cluster.
Background
The existing ultra-short-term prediction method of the wind power cluster is mainly an accumulative method, which is derived from power prediction of a single wind power plant.
The accumulation method is a very intuitive prediction method, the calculation principle of which is completely consistent with the wind power field prediction, but the method has defects in two aspects. Firstly, the wind power field has stronger volatility than a wind power cluster, the wind power cluster prediction precision based on the accumulation method is easily limited by the wind power field prediction precision, particularly, the prediction quality of a part of wind power fields is greatly influenced by the output data with stronger volatility due to sudden change of the weather process. And secondly, each wind power plant in the wind power cluster is independently opened by the accumulative method and is respectively predicted, the correlation of the output of the wind power plant in the cluster and the smoothness of regional wind power output are not considered, and original abundant data information in the wind power cluster is abandoned.
Therefore, a method and a system for predicting ultra-short term output of a wind power cluster are needed to solve the above problems.
Disclosure of Invention
Aiming at the problems in the prior art, the embodiment of the invention provides an ultra-short term output prediction method and system for a wind power cluster.
In a first aspect, an embodiment of the present invention provides a method for predicting an ultra-short term output of a wind power cluster, including:
the method comprises the steps that a wind power field of a wind power cluster to be predicted is divided into sub-regions, and all sub-region division combinations of the wind power cluster to be predicted are obtained;
acquiring a power real value of each subregion in all subregion partition combinations in a first preset historical time period and a power real value of each subregion in a second preset historical time period, wherein the first preset historical time period is a recent historical time period at the current moment, and the second preset historical time period is a recent historical time period of the first preset historical time period;
predicting the power of each sub-region of the first preset historical time period according to the real power value of the second preset historical time period to obtain a cluster power predicted value of the first preset historical time period in various division combination forms;
and acquiring errors between the real power value of the first preset historical time period and the cluster power predicted values in various partition combination forms, and taking the sub-region partition combination corresponding to the minimum error as an optimal sub-region partition combination to obtain the ultra-short term output predicted value of the wind power cluster to be predicted according to the optimal sub-region partition combination.
Further, before predicting the power of each sub-region of the first preset historical time period according to the real power value of the second preset historical time period to obtain a cluster power predicted value in various partition combination forms of the first preset historical time period, the method further includes:
acquiring a training data set, wherein the training data set comprises sample power data of each sub-region of a wind power cluster;
and training a support vector machine model through the training data set to obtain a fitting model function for predicting the output of each sub-region, so as to predict the power of each sub-region in the first preset historical time period through the fitting model function according to the real power value of the second preset historical time period, and obtain the cluster power predicted value in various division combination forms of the first preset historical time period.
Further, the acquiring the training data set specifically includes:
dividing the sample power data through a sliding time window to obtain sample power output characteristic data of the sample power data based on the sliding time window;
and constructing a training data set according to the sample power data and the sample power output characteristic data corresponding to the sample power data.
Further, after the obtaining of the error between the real power value of the first preset historical time period and the predicted cluster power value in various partition combination forms and taking the sub-region partition combination corresponding to the minimum error as the optimal sub-region partition combination, the method further includes:
acquiring an optimal subregion partition combination corresponding to a prediction time scale at each historical time point in the first preset historical time period;
according to the real power value and the fitting model function of each sub-area under the first historical moment point, acquiring the power predicted value of each sub-area at the future moment corresponding to each historical moment point;
and according to the power predicted value of each sub-region at the future moment, taking the optimal sub-region division combination corresponding to the predicted time scale at each historical moment as the division combination mode of the current predicted moment so as to obtain the ultra-short-term output predicted value of the wind power cluster to be predicted.
Further, the number of the historical time points in the first preset historical time period is not more than 16, and the time interval of each historical time point is 15 minutes.
Further, the sub-region division is performed on the wind power plant of the wind power cluster to be predicted to obtain all sub-region division combinations of the wind power cluster to be predicted, and the method comprises the following steps:
and on the basis of a Bell number formula, performing sub-region division on the wind power plant of the wind power cluster to be predicted to obtain all sub-region division combinations.
In a second aspect, an embodiment of the present invention provides an ultra-short term output prediction system for a wind power cluster, including:
the wind power cluster sub-region division module is used for performing sub-region division on a wind power plant of a wind power cluster to be predicted to obtain all sub-region division combinations of the wind power cluster to be predicted;
the device comprises an acquisition module and a control module, wherein the acquisition module is used for acquiring a power real value of each subregion in all subregion partition combinations in a first preset historical time period and a power real value in a second preset historical time period, the first preset historical time period is the latest historical time period at the current moment, and the second preset historical time period is the latest historical time period of the first preset historical time period;
the power prediction module is used for predicting the power of each sub-region of the first preset historical time period according to the real power value of the second preset historical time period to obtain a cluster power prediction value of the first preset historical time period in various division combination forms;
and the ultra-short-term output prediction module is used for acquiring errors between the real power value of the first preset historical time period and the cluster power prediction values in various partition combination forms, taking the sub-region partition combination corresponding to the minimum error as the optimal sub-region partition combination, and obtaining the ultra-short-term output prediction value of the wind power cluster to be predicted according to the optimal sub-region partition combination.
Further, the system further comprises:
the training data set construction module is used for acquiring a training data set, and the training data set comprises sample power data of each sub-region of the wind power cluster;
