CN105389625B - Active power distribution network ultra-short term load prediction method - Google Patents

Active power distribution network ultra-short term load prediction method Download PDF

Info

Publication number
CN105389625B
CN105389625B CN201510705770.0A CN201510705770A CN105389625B CN 105389625 B CN105389625 B CN 105389625B CN 201510705770 A CN201510705770 A CN 201510705770A CN 105389625 B CN105389625 B CN 105389625B
Authority
CN
China
Prior art keywords
value
data
curve
predicted
distribution network
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
Application number
CN201510705770.0A
Other languages
Chinese (zh)
Other versions
CN105389625A (en
Inventor
徐士华
王军锋
梁安韬
赵晖
吴建辉
吴孝彬
赵建勋
孙振业
许为钤
梁丽军
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.)
Fujian Automation Electric Power Technology Co ltd
Original Assignee
Fujian Automation Electric Power Technology Co ltd
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 Fujian Automation Electric Power Technology Co ltd filed Critical Fujian Automation Electric Power Technology Co ltd
Priority to CN201510705770.0A priority Critical patent/CN105389625B/en
Publication of CN105389625A publication Critical patent/CN105389625A/en
Application granted granted Critical
Publication of CN105389625B publication Critical patent/CN105389625B/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
    • 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—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
    • G06Q50/06—Electricity, gas 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

Abstract

The invention provides a method for forecasting the ultra-short term load of an active power distribution network, which comprises the following steps: 1. defining an optimal local shape similarity coefficient to express the similarity of two local shape similarity curve arrays in the active power distribution network; determining the optimal local shape similarity curve array which is most similar to the variation trend of the real-time curve array to be predicted according to the optimal local shape similarity coefficient; 2. obtaining a predicted value of the active power distribution network ultra-short-term load according to the optimal local similarity curve sequence and real-time data of the active power distribution network ultra-short-term load; 3. correcting the predicted value of the ultra-short-term load of the active power distribution network to obtain a corrected final load predicted value; therefore, the prediction of the microgrid ultra-short term load is completed. The prediction method provided by the invention has the advantages of good fitting degree on the continuity data, high accuracy and good quick tracking effect on the mutation data after the operation mode is changed.

