CN107453346B - Load curve time interval division method suitable for power distribution network reconstruction - Google Patents
Load curve time interval division method suitable for power distribution network reconstruction Download PDFInfo
- Publication number
- CN107453346B CN107453346B CN201610369284.0A CN201610369284A CN107453346B CN 107453346 B CN107453346 B CN 107453346B CN 201610369284 A CN201610369284 A CN 201610369284A CN 107453346 B CN107453346 B CN 107453346B
- Authority
- CN
- China
- Prior art keywords
- load
- equivalent
- time
- distribution network
- value
- 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
- 238000000034 method Methods 0.000 title claims abstract description 36
- 238000012545 processing Methods 0.000 claims abstract description 9
- 239000011159 matrix material Substances 0.000 claims description 26
- 239000013598 vector Substances 0.000 claims description 9
- 238000000638 solvent extraction Methods 0.000 claims description 4
- 238000004364 calculation method Methods 0.000 claims description 2
- 238000012216 screening Methods 0.000 claims description 2
- 239000000126 substance Substances 0.000 claims description 2
- 102000001708 Protein Isoforms Human genes 0.000 claims 2
- 108010029485 Protein Isoforms Proteins 0.000 claims 2
- SYSZENVIJHPFNL-UHFFFAOYSA-N (alpha-D-mannosyl)7-beta-D-mannosyl-diacetylchitobiosyl-L-asparagine, isoform B (protein) Chemical compound COC1=CC=C(I)C=C1 SYSZENVIJHPFNL-UHFFFAOYSA-N 0.000 claims 1
- 230000009471 action Effects 0.000 abstract description 9
- 238000012567 pattern recognition method Methods 0.000 abstract description 2
- 230000008859 change Effects 0.000 description 8
- 230000011218 segmentation Effects 0.000 description 7
- 238000005457 optimization Methods 0.000 description 5
- 238000004458 analytical method Methods 0.000 description 3
- 230000008901 benefit Effects 0.000 description 3
- 238000012423 maintenance Methods 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 238000011160 research Methods 0.000 description 3
- 238000004422 calculation algorithm Methods 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000003909 pattern recognition Methods 0.000 description 2
- 241000249931 Doronicum maximum Species 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000011217 control strategy Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 230000002068 genetic effect Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
Images
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 invention provides a load curve time interval division method suitable for power distribution network reconstruction, which comprises load per unit; dividing a load curve time interval; calculating the equivalent load of the load curve time interval; calculating the mode closeness; the equivalent mode of the load curve is determined. According to the technical scheme provided by the invention, the load curve is divided into time periods by adopting a fuzzy clustering and pattern recognition method, so that equivalent processing of complex power grid load data is completed, the load data volume is effectively compressed, the switching action time of dynamic reconstruction of the power distribution network is reasonably determined, and the economical efficiency and the reliability of operation of the power distribution system can be improved.
Description
Technical Field
The invention relates to a load characteristic analysis method, in particular to a load curve time interval division method suitable for power distribution network reconstruction.
Background
In the operation process of the power distribution network, the load always changes constantly along with time, two identical operation states cannot exist in the operation process of the power distribution network due to the randomness of the load change, and if each operation state is analyzed respectively, the calculation amount is huge; on the other hand, as the work and rest modes and the power consumption habits of people are stable, the load change in the power distribution network is regular and has certain correlation and reproducibility, the time with larger load fluctuation in a period of time (1 d) is divided into different time periods respectively according to the characteristics of a load curve in the power distribution network, and the time with smaller load fluctuation is classified into the same time period, so that the load levels in the same time period are close to each other, the division of the load curve is realized, the operation optimization difficulty of the power distribution network can be simplified, the calculated amount can be reduced, and the operation state of the power distribution network can be reflected really.