and the fitting model training module is used for training a support vector machine model through the training data set to obtain a fitting model function for predicting each sub-region, so as to predict the power of each sub-region in the first preset historical time period through the fitting model function according to the real power value in the second preset historical time period, and obtain the cluster power predicted value in various division and combination modes in the first preset historical time period.
In a third aspect, an embodiment of the present invention provides an electronic device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the steps of the method provided in the first aspect when executing the program.
In a fourth aspect, an embodiment of the present invention provides a non-transitory computer readable storage medium, on which a computer program is stored, which when executed by a processor, implements the steps of the method as provided in the first aspect.
According to the ultra-short term output prediction method and system for the wind power cluster, prediction modeling is performed based on the sub-region formed by the plurality of wind power plants, the correlation among the wind power plants and the characteristic property of regional output inside the wind power cluster are fully considered, the ultra-short term output of the wind power cluster is predicted, and the accuracy of output prediction is improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and those skilled in the art can also obtain other drawings according to the drawings without creative efforts.
Fig. 1 is a schematic flow chart of an ultra-short term output prediction method for a wind power cluster according to an embodiment of the present invention;
FIG. 2 is a timing diagram of the optimal partitioning according to an embodiment of the present invention;
FIG. 3 is a timing diagram illustrating the continuous partitioning under ultra-short term prediction according to an embodiment of the present invention;
fig. 4 is a schematic structural diagram of an ultra-short term output prediction system for a wind power cluster according to an embodiment of the present invention;
fig. 5 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
In the field of wind power output prediction, wind power output is predicted through three basic classification directions, namely a prediction method, a prediction object and a prediction time range. According to the angle of a prediction method, wind power prediction can be divided into physical model prediction and statistical model prediction, wherein the physical model prediction is that related weather information at a wind generating set in a region is obtained through iteration according to a partial differential equation set by utilizing the idea of hydrodynamics and through information such as wind speed, air pressure and temperature on a boundary in a certain region, so that future wind power output is obtained; and the statistical model prediction is to learn by using past historical data of a prediction object, to train a model by using a basic regression model or an artificial intelligence algorithm, to input historical output data of a period of time before the current moment in the real-time prediction, and to obtain a future prediction result in a data-driven mode. Due to the influence of terrain and data precision, accurate results are not easy to obtain in the physical model prediction, and in contrast, statistical model prediction often has higher precision and faster calculation speed on a short time scale.
According to the angle of a prediction object, wind power prediction can be divided into wind power cluster prediction, wind power plant prediction and wind power generation unit prediction, and the corresponding wind power installed capacity is gradually reduced. Because wind power output is influenced by spatial distribution, when the installed capacity is large, the fluctuation of the output is reduced due to the diversity of a space weather system, and a smooth effect is achieved. Therefore, the output of the wind power cluster tends to have a lower fluctuation level, and the prediction accuracy of the output tends to be higher than that of the output of the wind power cluster.
For the angle of the prediction time range, wind power prediction can be divided into ultra-short-term prediction, short-term prediction and medium-and-long-term prediction. According to the specification of the existing wind power prediction standard, ultra-short-term prediction refers to prediction of 0-4 hours in the future, the prediction interval is 15min, and a 16-step prediction result is output by one-time prediction; short term prediction is future 0-72 hours prediction; medium-long term prediction is a longer prediction. Because the weather system is a chaotic system, the prediction precision is correspondingly reduced along with the increase of the prediction time range. Therefore, the wind power prediction based on the ultra-short term has higher prediction precision, and the main purpose is to provide a wind power output basis for a real-time scheduling system.
Support vector machine regression is a regression method with e-insensitive loss. Compared with the traditional regression method, the support vector machine regression method greatly improves the sparsity of solutions, reduces the calculation difficulty, improves the calculation speed and can ensure the generalization capability of the model. The support vector machine adopted by the embodiment of the invention corresponds to the target function as follows:
Figure BDA0002288090500000061
s.t.|yi-(wTxi+b)|≤ε+ξi；
ξi≥0
wherein the content of the first and second substances,
Figure BDA0002288090500000062
representing an error term which only considers the case that the absolute value is larger than epsilon;
Figure BDA0002288090500000063
representing a regularization term for ensuring generalization capability of the prediction model; c represents the accepting or rejecting relation between two terms of the objective function, and w and b represent weight parameters of the objective function to be solved; e represents the size of an error band and is a parameter to be optimized; x is the number ofi，yiRepresenting the sample characteristics. According to the embodiment of the invention, based on the prediction model constructed by the regression of the support vector machine, the historical power of the sub-regions of the wind power cluster under various combinations in the wind power cluster is predicted, and the optimal sub-region division combinations at different historical moments are obtained by analyzing and judging, so that the ultra-short term output prediction is carried out on the wind power cluster according to the optimal sub-region division combinations.
Fig. 1 is a schematic flow diagram of a method for predicting ultra-short term output of a wind power cluster according to an embodiment of the present invention, and as shown in fig. 1, the embodiment of the present invention provides a method for predicting ultra-short term output of a wind power cluster, including:
101, performing sub-region division on a wind power plant of a wind power cluster to be predicted to obtain all sub-region division combinations of the wind power cluster to be predicted;
in the embodiment of the invention, the wind power plant is influenced by a terrain and a regional weather system, and regional characteristics with sub-region wind power output as a characteristic main body exist in a wind power cluster formed by the wind power plant. Because the weather system in a larger area has certain inertia, the total wind power output in the sub-area often has more regular output period characteristics, and the wind power output modeling prediction precision of the whole area is often better than the precision of the sum of individual predictions in the area. In consideration of the limitation of the actual wind power cluster construction and data communication, the embodiment of the invention takes the wind power plant as a basic unit, the sub-area is expressed in the form of a plurality of wind power plant combinations, and the wind power cluster is expressed in a plurality of sub-areasDue to the domain combination mode, the wind power field, the sub-regions and the wind power cluster form a three-level main body and form multiple division modes of the wind power cluster. Specifically, in the embodiment of the present invention, the sub-region c is definediIs formed by a wind farm jipThe composition, expressed as:
