CN105243451B - Based on the similar microgrid very Short-Term Load Forecasting Method of optimal partial shape - Google Patents
Based on the similar microgrid very Short-Term Load Forecasting Method of optimal partial shape Download PDFInfo
- Publication number
- CN105243451B CN105243451B CN201510708621.XA CN201510708621A CN105243451B CN 105243451 B CN105243451 B CN 105243451B CN 201510708621 A CN201510708621 A CN 201510708621A CN 105243451 B CN105243451 B CN 105243451B
- Authority
- CN
- China
- Prior art keywords
- numbers
- ordered series
- partial shape
- microgrid
- data
- 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.)
- Active
Links
Classifications
-
- 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
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P80/00—Climate change mitigation technologies for sector-wide applications
- Y02P80/10—Efficient use of energy, e.g. using compressed air or pressurized fluid as energy carrier
- Y02P80/14—District level solutions, i.e. local energy networks
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention provides a kind of based on the similar microgrid very Short-Term Load Forecasting Method of optimal partial shape, this method are as follows: 1, the similar microgrid ultra-short term model of one optimal partial shape of acquisition；2, from microgrid ultra-short term model, with the determination of optimal partial shape similarity factor and the most like optimal partial shape similar curves ordered series of numbers of real-time curve ordered series of numbers variation tendency to be predicted；3, the predicted value of microgrid super short period load is obtained according to the real time data of optimal partial shape similar curves ordered series of numbers and microgrid super short period load；The prediction technique of the invention is high to the good accuracy rate of continuity data fitting degree, has important practical significance to the large area ultra-short term in current microgrid.
Description
Technical field
The present invention relates to microgrid ultra-short term technical field, more particularly to one kind are similar based on optimal partial shape
Microgrid very Short-Term Load Forecasting Method.
Background technique
With the propulsion that intelligent network distribution is built, power supply, energy storage device, microgrid are that the active distribution network of support will in a distributed manner
It is greatly developed.It, need to be previously according to load variations for defence control, optimal control and the emergent control for realizing active distribution network
The hair storage of dynamic adjustment distributed generation resource and energy storage device changes grid operation mode peak load shifting and lower turn of state of emergency confession
Electricity.
Distribution ultra-short term can predict following -15 minutes 5 minutes load datas of system, to actively
The safety running control of power distribution network is of great significance.
Distribution ultra-short term is different from major network load prediction, there is its own feature, except forecasting system and region are negative
Outside lotus data, the load data of each feeder switch need to be predicted, be distribution real-time early warning and status assessment and distribution dynamic network
Reconstruct provides data source.And the distribution method of operation is flexible and changeable, if not considering the variation of mode, the regularity of historical data is by nothing
Method guarantees, the randomness of prediction result and error range is caused to expand.Although some loads with study and adaptation function
Prediction technique (such as artificial neural network, expert system and gray prediction method) have after changes of operating modes it is certain with
Track ability, but transition time is longer, and tracking effect is poor.
Therefore, a kind of very Short-Term Load Forecasting Method for being suitable for distribution large-scale data prediction characteristic is studied, is effectively referred to
Distribution Forewarn evaluation and dynamic restructuring are led, realizes that active distribution network safety running control etc. all has important practical significance.
Prior art discloses a paper document " being based on the similar very Short-Term Load Forecasting Method of partial shape " --- 1. sieve
Yunnan is raw, Li Weiwei, He Hongying (Hunan University electrically and School of Information Technology, Changsha 410082)；Power transmission network is mainly described
A kind of technology of overall load ultra-short term prediction, and this patent is the ultra-short term technology for describing microgrid, the two prediction
The order of magnitude of data is different, at least poor 3 or more the orders of magnitude, that is to say, that gap data is at 1000 times or so, so paper is literary
It offers higher to the reliability of algorithm requirement.
