CN107453346A - A kind of load curve Time segments division method suitable for power distribution network reconfiguration - Google Patents

A kind of load curve Time segments division method suitable for power distribution network reconfiguration Download PDF

Info

Publication number
CN107453346A
CN107453346A CN201610369284.0A CN201610369284A CN107453346A CN 107453346 A CN107453346 A CN 107453346A CN 201610369284 A CN201610369284 A CN 201610369284A CN 107453346 A CN107453346 A CN 107453346A
Authority
CN
China
Prior art keywords
mtd
mrow
msub
load
mtr
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
CN201610369284.0A
Other languages
Chinese (zh)
Other versions
CN107453346B (en
Inventor
宋晓辉
李建芳
高菲
张瑜
常松
赵珊珊
唐巍
王雨婷
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.)
Electric Power Research Institute of State Grid Chongqing Electric Power Co Ltd
State Grid Corp of China SGCC
China Electric Power Research Institute Co Ltd CEPRI
Original Assignee
Electric Power Research Institute of State Grid Chongqing Electric Power Co Ltd
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 Electric Power Research Institute of State Grid Chongqing Electric Power Co Ltd, State Grid Corp of China SGCC, China Electric Power Research Institute Co Ltd CEPRI filed Critical Electric Power Research Institute of State Grid Chongqing Electric Power Co Ltd
Priority to CN201610369284.0A priority Critical patent/CN107453346B/en
Publication of CN107453346A publication Critical patent/CN107453346A/en
Application granted granted Critical
Publication of CN107453346B publication Critical patent/CN107453346B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J3/00—Circuit arrangements for ac mains or ac distribution networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/30—Circuit design
    • G06F30/36—Circuit design at the analogue level
    • G06F30/367—Design verification, e.g. using simulation, simulation program with integrated circuit emphasis [SPICE], direct methods or relaxation methods
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J2203/00—Indexing scheme relating to details of circuit arrangements for AC mains or AC distribution networks
    • H02J2203/20—Simulating, e g planning, reliability check, modelling or computer assisted design [CAD]
    • 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
    • Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
    • Y02E60/00—Enabling technologies; Technologies with a potential or indirect contribution to GHG emissions mitigation

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Hardware Design (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Power Engineering (AREA)
  • Microelectronics & Electronic Packaging (AREA)
  • Evolutionary Computation (AREA)
  • Geometry (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Supply And Distribution Of Alternating Current (AREA)

Abstract

The present invention provides a kind of load curve Time segments division method suitable for power distribution network reconfiguration, and this method includes load standardization；Divide the load curve period；The duty value of calculated load curve period；Computation schema approach degree；Determine the equivalent mode of load curve.Technical scheme provided by the invention carries out Time segments division using the method for fuzzy clustering and pattern-recognition to load curve, complete and the equivalence of complex electric network load data is handled, it has been effectively compressed load data amount, the switch motion time of power distribution network dynamic restructuring is rationally determined, the economy and reliability of distribution system operation can be improved.