In view of dynamic reconstruction of the power distribution network with load fluctuation, the purpose of changing the network structure and realizing control of the operation mode of the power distribution network is achieved by adjusting the state of an operable switch in the network. The frequent opening and closing of the switches of the electrical equipment in the power distribution network can seriously shorten the electrical service life of the electrical equipment, reduce the reliability of equipment action, increase the equipment updating frequency, improve the operation and maintenance cost and the switch operation cost of the power distribution network, and reduce the network economy and reliability. The switching action frequency is controlled within a certain range, the actual network structure, the overall fluctuation condition of the load, the current weather temperature and other factors need to be combined under different time scales, and meanwhile, the frequency of the switching action is accurately limited according to relevant operation maintenance rules, artificial rules and the like, so that the load curve needs to be divided in reasonable time intervals according to the fluctuation trend, the switching state is considered to be kept unchanged in the time intervals, and the network structure does not change.
On the basis, the method for dividing the load curve is that the daily load curve is divided into a plurality of time intervals, the load value subjected to equivalent processing is used for replacing the load in each time interval, so that the action time of the switch of the power distribution network is determined, the electrical service life of the switch is protected, the maintenance cost of the operation of the power distribution system is reduced, the safety and reliability degree and the economical efficiency of the power distribution system are improved, good economic benefits and social benefits are created, and the method has important theoretical value and practical significance.
The inventor of the invention has found through a great deal of research that the invention is disclosed in the field of load curve segmentation in recent years; the method comprises the steps of segmenting according to the change trend of a load curve, firstly carrying out primary segmentation according to monotonicity, and then sequentially fusing primary segmentation points by adopting a fusion idea until the number of segments meets the action time constraint of equipment. The invention discloses a time-varying reactive power optimization algorithm for a power system (the journal of the power system and the automation thereof, 2007), wherein load segmentation is considered as a discrete optimization problem, and the method comprises the steps of intelligently segmenting a load curve by using a genetic algorithm, and determining starting and stopping time points of each segment so as to solve the problem of constraint of the action times of equipment. The research on the multi-period reactive power optimization of the power system (journal of the university of transport in western security, 2008) applies a statistical principle to a load sequence curve, describes the fluctuation change of the load curve by utilizing range and standard deviation, analyzes the sequence characteristic and divides the load periods, provides an enlightening iterative type segmentation method, and solves the value taking problem of the maximum range and the maximum standard deviation. And 2010 in the document of 'reactive power optimization control strategy research' discloses that after segmentation is carried out under the condition of ensuring that the voltage of a system node is qualified, the number of segments is compared with the number of equipment adjustment times, and if the number of segments is greater than the number of equipment adjustment times, the number of segments is taken as a final result; otherwise, the number of segments is increased until the number of device adjustments is approached.
Theoretically, in the prior art, the load curves are classified only by using a clustering method, typical daily load curves are extracted, and the load curves cannot be divided in time periods; when calculating a typical load curve, only the average value of all curves of each type at each point is obtained, and the load in a time period is not equalized, so that the effective utilization of load data is realized.
In order to meet the requirements of the prior art, the invention provides a load curve time interval division method which classifies the coincidence by a fuzzy clustering method and is suitable for power distribution network reconstruction.
Disclosure of Invention
To meet the needs in the art, the present invention provides a method for effectively partitioning load curve time intervals, which effectively compresses the load data volume.
The invention provides a load curve time interval dividing method suitable for power distribution network reconstruction, which is improved in that the method comprises the following steps:
step 1, per unit load;
step 2, preliminary division of load curve time intervals;
step 3, calculating the equivalent load of the load curve period;
step 4, calculating the mode closeness;
and 5, determining the equivalent mode of the load curve.
Further, the step 1 comprises:
(1-1) load per unit value: load per unit value a of ith line at time ti，tAs shown in the following formula (1):
in the formula, Pi，t: the head end load value of the ith line at the time t, i is 1,2 … n; pi，max: the annual maximum load value of the ith line;
(1-2) constructing a per unit coincidence matrix: on the basis of the per-unit load, the per-unit load matrix shown in the following formula (2) is formed by 24-hour load data of each line per day:
in the formula, ai，j: a load per unit value, wherein i is 1,2, …, n is each line serial number; j is 1,2, …,24, and is the time corresponding to each line load.
