CN104680246B - A kind of wind power plant realtime power Forecasting Methodology based on data-driven - Google Patents
A kind of wind power plant realtime power Forecasting Methodology based on data-driven Download PDFInfo
- Publication number
- CN104680246B CN104680246B CN201510044845.5A CN201510044845A CN104680246B CN 104680246 B CN104680246 B CN 104680246B CN 201510044845 A CN201510044845 A CN 201510044845A CN 104680246 B CN104680246 B CN 104680246B
- Authority
- CN
- China
- Prior art keywords
- moment
- real
- recursion
- power plant
- wind power
- 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.)
- Expired - Fee Related
Links
- 238000000034 method Methods 0.000 title claims abstract description 62
- 238000004146 energy storage Methods 0.000 claims abstract description 28
- 238000005096 rolling process Methods 0.000 claims abstract description 21
- 238000005259 measurement Methods 0.000 claims abstract description 13
- 238000012937 correction Methods 0.000 claims abstract description 6
- 230000005611 electricity Effects 0.000 abstract description 18
- 238000010248 power generation Methods 0.000 description 5
- 230000005540 biological transmission Effects 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 238000000205 computational method Methods 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000005684 electric field Effects 0.000 description 2
- 230000000630 rising effect Effects 0.000 description 2
- 238000012935 Averaging Methods 0.000 description 1
- 241001269238 Data Species 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000011217 control strategy Methods 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000013277 forecasting method Methods 0.000 description 1
- 230000035772 mutation Effects 0.000 description 1
- 238000012797 qualification Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
Abstract
A kind of wind power plant realtime power Forecasting Methodology based on data-driven, this method comprise the following steps:S1：According to the actual measurement power output numerical value of wind power plant last moment, applied mathematics point slope form method calculates the estimate of the real-time power output of wind power plant subsequent time, and using the estimate as wind power plant the m moment real-time recursion magnitude of power；S2：The estimate of the real-time power output of wind power plant obtained to step S1 is modified using rolling average method, and the correction value of estimate is predicted performance number as the real-time recursion at wind power plant m moment；S3：Calculate the expectation generated output value of energy-storage system.The inventive method is dispatched wind electricity storage station method with traditional delay and contrasted afterwards, wind electricity storage station can be enable to quickly respond the demand of state's net scheduling, and this method is simple, it is easy to accomplish.
Description
Technical field
The invention belongs to wind electricity storage station technical field of power generation, and in particular to a kind of real-time work(of wind power plant based on data-driven
Rate Forecasting Methodology.
Background technology
National wind-light storage transmission demonstration project is located in Hebei province Zhangjiakou City Zhangbei County and Shangyi County, planning construction 500MW
The energy-accumulating power station of wind power plant, 100MW photo-voltaic power generation stations and respective volume.Due to the intrinsic randomness of wind, the light first time energy, ripple
For dynamic property with intermittent, the concentration that large-scale wind Generate, Generation, Generator volt generates electricity is grid-connected, certainly will be to the operation, scheduling and control of power system
System etc. brings lot of challenges.In the research for exploring solution route, extensive energy storage technology arises at the historic moment.It is external to carry out in succession
Multinomial demonstration project, establish improves fitful power controllability by energy storage technology, improves its grid-connected application power.
The content of the invention
The present invention operates in tracking generation schedule pattern for Zhangbei County's wind-light storage transmission demonstration project-wind storing cogeneration, in advance
The real-time power output of wind power plant is surveyed, and then calculates and controls the generating of energy-storage system.
In order to realize the above object the present invention is achieved through the following technical solutions：
First according to the actual measurement power output data of wind power plant, using slope variation Forecasting Methodology and rolling average method,
Prediction calculates the real-time output power value of wind power plant subsequent time；Then according to the dispatch command of national grid, energy storage system is calculated
The expectation power generation values of system, make wind storing cogeneration meet dispatching of power netwoks demand.
The computational methods of the real-time power output of wind power plant are predicted, are the actual measurement power output numbers using the wind power plant m-1 moment
Wind power plant m moment recursion power datas are obtained according to recursion is carried out, Fig. 1 is real time sequence figure.This method includes step in detail below:
S1：According to the actual measurement power output numerical value of wind power plant last moment, applied mathematics point slope form method calculates wind power plant
The estimate of the real-time power output of subsequent time, and using the estimate as wind power plant the m moment real-time recursion magnitude of power.