Description

A kind of load curve Time segments division method suitable for power distribution network reconfiguration
Technical field
The present invention relates to Load Characteristic Analysis method, during in particular to a kind of load curve suitable for power distribution network reconfiguration Section division methods.
Background technology
Power distribution network in the process of running, always load is continually changing with the time, because load variations have randomness, is matched somebody with somebody There can not possibly be identical two kinds of running statuses during operation of power networks, if analyzing each running status respectively, meter Calculation amount is very huge；On the other hand, because the work and rest mode and consumption habit of people are more stable, the load variations in power distribution network It is regular, there is certain correlation and reappearance, so according to the characteristic of load curve in power distribution network, will a period of time The fluctuation of section (1d) internal loading is respectively divided in different time sections at the time of larger, and the load fluctuation less moment is included into together In one period, the load level in the same period is approached, realize the division to load curve, this can simplify power distribution network Running optimizatin difficulty, amount of calculation is reduced, and can truly reflects the running status of power distribution network.
In view of the power distribution network dynamic restructuring of load fluctuation, by adjusting the operable switch state in network, to be changed Become the purpose that network structure realizes the method for operation of control power distribution network.The switch of electrical equipment in frequent folding power distribution network can be tight Shorten its electrical endurance again, reduce the reliability of device action, increase renewal of the equipment frequency, improve power distribution network operation and maintenance cost And switching manipulation cost, reduce network economy and reliability.By switch motion number control within the specific limits, it is necessary to The factor such as situation and current weather temperature is integrally fluctuated with reference to actual network structure, load under different time scales, simultaneously According to related operation maintenance rule and artificial regulation etc., pair determine that the number of switch motion makes accurate limitation respectively, therefore Need to carry out reasonable time period division to load curve according to fluctuation tendency, it is believed that on off state keeps constant, network in the period Structure does not change.
On this basis, dividing the method for load curve is, daily daily load curve is divided into some periods, with warp The load value for crossing equivalent processing replaces each period internal loading, and to determine actuation time that power distribution network switchs, protection switch is electric Life-span, reduce the maintenance cost of distribution system operation, lift the economy of power distribution system secure degree of reliability sum, create good Economic benefit and social benefit, have most important theories value and realistic meaning.
The present inventor has found through substantial amounts of research, is segmented this field in load curve in recent years and discloses；" during power distribution network Become reactive Voltage Optimum method " (《Automation of Electric Systems》, 2005), according to the variation tendency of load curve point in this method Section, is first according to monotonicity and is tentatively segmented, and then preliminary waypoint is merged successively using the thought of fusion until segmentation Number meets the switching number constraint of equipment." Algorithm for Time-varying Reactive Power Optimization " (《Power System and its Automation journal》, 2007), think that load dividing is a kind of discrete optimization problem in the algorithm, it is disclosed that using genetic algorithm to load Curve carry out intelligent segmental, determine each section rise, only time point, to solve number of equipment action restricted problem." power system Multi-period idle work optimization research " (XI AN JIAOTONG UNIVERSITY Subject Index, 2008), wherein it is bent that Principle of Statistics is applied into load sequence On line, using extreme difference and standard deviation describe load curve fluctuating change analyze the sequence signature and carry out the load period draw Point, it is proposed that one kind inspires iterative segmentation method, solves maximum extreme difference and maximum standard deviation problems of value.And 2010 It is disclosed herein in " the control strategy research of idle work optimization " one after being segmented in the case of ensureing that system node voltage is qualified, then Compare segments and equipment regulation number, if segments is more than equipment regulation number, take segments as final result；Otherwise Increase segments, until proximity device adjusts number.
Theoretically, these prior arts are classified just with clustering method to load curve, extract typical day Load curve, and Time segments division can not be carried out to load curve；When calculating typical load curve, simply obtain per a kind of all Curve in the average value of each point, the load in the period is carried out it is equivalent, to realize effective utilization to load data.
To meet the needs of prior art, the present invention is provided one kind and classified with fuzzy clustering method to meeting, is applicable In the load curve Time segments division method of power distribution network reconfiguration.
The content of the invention
To meet the needs in prior art, the present invention provides a kind of method that can effectively divide the load curve period, had Effect have compressed load data amount.
Load curve Time segments division method provided by the invention suitable for power distribution network reconfiguration, it is theed improvement is that, institute The method of stating includes：
Step 1, load standardization；
Step 2, the Preliminary division of load curve period；
Step 3, the duty value of calculated load curve period；
Step 4, computation schema approach degree；
Step 5, the equivalent mode for determining load curve.
Further, the step 1 includes：
(1-1) seeks load perunit value：Load perunit value a of the i-th line road in tI, tAs shown in following formula (1)：
In formula, PI, t：I-th line road is in the head end load value of t, i=1,2 ... n；PI, max：The whole year on i-th line road Peak load value；
(1-2) construction perunit meets matrix：On the basis of the load standardization, by 24 hours of daily each bar circuit Load data forms the perunit matrix of loadings as shown in following formula (2)：
In formula, aI, j：Load perunit value, wherein, i=1,2 ..., n, it is each circuit sequence number；J=1,2 ..., 24, it is each line At the time of road load corresponds to.
Further, the step (2) includes：
(2-1) screening extreme difference, standard deviation are with meeting weight, the excessively gentle and small load circuit of superseded curve fluctuation；
(2-2) builds the sample matrix X as shown in following formula (3)：：