Further, the step (2) comprises:
(2-1) screening out extreme differences, standard differences and coincidence weights, and eliminating lines with too gentle curve fluctuation and small load;
(2-2) constructing a sample matrix X represented by the following formula (3): :
wherein the first column is the corresponding time of the load data, q ═ 12 … 24](ii) a The element in matrix a is used as the second column data of matrix X, and p ═ ai,1 ai,2 … ai,24]Wherein i is 1,2, …, n;
(2-3) initializing a fuzzy distribution membership matrix U as shown in the following formula (4):
in the formula ui,kThe degree of membership of the kth load data to the ith class, wherein k: the number of load data per unit, k is 1,2, …, n is 24; i: the number of clusters, i ═ 1,2, …, c;
(2-4) calculating the membership degree u of the c cluster centers and the kth load data to the ith classi,k；
And (2-5) modifying the clustering centers, minimizing the weighted sum of the distances from all data points to all the clustering centers and the membership degrees, and finishing the fuzzy clustering division of the load curves.
Further, in the step (2-1), the daily load at the head end of the ith line is very different from RiAs shown in the following formula (5):
Ri＝max(ai,j)-min(ai,j) (5)；
the standard deviation S of the daily load of the head end of the ith lineiAs shown in the following formula (6):
wherein the content of the first and second substances,
load average at the head end of the ith line.
Further, in the step 3, the total load of the power distribution network in each time period is equivalently processed in a mode that the representative load value is used for replacing the total load in each time period;
the representative load value in each time segment contains four equivalent patterns as follows:
1) corresponding load data of the clustering center time point;
2) average load per time period;
3) median load value in each time segment;
4) maximum load per time period.
Further, the method for calculating the equivalent load in each time period comprises the following steps:
(3-1) calculating the actual total load of the power distribution network at each moment of 24 hours according to the following formula (7):
in the formula, ai,t: the actual load value of the ith line at the time t;
(3-2) constructing a fuzzy model base of the actual total load data at each moment in 24 hours according to the following formula (8):
in the formula, ai,jThe distribution network equivalent total load of the j time slot under the ith equivalent mode, wherein i is 1,2,3,4, j is 1,2, … c;
a1,j＝xcenter(j)the total load value of the power distribution network corresponding to the clustering center time point in the jth time period;
a3,j＝mediam(xj) The median value of the total load of the power distribution network in the j time period;
a4,j＝max(xj): the maximum value of the total load of the power distribution network in the j time period;
(3-3) horizontally extending the data of the four groups of equivalent modes to generate total daily load data a of the power distribution network consisting of 24 equivalent loadsi′,jA fuzzy model feature matrix a' shown in the following equation (9) is formed:
further, in step 4, the method for calculating the closeness of the four sets of equivalent modes and determining the equivalent mode of the load curve includes:
(4-1) the row vectors of the fuzzy model feature matrix a' shown in the formula (9) are respectively used as fuzzy model feature vectors, and the actual total load of the power distribution network for 24 hours is used as an object B to be identified, wherein the object B is [ x ]1，x2，…，x24]Respectively calculating the closeness of the pattern to be recognized and the four groups of equivalent patterns;
and (4-2) comparing the closeness of the four groups of equivalent modes, and selecting the equivalent mode with the largest closeness for equivalent processing of the load data after the load curve is subjected to equivalent division.
Compared with the closest prior art, the invention also has the following excellent effects:
(1) the technical scheme provided by the invention considers the fluctuation change condition of the load of the power distribution network along with time, combines the change trend characteristic of the load curve, and divides the daily load curve of each line in the power distribution network into time intervals according to the clustering analysis principle, so that the initial time of each time interval of the division result can be determined, the annual development and change characteristics of the actual load can be reflected, and the method is used for load characteristic analysis, operation management and load prediction.