Description

Active power distribution network ultra-short term load prediction method
Technical Field
The invention relates to the technical field of active power distribution network ultra-short-term load prediction, in particular to an active power distribution network ultra-short-term load prediction method.
Background
With the advance of the construction of intelligent distribution networks, active distribution networks supported by distributed power supplies, energy storage devices and micro-networks are developed vigorously. In order to realize defense control, optimal control and emergency control of the active power distribution network, the power generation and storage of the distributed power supply and the energy storage device are dynamically adjusted in advance according to load changes, and the operation mode of the power grid is changed, so that peak clipping and valley filling and power supply in an emergency state are changed.
The distribution network ultra-short term load prediction can predict the load data of the system in the future of 5-15 minutes, and has important significance on the safe operation control of the active distribution network.
The distribution network ultra-short-term load prediction is different from the main network load prediction, has the characteristics of prediction, needs to predict load data of each feeder switch besides the load data of a prediction system and an area, and provides a data source for real-time early warning and state evaluation of a distribution network and dynamic network reconstruction of the distribution network. And the operation mode of the distribution network is flexible and changeable, and if the change of the mode is not considered, the regularity of historical data cannot be ensured, so that the randomness and the error range of the prediction result are enlarged. Although some load prediction methods (such as an artificial neural network, an expert system, a gray prediction method and the like) with learning and self-adapting functions have certain tracking capability after the operation mode is changed, the transition time is long, and the tracking effect is poor.
Therefore, the research of the ultra-short-term load prediction method suitable for the large-scale data prediction characteristic of the distribution network effectively guides the early warning evaluation and dynamic reconstruction of the distribution network, and has important practical significance in realizing the safe operation control of the active distribution network and the like.
The prior art discloses a paper document 'ultrashort term load prediction method based on local shape similarity' -Luoyuan, Liweiwei, Hoohung (university of Hunan, institute of Electrical and information engineering, Changsha 410082); the patent mainly describes a technology for ultra-short term prediction of total load of a power transmission network, and the patent describes an ultra-short term load prediction technology of an active power distribution network, wherein the two prediction data have different orders of magnitude and differ by at least more than 3 orders of magnitude, namely, the data difference is about 1000 times, so that the reliability of the requirement of a paper document on an algorithm is higher.
The paper document proposes a method for calculating a load prediction value by linear extrapolation, and the patent also proposes a method for applying local linear extrapolation, but the document only uses a single point before the point to be predicted. In the patent, all points of the optimal local shape similarity curve are calculated so as to predict values, and the predicted values are weighted, wherein the weight is a similar arithmetic series variable weight method.
In addition, the thesis documents indicate a shape coefficient calculation method, and the patent also proposes a character similarity coefficient calculation method, but the shape similarity coefficient of the patent has no weight. And the patent provides an optimal local shape similarity coefficient method taking the maximum value of the local shape similarity coefficients of all samples as the prediction curve. The two technical schemes are different.
Disclosure of Invention
The technical problem to be solved by the invention is to provide a method for predicting the ultra-short-term load of the active power distribution network, so that the large-scale ultra-short-term load prediction problem in the active power distribution network is solved, the prediction accuracy is improved, and the tracking effect is good.
The invention is realized by the following steps: an active power distribution network ultra-short-term load prediction method comprises the following steps:
step 1, defining an optimal local shape similarity coefficient to express the similarity of two local shape similarity curve arrays in an active power distribution network; determining the optimal local shape similarity curve array which is most similar to the variation trend of the real-time curve array to be predicted according to the optimal local shape similarity coefficient;
step 2, obtaining a predicted value of the ultra-short-term load of the active power distribution network according to the optimal local shape similarity curve series and the real-time data of the ultra-short-term load of the active power distribution network;
step 3, correcting the predicted value of the ultra-short term load of the active power distribution network to obtain a corrected final load predicted value; therefore, the prediction of the microgrid ultra-short term load is completed.
Further, the step 1 specifically comprises: taking data of d points before a data point x at a moment to be predicted to form a curve sequence LxComprises the following steps:
Lx＝{lx-d,lx-d-1,…,lx-1}
is provided with LxnIs LxThe nth element, n ═ 1,2, …, d; lx-dThe smallest data value in the data d points before the data point x;
taking data of d points before a data point x at a moment to be predicted to form an initial historical optimal local shape similar curve sequence MixAnd selecting the ith day from the initial historical similar curve series MixComprises the following steps:
Mix＝{mix-d,mix-d-1,…,mix-1}
let MixnIs MixThe nth element, n ═ 1,2, …, d; m isix-dThe initial historical local shape similarity curve array of the data d before the data point x is taken as the minimum data value in the ith day;