Wherein, first row is respectively the corresponding moment of load data, q=[1 2 ... 24]；Element conduct in matrix A Matrix X the second column data, p=[ai,1 ai,2 … ai,24], wherein i=1,2 ..., n；
(2-3) initializes fuzzy allocation subordinated-degree matrix U, as shown in following formula (4)：
In formula, ui,k:K-th of load data for the i-th class degree of membership, wherein, k：Perunit load data number, k=1, 2,…,n*24；i：Cluster numbers, i=1,2 ..., c；
(2-4) calculates the degree of membership u of c cluster centre and k-th of load data for the i-th classi,k；
(2-5) corrects cluster centre, the weighting of the distance and degree of membership of all data points of minimization to each cluster centre With the fuzzy clustering division of completion load curve.
Further, in the step (2-1), the daily load extreme difference R of the i-th line head endiAs shown in following formula (5)：
Ri=max (ai,j)-min(ai,j) (5)；
The daily load standard deviation S of the i-th line head endiAs shown in following formula (6)：
Wherein,The load average value of i-th line road head end.
Further, in the step 3, with the mode for representing load value and replacing the total load in each period, at equivalence Manage the power distribution network total load in each period；
The load value that represents in each period includes four kinds of equivalent patterns as follows：
1) cluster centre time point corresponding load data；
2) the load average value in each period；
3) the load intermediate value in each period；
4) the load maximum in each period.
Further, the computational methods of the day part duty value include：
(3-1) as the following formula (7), calculates actual total load of the power distribution network 24 hours each moment：
In formula, ai,t：Actual negative charge values of the i-th line road in t；
(3-2) as the following formula (8) build 24 hours in each moment actual total load data fuzzy model storehouse：
In formula, ai,j:The power distribution network equivalence total load of jth period under i-th kind of equivalent pattern, i=1,2,3,4, j=1, 2 ... c；
a1,j=xcenter(j):Power distribution network total load value corresponding to cluster centre time point in the jth period；
The total load average value of jth period power distribution network, tjFor the number of, lines in the period； xk,j:Load of k-th of load data in period j；
a3,j=mediam (xj):The total load intermediate value of jth period power distribution network；
a4,j=max (xj)：The total load maximum of jth period power distribution network；
The data of four groups of equivalence patterns of (3-3) horizontal extension, generate and bear the power distribution network being made up of 24 duty values total day Lotus data a 'i,j, form the fuzzy model eigenmatrix A ' as shown in following formula (9)：
Further, in the step 4, the approach degree of four groups of equivalence patterns is calculated, determines the equivalence of load curve The method of mode includes：
(4-1) makees fuzzy model characteristic vector with the row vector of the fuzzy model eigenmatrix A ' shown in formula (9) respectively, with The power distribution network actual total load of 24 hours is as object B=[x to be identified1, x2..., x24], calculate respectively pattern to be identified with The approach degree of four groups of equivalence patterns；
The approach degree of (4-2) more described four groups of equivalence patterns, the maximum equivalent pattern of approach degree is chosen, it is bent for load The equivalent processing of load data after the equivalent division of line.
With immediate prior art ratio, the present invention also has following excellent effect：
(1) technical scheme provided by the invention considers distribution network load with time fluctuating change situation, with reference to load curve Variation tendency feature, Time segments division is carried out to the daily load curve of each bar circuit in power distribution network according to cluster analysis principle, can be with The initial time of division result day part is determined, and reflects the annual development of actual load, variation characteristic, for part throttle characteristics point Analysis, operational management and load prediction.
(2) Similarity Principle in technical scheme Land use models provided by the invention identification, relatively identified by approach degree The equivalent pattern pressed close to the most with realized load curve, the duty value value of each period is obtained, instead of each in this time The actual negative charge values at moment, complete and the equivalence of complex electric network load data is handled, effectively have compressed load data amount, letter Change daily load curve.
(3) technical scheme provided by the invention to the daily load curve of power distribution network by carrying out Time segments division, when obtaining each Between section duty value, it may be determined that power distribution network reconfiguration switch actuation time, realize power distribution network optimization operation, improve distribution The economic benefit of system.
Brief description of the drawings
Fig. 1 is the design flow diagram of load curve Time segments division method provided by the invention.
Embodiment
Clearly to illustrate technical scheme provided by the invention, below with reference to the design flow diagram in Figure of description, Specifically describe load curve Time segments division method provided by the invention.
Load curve is classified just with clustering method for current most of invention, it is bent to extract typical daily load Line, and the problems such as Time segments division can not be carried out to load curve, it is contemplated that overcome the deficiencies in the prior art, using fuzzy poly- Class and the method for pattern-recognition carry out Time segments division to load curve, complete and the equivalence of complex electric network load data is handled, Load data amount is effectively have compressed, the switch motion time of power distribution network dynamic restructuring is rationally determined, distribution system can be improved The economy and reliability of operation.
In order to reach goal of the invention, the present invention adopts the technical scheme that bent using the load of fuzzy clustering and pattern-recognition Line segmentation method is as shown in figure 1, comprise the following steps：
Step 1：Load standardization
(1) in the power distribution network containing n bar circuits, perunit benchmark is used as using the peak load value on each bar circuit in 1 year Value, the perunit value by the use of the ratio of actual value and a reference value as this circuit, i-th line road are public in the load perunit value of t Formula is as follows：
ai,t=Pi,t/Pi,max (1)
In formula：Pi,tHead end load value of (i=1,2 ... n) --- the i-th line road in t；
Pi,max--- the annual peak load value on i-th line road；ai,t--- load perunit of the i-th line road in t Value.
(2) after the standardization of load is completed, using 24 hours load perunit values of daily each bar circuit as row vector structure It is as follows into perunit matrix of loadings：
In formula：ai,j--- load perunit value；I=1,2 ..., n --- each circuit sequence number；J=1,2 ..., 24 --- load At the time of corresponding.