(2) The technical scheme provided by the invention utilizes a near selection principle in pattern recognition, identifies the closest equivalent pattern to the actual load curve through the proximity comparison, obtains the equivalent load value of each time period, replaces the actual load value of each moment in the time period, completes the equivalent processing of the complex power grid load data, effectively compresses the load data volume and simplifies the daily load curve.
(3) According to the technical scheme provided by the invention, the daily load curve of the power distribution network is divided into time periods to obtain equivalent loads of each time period, so that the action time of the power distribution network reconfiguration switch can be determined, the optimized operation of the power distribution network is realized, and the economic benefit of a power distribution system is improved.
Drawings
Fig. 1 is a design flowchart of the load curve time interval division method provided by the present invention.
Detailed Description
In order to clearly illustrate the technical solution provided by the present invention, the load curve time interval dividing method provided by the present invention will be specifically described below with reference to the design flow chart in the drawings of the specification.
Aiming at solving the problems that most of the prior art only classifies the load curves by using a clustering method, extracts typical daily load curves, cannot divide the load curves by time periods and the like, the invention aims to overcome the defects of the prior art, adopts a fuzzy clustering and pattern recognition method to divide the load curves by time periods, completes equivalent processing on the load data of a complex power grid, effectively compresses the load data volume, reasonably determines the switching action time of dynamic reconstruction of the power distribution network, and can improve the economical efficiency and the reliability of the operation of the power distribution system.
In order to achieve the purpose of the invention, the technical scheme adopted by the invention is that a load curve segmentation method adopting fuzzy clustering and pattern recognition is shown in figure 1 and comprises the following steps:
step 1: load per unit
(1) In a power distribution network containing n lines, the maximum load value in one year on each line is taken as a per-unit reference value, the ratio of an actual value to the reference value is taken as the per-unit value of the line, and the load per-unit value formula of the ith line at the time t is as follows:
ai,t＝Pi,t/Pi,max (1)
in the formula: pi,t(i ═ 1,2 … n) — head end load value of the ith line at time t;
Pi,max-annual maximum load value of ith line; a isi,t-load per unit value of ith line at time t.
(2) After the load per unit is finished, the per unit load matrix is formed by taking the 24-hour load per unit value of each line per day as a row vector as follows:
in the formula: a isi,j-a load per unit value; i is 1,2, …, n-each line serial number; j is 1,2, …,24, the time corresponding to the load.
Step 2: preliminary division of load curve time period
(1) And reflecting the fluctuation conditions of the load curves of different nodes by using the standard deviation and the range.
Before time division is carried out on the load curves of all lines, extreme differences, standard deviations and load weights are screened, and the fact that the daily load curve of the line fluctuates greatly is shown due to the fact that the numerical values of the extreme differences and the standard deviations are large. And lines with too gentle curve fluctuation and small load are eliminated, the influence of the load curves of the lines on the whole dividing mode is not considered, and the data volume is effectively compressed. The formula of range, standard deviation is as follows:
Ri＝max(ai,j)-min(ai,j) (3)
in the formula:
-load average at the head end of the ith line; riThe daily load at the head end of the ith line is extremely poor;
Si-standard deviation of daily load at the head end of the ith line.
(2) Constructing a sample matrix X;
the first column indicates the corresponding time of the load data, q ═ 12 … 24](ii) a The element in the per unit load matrix A is used as the second column data of the matrix X, and p is [ a ]i,1 ai,2 … ai,24]I is 1,2, …, n, to yield
(3) Initializing a fuzzy distribution membership matrix U, enabling the U to satisfy the column summation to be 1, and setting the number of clustering centers, wherein the membership matrix is shown as the following formula (5):
in the formula: u. ofik-degree of membership of the kth load data to class i;
k is 1,2, …, n is 24, the number of load data per unit, and reflects the distance from the load data to each cluster center;
i-1, 2, …, c-cluster number.
Through a large number of experiments and load curve characteristics, the number of cluster centers is set to c equal to 6.