ci＝{ji1,ji2,...,jim},jip∈N；
wherein, ciDenotes the ith sub-region, jipRepresenting the p wind power plant of m wind power plants in the ith sub-region, wherein p is more than 0 and less than m; and N is a wind power cluster and indicates that N wind power plants exist in the wind power cluster. The wind power output of the sub-region is the sum of the outputs of all wind power plants in the sub-region, and is represented as:
Figure BDA0002288090500000071
wherein the content of the first and second substances,
Figure BDA0002288090500000072
represents a sub-region ciAnd wind power output at the moment t. In the embodiment of the invention, the sub-regions of the wind power cluster belong to a non-empty subset of the wind power cluster, and for the wind power cluster with N wind power plants, the number of the possibly formed sub-regions is 2N-1.
Further, in the embodiment of the invention, the sub-area division of the wind power cluster is defined. The wind power cluster is composed of sub-regions, the division of the wind power cluster is different combinations of the sub-regions, in one division combination, the sub-regions in the wind power cluster are not overlapped with each other, namely, each sub-region does not contain the same wind power field, and the wind power fields of all the sub-regions contain all the wind power fields of the wind power cluster N, and the division combination is expressed as follows:
Figure BDA0002288090500000073
Figure BDA0002288090500000074
Figure BDA0002288090500000075
wherein, CSDenotes the s seed region partition, csiRepresenting the ith sub-area in the division of the s-th sub-area of the wind power cluster, wherein the wind power output of the wind power cluster is the sum of the output of all the sub-areas in the wind power cluster, and the expression is as follows:
Figure BDA0002288090500000081
wherein, PtAnd representing the wind power output of the wind power cluster at the time t. Specifically, in the embodiment of the invention, the wind farm of the wind power cluster to be predicted is sub-divided based on the bell number formula, so that all sub-division combinations are obtained. As all the numbers of the wind power cluster division are one Bell number and the order divergence is presented, for the wind power cluster comprising N wind power plants, the Bell number formula is expressed as BNFor example, there are 3 wind power plants in a wind power cluster of a certain region, the division form of the wind power cluster can refer to the division form shown in FIG. 1,
TABLE 1
Cluster division numbering Cluster division form
1 {[1],[2],[3]}
2 {[1,2],[3]}
3 {[1],[2,3]}
4 {[1,3],[2]}
5 {[1,2,3]}
As can be seen from the cluster division modes in Table 1, the sub-regions in the wind power cluster are divided in various modes, and each division can sum up the processing predicted values of the sub-regions to obtain the overall wind power cluster predicted value. Since the selection of different sub-regions represents the use of different cluster-related characteristics, it is only possible to select the most appropriate cluster partitioning method to improve the accuracy. Therefore, the embodiment of the invention converts the problem of wind power output prediction into the selection of a proper cluster division mode. In the embodiment of the invention, based on the prediction algorithm of wind power cluster division, a prediction model f is constructed by exerting force on the sub-region in the wind power clusterciAnd then summing the wind power output prediction values of the sub-regions to obtain the output prediction of the wind power cluster, wherein the specific formula is as follows:
Figure BDA0002288090500000082
Figure BDA0002288090500000083
wherein the content of the first and second substances,
Figure BDA0002288090500000084
represents t0The predicted value of the wind power output of the ith sub-area at the kth moment in the future represents t0And (4) the predicted value of the wind power cluster of the moment wind power cluster at the kth moment in the future, wherein k represents the prediction time range.
According to the embodiment of the invention, the output characteristics of the sub-regions of the wind power cluster are fully considered, the periodicity of the rules, namely the region characteristics, is embodied based on the division of the sub-regions, the smooth processing effect is achieved through a larger space range of the sub-regions, and compared with the prediction of a single wind power plant, the fluctuation of the output is reduced, and the prediction precision is improved.
102, acquiring a power real value of each sub-region in all sub-region division combinations in a first preset historical time period and a power real value of each sub-region in a second preset historical time period, wherein the first preset historical time period is a latest historical time period at the current moment, and the second preset historical time period is a latest historical time period of the first preset historical time period;
103, predicting the power of each sub-region of the first preset historical time period according to the real power value of the second preset historical time period to obtain a cluster power predicted value of the first preset historical time period in various division combination forms;
and 104, acquiring errors between the real power value of the first preset historical time period and the cluster power predicted values in various partition combination forms, and taking the sub-region partition combination corresponding to the minimum error as an optimal sub-region partition combination to obtain the ultra-short-term output predicted value of the wind power cluster to be predicted according to the optimal sub-region partition combination.
According to the ultra-short term output prediction method for the wind power cluster, prediction modeling is carried out based on the sub-region formed by the plurality of wind power plants, the correlation among the wind power plants and the characteristic property of regional output inside the wind power cluster are fully considered, the ultra-short term output of the wind power cluster is predicted, and the accuracy of output prediction is improved.
On the basis of the above embodiment, before predicting the power of each sub-region in the first preset historical time period according to the real power value in the second preset historical time period to obtain the cluster power predicted value in the form of various division combinations in the first preset historical time period, the method further includes:
acquiring a training data set, wherein the training data set comprises sample power data of each sub-region of a wind power cluster;
and training a support vector machine model through the training data set to obtain a fitting model function for predicting the output of each sub-region, so as to predict the power of each sub-region in the first preset historical time period through the fitting model function according to the real power value of the second preset historical time period, and obtain the cluster power predicted value in various division combination forms of the first preset historical time period.
On the basis of the above embodiment, the acquiring the training data set specifically includes:
dividing the sample power data through a sliding time window to obtain sample power output characteristic data of the sample power data based on the sliding time window;
and constructing a training data set according to the sample power data and the sample power output characteristic data corresponding to the sample power data.
In the embodiment of the invention, firstly, a support vector machine model is trained by constructing a sample data set, and the specific steps are as follows:
step S10, obtaining sample power historical data of each wind power plant in the wind power cluster, and calculating sub-regions c in different division formsiSample power data of
Figure BDA0002288090500000101
Wherein the content of the first and second substances,
Figure BDA0002288090500000102
representing periods of model training
Figure BDA0002288090500000103
Time of day;
step S11, setting the length of the sliding time window as L, and training set by the sliding time window
Figure BDA0002288090500000104
Sub-area c of the momentiSample power data of
Figure BDA0002288090500000105
Dividing the sample data set after rolling to obtain
Figure BDA0002288090500000106
Samples of time of day
Figure BDA0002288090500000107
Wherein the content of the first and second substances,
Figure BDA0002288090500000108
the characteristic data representing the input to the sliding time window,
Figure BDA0002288090500000109
sample power characteristic data representing a sliding time window output;
step S12, the time is obtained by traversing the sample power historical data according to time and all the divisions of the sub-regions
Figure BDA00022880905000001010
In the range of 1 to T0All sub-regions of the sample in between, wherein sub-region ciIs represented as
Figure BDA00022880905000001011
Step S13, respectively inputting all possible sub-region sample data into a support vector machine model for training, thereby obtaining a trained model, namely obtaining a fitting model function for sub-region output prediction; wherein for sub-region ciSample data set of (2)
Figure BDA00022880905000001012
The obtained fitting model function is
Figure BDA00022880905000001013
In the embodiment of the invention, a data set of historical power characteristic data is constructed by using a sliding time window, the length of the sliding time window represents the number of characteristics of an input model, and the model outputs 16 points which represent 16 steps corresponding to 0-4h advanced ultra-short-term prediction. And (3) taking L pieces of output data before the t moment of the sub-region as the input of the model through the fitting model function to obtain a prediction result of 16 steps in the future at the t moment. And adding the prediction results of the corresponding prediction time ranges of the sub-regions in the cluster division to obtain the super-short-term prediction result of the wind power cluster.
On the basis of the above embodiment, after the obtaining of the error between the real power value of the first preset historical time period and the predicted cluster power value in various partition combination forms, and taking the sub-region partition combination corresponding to the minimum error as the optimal sub-region partition combination, the method further includes:
acquiring an optimal subregion partition combination corresponding to a prediction time scale at each historical time point in the first preset historical time period;
acquiring a power predicted value of each sub-region at a future moment corresponding to each historical moment according to the real power value and the fitting model function of each sub-region at the first preset historical time period;
and according to the power predicted value of each sub-region at the future moment, taking the optimal sub-region division combination corresponding to the predicted time scale at each historical moment as the division combination mode of the current predicted moment so as to obtain the ultra-short-term output predicted value of the wind power cluster to be predicted.
In the embodiment of the invention, because the relevance of the wind power plant in the wind power cluster is complex, the optimal cluster division mode is generally difficult to obtain directly through external conditions. The embodiment of the invention adopts a continuous division mode, calculates the optimal cluster division mode at the previous moment on the basis of the known actual cluster output, and takes the optimal division as the optimal division at the current moment. Since this method uses the previous time result as the current prediction, it is similar to the persistence method, and is called persistence partitioning. According to the method, the corresponding optimal division is basically consistent according to the characteristic that a weather system always keeps certain inertia and the correlation properties of wind power plants in the cluster corresponding to two time points which are close to each other.
In particular, in numerous setsAnd in the cluster division mode, taking a group of division modes with the minimum prediction error of the first preset historical time period as the optimal division of the current cluster. In wind power ultra-short-term cluster prediction, optimization on two scales exists in optimal division, namely sample optimization and prediction time range optimization. The sample optimization means that at each moment, the sample at each time point is subjected to calculation of optimal division once, and the optimal division at two adjacent time points may be different; the optimization of the prediction time range means that in the ultra-short term 16-step prediction, the prediction of each step corresponds to an optimal partition, and the partitions corresponding to different prediction time ranges at the same time point may be different. Therefore, at a time point, the optimization of 16 divisions will be performed
Figure BDA0002288090500000111
And the optimal division of k time ranges is predicted in advance when the wind power cluster is at the time t. Fig. 2 is a timing diagram of optimal partitioning according to an embodiment of the present invention, and as shown in fig. 2, the optimal sub-area partition obtained by predicting the first preset historical time period is used as the sub-area partition at this time, so as to be used for subsequent wind power cluster output prediction.
Further, the optimal partitioning manner provided by the above embodiment is a concept of a posterior calculation, that is, only after the real cluster contribution at the predicted time is known, the error can be calculated to obtain the optimal partitioning at the corresponding time. In order to meet the actual application requirements, the embodiment of the invention provides a continuous partition prediction method, and under the current time, optimal partition of a sub-region predicted by k steps ahead at a historical time t-k closest to the current time t is calculated
Figure BDA0002288090500000121
And optimally dividing the sub-region
Figure BDA0002288090500000122
Cluster partitioning used in a k-step-ahead prediction as the current time prediction t
Figure BDA0002288090500000123
It should be noted that, in the embodiment of the present invention, for different prediction time ranges k, so-called recent historical times are not the same, because the longer the prediction time range is, the longer the real cluster effort is required to be determined, so that the larger k is, the farther the corresponding historical time is from the current time, but the optimal division of the selected historical time sub-region is the same as the leading time k of the current cluster division. Fig. 3 is a timing diagram of the continuous partitioning under ultra-short term prediction according to an embodiment of the present invention, and reference is made to fig. 3, where each row represents a different prediction time range. In each row, green represents the historical time of the optimal partition in the known corresponding time range, and for the prediction of leading k steps, the historical time is cut off to t-k, namely, the persistence method is to use the optimal partition in the nearest known corresponding time range as the cluster partition of the prediction time range corresponding to the current time.
On the basis of the above embodiment, the number of the historical time points in the first preset historical time period is not more than 16, and the time interval of each historical time point is 15 minutes.