the curve series LxNumber series M of curves similar to local shape of initial history of day iixHas a shape similarity coefficient of SixThe method specifically comprises the following steps:
Figure GDA0001963062900000031
wherein:
eixn＝lxn-mixn，
Figure GDA0001963062900000032
the maximum value of the shape similarity coefficients of all the initial historical local shape similarity curve series and the real-time curve series is defined as the optimal local shape similarity coefficient S of the real-time curve series at the momentxThe method specifically comprises the following steps:
Sx＝max(S1x,S2x,…,Six,…SNx)
wherein i is 1,2, …, N, N is the initial history similarity curve number;
the optimal local shape similarity coefficient SxCorresponding initial historical local shape similarity curve array MixI.e. the real-time curve sequence L to be predictedxThe optimal local similarity curve sequence.
Further, the step 2 of obtaining the predicted value of the ultra-short-term load of the active power distribution network specifically includes:
by using linear extrapolation, the data at the future time can be obtained by summing the value at the current time and a certain relative error, that is, the predicted value at the xth time can be expressed as:
Figure GDA0001963062900000033
wherein
Figure GDA0001963062900000034
For the predicted value at time x, e, obtained from the nth dataxnThe difference value of the nth data before the xth time is the x time;
due to MixAnd LxThe change trend of the optimal local shape similarity curve sequence is the same, and the local shape similarity curve sequence M can be obtained from the initial historyixIs approximately represented by the curve sequence LxE, thenxnExpressed as:
exn＝mix-mixn
knowing that d predicted values at the x moment are to be obtained, performing weighted average on the d data to obtain predicted value l 'of the ultra-short-term load of the active power distribution network at the moment'xComprises the following steps:
Figure GDA0001963062900000041
wherein alpha is the weight of each initial predicted value, all weights are set as an arithmetic progression, and the initial value is alpha1And the tolerance is k, then:
Figure GDA0001963062900000042
to determine alphanThe value of (c).
Further, the step 3 specifically includes: each moment of timeThe load predicted value forms an initial load predicted value number line l 'in a certain time period'xThe method specifically comprises the following steps:
L′x＝{l′x-d,l′x-d-1,…,l′x-1}
l 'is'xnIs l'xThe nth element, n ═ 1,2, …, d;
real time curve series LxAnd initial load predicted value line l'xThe error between: array HxSpecifically, the method comprises the following steps:
Hx＝Lx-L′x
let HxnIs HxThe nth element, n ═ 1,2, …, d;
array HxThe values of (A) are in three conditions, namely, zero is greater than or equal to zero, zero is less than or equal to zero, and positive and negative zero are alternated; for the situations of being more than or equal to zero and being less than or equal to zero, carrying out weighted average on all error values in local time, and taking the error as a correction error of a predicted value and a real-time value of the moment to be predicted, specifically:
Figure GDA0001963062900000043
wherein l "xPredicting a correction value for the load;
for the case that the sequence is between positive and negative zero, the error value H of the downstream adjacent point of the moment to be predicted, namely the x-1 point, needs to be obtainedxnIn the error sequence HxSearching error values with the same sign as the value in the middle sequence, stopping searching if the signs are different, and counting the searched error values as k; if the error array value is zero, sequentially shifting down one bit of adjacent data, i "xThe predicted correction value for the load is:
Figure GDA0001963062900000044
where n is 1,2, …, k.
The invention has the following advantages: the invention determines the similarity of the number series of two local similarity curves expressed by the local similarity coefficient; the method for forecasting the ultra-short-term load of the active power distribution network with the similarity of the optimal local shape based on the multi-level weighted average is realized, a self-adaptive ultra-short-term load forecasting and correcting method is provided by using the difference value between real-time data and an initial ultra-short-term load forecasting value, and a detailed calculation formula of the method is determined. The invention establishes a self-adaptive active power distribution network ultra-short term load prediction method, which not only has good fitting degree and high accuracy on continuity data, but also has good and quick tracking effect on mutation data after the operation mode is changed.
Drawings
FIG. 1 is a schematic flow chart of the method of the present invention.
Detailed Description
Referring to fig. 1, the method for predicting the ultra-short term load of the active distribution network of the present invention includes the following steps:
step 1, defining an optimal local shape similarity coefficient to express the similarity of two local shape similarity curve arrays in an active power distribution network; determining the optimal local shape similarity curve array which is most similar to the variation trend of the real-time curve array to be predicted according to the optimal local shape similarity coefficient;
the step 1 specifically comprises the following steps: taking data of d points before a data point x at a moment to be predicted to form a curve sequence LxComprises the following steps:
Lx＝{lx-d,lx-d-1,…,lx-1}
is provided with LxnIs LxThe nth element, n ═ 1,2, …, d; lx-dThe smallest data value in the data d points before the data point x;
taking data of d points before a data point x at a moment to be predicted to form an initial historical local shape similar curve sequence MixAnd selecting the ith day from the initial historical similar curve series MixComprises the following steps:
Mix＝{mix-d,mix-d-1,…,mix-1}
let MixnIs MixThe nth element, n ═ 1,2, …, d; m isix-dThe initial historical optimal local shape similarity curve array of data d before the data point x is taken as the minimum data value in the ith day;
the curve series LxNumber series M of curves similar to local shape of initial history of day iixHas a shape similarity coefficient of SixThe method specifically comprises the following steps:
Figure GDA0001963062900000051