(4) Respectively calculating c clustering centers { v1,v2,…,vc…, vc } and kthMembership u of load data to class iik。
(5) And continuously correcting the clustering centers by an iteration method, minimizing the weighted sum of the distances from all data points to each clustering center and the membership degree, and theoretically obtaining the clustering center of each class, the time point corresponding to the clustering center and the time point corresponding to the boundary of each class when the weighted sum is minimized, thereby completing the fuzzy clustering division of the load curve.
For some load data in a time segment boundary, loads at the same time point of different lines may belong to two clusters respectively, and under the condition that the same time point appears in both the two clusters, the cluster numbers of the different loads at each moment and the occurrence times are counted, the load curve is further divided by calculating the occurrence probability, the class to which the moment belongs is judged according to the probability, and finally the conversion from the cluster division of the load points to the curve time segment division is realized.
And step 3: establishing a fuzzy library and calculating the equivalent load of each time interval
After the load curve is segmented, the load curve segments of the power distribution network are consistent, so that the representative load value is adopted to replace the total load in the time period, and the equivalent processing is carried out on the total load of the power distribution network in each time period.
The representative load value for each time segment can have 4 equivalent patterns:
1) corresponding load data of the clustering center time point;
2) average load per time period;
3) median load value in each time segment;
4) maximum load per time period.
The results calculated by the 4 methods are used as 4 known modes, the actual total line load of 24 hours is used as an object to be identified, and the optimal equivalent processing method is selected by using the mode identification method.
(1) Actual 24 hour total load data for the distribution grid is calculated.
the formula of the actual total load value of the power distribution network at the moment t is as follows:
in the formula: a isi,t-actual load value of the ith line at time t.
(2) The fuzzy model library is constructed as follows:
in the formula: a isi,j-the distribution network equivalent total load of the j time slot in the i-th equivalent mode, i being 1,2,3,4, j being 1,2, …, c;
a1,j＝xcenter(j) -the total load value of the distribution network corresponding to the cohesive center time point of the j time period;
a3,j＝median（xk,j) The total load median of the power distribution network in the j time period is also called as a median, namely, the variable values in the middle position of the array are arranged from small to large, if the variable number in the array is an odd number, the median is a variable value of the middle position, and if the variable number in the array is an even number, the median is an average value of two adjacent variables in the middle position;
a4,j＝max（xk,j) -maximum value of total load of distribution network at j time period.
(3) Horizontally extending the data of 4 groups of modes, namely, the number of c in each row is determined according to the time length t of each time periodjRespectively extending to generate total daily load data a of the power distribution network consisting of 24 equivalent loadsi′jForming a feature matrix of the fuzzy model
And 4, step 4: calculating the closeness of 4 groups of equivalent modes and determining the equivalent mode of the load curve
(1) Respectively using the row vector A of the fuzzy model characteristic matrix Ai′＝[ai′,1 ai′,2 … ai′,24]Making fuzzy model characteristic vector, using actual total load of 24 hours of power distribution network as object B ═ x to be identified1 x2 … x24]And respectively calculating the closeness of the to-be-recognized mode and 4 groups of equivalent modes according to a closeness formula.
The following definition formulas for 5 closeness are given:
a. degree of closeness of grid
b. Proximity of Haiming
c. Degree of closeness of Ou
d. Maximum minimum closeness
e. Arithmetic mean minimum closeness
In the formula: n-time scale 24.
In the above process of calculating closeness, A corresponds to the row vector of the fuzzy model feature matrix A', A (u)i) Corresponding to i elements in the row vector; b corresponds to the actual total load matrix, B (u), at each moment in 24 hours of the distribution networki) Corresponding to the actual total load at each moment.
(2) And (5) comparing and calculating the closeness of the 4 groups of modes.
The greater the closeness, the closest the equivalent load data to the actual load data is, the best equivalent processing effect is, and the method can be used for the equivalent processing of the load data after the load curve time period division.
Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art can make modifications and equivalents to the embodiments of the present invention without departing from the spirit and scope of the present invention, which is set forth in the claims of the present application.