In the embodiment of the invention, the length of the advanced prediction can be preset according to the actual output prediction requirement of the wind power cluster, so that the sub-region division corresponding to the minimum error between the real power value and the predicted power value in the first preset historical time period is used as the optimal sub-region division combination by setting the number of historical time points of the latest historical time at the current moment, so as to predict the ultra-short-term output of the wind power cluster. For example, only the output of the wind power cluster in the future of 1 hour at the current moment needs to be predicted, and only the ultra-short term output prediction needs to be performed according to 4 historical time points of the recent historical time at the current moment, so that the output prediction value of the wind power cluster in the future of 1 hour at the current moment is obtained.
In another embodiment of the invention, a prediction model is established based on historical output of all possible sub-regions in the wind power cluster, and the part can be calculated and obtained at an off-line moment; in the online operation of the system, an optimal division algorithm is adopted to calculate the optimal division of the recent historical moment at the current moment; and finally, adopting a prediction method of continuous division, taking the optimal division of the latest historical moment as the division adopted by the prediction of the current moment, calculating the predicted power of each sub-area, and then adding the power of the sub-areas in the optimal division to obtain the predicted value of the cluster power. The method comprises the following specific steps:
step S20, obtaining sample power historical data of each wind power plant in the wind power cluster, and calculating sub-regions c in different division formsiSample power data of
Figure BDA0002288090500000131
Wherein the content of the first and second substances,
Figure BDA0002288090500000132
representing periods of model training
Figure BDA0002288090500000133
Time of day;
step S21, setting the length of the sliding time window as L, and training set by the sliding time window
Figure BDA0002288090500000134
Sub-area c of the momentiSample power data of
Figure BDA0002288090500000135
Dividing the sample data set after rolling to obtain
Figure BDA0002288090500000136
Samples of time of day
Figure BDA0002288090500000137
Wherein the content of the first and second substances,
Figure BDA0002288090500000138
the characteristic data representing the input to the sliding time window,
Figure BDA0002288090500000139
sample power characteristic data representing a sliding time window output;
step S22, the time is obtained by traversing the sample power historical data according to time and all the divisions of the sub-regions
Figure BDA00022880905000001310
In the range of 1 to T0All sub-regions of the sample in between, wherein sub-region ciIs represented as
Figure BDA00022880905000001311
Step S23, respectively inputting all possible sub-region sample data into a support vector machine model for training, thereby obtaining a trained model, namely obtaining a fitting model function for sub-region output prediction; wherein for sub-region ciSample data set of (2)
Figure BDA00022880905000001312
The obtained fitting model function is
Figure BDA00022880905000001313
Step S24, historical power data of the previous sliding time window (namely the second preset historical time period) of each sub-area in the last 16 historical time points (namely the historical time points in the first preset historical time period) of the current time T of the wind power cluster are obtained
Figure BDA00022880905000001314
Inputting the prediction range into a fitting model function, initializing the prediction range, setting the prediction range to be k ', setting a history time point T ' in a first preset history time period to be T-k ', and setting a sub-region number i to be 1;
step S25, obtaining the predicted power values of 16 time points of all sub-areas in the first preset historical time period through the historical power data of the second preset historical time period by fitting the model function
Figure BDA0002288090500000141
Step S26, obtaining the division C of the wind power cluster in the s seed regionSThe sum of the sub-region power predictors in each partition below,
Figure BDA0002288090500000142
after the sum of the sub-region power predicted values under the division combination is obtained, performing step S27;
step S27, the partition with the minimum error is obtained through the following formula, that is, the optimal sub-area partition combination, where the formula is:
Figure BDA0002288090500000143
step S28, the optimal sub-region partition combination of each historical time point in the first preset historical time period is used as the prediction partition corresponding to the future k 'time at the current time, that is, the optimal sub-region partition combination is used as the prediction partition corresponding to the future k' time at the current time
Figure BDA0002288090500000144
Step S29, according to the optimal subregion partition combination of each historical time point in the first preset historical time period, and the historical power data of the sliding time window before the current time T of the corresponding subregion
Figure BDA0002288090500000145
Calculating a predicted value of the power at a future time k' corresponding to the current time T of the sub-region, i.e.
Figure BDA0002288090500000146
Step S30, calculating the sum of the predicted power values of the wind power cluster at the future k 'moment according to the predicted power values at the future k' moment of the current moment T of the sub-region, wherein the formula is as follows:
Figure BDA0002288090500000147
step S31, obtaining the output prediction of the wind power cluster at the current time T in 0-4 hours in the future
Figure BDA0002288090500000148
In an embodiment of the invention, the method for predicting the ultra-short term output of the wind power cluster provided by the embodiment of the invention is compared with the existing method by predicting the wind power cluster data of a certain area. Specifically, two time points of spring and summer in a certain area are adopted, 1000 time points are selected as training samples and 500 time points are selected as testing samples for each time point. And taking 7 as the length L of the sliding time window, and setting the error root mean square error as an evaluation index according to the existing wind power marking regulation. Since the ultra-short term prediction is rolling prediction in multiple time ranges of 0-4h, the k time range prediction error root mean square E is givenRMSE,k forwardCombined prediction error root mean square E with 16 time rangesRMSE,k allThe specific definition is shown as the following formula:
Figure BDA0002288090500000151
Figure BDA0002288090500000152
for explaining the wind power cluster in spring, referring to table 2, the prediction errors of the prediction method provided by the embodiment of the invention in each time range and comprehensive range are smaller than those of the existing cumulative method, so that the prediction precision is improved.
TABLE 2
Existing addition method The method provided by the invention
ERMSE，1h forward/％ 8.98 7.92
ERMSE，2h forward/％ 11.78 10.17
ERMSE，3h forward/％ 13.52 11.81
ERMSE，4h forward/％ 14.73 12.85
ERMSE，4h all/％ 11.68 10.20
For the description of the summer wind power cluster, as shown in table 3, the prediction error of the prediction method provided by the embodiment of the invention in each time range and comprehensive range is smaller than that of the existing cumulative method, and the prediction precision is improved. Due to the fact that prediction accuracy in spring and summer is improved, generalization of the method provided by the embodiment of the invention is demonstrated, and the method can be more suitable for multiple event scenes.
TABLE 3
Existing addition method The method provided by the invention
ERMSE，1h forward/％ 12.56 10.62
ERMSE，2h forward/％ 13.76 12.48
ERMSE，3h forward/％ 14.24 12.60
ERMSE，4h forward/％ 15.03 13.05
ERMSE，4h all/％ 13.57 11.86