The paper document proposes the method with linear extrapolation calculated load predicted value, and this patent is it is also proposed that a kind of application office
Portion's linear extrapolation, but document is only calculated with the point list before point moment to be predicted.And this patent is the similar song of optimal partial shape
The all the points of line all calculate so as to predicted value, and are weighted processing to predicted value, and weight is a type arithmetic progression variable weight
Weighing method.The two technical solution is different.And paper document does not provide the adaptive load prediction amendment follower method of one kind
Enlightenment, and this patent gives the adaptive load prediction of one kind and follows amendment scheme.And give initial prediction and reality
Error ordered series of numbers between value；According to each value of error ordered series of numbers compared with 0, it is greater than zero, be equal to zero and less than zero, is set forth
Error correcting method and correction formula under different situations.This patent has also gathered the similar load prediction technology of optimal partial shape,
Application condition analytical technology, the synthetic load adaptive forecasting method of the one such as load follow correction technique.
Summary of the invention
The technical problem to be solved in the present invention is to provide a kind of based on the similar microgrid super short period load of optimal partial shape
Prediction technique solves the problems, such as large-scale ultra-short term in microgrid, improves the accuracy rate of prediction, tracking effect
It is good.
The present invention is implemented as follows: a kind of be based on the similar microgrid very Short-Term Load Forecasting Method of optimal partial shape, institute
The method of stating includes the following steps:
Step 1 obtains the similar microgrid ultra-short term model of an optimal partial shape；
Step 2, from microgrid ultra-short term model, with optimal partial shape similarity factor it is determining with it is to be predicted in real time
The most like optimal partial shape similar curves ordered series of numbers of curve ordered series of numbers variation tendency；
Step 3 obtains microgrid according to the real time data of optimal partial shape similar curves ordered series of numbers and microgrid super short period load
The predicted value of super short period load.
Further, the step 2 specifically: take the data composition curve ordered series of numbers L of d point before time data point x to be predictedx
Are as follows:
Lx={ lx-d,lx-d-1,…,lx-1}
If LxnFor LxNth elements, n=1,2 ..., d；lx-dFor the smallest data in the data of d point before data point x
Value；
D point data before time data point x to be predicted is taken to constitute initial history partial shape similar curves ordered series of numbers Mix, and initial
It is chosen i-th in history similar curves ordered series of numbers M, curve ordered series of numbers MixAre as follows:
Mix={ mix-d,mix-d-1,…,mix-1}
If MixnFor MixNth elements, n=1,2 ..., d；mix-dFor the initial history partial shape of d point data before data point x
Similar curves ordered series of numbers, and the curve ordered series of numbers takes a smallest data value in i-th day；
Then curve ordered series of numbers LxWith initial history partial shape similar curves ordered series of numbers M on the i-ththixShape similarity factor be Six, specifically
Are as follows:
Wherein:
The minimum value of the shape similarity factor of all initial history partial shape similar curves ordered series of numbers and real-time curve ordered series of numbers defines
For the moment optimal partial shape similarity factor S of the real-time curve ordered series of numbersx, specifically:
Sx=min (S1x,S2x,…,Six,…SNx)
Wherein i=1,2 ..., N, N are initial history similar curves number；
Then optimal partial shape similarity factor SxCorresponding initial history partial shape similar curves ordered series of numbers MixAs this is to be predicted
Real-time curve ordered series of numbers LxOptimal partial shape similar curves ordered series of numbers Mix’。
Further, the predicted value of microgrid super short period load is obtained in the step 3 specifically:
With linear extrapolation, then the data of future time instance can be obtained by the value at current time with certain relative error summation
, i.e. the predicted value at xth moment may be expressed as:
WhereinFor the x moment predicted value acquired with nth data, exnFor the difference at xth moment and its preceding nth data
Value；
Due to Mix' it is LxOptimal partial shape similar curves ordered series of numbers, then its variation tendency is identical, can be by Mix' variable quantity
Approximate representation curve ordered series of numbers LxVariable quantity, then exnIt indicates are as follows:
exn=mix’-mixn', mixn' it is Mix' nth elements；
From the above, it can be seen that the predicted value at x moment there will be d, then weighted average is done to the d data and obtain the final of the moment
Predicted value l'xAre as follows:
Wherein, αnFor the weight of each initial prediction, if all weights are arithmetic progression, initial value α1, tolerance k,
Then have:
To determine αnValue.