Claims (6)
1. A load curve time interval division method suitable for power distribution network reconstruction is characterized by comprising the following steps:
step 1, per unit load;
step 2, preliminary division of load curve time intervals;
step 3, calculating the equivalent load of the load curve period;
step 4, calculating the mode closeness;
step 5, determining the equivalent mode of the load curve;
in the step 3, the total load of the power distribution network in each time period is equivalently processed in a mode of replacing the total load in each time period with the representative load value;
the representative load value in each time segment contains four equivalent patterns as follows:
1) corresponding load data of the clustering center time point;
2) average load per time period;
3) median load value in each time segment;
4) maximum load per time period.
2. The partitioning method according to claim 1, wherein said step 1 comprises:
step 1-1, load per unit value is obtained: load per unit value a of ith line at time ti，tAs shown in the following formula (1):
in the formula, Pi，t: the head end load value of the ith line at the time t, i is 1,2 … n; pi，max: the annual maximum load value of the ith line;
step 1-2, constructing a per unit load matrix: on the basis of the per-unit load, the per-unit load matrix shown in the following formula (2) is formed by 24-hour load data of each line per day:
in the formula, ai，j: a load per unit value, wherein i is 1,2, …, n is each line serial number; j is 1,2, …,24, and is the time corresponding to each line load.
3. The partitioning method according to claim 1, wherein said step 2 comprises:
step 2-1, screening range error, standard deviation and load weight, and eliminating lines with too gentle curve fluctuation and small load;
step 2-2 constructs a sample matrix X as shown in the following formula (3):
wherein, the firstOne row is the corresponding time of the load data, q is [ 12 … 24 ]](ii) a The elements in the per-unit load matrix H are used as the second column data, p, of the matrix Xi＝[ai,1 ai,2 … ai,24]Wherein i is 1,2, …, n;
step 2-3 initializes a fuzzy distribution membership matrix U as shown in the following formula (4):
in the formula ui,kThe degree of membership of the kth load data to the ith class, wherein k: the number of load data per unit, k is 1,2, …, n is 24; i: the number of clusters, i ═ 1,2, …, c;
step 2-4, calculating the membership u of the c clustering centers and the kth load data to the ith classi,k；
And 2-5, modifying the clustering centers, minimizing the weighted sum of the distances from all data points to all the clustering centers and the membership degrees, and finishing the fuzzy clustering division of the load curves.
4. The division method as claimed in claim 3, wherein in said step 2-1, the daily load of the head end of the i-th line is very poor RiAs shown in the following formula (5):
Ri＝max(ai,j)-min(ai,j) (5)；
the standard deviation S of the daily load of the head end of the ith lineiAs shown in the following formula (6):
wherein the content of the first and second substances,
load average at the head end of the ith line.
5. The division method according to claim 1, wherein said calculation method of the equivalent load for each period comprises:
step 3-1, calculating the actual total load of the power distribution network at each moment of 24 hours according to the following formula (7):
in the formula, ai,t: the actual load value of the ith line at the time t;
step 3-2, constructing a fuzzy model base of actual total load data at each moment in 24 hours according to the following formula (8):
in the formula, ai,jThe distribution network equivalent total load of the j time slot under the ith equivalent mode, wherein i is 1,2,3,4, j is 1,2, … c;
a1,j＝xcenter(j)the total load value of the power distribution network corresponding to the clustering center time point in the jth time period;
total load x of distribution network in j time periodjAverage value of (d), tjThe number of lines in the time period; x is the number ofk,jThe load capacity of the kth load data in the time period j;
a3,j＝mediam(xj) Total load x of distribution network in j time periodjThe median value of (d);
a4,j＝max(xj): total load x of distribution network in j time periodjMaximum value of (d);
step 3-3, horizontally extending the data of the four equivalent modes, namely, the number of c in each row is determined according to the time length t of each time periodjRespectively extended to generate a load composed of 24 equivalent loadsTotal daily load data a 'of power distribution network'i,jA fuzzy model feature matrix a' shown in the following equation (9) is formed:
6. the partitioning method as claimed in claim 5, wherein in said step 4, the closeness of said four iso-forms is calculated, and the method for determining the iso-form of the load curve includes:
step 4-1 uses the row vectors of the fuzzy model feature matrix a' shown in formula (9) as fuzzy model feature vectors, and uses the actual total load of the distribution network for 24 hours as the object B to be identified ═ x1，x2，…，x24]Respectively calculating the closeness of the pattern to be recognized and the four groups of equivalent patterns;
and 4-2, comparing the closeness of the four groups of equivalent modes, selecting the equivalent mode with the largest closeness, and using the equivalent mode for equivalent processing of the load data after the load curve is subjected to equivalent division.