Fig. 4 is a schematic structural diagram of an ultra-short term output prediction system for a wind power cluster according to an embodiment of the present invention, and as shown in fig. 4, an embodiment of the present invention provides an ultra-short term output prediction system for a wind power cluster, including a wind power cluster sub-region partitioning module 401, an obtaining module 402, a power prediction module 403, and an ultra-short term output prediction module 404, where the wind power cluster sub-region partitioning module 401 is configured to perform sub-region partitioning on a wind farm of a wind power cluster to be predicted, so as to obtain all sub-region partition combinations of the wind power cluster to be predicted; the obtaining module 402 is configured to obtain a power real value of each sub-region in all sub-region division combinations in a first preset historical time period and a power real value in a second preset historical time period, where the first preset historical time period is a recent historical time period at a current moment, and the second preset historical time period is a recent historical time period of the first preset historical time period; the power prediction module 403 is configured to predict, according to the real power value of the second preset historical time period, the power of each sub-region of the first preset historical time period, so as to obtain a cluster power prediction value in various division combination forms of the first preset historical time period; the ultra-short-term output prediction module 404 is configured to obtain an error between the real power value of the first preset historical time period and the cluster power prediction value in each partition combination form, use a sub-region partition combination corresponding to the minimum error as an optimal sub-region partition combination, and obtain the ultra-short-term output prediction value of the wind power cluster to be predicted according to the optimal sub-region partition combination.
According to the ultra-short term output prediction system for the wind power cluster, prediction modeling is performed based on the sub-region formed by the plurality of wind power plants, the correlation among the wind power plants and the characteristic property of regional output inside the wind power cluster are fully considered, the ultra-short term output of the wind power cluster is predicted, and the accuracy of output prediction is improved.
On the basis of the above embodiment, the system further includes:
the training data set construction module is used for acquiring a training data set, and the training data set comprises sample power data of each sub-region of the wind power cluster;
and the fitting model training module is used for training a support vector machine model through the training data set to obtain a fitting model function for predicting the output of each sub-region, so that the power of each sub-region in the first preset historical time period is predicted through the fitting model function according to the real power value in the second preset historical time period, and the cluster power predicted value in each division form in the first preset historical time period is obtained.
The system provided by the embodiment of the present invention is used for executing the above method embodiments, and for details of the process and the details, reference is made to the above embodiments, which are not described herein again.
Fig. 5 is a schematic structural diagram of an electronic device according to an embodiment of the present invention, and referring to fig. 5, the electronic device may include: a processor (processor)501, a communication Interface (Communications Interface)502, a memory (memory)503, and a communication bus 504, wherein the processor 501, the communication Interface 502, and the memory 503 are configured to communicate with each other via the communication bus 504. The processor 501 may call logic instructions in the memory 503 to perform the following method: the method comprises the steps that a wind power field of a wind power cluster to be predicted is divided into sub-regions, and all sub-region division combinations of the wind power cluster to be predicted are obtained; acquiring a power real value of each subregion in all subregion partition combinations in a first preset historical time period and a power real value of each subregion in a second preset historical time period, wherein the first preset historical time period is a recent historical time period at the current moment, and the second preset historical time period is a recent historical time period of the first preset historical time period; predicting the power of each sub-region of the first preset historical time period according to the real power value of the second preset historical time period to obtain a cluster power predicted value of the first preset historical time period in various division combination forms; and acquiring errors between the real power value of the first preset historical time period and the cluster power predicted values in various partition combination forms, and taking the sub-region partition combination corresponding to the minimum error as an optimal sub-region partition combination to obtain the ultra-short term output predicted value of the wind power cluster to be predicted according to the optimal sub-region partition combination.
In addition, the logic instructions in the memory 503 may be implemented in the form of software functional units and stored in a computer readable storage medium when the logic instructions are sold or used as independent products. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
In another aspect, an embodiment of the present invention further provides a non-transitory computer-readable storage medium, on which a computer program is stored, where the computer program is implemented by a processor to perform the ultra-short term output prediction method for a wind power cluster provided in the foregoing embodiments, for example, the method includes: the method comprises the steps that a wind power field of a wind power cluster to be predicted is divided into sub-regions, and all sub-region division combinations of the wind power cluster to be predicted are obtained; acquiring a power real value of each subregion in all subregion partition combinations in a first preset historical time period and a power real value of each subregion in a second preset historical time period, wherein the first preset historical time period is a recent historical time period at the current moment, and the second preset historical time period is a recent historical time period of the first preset historical time period; predicting the power of each sub-region of the first preset historical time period according to the real power value of the second preset historical time period to obtain a cluster power predicted value of the first preset historical time period in various division combination forms; and acquiring errors between the real power value of the first preset historical time period and the cluster power predicted values in various partition combination forms, and taking the sub-region partition combination corresponding to the minimum error as an optimal sub-region partition combination to obtain the ultra-short term output predicted value of the wind power cluster to be predicted according to the optimal sub-region partition combination.
The above-described embodiments of the apparatus are merely illustrative, and the units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.
Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware. With this understanding in mind, the above-described technical solutions may be embodied in the form of a software product, which can be stored in a computer-readable storage medium such as ROM/RAM, magnetic disk, optical disk, etc., and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in the embodiments or some parts of the embodiments.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present 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 solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.