Step 2：The Preliminary division of load curve period
(1) the fluctuation situation of standard deviation node load curve different from extreme difference reflection is utilized.
Before time division is carried out to the load curve of all circuits, by screening extreme difference, standard deviation and load weight, pole Difference and big explanation this circuit daily load curve fluctuating change of standard deviation are big.Superseded curve fluctuation is excessively gentle and load is small Circuit, do not consider influence of the load curve of these circuits to overall dividing mode, be effectively compressed data volume.Extreme difference, standard The formula of difference is as follows：
Ri=max (ai,j)-min(ai,j) (3)
In formula：--- the load average value of i-th line road head end；Ri--- the daily load extreme difference of i-th line head end；
Si--- the daily load standard deviation of i-th line head end.
(2) sample matrix X is built；
First row is respectively the corresponding moment of load data, q=[1 2 ... 24]；Element in perunit matrix of loadings A is made For matrix X the second column data, p=[ai,1 ai,2 … ai,24], i=1,2 ..., n, obtain
(3) fuzzy allocation subordinated-degree matrix U is initialized, makes U meet to add up to 1 on row, sets the number of cluster centre, be subordinate to Shown in category degree matrix such as following formula (5)：
In formula：uik--- degree of membership of k-th of load data for the i-th class；
K=1,2 ..., n*24 --- perunit load data number, the distance of reflection load data to each cluster centre；
I=1,2 ..., c --- cluster numbers.
By many experiments and load curve feature, cluster centre number is set as c=6.
(4) c cluster centre { v is calculated respectively1,v2,…,vcAnd k-th of load data for the i-th class degree of membership uik。
(5) cluster centre, the distance of all data points of minimization to each cluster centre are constantly corrected by alternative manner With the weighted sum of degree of membership, when weighted sum takes it is minimum when, just obtain the cluster centre of each class in theory, corresponding to cluster centre Time point, and per time point corresponding to a kind of border, complete load curve fuzzy clustering division.
For some load datas in time slice border, in fact it could happen that the same time point load of different circuits point Not belonging to two clusters, the situation at same time point occur in two classes, it is necessary to the different load inscribed during by counting each Residing cluster numbering, and the number occurred, period division, root further are carried out to load curve by calculating probability of occurrence Judge the class belonging to the moment according to probability size, finally realize the conversion that load point clustering divides to plot against time section.
Step 3：Fuzzy type storehouse is established, and calculates the duty value of day part
After completing to the segmentation of load curve, because the load curve segmentation of the power distribution network is consistent, so using generation Table load value replaces the total load in the period, and equivalent processing is carried out to the power distribution network total load in each period.
The representative load value of each period, can there is 4 kinds of equivalent patterns：
1) cluster centre time point corresponding load data；
2) the load average value in each period；
3) the load intermediate value in each period；
4) the load maximum in each period.
The result that this 4 kinds of methods are calculated as pattern known to 4 kinds, and using actual 24 hours circuit total load as Optimal equivalent processing method is chosen in object to be identified, Land use models recognition methods.
(1) actual 24 hours total load data of power distribution network are calculated.
The actual total load value formula of power distribution network of t is as follows：
In formula：ai,t--- actual negative charge values of the i-th line road in t.
(2) it is as follows to build fuzzy model storehouse：
In formula：ai,j--- the power distribution network equivalence total load of jth period under i-th kind of equivalent pattern, i=1,2,3,4, j= 1,2,…,c；
a1,j=xcenter(j)--- power distribution network total load value corresponding to the interior cluster center time point of jth period；
--- the total load average value of jth period power distribution network, tjFor the line in the period Way amount；
a3,j=median (xk,j) --- the total load intermediate value of jth period power distribution network, also referred to as median, i.e., by variable After being worth the arrangement of ascending order, the variable value in ordered series of numbers centre position, if variable number is odd number in ordered series of numbers, middle position Number is intermediate location variable numerical value, if variable number is even number in ordered series of numbers, median is two, centre position adjacent variable Average value；
a4,j=max (xk,j) --- the total load maximum of jth period power distribution network.
(3) horizontal extension is carried out to the data of 4 groups of patterns, c number is according to each period duration t in will often goingjPoint Do not extended, generate the total daily load data a ' of the power distribution network being made up of 24 duty valuesi,j, form fuzzy model feature square Battle array
Step 4：The approach degree of 4 groups of equivalence patterns is calculated, determines the equivalent mode of load curve
(1) respectively with fuzzy model eigenmatrix A ' row vector Ai'=[a 'i,1 a′i,2 … a′i,24] make fuzzy model Characteristic vector, object B=[x to be identified are used as using the power distribution network actual total load of 24 hours1 x2 … x24], according to pressing close to The formula of degree, the approach degree of pattern to be identified and 4 groups of equivalence patterns is calculated respectively.
The defined formula of 5 kinds of approach degrees given below：
A. lattice close-degree
B. Hamming approach degree
C. euclidean planes
D. minimax approach degree
E. arithmetic average minimum approach degree
In formula：N --- time scale 24.
During above-mentioned calculating approach degree, A corresponds to fuzzy model eigenmatrix A ' row vector, A (ui) corresponding row to I element in amount；The actual total load matrix at B each moment in corresponding to power distribution network 24 hours, B (ui) correspond to each moment Actual total load.
(2) approach degree that 4 groups of patterns are calculated is compared.
Approach degree is bigger, illustrates that the duty value data and actual load data are closest, and equivalent treatment effect is best, The equivalent processing of load data after being divided available for the load curve period.
The above embodiments are merely illustrative of the technical scheme of the present invention and are not intended to be limiting thereof, although with reference to above-described embodiment pair The present invention is described in detail, and those of ordinary skill in the art can still enter to the embodiment of the present invention Row modification or equivalent substitution, these are applying without departing from any modification of spirit and scope of the invention or equivalent substitution Within pending claims of the invention.