This step is to carry out recursion according to the actual measurement power output numerical value at m-2, m-1 moment using mathematics point slope form method to obtain
To the estimate of m moment real-time power output.
Shown in mathematical formulae such as formula (1), (2) according to the real-time recursion of mathematics point slope form
y'm=ym-1+km-1,m*Δt (1)
Wherein：Y ' m are the wind power plants obtained using slope recursion in the estimation power output at m moment, ym-1It is wind power plant m-
The measured data at 1 moment, KM-1, mIt is the slope between wind power plant m-1 moment and the measured data at m-2 moment, Δ t=1min.
Problems with occurs with the data of mathematics slope recurrence method prediction subsequent time：Firstly, because slope recursion
The method of prediction is that the measured data of recursion prediction subsequent time, institute are gone in the change of measured data and slope based on previous moment
It is pre- using this method recursion with when larger fluctuation occurs in the case that measured power data are constant in the trend that rises or falls
Error between the real time data and measured data that measure will be bigger；Secondly, when measured power data variation tendency by
When rising is changed into declining or is changed into rising from decline, the delay and mutation of recursion prediction data can be caused.If measured power number
According to change it is relatively steady slower when, the method for slope recursion prediction can be used to carry out the data of recursion actual measurement.
Because slope recursion Forecasting Methodology is an influence of the nearest data of consideration history to Future Data, institute is in this way
Precision of prediction will not be too high.But this method is fairly simple to be easy to apply in engineering practice, so needing to mathematics slope
The real time data estimate that recursion is predicted to obtain moves average treatment amendment.
S2：The estimate of the real-time power output of wind power plant obtained to step S1 recursion is repaiied using rolling average method
Just, the correction value of estimate is predicted performance number as the real-time recursion at wind power plant m moment.
Rolling average method is that the real time data for using slope recursion Forecasting Methodology to obtain in step S1 is entered at line slip
Reason.Due to slope recursion Forecasting Methodology precision of prediction the problem of can increase the fluctuation of wind power plant measured power data, so adopting
The fluctuation of wind power plant recursion prediction realtime power can also effectively be reduced with moving average method.
It is to repairing using the history real time data obtained after slope recursion prediction+rolling average using rolling average method
On the occasion of yiThe real-time estimate y' only obtained with the m moment by slope recursion Forecasting MethodologymSum-average arithmetic calculating is carried out, is obtained
As a result the real-time estimated data of mathematics slope recursion is used in step of replacing S1.The expression formula of rolling average method such as formula (3) institute
Show.
Wherein, N is using the number of the history real time data obtained after the prediction of slope recursion and rolling average, ym-i, i=
0,1 ..., N are using the correction value of the history real time data obtained after slope recursion prediction+rolling average, y'mIt is to pass through slope
The estimate of the real-time recursion data at the m moment that recursion Forecasting Methodology obtains.
It is higher than the prediction only with slope recursion Forecasting Methodology using the precision of prediction of slope recursion+moving average method
Precision.
S3：It is calculated according to generation schedule power and process step S1, S2 of the wind storing cogeneration power station at the m moment
The real-time power output at wind power plant m moment, calculate the expectation generated output value of energy-storage system.Expectation of the energy-storage system at the m moment
Generated output is to do algebraically phase by the generation schedule performance number in m moment power stations and the predicted value of the power output at wind power plant m moment
Subtract to obtain.
To extend the service life of energy-storage system, it is necessary to according to the depth of discharge of energy-storage system, rated power, specified appearance
Amount etc. influences the Real-time Feedback value of the constraints of energy-storage system service life, and the expectation generated output value to energy-storage system is carried out
Amendment in real time, and then avoid energy-storage system from occurring breaking through the phenomenon put.
The inventive method is dispatched wind electricity storage station method with traditional delay and contrasted afterwards, and wind electricity storage station can be enable fast
The demand of state's net scheduling is responded fastly, and this method is simple, it is easy to accomplish.
Brief description of the drawings
Fig. 1 real time sequence figures of the present invention.