Claims (10)

1. An ultra-short term output prediction method for a wind power cluster is characterized by comprising the following steps:
the method comprises the steps that a wind power field of a wind power cluster to be predicted is divided into sub-regions, and all sub-region division combinations of the wind power cluster to be predicted are obtained;
acquiring a power real value of each subregion in all subregion partition combinations in a first preset historical time period and a power real value of each subregion in a second preset historical time period, wherein the first preset historical time period is a recent historical time period at the current moment, and the second preset historical time period is a recent historical time period of the first preset historical time period;
predicting the power of each sub-region of the first preset historical time period according to the real power value of the second preset historical time period to obtain a cluster power predicted value of the first preset historical time period in various division combination forms;
and acquiring errors between the real power value of the first preset historical time period and the cluster power predicted values in various partition combination forms, and taking the sub-region partition combination corresponding to the minimum error as an optimal sub-region partition combination to obtain the ultra-short term output predicted value of the wind power cluster to be predicted according to the optimal sub-region partition combination.
2. The ultra-short term output prediction method for the wind power cluster according to claim 1, wherein before the predicting the power of each sub-region of the first preset historical time period according to the real power value of the second preset historical time period to obtain the cluster power prediction value in various division combinations of the first preset historical time period, the method further comprises:
acquiring a training data set, wherein the training data set comprises sample power data of each sub-region of a wind power cluster;
and training a support vector machine model through the training data set to obtain a fitting model function for predicting the output of each sub-region, so as to predict the power of each sub-region in the first preset historical time period through the fitting model function according to the real power value of the second preset historical time period, and obtain the cluster power predicted value in various division combination forms of the first preset historical time period.
3. The ultra-short term output prediction method for a wind power cluster according to claim 2, wherein the obtaining of the training data set specifically comprises:
dividing the sample power data through a sliding time window to obtain sample power output characteristic data of the sample power data based on the sliding time window;
and constructing a training data set according to the sample power data and the sample power output characteristic data corresponding to the sample power data.
4. The ultra-short term output prediction method for the wind power cluster according to claim 2, wherein after the obtaining of the error between the real power value of the first preset historical time period and the predicted cluster power value in various partition combination forms, and taking the sub-region partition combination corresponding to the minimum error as the optimal sub-region partition combination, the method further comprises:
acquiring an optimal subregion partition combination corresponding to a prediction time scale at each historical time point in the first preset historical time period;
acquiring a power predicted value of each sub-region at a future moment corresponding to each historical moment according to the real power value and the fitting model function of each sub-region at the first preset historical time period;
and according to the power predicted value of each sub-region at the future moment, taking the optimal sub-region division combination corresponding to the predicted time scale at each historical moment as the division combination mode of the current predicted moment so as to obtain the ultra-short-term output predicted value of the wind power cluster to be predicted.
5. The ultra-short term output prediction method for a wind power cluster according to claim 4, wherein the number of historical time points in the first preset historical time period is not more than 16, and the time interval of each historical time point is 15 minutes.
6. The ultra-short term output prediction method for the wind power cluster according to claim 1, wherein the sub-region division is performed on the wind power plant of the wind power cluster to be predicted to obtain all sub-region division combinations of the wind power cluster to be predicted, and the method comprises the following steps:
and on the basis of a Bell number formula, performing sub-region division on the wind power plant of the wind power cluster to be predicted to obtain all sub-region division combinations.
7. An ultra-short term output prediction system for a wind power cluster, comprising:
the wind power cluster sub-region division module is used for performing sub-region division on a wind power plant of a wind power cluster to be predicted to obtain all sub-region division combinations of the wind power cluster to be predicted;
the device comprises an acquisition module and a control module, wherein the acquisition module is used for acquiring a power real value of each subregion in all subregion partition combinations in a first preset historical time period and a power real value in a second preset historical time period, the first preset historical time period is the latest historical time period at the current moment, and the second preset historical time period is the latest historical time period of the first preset historical time period;
the power prediction module is used for predicting the power of each sub-region of the first preset historical time period according to the real power value of the second preset historical time period to obtain a cluster power prediction value of the first preset historical time period in various division combination forms;
and the ultra-short-term output prediction module is used for acquiring errors between the real power value of the first preset historical time period and the cluster power prediction values in various partition combination forms, taking the sub-region partition combination corresponding to the minimum error as the optimal sub-region partition combination, and obtaining the ultra-short-term output prediction value of the wind power cluster to be predicted according to the optimal sub-region partition combination.
8. The ultra-short term output prediction system for a wind power cluster of claim 7, further comprising:
the training data set construction module is used for acquiring a training data set, and the training data set comprises sample power data of each sub-region of the wind power cluster;
and the fitting model training module is used for training a support vector machine model through the training data set to obtain a fitting model function for predicting the output of each sub-region, so that the power of each sub-region in the first preset historical time period is predicted through the fitting model function according to the real power value in the second preset historical time period, and the cluster power predicted value in various division and combination modes of the first preset historical time period is obtained.
9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor when executing the program implements the steps of the ultra-short term output prediction method for a wind power cluster as claimed in any one of claims 1 to 6.
10. A non-transitory computer readable storage medium, on which a computer program is stored, which, when being executed by a processor, performs the steps of the ultra-short term output prediction method for a wind power cluster according to any one of claims 1 to 6.
CN201911168558.XA 2019-11-25 2019-11-25 Ultra-short-term output prediction method and system for wind power cluster Active CN111191815B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201911168558.XA CN111191815B (en) 2019-11-25 2019-11-25 Ultra-short-term output prediction method and system for wind power cluster