Claims (7)

  1. A kind of 1. load curve Time segments division method suitable for power distribution network reconfiguration, it is characterised in that methods described includes：
    Step 1, load standardization；
    Step 2, the Preliminary division of load curve period；
    Step 3, the duty value of calculated load curve period；
    Step 4, computation schema approach degree；
    Step 5, the equivalent mode for determining load curve.
  2. 2. division methods as claimed in claim 1, it is characterised in that the step 1 includes：
    (1-1) seeks load perunit value：Load perunit value a of the i-th line road in tI, tAs shown in following formula (1)：
    <mrow> <msub> <mi>a</mi> <mrow> <mi>i</mi> <mo>,</mo> <mi>t</mi> </mrow> </msub> <mo>=</mo> <mfrac> <msub> <mi>P</mi> <mrow> <mi>i</mi> <mo>,</mo> <mi>t</mi> </mrow> </msub> <msub> <mi>P</mi> <mrow> <mi>i</mi> <mo>,</mo> <mi>max</mi> </mrow> </msub> </mfrac> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>1</mn> <mo>)</mo> </mrow> </mrow>
    In formula, PI, t：I-th line road is in the head end load value of t, i=1,2 ... n；PI, max：The whole year on i-th line road is maximum Load value；
    (1-2) construction perunit meets matrix：On the basis of the load standardization, by 24 hours loads of daily each bar circuit Data form the perunit matrix of loadings as shown in following formula (2)：
    <mrow> <mi>H</mi> <mo>=</mo> <mfenced open = "[" close = "]"> <mtable> <mtr> <mtd> <msub> <mi>a</mi> <mrow> <mn>1</mn> <mo>,</mo> <mn>1</mn> </mrow> </msub> </mtd> <mtd> <msub> <mi>a</mi> <mrow> <mn>1</mn> <mo>,</mo> <mn>2</mn> </mrow> </msub> </mtd> <mtd> <mn>...</mn> </mtd> <mtd> <msub> <mi>a</mi> <mrow> <mn>1</mn> <mo>,</mo> <mn>24</mn> </mrow> </msub> </mtd> </mtr> <mtr> <mtd> <msub> <mi>a</mi> <mrow> <mn>2</mn> <mo>,</mo> <mn>1</mn> </mrow> </msub> </mtd> <mtd> <msub> <mi>a</mi> <mrow> <mn>2</mn> <mo>,</mo> <mn>2</mn> </mrow> </msub> </mtd> <mtd> <mn>...</mn> </mtd> <mtd> <msub> <mi>a</mi> <mrow> <mn>2</mn> <mo>,</mo> <mn>24</mn> </mrow> </msub> </mtd> </mtr> <mtr> <mtd> <mo>.</mo> </mtd> <mtd> <mo>.</mo> </mtd> <mtd> <mo>.</mo> </mtd> <mtd> <mo>.</mo> </mtd> </mtr> <mtr> <mtd> <mo>.</mo> </mtd> <mtd> <mo>.</mo> </mtd> <mtd> <mo>.</mo> </mtd> <mtd> <mo>.</mo> </mtd> </mtr> <mtr> <mtd> <mo>.</mo> </mtd> <mtd> <mo>.</mo> </mtd> <mtd> <mo>.</mo> </mtd> <mtd> <mo>.</mo> </mtd> </mtr> <mtr> <mtd> <msub> <mi>a</mi> <mrow> <mi>n</mi> <mo>,</mo> <mn>1</mn> </mrow> </msub> </mtd> <mtd> <msub> <mi>a</mi> <mrow> <mi>n</mi> <mo>,</mo> <mn>2</mn> </mrow> </msub> </mtd> <mtd> <mn>...</mn> </mtd> <mtd> <msub> <mi>a</mi> <mrow> <mi>n</mi> <mo>,</mo> <mn>24</mn> </mrow> </msub> </mtd> </mtr> </mtable> </mfenced> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>2</mn> <mo>)</mo> </mrow> </mrow>
    In formula, aI, j：Load perunit value, wherein, i=1,2 ..., n, it is each circuit sequence number；J=1,2 ..., 24, born for each circuit At the time of lotus corresponds to.
  3. 3. division methods as claimed in claim 1, it is characterised in that the step (2) includes：
    (2-1) screening extreme difference, standard deviation are with meeting weight, the excessively gentle and small load circuit of superseded curve fluctuation；
    (2-2) builds the sample matrix X as shown in following formula (3)：：
    <mrow> <mi>X</mi> <mo>=</mo> <mfenced open = "[" close = "]"> <mtable> <mtr> <mtd> <msup> <mi>q</mi> <mi>T</mi> </msup> </mtd> <mtd> <msubsup> <mi>p</mi> <mn>1</mn> <mi>T</mi> </msubsup> </mtd> </mtr> <mtr> <mtd> <msup> <mi>q</mi> <mi>T</mi> </msup> </mtd> <mtd> <msubsup> <mi>p</mi> <mn>2</mn> <mi>T</mi> </msubsup> </mtd> </mtr> <mtr> <mtd> <mo>.</mo> </mtd> <mtd> <mo>.</mo> </mtd> </mtr> <mtr> <mtd> <mo>.</mo> </mtd> <mtd> <mo>.</mo> </mtd> </mtr> <mtr> <mtd> <mo>.</mo> </mtd> <mtd> <mo>.</mo> </mtd> </mtr> <mtr> <mtd> <msup> <mi>q</mi> <mi>T</mi> </msup> </mtd> <mtd> <msubsup> <mi>p</mi> <mn>24</mn> <mi>T</mi> </msubsup> </mtd> </mtr> </mtable> </mfenced> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>3</mn> <mo>)</mo> </mrow> </mrow>