wherein:
eixn＝lxn-mixn，
Figure GDA0001963062900000061
the maximum value of the shape similarity coefficients of all the initial historical local shape similarity curve series and the real-time curve series is defined as the optimal local shape similarity coefficient S of the real-time curve series at the momentxThe method specifically comprises the following steps:
Sx＝max(S1x,S2x,…,Six,…SNx)
wherein i is 1,2, …, N, N is the initial history similarity curve number; (since S is used in the present inventionx＝max(S1x,S2x,…,Six,…SNx) The optimal local shape similarity coefficient S is obtained by the formulaxTherefore, the local shape similarity curve array is also optimal. )
The optimal local shape similarity coefficient SxCorresponding initial historical local shape similarity curve array MixI.e. the real-time curve sequence L to be predictedxThe optimal local similarity curve sequence.
Step 2, obtaining a predicted value of the ultra-short-term load of the active power distribution network according to the optimal local shape similarity curve series and the real-time data of the ultra-short-term load of the active power distribution network;
the step 2 of obtaining the predicted value of the ultra-short-term load of the active power distribution network specifically comprises the following steps:
by using linear extrapolation, the data at the future time can be obtained by summing the value at the current time and a certain relative error, that is, the predicted value at the xth time can be expressed as:
Figure GDA0001963062900000062
wherein
Figure GDA0001963062900000063
For the predicted value at time x, e, obtained from the nth dataxnThe difference value of the nth data before the xth time is the x time;
due to MixAnd LxThe change trend of the optimal local shape similarity curve sequence is the same, and the local shape similarity curve sequence M can be obtained from the initial historyixIs approximately represented by the curve sequence LxE, thenxnExpressed as:
exn＝mix-mixn
knowing that d predicted values at the x moment are to be obtained, performing weighted average on the d data to obtain predicted value l 'of the ultra-short-term load of the active power distribution network at the moment'xComprises the following steps:
Figure GDA0001963062900000064
wherein alpha is the weight of each initial predicted value, all weights are set as an arithmetic progression according to the principle of 'big-small-big-near', and the initial value is alpha1And the tolerance is k, then:
Figure GDA0001963062900000065
specifying an initial value of alpha1Alpha can be determined by determining the arithmetic progression of the whole weight with any one of the tolerances knThe value of (c).
Step 3, correcting the predicted value of the ultra-short term load of the active power distribution network to obtain a corrected final load predicted value; therefore, the prediction of the microgrid ultra-short term load is completed.
The step 3 specifically comprises the following steps: the load predicted value at each moment forms an initial load predicted value sequence l 'in a certain time period'xThe method specifically comprises the following steps:
L′x＝{l′x-d,l′x-d-1,…,l′x-1}
l 'is'xnIs l'xThe nth element, n ═ 1,2, …, d;
real time curve series LxAnd initial load predicted value line l'xThe error between: array HxSpecifically, the method comprises the following steps:
Hx＝Lx-L′x
let HxnIs HxThe nth element, n ═ 1,2, …, d;
array HxThe values of (A) are in three conditions, namely, zero is greater than or equal to zero, zero is less than or equal to zero, and positive and negative zero are alternated; for the situations of being greater than or equal to zero and being less than or equal to zero, a method of "far, small and near, may be used to perform weighted average on all error values in the local time, and the weighted average is used as a correction error of the predicted value and the real-time value at the time to be predicted, specifically:
Figure GDA0001963062900000071
wherein l "xPredicting a correction value for the load;
for the case that the sequence is between positive and negative zero, the error value H of the downstream adjacent point of the moment to be predicted, namely the x-1 point, needs to be obtainedxnIn the error sequence HxSearching error values with the same sign as the value in the middle sequence, stopping searching if the signs are different, and counting the searched error values as k; if the error array value is zero, sequentially shifting down one bit of adjacent data, i "xThe predicted correction value for the load is:
Figure GDA0001963062900000072
where n is 1,2, …, k.
The invention is further illustrated by the following specific examples in which:
firstly, generating a human body comfort index of each day according to historical meteorological data of a day to be predicted and 30 days before, and selecting 7 days which are most similar to the human body comfort index of the day to be predicted to form an initial historical similarity curve sequence M.
And secondly, performing data identification, completion correction and other processing on the initial historical similarity curve array M to obtain complete historical mature data.
Then, selecting d points before the time x to be predicted to form a real-time sequence LxAnd selecting the point d in the same time period in each calendar history similarity curve sequence M to form an initial history local similarity sequence.
Thirdly, obtaining and predicting a real-time sequence L according to the mode in the step 1xInitial history local similarity series M with most similar shape similarity coefficientix。
Then, obtaining an initial load predicted value l 'of the time to be predicted according to the mode in the step 2'x。
Finally, predicting the initial load according to the mode in the step 3 to obtain a predicted value l'xCorrecting to obtain the final load predicted value l after correction "x。
In a word, the invention establishes a self-adaptive active power distribution network ultra-short term load prediction method, which not only has good fitting degree and high accuracy on continuity data, but also has good and quick tracking effect on mutation data after the operation mode is changed.
The above description is only a preferred embodiment of the present invention, and all equivalent changes and modifications made in accordance with the claims of the present invention should be covered by the present invention.