The present invention has the advantage that the present invention establishes the similar microgrid ultra-short term of the optimal shape of optimal partial
Model；It is proposed the similarity that the similar curve number column in two parts are stated with shape similarity factor；It realizes a kind of based on average weighted
The similar microgrid very Short-Term Load Forecasting Method of optimal partial shape, and the detailed calculation formula of this method has been determined, according to two songs
The real time data of the shape similarity factors of line number column and microgrid super short period load obtains the predicted value of microgrid super short period load.This hair
Bright prediction technique is high to the good accuracy rate of continuity data fitting degree, and tracking effect is good, super to the large area in current microgrid
Short-term load forecasting has important practical significance.
Detailed description of the invention
Fig. 1 is the method for the present invention flow diagram.
Specific embodiment
Refering to Figure 1, one kind of the invention is based on the similar microgrid very Short-Term Load Forecasting Method of optimal partial shape,
Described method includes following steps:
Step 1 obtains the similar microgrid ultra-short term model of an optimal partial shape；
Step 2, from microgrid ultra-short term model, with optimal partial shape similarity factor it is determining with it is to be predicted in real time
The most like optimal partial shape similar curves ordered series of numbers of curve ordered series of numbers variation tendency；
The step 2 specifically: take the data composition curve ordered series of numbers L of d point before time data point x to be predictedxAre as follows:
Lx={ lx-d,lx-d-1,…,lx-1}
If LxnFor LxNth elements, n=1,2 ..., d；lx-dFor the smallest data in the data of d point before data point x
Value；
D point data before time data point x to be predicted is taken to constitute initial history partial shape similar curves ordered series of numbers Mix, and initial
It is chosen i-th in history similar curves ordered series of numbers M, curve ordered series of numbers MixAre as follows:
Mix={ mix-d,mix-d-1,…,mix-1}
If MixnFor MixNth elements, n=1,2 ..., d；mix-dFor the initial history partial shape of d point data before data point x
Similar curves ordered series of numbers, and the curve ordered series of numbers takes a smallest data value in i-th day；
Then curve ordered series of numbers LxWith initial history partial shape similar curves ordered series of numbers M on the i-ththixShape similarity factor be Six, specifically
Are as follows:
Wherein:
The minimum value of the shape similarity factor of all initial history partial shape similar curves ordered series of numbers and real-time curve ordered series of numbers defines
For the moment optimal partial shape similarity factor S of the real-time curve ordered series of numbersx, specifically:
Sx=min (S1x,S2x,…,Six,…SNx)
Wherein i=1,2 ..., N, N are initial history similar curves number；(due to using S in the present inventionx=min (S1x,
S2x,…,Six,…SNx) formula, obtain optimal partial shape similarity factor Sx, therefore, which is also
Optimal.)
Then optimal partial shape similarity factor SxCorresponding initial history partial shape similar curves ordered series of numbers MixAs this is to be predicted
Real-time curve ordered series of numbers LxOptimal partial shape similar curves ordered series of numbers Mix’。
Step 3 obtains microgrid according to the real time data of optimal partial shape similar curves ordered series of numbers and microgrid super short period load
The predicted value of super short period load.
The predicted value of microgrid super short period load is obtained in the step 3 specifically:
With linear extrapolation, then the data of future time instance can be obtained by the value at current time with certain relative error summation
, i.e. the predicted value at xth moment may be expressed as:
WhereinFor the x moment predicted value acquired with nth data, exnFor the difference at xth moment and its preceding nth data
Value；
Due to Mix' it is LxOptimal partial shape similar curves ordered series of numbers, then its variation tendency is identical, can be by Mix' variable quantity
Approximate representation curve ordered series of numbers LxVariable quantity, then exnIt indicates are as follows:
exn=mix’-mixn', mixn' it is Mix' nth elements；
From the above, it can be seen that the predicted value at x moment there will be d, then weighted average is done to the d data and obtain the final of the moment
Predicted value l'xAre as follows:
Wherein, αnFor the weight of each initial prediction, according to the principle of " remote small close big ", if all weights are equal differences
Column, initial value α1, tolerance k then has:
Specified initial value α1With any one in tolerance k, that is, it can determine the arithmetic progression of entire weight, to determine αnValue.