    Wherein, first row is respectively the corresponding moment of load data, q=[1 2 ... 24]；Element in matrix A is as matrix X The second column data, p=[ai,1 ai,2 … ai,24], wherein i=1,2 ..., n；
    (2-3) initializes fuzzy allocation subordinated-degree matrix U, as shown in following formula (4)：
    <mrow> <mi>U</mi> <mo>=</mo> <mfenced open = "[" close = "]"> <mtable> <mtr> <mtd> <msub> <mi>u</mi> <mrow> <mn>1</mn> <mo>,</mo> <mn>1</mn> </mrow> </msub> </mtd> <mtd> <msub> <mi>u</mi> <mrow> <mn>1</mn> <mo>,</mo> <mn>2</mn> </mrow> </msub> </mtd> <mtd> <mn>...</mn> </mtd> <mtd> <msub> <mi>u</mi> <mrow> <mn>1</mn> <mo>,</mo> <mi>n</mi> <mo>*</mo> <mn>24</mn> </mrow> </msub> </mtd> </mtr> <mtr> <mtd> <msub> <mi>u</mi> <mrow> <mn>2</mn> <mo>,</mo> <mn>1</mn> </mrow> </msub> </mtd> <mtd> <msub> <mi>u</mi> <mrow> <mn>2</mn> <mo>,</mo> <mn>2</mn> </mrow> </msub> </mtd> <mtd> <mn>...</mn> </mtd> <mtd> <msub> <mi>u</mi> <mrow> <mn>2</mn> <mo>,</mo> <mi>n</mi> <mo>*</mo> <mn>24</mn> </mrow> </msub> </mtd> </mtr> <mtr> <mtd> <mo>.</mo> </mtd> <mtd> <mo>.</mo> </mtd> <mtd> <mn>...</mn> </mtd> <mtd> <mo>.</mo> </mtd> </mtr> <mtr> <mtd> <mo>.</mo> </mtd> <mtd> <mo>.</mo> </mtd> <mtd> <mo>.</mo> </mtd> <mtd> <mo>.</mo> </mtd> </mtr> <mtr> <mtd> <mo>.</mo> </mtd> <mtd> <mo>.</mo> </mtd> <mtd> <mo>.</mo> </mtd> <mtd> <mo>.</mo> </mtd> </mtr> <mtr> <mtd> <msub> <mi>u</mi> <mrow> <mi>c</mi> <mo>,</mo> <mn>1</mn> </mrow> </msub> </mtd> <mtd> <msub> <mi>u</mi> <mrow> <mi>c</mi> <mo>,</mo> <mn>2</mn> </mrow> </msub> </mtd> <mtd> <mn>...</mn> </mtd> <mtd> <msub> <mi>u</mi> <mrow> <mi>c</mi> <mo>,</mo> <mi>n</mi> <mo>*</mo> <mn>24</mn> </mrow> </msub> </mtd> </mtr> </mtable> </mfenced> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>4</mn> <mo>)</mo> </mrow> </mrow>
    In formula, ui,k:K-th of load data for the i-th class degree of membership, wherein, k：Perunit load data number, k=1, 2,…,n*24；i：Cluster numbers, i=1,2 ..., c；
    (2-4) calculates the degree of membership u of c cluster centre and k-th of load data for the i-th classi,k；
    (2-5) corrects cluster centre, and all data points of minimization are complete to the distance of each cluster centre and the weighted sum of degree of membership Divided into load curve fuzzy clustering.
  4. 4. division methods as claimed in claim 3, it is characterised in that in the step (2-1), the i-th line head end Daily load extreme difference RiAs shown in following formula (5)：
    Ri=max (ai,j)-min(ai,j) (5)；
    The daily load standard deviation S of the i-th line head endiAs shown in following formula (6)：
    <mrow> <msub> <mi>S</mi> <mi>i</mi> </msub> <mo>=</mo> <msup> <mrow> <mo>(</mo> <mfrac> <mn>1</mn> <mn>23</mn> </mfrac> <msubsup> <mo>&amp;Sigma;</mo> <mrow> <mi>j</mi> <mo>=</mo> <mn>1</mn> </mrow> <mn>24</mn> </msubsup> <msup> <mrow> <mo>(</mo> <mrow> <msub> <mi>a</mi> <mrow> <mi>i</mi> <mo>,</mo> <mi>j</mi> </mrow> </msub> <mo>-</mo> <msub> <mover> <mi>a</mi> <mo>&amp;OverBar;</mo> </mover> <mi>i</mi> </msub> </mrow> <mo>)</mo> </mrow> <mn>2</mn> </msup> <mo>)</mo> </mrow> <mfrac> <mn>1</mn> <mn>2</mn> </mfrac> </msup> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>6</mn> <mo>)</mo> </mrow> </mrow>
    Wherein,The load average value of i-th line road head end.
  5. 5. division methods as claimed in claim 1, it is characterised in that in the step 3, each time is replaced with load value is represented The mode of total load in section, equivalence handle the power distribution network total load in each period；
    The load value that represents in each period includes four kinds of equivalent patterns as follows：
    1) cluster centre time point corresponding load data；
    2) the load average value in each period；
    3) the load intermediate value in each period；
    4) the load maximum in each period.
  6. 6. division methods as claimed in claim 5, it is characterised in that the computational methods of the day part duty value include：
    (3-1) as the following formula (7), calculates actual total load of the power distribution network 24 hours each moment：