Claims (2)

1. An active power distribution network ultra-short-term load prediction method is characterized by comprising the following steps: the method comprises the following steps:
step 1, obtaining data of d points before a data point x at a moment to be predicted to form a curve arrayLxComprises the following steps:
Lx＝{lx-d,lx-(d-1),…,lx-1}
is provided with LxnIs LxThe nth element, n ═ 1,2, …, d; lx-dThe smallest data value in the data d points before the data point x;
taking data of d points before a data point x at a moment to be predicted to form an initial historical local shape similar curve sequence MixAnd selecting the ith day from the initial historical similar curve series MixComprises the following steps:
Mix＝{mix-d,mix-(d-1),…,mix-1}
let MixnIs MixThe nth element, n ═ 1,2, …, d; m isix-dThe initial historical local shape similarity curve array of the data d before the data point x is taken as the minimum data value in the ith day;
series of curves LxSimilar curve array M with optimal local shape of initial history on ith dayixHas a shape similarity coefficient of SixThe method specifically comprises the following steps:
Figure FDA0002984121380000011
wherein:
eixn＝lxn-mixn，
Figure FDA0002984121380000012
the maximum value of the shape similarity coefficients of all the initial historical local shape similarity curve series and the real-time curve series is defined as the optimal local shape similarity coefficient S of the real-time curve series at the momentxThe method specifically comprises the following steps:
Sx＝max(S1x,S2x,…,Six,…SNx)
wherein i is 1,2, …, N, N is the initial history similarity curve number;
then the optimal local shapeSimilarity coefficient SxCorresponding initial historical local shape similarity curve array MixI.e. the real-time curve sequence L to be predictedxThe optimal local shape similarity curve sequence;
step 2, obtaining a predicted value of the ultra-short-term load of the active power distribution network according to the optimal local shape similarity curve series and the real-time data of the ultra-short-term load of the active power distribution network; the method comprises the following steps:
by using linear extrapolation, the data at the future time can be obtained by summing the value at the current time and a certain relative error, that is, the predicted value at the xth time can be expressed as:
Figure FDA0002984121380000021
wherein
Figure FDA0002984121380000022
For the predicted value at time x, e, obtained from the nth dataxnThe difference value of the nth data before the xth time is the x time;
due to MixAnd LxThe change trend of the optimal local shape similarity curve sequence is the same, and the local shape similarity curve sequence M can be obtained from the initial historyixIs approximately represented by the curve sequence LxE, thenxnExpressed as:
exn＝mix-mixn
knowing that d predicted values at the x moment are to be obtained, performing weighted average on the d data to obtain predicted value l 'of the ultra-short-term load of the active power distribution network at the moment'xComprises the following steps:
Figure FDA0002984121380000023
wherein alpha is the weight of each initial predicted value, all weights are set as an arithmetic progression, and the initial value is alpha1And the tolerance is k, then:
Figure FDA0002984121380000024
to determine alphanA value of (d);
step 3, correcting the predicted value of the ultra-short term load of the active power distribution network to obtain a corrected final load predicted value; therefore, the prediction of the microgrid ultra-short term load is completed.
2. The ultra-short term load prediction method for the active power distribution network according to claim 1, wherein: the step 3 specifically comprises the following steps: the load predicted value at each moment forms an initial load predicted value sequence l 'in a certain time period'xThe method specifically comprises the following steps:
L'x＝{l'x-d,l'x-d-1,…,l'x-1}
l 'is'xnIs l'xThe nth element, n ═ 1,2, …, d;
real time curve series LxAnd initial load predicted value line l'xThe error between: array HxSpecifically, the method comprises the following steps:
Hx＝Lx-L′x
let HxnIs HxThe nth element, n ═ 1,2, …, d;
array HxThe values of (A) are in three conditions, namely, zero is greater than or equal to zero, zero is less than or equal to zero, and positive and negative zero are alternated; for the situations of being more than or equal to zero and being less than or equal to zero, carrying out weighted average on all error values in local time, and taking the error as a correction error of a predicted value and a real-time value of the moment to be predicted, specifically:
Figure FDA0002984121380000031
wherein l "xPredicting a correction value for the load;
for the case that the sequence is between positive and negative zero, the error value H of the downstream adjacent point of the moment to be predicted, namely the x-1 point, needs to be obtainedxnIn the error sequence HxSearching error values with the same sign as the value in the middle sequence, stopping searching if the signs are different, and counting the searched error values as k; if the error array value is zero, sequentially shifting down one bit of adjacent data, i "xThe predicted correction value for the load is:
Figure FDA0002984121380000032
where q is 1,2, …, k.
CN201510705770.0A 2015-10-27 2015-10-27 Active power distribution network ultra-short term load prediction method Active CN105389625B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201510705770.0A CN105389625B (en) 2015-10-27 2015-10-27 Active power distribution network ultra-short term load prediction method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201510705770.0A CN105389625B (en) 2015-10-27 2015-10-27 Active power distribution network ultra-short term load prediction method