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201911168558.XA CN111191815B (en) 2019-11-25 2019-11-25 Ultra-short-term output prediction method and system for wind power cluster

Publications (2)

Publication Number Publication Date
CN111191815A true CN111191815A (en) 2020-05-22
CN111191815B CN111191815B (en) 2022-08-16

Family

ID=70710955

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201911168558.XA Active CN111191815B (en) 2019-11-25 2019-11-25 Ultra-short-term output prediction method and system for wind power cluster

Country Status (1)

Country Link
CN (1) CN111191815B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114819374A (en) * 2022-05-10 2022-07-29 南京南瑞水利水电科技有限公司 Regional new energy ultra-short term power prediction method and system

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106875033A (en) * 2016-12-26 2017-06-20 华中科技大学 A kind of wind-powered electricity generation cluster power forecasting method based on dynamic self-adapting
CN108734359A (en) * 2018-06-08 2018-11-02 上海电机学院 A kind of wind power prediction data preprocessing method
US20180320658A1 (en) * 2017-05-03 2018-11-08 Uptake Technologies, Inc. Computer System & Method for Predicting an Abnormal Event at a Wind Turbine in a Cluster
CN108876039A (en) * 2018-06-21 2018-11-23 浙江工业大学 A kind of prediction technique of power quality containing distributed power distribution network based on support vector machines

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106875033A (en) * 2016-12-26 2017-06-20 华中科技大学 A kind of wind-powered electricity generation cluster power forecasting method based on dynamic self-adapting
US20180320658A1 (en) * 2017-05-03 2018-11-08 Uptake Technologies, Inc. Computer System & Method for Predicting an Abnormal Event at a Wind Turbine in a Cluster
CN108734359A (en) * 2018-06-08 2018-11-02 上海电机学院 A kind of wind power prediction data preprocessing method
CN108876039A (en) * 2018-06-21 2018-11-23 浙江工业大学 A kind of prediction technique of power quality containing distributed power distribution network based on support vector machines

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
CHENYU LIU ET AL.: ""Local-pattern-aware forecast of regional wind power: Adaptive partition and long-short-term matching"", 《ENERGY CONVERSION AND MANAGEMENT》 *
王勃等: "基于集群划分的短期风电功率预测方法", 《高电压技术》 *

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114819374A (en) * 2022-05-10 2022-07-29 南京南瑞水利水电科技有限公司 Regional new energy ultra-short term power prediction method and system

Also Published As

Publication number Publication date
CN111191815B (en) 2022-08-16

Similar Documents

Publication Publication Date Title
CN108564326B (en) Order prediction method and device, computer readable medium and logistics system
CN108764540B (en) Water supply network pressure prediction method based on parallel LSTM series DNN
CN110751318B (en) Ultra-short-term power load prediction method based on IPSO-LSTM
CN108764568B (en) Data prediction model tuning method and device based on LSTM network
CN111047085B (en) Hybrid vehicle working condition prediction method based on meta-learning
CN110942194A (en) Wind power prediction error interval evaluation method based on TCN
CN103778482B (en) Aquaculture dissolved oxygen short term prediction method based on multiscale analysis
CN103489039B (en) There is the freeway traffic flow amount fusion forecasting method of online self-tuning optimization ability
CN104091216A (en) Traffic information predication method based on fruit fly optimization least-squares support vector machine
JP2013074695A (en) Device, method and program for predicting photovoltaic generation
CN110197307B (en) Regional sea surface temperature prediction method combined with attention mechanism
Chen et al. Short-Term Wind Power Forecasting Using Gaussian Processes.
CN114362175B (en) Wind power prediction method and system based on depth certainty strategy gradient algorithm
Lv et al. Novel deterministic and probabilistic combined system based on deep learning and self-improved optimization algorithm for wind speed forecasting
CN107092987B (en) Method for predicting autonomous landing wind speed of small and medium-sized unmanned aerial vehicles
CN106384122A (en) Device fault mode identification method based on improved CS-LSSVM
CN109460874B (en) Sense wave height prediction method based on deep learning
CN109934422A (en) Neural network wind speed prediction method based on time series data analysis
CN110516889B (en) Load comprehensive prediction method based on Q-learning and related equipment
CN112215412A (en) Dissolved oxygen prediction method and device
CN111191815B (en) Ultra-short-term output prediction method and system for wind power cluster
CN115879369A (en) Coal mill fault early warning method based on optimized LightGBM algorithm
CN112836885B (en) Combined load prediction method, combined load prediction device, electronic equipment and storage medium
CN114239396A (en) Fan gear box state prediction method and system
Chi et al. Comparison of two multi-step ahead forecasting mechanisms for wind speed based on machine learning models

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
GR01 Patent grant
GR01 Patent grant