    <mrow> <msub> <mi>a</mi> <mi>t</mi> </msub> <mo>=</mo> <msubsup> <mo>&amp;Sigma;</mo> <mrow> <mi>i</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>n</mi> </msubsup> <msub> <mi>a</mi> <mrow> <mi>i</mi> <mo>,</mo> <mi>t</mi> </mrow> </msub> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>7</mn> <mo>)</mo> </mrow> </mrow>
    In formula, ai,t：Actual negative charge values of the i-th line road in t；
    (3-2) as the following formula (8) build 24 hours in each moment actual total load data fuzzy model storehouse：
    <mrow> <mi>A</mi> <mo>=</mo> <mfenced open = "[" close = "]"> <mtable> <mtr> <mtd> <msub> <mi>a</mi> <mrow> <mn>1</mn> <mo>,</mo> <mn>1</mn> </mrow> </msub> </mtd> <mtd> <msub> <mi>a</mi> <mrow> <mn>1</mn> <mo>,</mo> <mn>2</mn> </mrow> </msub> </mtd> <mtd> <mn>...</mn> </mtd> <mtd> <msub> <mi>a</mi> <mrow> <mn>1</mn> <mo>,</mo> <mi>c</mi> </mrow> </msub> </mtd> </mtr> <mtr> <mtd> <msub> <mi>a</mi> <mrow> <mn>2</mn> <mo>,</mo> <mn>1</mn> </mrow> </msub> </mtd> <mtd> <msub> <mi>a</mi> <mrow> <mn>2</mn> <mo>,</mo> <mn>2</mn> </mrow> </msub> </mtd> <mtd> <mn>...</mn> </mtd> <mtd> <msub> <mi>a</mi> <mrow> <mn>2</mn> <mo>,</mo> <mi>c</mi> </mrow> </msub> </mtd> </mtr> <mtr> <mtd> <msub> <mi>a</mi> <mrow> <mn>3</mn> <mo>,</mo> <mn>1</mn> </mrow> </msub> </mtd> <mtd> <msub> <mi>a</mi> <mrow> <mn>3</mn> <mo>,</mo> <mn>2</mn> </mrow> </msub> </mtd> <mtd> <mn>...</mn> </mtd> <mtd> <msub> <mi>a</mi> <mrow> <mn>3</mn> <mo>,</mo> <mi>c</mi> </mrow> </msub> </mtd> </mtr> <mtr> <mtd> <msub> <mi>a</mi> <mrow> <mn>4</mn> <mo>,</mo> <mn>1</mn> </mrow> </msub> </mtd> <mtd> <msub> <mi>a</mi> <mrow> <mn>4</mn> <mo>,</mo> <mn>2</mn> </mrow> </msub> </mtd> <mtd> <mn>...</mn> </mtd> <mtd> <msub> <mi>a</mi> <mrow> <mn>4</mn> <mo>,</mo> <mi>c</mi> </mrow> </msub> </mtd> </mtr> </mtable> </mfenced> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>8</mn> <mo>)</mo> </mrow> </mrow>
    In formula, ai,j:The power distribution network equivalence total load of jth period under i-th kind of equivalent pattern, i=1,2,3,4, j=1,2 ... c；
    a1,j=xcenter(j):Power distribution network total load value corresponding to cluster centre time point in the jth period；
    The total load x of jth period power distribution networkjAverage value, tjFor the number of, lines in the period； xk,j:Load of k-th of load data in period j；
    a3,j=mediam (xj):The total load x of jth period power distribution networkjIntermediate value；
    a4,j=max (xj)：The total load x of jth period power distribution networkjMaximum；
    The data of four groups of equivalence patterns of (3-3) horizontal extension, generate the total daily load number of the power distribution network being made up of 24 duty values According to a 'i,j, form the fuzzy model eigenmatrix A ' as shown in following formula (9)：
  7. 7. division methods as claimed in claim 1, it is characterised in that in the step 4, calculate four groups of equivalence patterns Approach degree, determining the method for the equivalent mode of load curve includes：
    (4-1) makees fuzzy model characteristic vector with the row vector of the fuzzy model eigenmatrix A ' shown in formula (9) respectively, with distribution The net actual total load of 24 hours is as object B=[x to be identified1, x2..., x24], calculate respectively pattern to be identified with it is described The approach degree of four groups of equivalence patterns；
    The approach degree of (4-2) more described four groups of equivalence patterns, the maximum equivalent pattern of approach degree is chosen, for load curve etc. The equivalent processing of load data after value division.
CN201610369284.0A 2016-05-30 2016-05-30 Load curve time interval division method suitable for power distribution network reconstruction Active CN107453346B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201610369284.0A CN107453346B (en) 2016-05-30 2016-05-30 Load curve time interval division method suitable for power distribution network reconstruction