In short, the present invention realizes one kind based on the similar microgrid ultra-short term side of average weighted optimal partial shape
Method, and the detailed calculation formula of this method has been determined, prediction technique of the invention is to the good accuracy rate of continuity data fitting degree
Height, tracking effect is good, has important practical significance to the large area ultra-short term in current microgrid.
The foregoing is merely presently preferred embodiments of the present invention, all equivalent changes done according to scope of the present invention patent with
Modification, is all covered by the present invention.
Claims (1)
1. one kind is based on the similar microgrid very Short-Term Load Forecasting Method of optimal partial shape, it is characterised in that: the method includes
Following steps:
Step 1 obtains the similar microgrid ultra-short term model of an optimal partial shape；The step 1 specifically include step 2 and
Step 3；
Step 2, from microgrid ultra-short term model, it is determining with real-time curve to be predicted with optimal partial shape similarity factor
The most like optimal partial shape similar curves ordered series of numbers of ordered series of numbers variation tendency；The step 2 specifically: take time data point to be predicted
The data of d point constitute curve ordered series of numbers L before xxAre as follows:
Lx={ lx-d,lx-d-1,…,lx-1}
If LxnFor LxNth elements, n=1,2 ..., d；lx-dFor a smallest data value in the data of d point before data point x；
D point data before time data point x to be predicted is taken to constitute initial history partial shape similar curves ordered series of numbers Mix, and in initial history
It is chosen i-th in similar curves ordered series of numbers M, curve ordered series of numbers MixAre as follows:
Mix={ mix-d,mix-d-1,…,mix-1}
If MixnFor MixNth elements, n=1,2 ..., d；mix-dInitial history partial shape for d point data before data point x is similar
Curve ordered series of numbers, and the curve ordered series of numbers takes a smallest data value in i-th day；
Then curve ordered series of numbers LxWith initial history partial shape similar curves ordered series of numbers M on the i-ththixShape similarity factor be Six, specifically:
Wherein:
eixn=lxn-mixn,
The minimum value of the shape similarity factor of all initial history partial shape similar curves ordered series of numbers and real-time curve ordered series of numbers is defined as this
The moment optimal partial shape similarity factor S of real-time curve ordered series of numbersx, specifically:
Sx=min (S1x,S2x,…,Six,…SNx)
Wherein i=1,2 ..., N, N are initial history similar curves number；
Then optimal partial shape similarity factor SxCorresponding initial history partial shape similar curves ordered series of numbers MixAs this is to be predicted in real time
Curve ordered series of numbers LxOptimal partial shape similar curves ordered series of numbers Mix'；
It is step 3, ultrashort to obtain microgrid according to the real time data of optimal partial shape similar curves ordered series of numbers and microgrid super short period load
The predicted value of phase load；
The predicted value of microgrid super short period load is obtained in the step 3 specifically:
With linear extrapolation, then the data of future time instance can be obtained by the value and certain relative error summation at current time,
That is the predicted value at xth moment may be expressed as:
WhereinFor the x moment predicted value acquired with nth data, exnFor the difference at xth moment and its preceding nth data；
Due to Mix' it is LxOptimal partial shape similar curves ordered series of numbers, then its variation tendency is identical, can be by Mix' variable quantity it is approximate
Indicate curve ordered series of numbers LxVariable quantity, then exnIt indicates are as follows:
exn=mix’-mixn', mixn' it is Mix' nth elements；
From the above, it can be seen that the predicted value at x moment there will be d, then the final prediction that weighted average obtains the moment is done to the d data
Value l'xAre as follows:
Wherein, αnFor the weight of each initial prediction, if all weights are arithmetic progression, initial value α1, tolerance k, then
Have:
To determine αnValue.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510708621.XA CN105243451B (en) | 2015-10-27 | 2015-10-27 | Based on the similar microgrid very Short-Term Load Forecasting Method of optimal partial shape |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510708621.XA CN105243451B (en) | 2015-10-27 | 2015-10-27 | Based on the similar microgrid very Short-Term Load Forecasting Method of optimal partial shape |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105243451A CN105243451A (en) | 2016-01-13 |