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 CN107453346A (en) | 2017-12-08 |
| CN107453346B true 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) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108256738B (en) * | 2017-12-22 | 2021-12-14 | 同济大学 | Turnout action reference curve selection method and application thereof |
| 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 |
| CN110097261B (en) * | 2019-04-17 | 2022-11-18 | 三峡大学 | Method for judging abnormal electricity utilization of user |
Citations (8)
| 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 |
-
2016
- 2016-05-30 CN CN201610369284.0A patent/CN107453346B/en active Active
Patent Citations (8)
| 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)
| Title |
|---|
| Dynamic reconfiguration of distribution network considering scheduling of DG active power outputs;Xiaoli Meng等;《2014 International Conference on Power System Technology》;20141222;第1433-1439页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107453346A (en) | 2017-12-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109063911A (en) | A kind of Load aggregation body regrouping prediction method based on gating cycle unit networks | |
| CN106600059A (en) | Intelligent power grid short-term load predication method based on improved RBF neural network | |
| CN110263995A (en) | Consider the distribution transforming heavy-overload prediction technique of load growth rate and user power utilization characteristic | |
| CN107453346B (en) | Load curve time interval division method suitable for power distribution network reconstruction | |
| Chang et al. | Enforced self‐organizing map neural networks for river flood forecasting | |
| CN102024179A (en) | Genetic algorithm-self-organization map (GA-SOM) clustering method based on semi-supervised learning | |
| CN102200759A (en) | Nonlinear kernelled adaptive prediction method | |
| CN111525587B (en) | Reactive load situation-based power grid reactive voltage control method and system | |
| CN110322075A (en) | A kind of scenic spot passenger flow forecast method and system based on hybrid optimization RBF neural | |
| Natarajan et al. | New SVM kernel soft computing models for wind speed prediction in renewable energy applications | |
| CN111461921B (en) | Load modeling typical user database updating method based on machine learning | |
| CN110083190A (en) | A kind of green pepper greenhouse intelligent monitor system based on subtractive clustering classifier | |
| CN109861211A (en) | A kind of power distribution network dynamic reconfiguration method based on data-driven | |
| CN111882114B (en) | Short-time traffic flow prediction model construction method and prediction method | |
| CN113298569A (en) | Power demand side management method and system based on artificial intelligence | |
| CN115329899A (en) | Clustering equivalent model construction method, system, equipment and storage medium | |
| CN111832836B (en) | Power distribution network reconstruction method and system considering load power utilization characteristics | |
| CN113627674A (en) | Distributed photovoltaic power station output prediction method and device and storage medium | |
| CN108038518A (en) | A kind of photovoltaic generation power based on meteorological data determines method and system | |
| Luo et al. | A reduced mixed representation based multi-objective evolutionary algorithm for large-scale overlapping community detection | |
| CN116523142A (en) | Photovoltaic cluster interval prediction method considering space-time characteristic clustering | |
| CN110852370A (en) | Clustering algorithm-based large-industry user segmentation method | |
| CN109615142A (en) | A kind of wind farm wind velocity combination forecasting method based on wavelet analysis | |
| CN114444763A (en) | Wind power prediction method based on AFSA-GNN | |
| Smith et al. | Time series prediction via two-step clustering |
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 |