Publications (2)

Publication Number Publication Date
CN105389625A CN105389625A (en) 2016-03-09
CN105389625B true CN105389625B (en) 2021-05-28

Family

ID=55421891

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201510705770.0A Active CN105389625B (en) 2015-10-27 2015-10-27 Active power distribution network ultra-short term load prediction method

Country Status (1)

Country Link
CN (1) CN105389625B (en)

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106650986A (en) * 2016-09-13 2017-05-10 云南电网有限责任公司 Method and device for forecasting regional maximum loads of power distribution network
CN110533216A (en) * 2019-07-19 2019-12-03 国网辽宁省电力有限公司 Ultra-short term correction technique based on regulation cloud
CN110377596A (en) * 2019-07-29 2019-10-25 合肥阳光新能源科技有限公司 Data correcting method and system

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104008432A (en) * 2014-06-03 2014-08-27 华北电力大学 Micro-grid short-term load forecasting method based on EMD-KELM-EKF
CN104881706A (en) * 2014-12-31 2015-09-02 天津弘源慧能科技有限公司 Electrical power system short-term load forecasting method based on big data technology

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20030182250A1 (en) * 2002-03-19 2003-09-25 Mohammad Shihidehpour Technique for forecasting market pricing of electricity

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104008432A (en) * 2014-06-03 2014-08-27 华北电力大学 Micro-grid short-term load forecasting method based on EMD-KELM-EKF
CN104881706A (en) * 2014-12-31 2015-09-02 天津弘源慧能科技有限公司 Electrical power system short-term load forecasting method based on big data technology

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
一种实用化的配电网超短期负荷预测方法;闫冬;《电力系统自动化》;20011125;第45-48页 *

Also Published As

Publication number Publication date
CN105389625A (en) 2016-03-09

Similar Documents

Publication Publication Date Title
Lin et al. An improved moth-flame optimization algorithm for support vector machine prediction of photovoltaic power generation
Ye et al. Predicting electricity consumption in a building using an optimized back-propagation and Levenberg–Marquardt back-propagation neural network: Case study of a shopping mall in China
Ghaderi et al. Deep forecast: Deep learning-based spatio-temporal forecasting
Wang et al. Probabilistic wind power forecasting based on spiking neural network
JP5888640B2 (en) Photovoltaic power generation prediction apparatus, solar power generation prediction method, and solar power generation prediction program
CN103489038A (en) Photovoltaic ultra-short-term power prediction method based on LM-BP neural network
CN104636985A (en) Method for predicting radio disturbance of electric transmission line by using improved BP (back propagation) neural network
CN108549960A (en) A kind of 24 hours Methods of electric load forecasting
CN102270309A (en) Short-term electric load prediction method based on ensemble learning
CN104636801A (en) Transmission line audible noise prediction method based on BP neural network optimization
CN105023070A (en) Output power prediction method of photovoltaic system
CN105389625B (en) Active power distribution network ultra-short term load prediction method
CN105631532A (en) Power system load prediction method using fuzzy decision-based neural network model
CN105956682A (en) Short-period electricity price prediction method based on BP neural network and Markov chain
CN109636059A (en) Electric heating distribution transformer load forecasting method and device
Tian et al. A network traffic hybrid prediction model optimized by improved harmony search algorithm
Zeng et al. Short-term load forecasting of smart grid systems by combination of general regression neural network and least squares-support vector machine algorithm optimized by harmony search algorithm method
CN107301478A (en) A kind of cable run short-term load forecasting method
Yu et al. Improved Elman neural network short-term residents load forecasting considering human comfort index
Gao et al. Day-ahead dynamic thermal line rating forecasting and power transmission capacity calculation based on ForecastNet
CN117200352A (en) Photovoltaic power generation regulation and control method and system based on cloud edge fusion
Kown et al. Short term load forecasting based on BPL neural network with weather factors
Gomes et al. Time series forecasting with neural networks and choquet integral
CN106682760A (en) Wind power climbing prediction method
CN105243451B (en) Based on the similar microgrid very Short-Term Load Forecasting Method of optimal partial shape

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