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201610369284.0A CN107453346B (en) 2016-05-30 2016-05-30 Load curve time interval division method suitable for power distribution network reconstruction

Publications (2)

Publication Number Publication Date
CN107453346A true CN107453346A (en) 2017-12-08
CN107453346B CN107453346B (en) 2022-03-18

Family

ID=60485633

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201610369284.0A Active CN107453346B (en) 2016-05-30 2016-05-30 Load curve time interval division method suitable for power distribution network reconstruction

Country Status (1)

Country Link
CN (1) CN107453346B (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108256738A (en) * 2017-12-22 2018-07-06 同济大学 Track switch action reference curve choosing method and its application
CN108596362A (en) * 2018-03-22 2018-09-28 国网四川省电力公司经济技术研究院 It polymerize approximate electric load curve form clustering method based on adaptive segmentation
CN110097261A (en) * 2019-04-17 2019-08-06 三峡大学 A method of judging user power utilization exception

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
FR961086A (en) * 1943-04-27 1950-05-03
CN103632031A (en) * 2013-10-22 2014-03-12 国家电网公司 Rural load type load modeling method based on load curve decomposition
CN103679555A (en) * 2013-12-16 2014-03-26 成都安健发科技有限公司 Time-of-use electricity price determining method based on load characteristic classification
CN103761574A (en) * 2014-01-02 2014-04-30 上海电力学院 Distributed power supply and region load matched feature matching method
CN104376402A (en) * 2014-10-31 2015-02-25 国家电网公司 Load characteristic classification and synthesis method based on frequency domain indexes
CN104917173A (en) * 2015-06-01 2015-09-16 国网天津市电力公司 Power distribution network optimization method adapting to power distribution network high capacity load transfer
CN105389633A (en) * 2015-12-01 2016-03-09 上海电力学院 Optimization planning method of substation considering distributed power supplies
CN105528660A (en) * 2016-03-09 2016-04-27 湖南大学 Substation load model parameter prediction method based on daily load curve

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
FR961086A (en) * 1943-04-27 1950-05-03
CN103632031A (en) * 2013-10-22 2014-03-12 国家电网公司 Rural load type load modeling method based on load curve decomposition
CN103679555A (en) * 2013-12-16 2014-03-26 成都安健发科技有限公司 Time-of-use electricity price determining method based on load characteristic classification
CN103761574A (en) * 2014-01-02 2014-04-30 上海电力学院 Distributed power supply and region load matched feature matching method
CN104376402A (en) * 2014-10-31 2015-02-25 国家电网公司 Load characteristic classification and synthesis method based on frequency domain indexes
CN104917173A (en) * 2015-06-01 2015-09-16 国网天津市电力公司 Power distribution network optimization method adapting to power distribution network high capacity load transfer
CN105389633A (en) * 2015-12-01 2016-03-09 上海电力学院 Optimization planning method of substation considering distributed power supplies
CN105528660A (en) * 2016-03-09 2016-04-27 湖南大学 Substation load model parameter prediction method based on daily load curve

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
XIAOLI MENG等: "Dynamic reconfiguration of distribution network considering scheduling of DG active power outputs", 《2014 INTERNATIONAL CONFERENCE ON POWER SYSTEM TECHNOLOGY》 *

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108256738A (en) * 2017-12-22 2018-07-06 同济大学 Track switch action reference curve choosing method and its application
CN108256738B (en) * 2017-12-22 2021-12-14 同济大学 Turnout action reference curve selection method and application thereof
CN108596362A (en) * 2018-03-22 2018-09-28 国网四川省电力公司经济技术研究院 It polymerize approximate electric load curve form clustering method based on adaptive segmentation
CN108596362B (en) * 2018-03-22 2021-12-28 国网四川省电力公司经济技术研究院 Power load curve form clustering method based on adaptive piecewise aggregation approximation
CN110097261A (en) * 2019-04-17 2019-08-06 三峡大学 A method of judging user power utilization exception
CN110097261B (en) * 2019-04-17 2022-11-18 三峡大学 Method for judging abnormal electricity utilization of user

Also Published As

Publication number Publication date
CN107453346B (en) 2022-03-18

Similar Documents

Publication Publication Date Title
Zhang et al. Typical wind power scenario generation for multiple wind farms using conditional improved Wasserstein generative adversarial network
Wang et al. Short-term wind power prediction based on multidimensional data cleaning and feature reconfiguration
Liu et al. Study on SVM compared with the other text classification methods
CN109063911A (en) A kind of Load aggregation body regrouping prediction method based on gating cycle unit networks
CN102231144B (en) A kind of power distribution network method for predicting theoretical line loss based on Boosting algorithm
CN103903189B (en) Low-voltage distribution network platform area clustering method based on fuzzy clustering
CN108446794A (en) One kind being based on multiple convolutional neural networks combination framework deep learning prediction techniques
CN106408008A (en) Load curve distance and shape-based load classification method
CN106845717A (en) A kind of energy efficiency evaluation method based on multi-model convergence strategy
CN109492748B (en) Method for establishing medium-and-long-term load prediction model of power system based on convolutional neural network
CN104850890A (en) Method for adjusting parameter of convolution neural network based on example learning and Sadowsky distribution
CN104809658A (en) Method for rapidly analyzing low-voltage distributing network area line loss
CN107301604A (en) Multi-model fusion estimation system
CN110717610A (en) Wind power prediction method based on data mining
CN109255726A (en) A kind of ultra-short term wind power prediction method of Hybrid Intelligent Technology
CN102750286A (en) Novel decision tree classifier method for processing missing data
CN107563451A (en) Running rate recognizing method under a kind of pumping plant steady state condition
CN107453346A (en) A kind of load curve Time segments division method suitable for power distribution network reconfiguration
CN108984830A (en) A kind of building efficiency evaluation method and device based on FUZZY NETWORK analysis
CN107748940A (en) A kind of energy conservation potential Quantitative prediction methods
CN109272058A (en) Integrated power load curve clustering method
Wang et al. Big data analytics for price forecasting in smart grids
CN105653670A (en) Intelligent electricity consumption data mining method based on manifold learning clustering algorithm
CN107093005A (en) The method that tax handling service hall&#39;s automatic classification is realized based on big data mining algorithm
CN108694475B (en) Short-time-scale photovoltaic cell power generation capacity prediction method based on hybrid model

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