The real-time the fundamentals of successive deduction figure of Fig. 2 mathematics point slope forms of the present invention.
Fig. 3 energy-storage system Generation Control strategy block diagrams of the present invention.
Fig. 4 is Zhangbei County's wind power plant in power output change on January 22 relatively more steady slowly curve and grid generation meter
The curve map drawn.
Fig. 5 be Zhangbei County's wind power plant January 9 output-power fluctuation change greatly relatively rapid curve and grid generation meter
The curve map drawn.
Fig. 6 is that the slope recursion+output of moving average method and slope recursion predicted method to wind power plant in Fig. 3 is respectively adopted
Power curve carries out the curve map that recursion is predicted to obtain.
Fig. 7 is that the slope recursion+output of moving average method and slope recursion predicted method to wind power plant in Fig. 5 is respectively adopted
Power curve carries out the curve map that recursion is predicted to obtain.
Fig. 8 is by controlling the generating of energy-storage system to make to generate electricity in the output power curve tracing figure 4 of wind storing cogeneration
The tracking effect figure of Plan Curve.
Fig. 9 is by controlling the generating of energy-storage system to make to generate electricity in the output power curve tracing figure 5 of wind storing cogeneration
The tracking effect figure of Plan Curve.
Embodiment
Below in conjunction with the accompanying drawings, the present invention is described in more detail.
The present invention provides the method and power station real-time tracking state net of a kind of real-time power output prediction of wind electricity storage station wind power plant
The computational methods that the energy-storage system of generation schedule generates electricity.In order to realize the adjustable of new energy wind/light/honourable electric field power output
Degree is, it is necessary to improve the power output of wind-powered electricity generation and photovoltaic generation using energy-storage system, with Zhangbei County's wind-light storage transmission demonstration project-energy storage
System is used to exemplified by real-time tracking generation schedule illustrate.
Extraction Zhangbei County's wind power plant January power output change relatively slowly and output-power fluctuation change greatly two
The actual measurement output power curve and generation schedulecurve of day is as shown in Figure 4,5.It can be seen that the actual measurement work(of wind power plant from Fig. 4,5
Deviation be present between rate curve and the generation schedulecurve of power network, it is therefore desirable to control the generating of energy-storage system made up or
Reduce deviation.Due to the limitation of the constraintss such as the rated capacity of energy-storage system and rated power, so wind storing cogeneration exists
The power curve for not catching up with generation schedule occurs in some periods.
Because wind electricity storage station is in actual motion, the real-time power output of wind-powered electricity generation can not be known in advance, so adopting respectively
With slope recursion and the real-time power output data of slope recursion+rolling average method prediction wind-powered electricity generation subsequent time.Fig. 6 is to figure
The real-time output power curve of 4 wind power plants is respectively adopted above two recursion and predicts what is obtained, and chain-dotted line therein is to use slope
Recursion+rolling average method predicts the obtained real-time output power curve of wind power plant；... line is pre- using slope recurrence method
The real-time output power curve of wind power plant measured；Solid line is the actual measurement output power curve of wind power plant.Fig. 7 is to Fig. 5 wind-powered electricity generations
Output power curve is respectively adopted above two recursion and predicts what is obtained in real time, chain-dotted line therein be using slope recursion+
Rolling average method predicts the obtained real-time output power curve of wind power plant；... line is to predict to obtain using slope recurrence method
The real-time output power curve of wind power plant；Solid line is the actual measurement output power curve of wind power plant.Comparison diagram 6, Fig. 7, which can be seen that, to be adopted
The obtained real-time output power curve of wind power plant and the actual measurement output work of wind power plant is predicted with slope recursion+rolling average method
Deviation between rate curve is smaller；The real-time output power curve of wind power plant and wind-powered electricity generation for predicting to obtain using slope recurrence method
Deviation between the actual measurement output power curve of field is bigger, and the deviation amplitude in Fig. 7 is greater than the deviation width in Fig. 6
Degree.Illustrate that either power output change relatively more steady slowly curve or output-power fluctuation change greatly relatively rapid song
Line can use slope recursion+rolling average method to carry out recursion prediction.