| CN105243451B true CN105243451B (en) | 2019-06-07 |
Family
ID=55041088
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510708621.XA Active CN105243451B (en) | 2015-10-27 | 2015-10-27 | Based on the similar microgrid very Short-Term Load Forecasting Method of optimal partial shape |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105243451B (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102930344A (en) * | 2012-10-09 | 2013-02-13 | 中国电力科学研究院 | Method for forecasting ultra-short term bus load based on load trend changes |
| CN103336891A (en) * | 2013-06-09 | 2013-10-02 | 广东电网公司佛山供电局 | Pseudo-measurement generation method applied to estimating condition of distribution network |
| CN103414173A (en) * | 2013-09-02 | 2013-11-27 | 国家电网公司 | Method for performing fault recovery on power distribution network based on ultra-short term load |
-
2015
- 2015-10-27 CN CN201510708621.XA patent/CN105243451B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102930344A (en) * | 2012-10-09 | 2013-02-13 | 中国电力科学研究院 | Method for forecasting ultra-short term bus load based on load trend changes |
| CN103336891A (en) * | 2013-06-09 | 2013-10-02 | 广东电网公司佛山供电局 | Pseudo-measurement generation method applied to estimating condition of distribution network |
| CN103414173A (en) * | 2013-09-02 | 2013-11-27 | 国家电网公司 | Method for performing fault recovery on power distribution network based on ultra-short term load |
Non-Patent Citations (1)
| Title |
|---|
| "基于局部形相似的超短期负荷预测方法";罗滇生;《电力系统及其自动化学报》;20080229;第20卷(第1期);75-79 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105243451A (en) | 2016-01-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Shi et al. | Direct interval forecast of uncertain wind power based on recurrent neural networks | |
| Yang et al. | Modelling a combined method based on ANFIS and neural network improved by DE algorithm: A case study for short-term electricity demand forecasting | |
| Chen et al. | Solar radiation forecast based on fuzzy logic and neural networks | |
| Vafaeipour et al. | Application of sliding window technique for prediction of wind velocity time series | |
| CN102102626B (en) | Method for forecasting short-term power in wind power station | |
| Rejc et al. | Short-term transmission-loss forecast for the slovenian transmission power system based on a fuzzy-logic decision approach | |
| Jain et al. | Analytical study of Wind power prediction system by using Feed Forward Neural Network | |
| Yin et al. | Relaxed deep learning for real-time economic generation dispatch and control with unified time scale | |
| Zhang et al. | A hybrid prediction model for forecasting wind energy resources | |
| Qiuyu et al. | Short-term load forecasting based on load decomposition and numerical weather forecast | |
| Zhang et al. | Short-term load forecasting for microgrids based on DA-SVM | |
| Salam et al. | Energy consumption prediction model with deep inception residual network inspiration and LSTM | |
| Peesapati et al. | GSA–FAPSO-based generators active power rescheduling for transmission congestion management | |
| Li et al. | A VVWBO-BVO-based GM (1, 1) and its parameter optimization by GRA-IGSA integration algorithm for annual power load forecasting | |
| Vlachogiannis et al. | Multi-objective based on parallel vector evaluated particle swarm optimization for optimal steady-state performance of power systems | |
| Mishra et al. | Application of neural networks in wind power (generation) prediction | |
| CN105389625B (en) | Active power distribution network ultra-short term load prediction method | |
| CN105243451B (en) | Based on the similar microgrid very Short-Term Load Forecasting Method of optimal partial shape | |
| Chi et al. | Comparison of two multi-step ahead forecasting mechanisms for wind speed based on machine learning models | |
| Li et al. | A double-stage hierarchical hybrid PSO-ANFIS model for short-term wind power forecasting | |
| Wang et al. | An ultra-short-term forecasting model for high-resolution solar irradiance based on SOM and deep learning algorithm | |
| Saravanan et al. | Generation scheduling with large-scale integration of renewable energy sources using grey wolf optimization | |
| Kerem et al. | Multi-step forward forecasting of electrical power generation in lignite-fired thermal power plant | |
| Ramırez-Rosado et al. | An advanced model for short-term forecasting of mean wind speed and wind electric power | |
| Karwade et al. | Review paper on load forecasting using neuro fuzzy system |
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 |