According to the generation schedule dispatch command of power network with using slope recursion+rolling average method recursion to predict obtained wind
The expectation power generation values of mathematic interpolation energy-storage system between the real-time power output of electric field, it is deep according to the discharge and recharge of energy-storage system
The constraintss such as degree, rated power, rated capacity correct the expectation power generation values of energy storage in real time.Fig. 8 is the tracking of wind storing cogeneration
The design sketch of generation schedulecurve in Fig. 2；Fig. 9 is the design sketch of generation schedulecurve in wind storing cogeneration tracing figure 5.Due to
Generating electricity for energy-storage system can be limited by constraintss such as the depth of discharge of energy-storage system, rated power, so wind storage connection
Closing the tracking effect to generate electricity can be affected as shown in Figure 9.
Table one be respectively adopted qualification rate more than 80% of the power output of the wind power plant that two methods are predicted to obtain and
Accuracy rate, when output power fluctuation of wind farm randomness is bigger as can be seen from Table I, using two kinds of power forecasting methods
Accuracy rate difference it is bigger.In the process of the present invention in slope recursion+shifting and averaging prediction method be practically applicable to very much wind power plant
The prediction of power output.
Table one
Claims (1)
1. a kind of wind power plant realtime power Forecasting Methodology based on data-driven, it is characterised in that this method comprises the following steps:
S1：According to the actual measurement power output numerical value of wind power plant last moment, it is next that applied mathematics point slope form method calculates wind power plant
The estimate of moment real-time power output, and using the estimate as wind power plant the m moment real-time recursion magnitude of power；It is described
Real-time recursion magnitude of power is calculated using following mathematical formulae：
y'm=ym-1+km-1,m*Δt (1)
Wherein：y′mIt is the wind power plant obtained using slope recursion in the estimation power output at m moment, ym-1It is the wind power plant m-1 moment
Measured data, KM-1, mIt is the slope between wind power plant m-1 moment and the measured data at m-2 moment, Δ t=1min；
S2：The estimate of the real-time power output of wind power plant obtained to step S1 recursion is modified using rolling average method,
The correction value of estimate is predicted performance number as the real-time recursion at wind power plant m moment；The rolling average method is to step
The real time data obtained in S1 using slope recursion Forecasting Methodology enters line slip processing, is to using oblique using rolling average method
The correction value y of the history real time data obtained after rate recursion prediction+rolling averageiOnly pass through slope recursion prediction side with the m moment
The real-time estimate y' that method obtainsmSum-average arithmetic calculating is carried out, mathematics slope recursion is used in obtained result step of replacing S1
Real-time estimated data；
Wherein, N is using the number of the history real time data obtained after the prediction of slope recursion and rolling average, ym-i, i=0,
1 ..., N are using the correction value of the history real time data obtained after slope recursion prediction+rolling average, y'mIt is to be passed by slope
Push away the estimate of the real-time recursion data at the m moment that Forecasting Methodology obtains；
S3：Calculate the expectation generated output value of energy-storage system；Wherein, according to wind storing cogeneration power station the m moment generating meter
The real-time power output at power and the wind power plant m moment being calculated by step S1, S2 is drawn, calculates the expectation hair of energy-storage system
Electrical power value；Expectation generated output of the energy-storage system at the m moment is the generation schedule performance number and wind power plant m by m moment power stations
The predicted value of the power output at moment does algebraically and subtracts each other to obtain.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510044845.5A CN104680246B (en) | 2015-01-29 | 2015-01-29 | A kind of wind power plant realtime power Forecasting Methodology based on data-driven |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510044845.5A CN104680246B (en) | 2015-01-29 | 2015-01-29 | A kind of wind power plant realtime power Forecasting Methodology based on data-driven |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104680246A CN104680246A (en) | 2015-06-03 |
| CN104680246B true CN104680246B (en) | 2017-12-12 |
Family
ID=53315255
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510044845.5A Expired - Fee Related CN104680246B (en) | 2015-01-29 | 2015-01-29 | A kind of wind power plant realtime power Forecasting Methodology based on data-driven |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104680246B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105552970A (en) * | 2016-02-25 | 2016-05-04 | 华北电力科学研究院有限责任公司 | Method and apparatus for improving accuracy of predicting power of wind power station |
| CN112103994B (en) * | 2020-08-25 | 2022-04-01 | 同济大学 | Layered coordination control method and device for wind-hydrogen coupling system based on MPC |
| CN113705862B (en) * | 2021-08-12 | 2024-02-13 | 内蒙古电力(集团)有限责任公司电力调度控制分公司 | Ultra-short-term new energy prediction data correction method in electric power spot market environment |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102521671A (en) * | 2011-11-29 | 2012-06-27 | 华北电力大学 | Ultrashort-term wind power prediction method |
| CN103927597A (en) * | 2014-04-22 | 2014-07-16 | 国家电网公司 | Ultra-short-term wind power prediction method based on autoregression moving average model |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140195159A1 (en) * | 2013-01-09 | 2014-07-10 | Iteris, Inc. | Application of artificial intelligence techniques and statistical ensembling to forecast power output of a wind energy facility |
-
2015
- 2015-01-29 CN CN201510044845.5A patent/CN104680246B/en not_active Expired - Fee Related
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102521671A (en) * | 2011-11-29 | 2012-06-27 | 华北电力大学 | Ultrashort-term wind power prediction method |
| CN103927597A (en) * | 2014-04-22 | 2014-07-16 | 国家电网公司 | Ultra-short-term wind power prediction method based on autoregression moving average model |
Non-Patent Citations (2)
| Title |
|---|
| A review of wind power and wind speed forecasting methods with different time horizons;SS Soman 等;《North American Power Symposium》;20100928;第1-8页 * |
| 基于时间序列的风电功率日前预测模型及其应用;唐波 等;《水电能源科学》;20141125;第32卷(第11期);第193-196页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104680246A (en) | 2015-06-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109103929B (en) | Power distribution network economic optimization scheduling method based on improved dynamic kriging model | |
| Ding et al. | A forecast-driven decision-making model for long-term operation of a hydro-wind-photovoltaic hybrid system | |
| Biegel et al. | Distributed low-complexity controller for wind power plant in derated operation | |
| Fu et al. | Research on joint optimal dispatching method for hybrid power system considering system security | |
| CN104993522A (en) | Active power distribution network multi-time scale coordinated optimization scheduling method based on MPC | |
| CN107317361B (en) | active power distribution network global optimization scheduling method considering regional autonomous capacity | |
| CN102419394B (en) | Wind/solar power prediction method with variable prediction resolution | |
| CN112583017B (en) | Hybrid micro-grid energy distribution method and system considering energy storage operation constraint | |
| CN104616069A (en) | Annual power generation plan rolled decomposition optimization method taking balance between plan finishing rate and load rate into consideration | |
| CN105048499A (en) | Wind power integration real-time scheduling method and device based on model prediction and control | |
| CN104680246B (en) | A kind of wind power plant realtime power Forecasting Methodology based on data-driven | |
| CN104124685A (en) | Sample fan method based wind power plant theoretical power computing method | |
| Qudaih et al. | Model predictive based load frequency control design in the presence of DFIG wind turbine | |
| CN103986193B (en) | A kind of method that maximum wind grid connection capacity obtains | |
| Li et al. | Optimized operation of hybrid system integrated with MHP, PV and PHS considering generation/load similarity | |
| Dieulot et al. | Economic supervisory predictive control of a hybrid power generation plant | |
| Li et al. | Optimal dispatching method for smoothing power fluctuations of the wind-photovoltaic-battery hybrid generation system | |
| CN105305463B (en) | The idle work optimization method based on probabilistic loadflow of meter and photovoltaic generation and harmonic pollution | |
| Tungadio et al. | Energy flow estimation-control of two interconnected microgrids | |
| Ma et al. | Coordinated control of micro-grid based on distributed moving horizon control | |
| CN104657786A (en) | Short-term wind power mixed predicting method based on Boosting algorithm | |
| CN105576712B (en) | A kind of active power of wind power field On-Line Control Method based on adaptive pole configuration | |
| Yang et al. | Assessing the dispatch flexibility of coordinated solar and hydro generation | |
| CN104184165A (en) | Cooperative control method applicable to photovoltaic generator set without main control center | |
| Alarfaj et al. | A controlled load estimator-based energy management system for water pumping systems |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20171212 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |