CN113886183B - Method for measuring and calculating occurrence time of voltage sag event - Google Patents
Method for measuring and calculating occurrence time of voltage sag event Download PDFInfo
- Publication number
- CN113886183B CN113886183B CN202111183809.9A CN202111183809A CN113886183B CN 113886183 B CN113886183 B CN 113886183B CN 202111183809 A CN202111183809 A CN 202111183809A CN 113886183 B CN113886183 B CN 113886183B
- Authority
- CN
- China
- Prior art keywords
- interval
- voltage sag
- time
- fuzzy
- domain
- 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 40
- 238000012544 monitoring process Methods 0.000 claims abstract description 43
- 230000002776 aggregation Effects 0.000 claims abstract description 19
- 238000004220 aggregation Methods 0.000 claims abstract description 19
- 239000011159 matrix material Substances 0.000 claims abstract description 13
- 238000004519 manufacturing process Methods 0.000 claims abstract description 9
- 238000012545 processing Methods 0.000 claims abstract description 6
- 238000011084 recovery Methods 0.000 claims abstract description 5
- 239000013598 vector Substances 0.000 claims description 13
- 238000004364 calculation method Methods 0.000 claims description 6
- 238000006116 polymerization reaction Methods 0.000 claims description 4
- 230000001174 ascending effect Effects 0.000 claims description 3
- 238000010606 normalization Methods 0.000 claims description 3
- 238000005457 optimization Methods 0.000 claims description 3
- NAWXUBYGYWOOIX-SFHVURJKSA-N (2s)-2-[[4-[2-(2,4-diaminoquinazolin-6-yl)ethyl]benzoyl]amino]-4-methylidenepentanedioic acid Chemical compound C1=CC2=NC(N)=NC(N)=C2C=C1CCC1=CC=C(C(=O)N[C@@H](CC(=C)C(O)=O)C(O)=O)C=C1 NAWXUBYGYWOOIX-SFHVURJKSA-N 0.000 claims description 2
- 230000017105 transposition Effects 0.000 claims description 2
- 230000008569 process Effects 0.000 description 6
- 238000000605 extraction Methods 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 238000005259 measurement Methods 0.000 description 2
- 230000008447 perception Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000002864 sequence alignment Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/30—Monitoring
- G06F11/3058—Monitoring arrangements for monitoring environmental properties or parameters of the computing system or of the computing system component, e.g. monitoring of power, currents, temperature, humidity, position, vibrations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/16—Matrix or vector computation, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Computing Systems (AREA)
- Computational Mathematics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Algebra (AREA)
- Quality & Reliability (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- Probability & Statistics with Applications (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
A method of measuring voltage sag event occurrence time, comprising: collecting voltage sag monitoring data of a target site, wherein the time from interruption of voltage sag to recovery of a production process is an aggregation target, and performing homologous aggregation on the collected voltage sag monitoring data; carrying out differential processing on the voltage sag monitoring data subjected to homologous aggregation to obtain sample data of the voltage sag interval time of the target site; establishing a domain of interval time sample data, and dividing the domain interval by adopting a fuzzy C-means algorithm; aiming at the divided domain interval, a membership function of voltage sag interval time is formulated, and an interval time sequence is converted into a fuzzy time sequence; establishing a fuzzy relation matrix between elements in a fuzzy time sequence based on the fuzzified sample data; prediction is performed on the basis of the 'maximum-minimum principle', and a weighted average method is adopted to deblur. The invention can calculate the occurrence time of the next voltage sag event and early warn future disturbance.
Description
Technical Field
The invention relates to the field of voltage monitoring, in particular to a method for measuring and calculating the occurrence time of a voltage sag event.
Background
Voltage sag events caused by grid disturbances cause significant economic losses to power consumers of industrial processes including ac contactors, programmable logic controllers, personal computers, and other sensitive equipment. The occurrence of the voltage sag event is predicted in advance, so that power users and power grid companies can be helped to reasonably avoid risks caused by the voltage sag. Along with the construction and improvement of a domestic electric energy quality monitoring system, a large amount of voltage sag monitoring data are collected in the electric power system in each region, and the information and knowledge contained in the data provide possibility for the prediction of the voltage sag. The existing method for predicting the monitoring data mainly comprises a gray prediction model, a random process model, a deep learning algorithm and the like, wherein the algorithm has certain universality for the monitoring data, simultaneously ignores the characteristic of the voltage sag monitoring data, and has certain influence on the prediction precision.
Disclosure of Invention
The invention aims to provide a method for measuring and calculating the occurrence time of a voltage sag event, so as to improve the prediction accuracy of the occurrence time of the voltage sag event.
The technical scheme of the invention is as follows:
a method for measuring and calculating the occurrence time of a voltage sag event, comprising the steps of:
a step of monitoring data homologous aggregation, in which voltage sag monitoring data of a target site are collected, the duration from interruption of voltage sag to recovery of a production process is taken as an aggregation target, and the collected voltage sag monitoring data are subjected to homologous aggregation;
blurring step of sample data of the sag interval time, namely differentially processing the voltage sag monitoring data subjected to homologous aggregation to obtain sample data of the voltage sag interval time of the target site; establishing a domain of interval time sample data, and dividing the domain interval by adopting a fuzzy C-means algorithm; aiming at the divided domain interval, a membership function of voltage sag interval time is formulated, and an interval time sequence is converted into a fuzzy time sequence;
a fuzzy relation establishing step of establishing a fuzzy relation matrix among elements in a fuzzy time sequence based on the fuzzified sample data;
and predicting future time, namely predicting by using a maximum-minimum principle, and defuzzifying by adopting a weighted average method to realize the calculation of the future voltage sag event.
Preferably, the voltage sag monitoring data includes an event sequence number, an occurrence time, a site name, a sag amplitude and a duration.
Preferably, the step of monitoring data homology aggregation includes: let s= (T, O) be the collected voltage sag monitoring dataset; t= (T 1 ,t 2 ,…,t n ) Is an occurrence time set and is arranged according to time ascending order; o= (O) 1 ,o 2 ,…,o n ) Is a sag characteristic value; classifying the occurrence time sets according to the rule of the formula (1);
wherein T is i Is a subset of the occurrence times; t is t p 、t q The p-th and q-th occurrence time data in T are boundary elements of the occurrence time subset; selecting T according to the principle of minimum amplitude and maximum duration i Corresponding monitoring data in the database is used as event characteristic values after subset aggregation;
after homologous polymerization, a new set of voltage sag occurrence times is obtained as Nt= (NT) 1 ,nt 2 ,…,nt m )。
Preferably, the blurring step of the dip interval time sample data includes:
differential processing NT yields an interval time set Dt= (DT) of voltage sag events 1 ,dt 2 ,…,dt m-1 )；
Define interval time domain u= [ min { DT } -D ] 1 ,max{DT}+D 2 ]Min { DT } and max { DT } are the minimum and minimum values in the set of interval times, D 1 And D 2 Is a proper positive value to perfect the domain boundary;
1) Voltage sag interval time domain division
Dividing the domain of voltage sag interval time by adopting a fuzzy C-means clustering method, firstly determining the number C of intervals expected to be divided, and randomly determining the initial interval center vectorEstablishing an optimization function J aiming at minimizing the geometric distance from each sample data to the nearest interval center m As shown in formula (2);
d ij ＝||dt i -c j || (3)
wherein (m-1) is the number of samples, u ij Membership of the ith sample to the center of the jth interval, d ij For the geometrical distance from the ith sample to the center of the jth interval, dt i For the ith sample value, c j The center value of the jth interval;
continuously and iteratively calculating through the formulas (4) and (5), and calculating an optimal interval center vector c= (c) according to the constraint condition of the formula (6) 1 ,…,c C )；
Wherein h is the iteration number and epsilon is the error threshold;
thus, C intervals ζ are obtained 1 ＝[min{DT}-D 1 ,b 1 ],ξ 2 ＝[b 1 ,b 2 ],…,ξ C ＝[b C-1 ,max{DT}+D 2 ]Wherein, the method comprises the steps of, wherein,ξ 1 is the left boundary of the domain, ζ C The right boundary of the domain is the right boundary of the domain, and other interval boundary calculation methods are shown in the formula (7);
b i ＝(c i+1 -c i )/2,i＝1,2,...,(C-1) (7)
2) Voltage sag interval time sample blurring
Definition of fuzzy sets { A|A ] in the Voltage sag Interval Domain U i ＝(μ 1 (dt i ),μ 2 (dt i ),…,μ C (dt i ) I=1, 2, …, (m-1) }, wherein a i For interval time sample dt i Corresponding fuzzy subset, mu 1 (dt i ),μ 2 (dt i ),…,μ C (dt i ) Dt respectively i In interval xi 1 ,ξ 2 ,…,ξ C The membership degree is calculated according to the formula (8);
preferably, the fuzzy relation establishing step includes: ambiguity of interval time of the t-th sequence number subset A t As a standard matrix, as shown in formula (9);
B(t)＝[μ 1 (dt t ) μ 2 (dt t ) … μ C (dt t )] (9)
will A t Constructing a fuzzy relation matrix of the first 5 fuzzy subsets as shown in a formula (10);
constructing a fuzzy relation matrix as shown in a formula (11);
wherein R is ij ＝μ j (dt (t-i) )×μ j (dt t ) 1.ltoreq.i.ltoreq.5, 1.ltoreq.j.ltoreq.C, "×" represents multiplication, "(t-i)" represents "t minus i".
Further preferably, the future time prediction step includes: calculating fuzzy subset A of (t+1) time points by' maximum-minimum principle t+1 As shown in formula (12);
pair A by adopting a weighted average method t+1 Deblurring, as shown in formula (13);
dt t+1 ＝map(A t+1 )·c T (13)
wherein map (A) t+1 ) Representation of pair A t+1 Vector normalization, c T Is the transposition of the interval center vector;
the interval time dt obtained by (13) t+1 Calculating the next time nt of occurrence of the sag according to the formula (14) t+1 Completion prediction
nt t+1 ＝nt t +dt t+1 (14)
The beneficial effects of the invention are as follows:
1. the invention provides a voltage sag homologous aggregation algorithm considering industrial sensitive process interruption recovery time, establishes a corresponding relation of 'monitoring data-process interruption', aims at user perception to reduce data redundancy, and provides an accurate data basis for sag prediction; the data blurring method considering the uncertainty of the interval time is provided, the corresponding relation of monitoring data, process interruption and power grid disturbance is initially established, the information content contained in single sample data is expanded by a membership method, and reasonable description of complex power grid disturbance is realized; the fuzzy time sequence-based voltage sag occurrence time algorithm is provided, the measurement and calculation of the occurrence time of the next voltage sag event are realized, the early warning of future disturbance of sensitive users and power grid companies is facilitated, the production plan is reasonably arranged, and the sag risk is avoided.
Drawings
FIG. 1 is a flow chart of a method of measuring the occurrence time of a voltage sag event;
FIG. 2 is a comparison of data before and after homologous polymerization;
FIG. 3 is a 10 site prediction sequence alignment.
Detailed Description
The present invention is described in the following embodiments in conjunction with the accompanying drawings to assist those skilled in the art in understanding and implementing the invention. The following examples and technical terms therein should not be construed to depart from the technical knowledge of the art unless otherwise indicated.
A method for measuring and calculating the occurrence time of a voltage sag event is shown in the accompanying figure 1, and comprises the following main steps: (1) monitoring a data homology aggregation step:
the power quality monitoring system collects node temporary steady state data, and simultaneously, each monitoring node does not consider the relevance between power disturbance events, so that data redundancy is caused, and the power quality monitoring system is mainly characterized in that: 1) The voltage sag event caused by the same disturbance can be recorded by a plurality of monitoring nodes nearby, so that the system collects a plurality of pieces of different sag data; 2) Due to the fact that multiple stages of sag are formed due to the fact that lines are reclosed for many times, different pieces of sag data are recorded for the same disturbance by adjacent monitoring nodes. Such redundant data, if used for prediction, inevitably leads to a decrease in prediction performance. Therefore, taking the user perception as a prediction target, the time from the step-down of the production process suffered by a typical sensitive industrial user to the recovery of the production process is 1 hour, and S= (T, O) is set as a collected voltage sag monitoring data set; t= (T 1 ,t 2 ,…,t n ) Is an occurrence time set and is arranged according to time ascending order; o= (O) 1 ,o 2 ,…,o n ) Is a sag characteristic value, such as sag amplitude, duration, etc. The set of occurrence times is classified according to the rule of equation (15).
Wherein T is i Is a subset of the occurrence times; t is t p 、t q Is the p-th and q-th time of occurrence data in T, and is the boundary element of the subset of times of occurrence. According to the principle of "1h" above, the division into m subsets of time of occurrence may be considered as each subset of time of occurrence T i Corresponds to a voltage sag event which can cause the interruption of the production process, and is identical with T i The primary time with the smallest intermediate amplitude and the longest duration is associated with T i Other voltage sag events in (a) may be perceived by the user as production breaks. Therefore, T is selected according to the principle of' minimum amplitude and maximum duration i As the event characteristic value after subset aggregation. Since the invention only researches the occurrence time of the dip, the magnitude and the duration are not excessively tired.
After homologous polymerization, a new set of voltage sag occurrence times is obtained as Nt= (NT) 1 ,nt 2 ,…,nt m ) Information extraction of the corresponding relation of the monitoring data and the process interruption is realized.
(2) Blurring the voltage sag interval time sample data:
to facilitate extraction of data features in NT, differential processing is performed to obtain an interval time set Dt= (DT) of voltage dip events 1 ,dt 2 ,…,dt m-1 ). Next, an interval domain u= [ min { DT } -D = [ min { DT } -D ] is defined 1 ,max{DT}+D 2 ]Min { DT } and max { DT } are the minimum and minimum values in the set of interval times, D 1 And D 2 Is a suitable positive value to perfect the domain boundaries.
1) Voltage sag interval time domain division
The essence of the predicted voltage sag interval time is the prediction of future grid disturbances, and due to the inherent uncertainty of the grid disturbances, it is determined that the occurrence time of each type of grid disturbance may vary in one interval, and the occurrence time intervals of different grid disturbances may also be different. Therefore, the domain of the voltage sag interval time needs to be divided to realize the corresponding relation of the monitoring data and the power grid disturbance. To achieve reasonable division of the domain, we adoptThe method of fuzzy C-means clustering is used, firstly, the number C of intervals expected to be divided is determined (C=8 is generally selected), and the initial interval center vector is randomly determined Establishing an optimization function J aiming at minimizing the geometric distance from each sample data to the nearest interval center m As shown in formula (16).
d ij ＝||dt i -c j || (17)
Wherein (m-1) is the number of samples, u ij Membership of the ith sample to the center of the jth interval, d ij For the geometrical distance from the ith sample to the center of the jth interval, dt i For the ith sample value, c j Is the jth interval center value.
Continuously and iteratively calculating through the formulas (18) and (19), and calculating an optimal interval center vector c= (c) according to the constraint condition of the formula (20) 1 ,…,c C )。
Where h is the number of iterations and ε is the error threshold.
Thus, C intervals ζ are obtained 1 ＝[min{DT}-D 1 ,b 1 ],ξ 2 ＝[b 1 ,b 2 ],…,ξ C ＝[b C-1 ,max{DT}+D 2 ]Wherein, xi 1 Is the left boundary of the domain, ζ C The right boundary of the domain is the right boundary of the domain, and other interval boundary calculation methods are shown in the formula (21).
b i ＝(c i+1 -c i )/2,i＝1,2,...,(C-1) (21)
2) Voltage sag interval time sample blurring
And blurring the voltage sag interval time samples aiming at the acquired interval set, so as to realize the preliminary establishment of the corresponding relation of monitoring data, process interruption and power grid disturbance. Definition of fuzzy sets { A|A ] in the Voltage sag Interval Domain U i ＝(μ 1 (dt i ),μ 2 (dt i ),…,μ C (dt i ) I=1, 2, …, (m-1) }, wherein a i For interval time sample dt i Corresponding fuzzy subset, mu 1 (dt i ),μ 2 (dt i ),…,μ C (dt i ) Dt respectively i In interval xi 1 ,ξ 2 ,…,ξ C The membership degree is calculated by the method shown in the formula (22).
Therefore, the voltage sag interval time samples in the numerical form are converted into (m-1) group C-dimensional vectors, extraction of power grid disturbance occurrence information contained in interval time sample data is achieved, and a corresponding relation of monitoring data, process interruption and power grid disturbance is initially established.
(3) Establishing a fuzzy relation:
based on the voltage sag interval time fuzzy set obtained in the upper section, if the interval time of the (t+1) th time sequence number is predicted, the interval time fuzzy subset A of the t th time sequence number t The standard matrix is represented by formula (23).
B(t)＝[μ 1 (dt t ) μ 2 (dt t ) … μ C (dt t )] (23)
Will A t The first 5 fuzzy subsets of (a) construct a fuzzy relation matrix as shown in equation (24).
Accordingly, a fuzzy relation matrix is constructed as shown in expression (25).
Wherein R is ij ＝μ j (dt (t-i) )×μ j (dt t ) 1.ltoreq.i.ltoreq.5, 1.ltoreq.j.ltoreq.C, "×" represents multiplication, "(t-i)" represents "t minus i".
(4) Future time prediction step:
the correlation degree and the change trend of the interval time and each interval in the past year can be seen from the fuzzy relation matrix, and then the fuzzy subset A of (t+1) time points is calculated through the maximum-minimum principle t+1 As shown in formula (26).
Pair A by adopting a weighted average method t+1 Deblurring, as shown in formula (27)
dt t+1 ＝map(A t+1 )·c T (27)
Wherein map (A) t+1 ) Representation of pair A t+1 Vector normalization, c T Is the transpose of the interval center vector.
The interval time dt obtained by (27) t+1 Calculating the next time nt of occurrence of the sag according to the formula (28) t+1 And (5) completing prediction.
nt t+1 ＝nt t +dt t+1 (28)
The following is a result of predicting voltage sag occurrence time data of 10 monitoring points in a certain province in China:
1) Data capacity of 10 sites
Table 1 example data capacity
2) The data before and after homologous aggregation of a certain site is shown in fig. 2.
3) 10 site prediction sequence pairs are shown in fig. 3.
4) Prediction accuracy of the proposed method
And (3) calculating a prediction error of the prediction result by applying the formula (29). Considering the error tolerance of the predicted variable to be +/-10%, and considering the prediction result to be correct when the prediction error eta is smaller than the error tolerance. Counting the number m of correctly predicted events in the total step size of the prediction η<0.1 And calculating the prediction accuracy rate of the corresponding method as the precision measurement of the event prediction result.
In dt (dt) pre Representing the predicted value, t obs Representing the sample value.
For sequence prediction results, root mean square error (root means square error, RMSE) is used to measure prediction accuracy. RMSE is sensitive to both maximum and minimum errors of the predicted values and can well reflect the prediction accuracy. The smaller the RMSE value, the higher the prediction accuracy, as shown in equation (30).
The prediction accuracy of the proposed method is shown in table 2. The proposed method shows good prediction performance in both the prediction accuracy for single event and RMSE of the predicted sequence, as a result of the prediction at the monitoring point # 2. In addition, the prediction performance of the method is good at monitoring points with a large number of samples, and the prediction accuracy of single events is basically more than 70%. The prediction accuracy at monitoring point #3 is lower, probably due to fewer data samples. RMSE is higher at monitoring points #3- #8 and #10 because the sequence trend is more complex, and although the method presented herein describes the trend substantially correctly, the variation in the sequence amplitude remains to be improved.
Prediction accuracy of the method presented in Table 2
The invention is described in detail above with reference to the drawings and examples. It should be understood that the description of all possible embodiments is not intended to be exhaustive or to limit the inventive concepts disclosed herein to the precise form disclosed. The technical characteristics of the above embodiments are selected and combined, specific parameters are experimentally changed by those skilled in the art, or the technical means disclosed in the present invention are conventionally replaced by the prior art in the technical field, which is not paid with creative work, and all the specific embodiments are implicitly disclosed in the present invention.
Claims (4)
1. A method for measuring and calculating the occurrence time of a voltage sag event, comprising the steps of:
a step of monitoring data homologous aggregation, in which voltage sag monitoring data of a target site are collected, the duration from interruption of voltage sag to recovery of a production process is taken as an aggregation target, and the collected voltage sag monitoring data are subjected to homologous aggregation;
blurring step of sample data of the sag interval time, namely differentially processing the voltage sag monitoring data subjected to homologous aggregation to obtain sample data of the voltage sag interval time of the target site; establishing a domain of interval time sample data, and dividing the domain interval by adopting a fuzzy C-means algorithm; aiming at the divided domain interval, a membership function of voltage sag interval time is formulated, and an interval time sequence is converted into a fuzzy time sequence;
a fuzzy relation establishing step of establishing a fuzzy relation matrix among elements in a fuzzy time sequence based on the fuzzified sample data;
predicting future time, namely predicting by using a maximum-minimum principle, and defuzzifying by adopting a weighted average method to realize the calculation of a future voltage sag event;
the step of monitoring data homologous aggregation comprises the following steps: let s= (T, O) be the collected voltage sag monitoring dataset; t= (T 1 ,t 2 ,…,t n ) Is an occurrence time set and is arranged according to time ascending order; o= (O) 1 ,o 2 ,…,o n ) Is a sag characteristic value; classifying the occurrence time sets according to the rule of the formula (1);
wherein T is i Is a subset of the occurrence times; t is t p 、t q The p-th and q-th occurrence time data in T are boundary elements of the occurrence time subset; selecting T according to the principle of minimum amplitude and maximum duration i Corresponding monitoring data in the database is used as event characteristic values after subset aggregation;
after homologous polymerization, a new set of voltage sag occurrence times is obtained as Nt= (NT) 1 ,nt 2 ,…,nt m )；
The step of blurring the dip interval time sample data comprises the following steps:
differential processing NT yields an interval time set Dt= (DT) of voltage sag events 1 ,dt 2 ,…,dt m-1 )；
Define interval time domain u= [ min { DT } -D ] 1 ,max{DT}+D 2 ]Min { DT } and max { DT } are the minimum and minimum values in the set of interval times, D 1 And D 2 Is a proper positive value to perfect the domain boundary;
1) Voltage sag interval time domain division
Dividing the domain of voltage sag interval time by adopting a fuzzy C-means clustering method, firstly determining the number C of intervals expected to be divided, and randomly determining the initial interval center vectorEstablishing an optimization function J aiming at minimizing the geometric distance from each sample data to the nearest interval center m As shown in formula (2);
d ij ＝||dt i -c j || (3)
wherein (m-1) is the number of samples, u ij Membership of the ith sample to the center of the jth interval, d ij For the geometrical distance from the ith sample to the center of the jth interval, dt i For the ith sample value, c j The center value of the jth interval;
continuously and iteratively calculating through the formulas (4) and (5), and calculating an optimal interval center vector c= (c) according to the constraint condition of the formula (6) 1 ,…,c C )；
Wherein h is the iteration number and epsilon is the error threshold;
thus, C intervals ζ are obtained 1 ＝[min{DT}-D 1 ,b 1 ],ξ 2 ＝[b 1 ,b 2 ],…,ξ C ＝[b C-1 ,max{DT}+D 2 ]Wherein, xi 1 Is the left boundary of the domain, ζ C The right boundary of the domain is the right boundary of the domain, and other interval boundary calculation methods are shown in the formula (7);
b i ＝(c i+1 -c i )/2,i＝1,2,...,(C-1) (7)
2) Voltage sag interval time sample blurring
Definition of fuzzy sets { A|A ] in the Voltage sag Interval Domain U i ＝(μ 1 (dt i ),μ 2 (dt i ),…,μ C (dt i ) I=1, 2, …, (m-1) }, wherein a i For interval time sample dt i Corresponding fuzzy subset, mu 1 (dt i ),μ 2 (dt i ),…,μ C (dt i ) Dt respectively i In interval xi 1 ,ξ 2 ,…,ξ C The membership degree is calculated according to the formula (8);
2. the method of measuring voltage sag event occurrence time of claim 1, wherein the voltage sag monitoring data includes an event number, an occurrence time, a site name, a sag amplitude, and a duration.
3. The method for measuring and calculating the occurrence time of a voltage sag event according to claim 1, wherein the fuzzy relation establishing step includes: ambiguity of interval time of the t-th sequence number subset A t As a standard matrix, as shown in formula (9);
B(t)＝[μ 1 (dt t ) μ 2 (dt t ) …μ C (dt t )] (9)
will A t Constructing a fuzzy relation matrix of the first 5 fuzzy subsets as shown in a formula (10);
constructing a fuzzy relation matrix as shown in a formula (11);
wherein R is ij ＝μ j (dt (t-i) )×μ j (dt t ) 1.ltoreq.i.ltoreq.5, 1.ltoreq.j.ltoreq.C, "×" represents multiplication, "(t-i)" represents "t minus i".
4. A method of measuring voltage sag event occurrence times as set forth in claim 3 wherein said future time predicting step comprises: calculating fuzzy subset A of (t+1) time points by' maximum-minimum principle t+1 As shown in formula (12);
pair A by adopting a weighted average method t+1 Deblurring, as shown in formula (13);
dt t+1 ＝map(A t+1 )·c T (13)
wherein map (A) t+1 ) Representation of pair A t+1 Vector normalization, c T Is the transposition of the interval center vector;
the interval time dt obtained by (13) t+1 Calculating the next time nt of occurrence of the sag according to the formula (14) t+1 Completion prediction
nt t+1 ＝nt t +dt t+1 (14)。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111183809.9A CN113886183B (en) | 2021-10-11 | 2021-10-11 | Method for measuring and calculating occurrence time of voltage sag event |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111183809.9A CN113886183B (en) | 2021-10-11 | 2021-10-11 | Method for measuring and calculating occurrence time of voltage sag event |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113886183A CN113886183A (en) | 2022-01-04 |
| CN113886183B true CN113886183B (en) | 2023-11-21 |
Family
ID=79006162
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111183809.9A Active CN113886183B (en) | 2021-10-11 | 2021-10-11 | Method for measuring and calculating occurrence time of voltage sag event |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113886183B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116014731B (en) * | 2023-03-27 | 2023-07-18 | 山东省工业技术研究院 | Method for short-term prediction of electric load by using fuzzy time sequence |
| CN116298482B (en) * | 2023-05-25 | 2023-08-01 | 常州满旺半导体科技有限公司 | Intelligent early warning system and method for voltage data monitoring |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101625381A (en) * | 2009-08-06 | 2010-01-13 | 江苏省电力试验研究院有限公司 | Fuzzy evaluation method of voltage sags |
| CN107462764A (en) * | 2017-09-25 | 2017-12-12 | 南京灿能电力自动化股份有限公司 | A kind of voltage dip detection and the automatic segmentation method portrayed |
| CN111220842A (en) * | 2020-01-16 | 2020-06-02 | 华南理工大学 | Voltage sag disturbance event discovery and positioning method, device, equipment and medium |
| CN112731049A (en) * | 2020-12-08 | 2021-04-30 | 深圳供电局有限公司 | Clock synchronization abnormity monitoring method and device and computer readable storage medium |
-
2021
- 2021-10-11 CN CN202111183809.9A patent/CN113886183B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101625381A (en) * | 2009-08-06 | 2010-01-13 | 江苏省电力试验研究院有限公司 | Fuzzy evaluation method of voltage sags |
| CN107462764A (en) * | 2017-09-25 | 2017-12-12 | 南京灿能电力自动化股份有限公司 | A kind of voltage dip detection and the automatic segmentation method portrayed |
| CN111220842A (en) * | 2020-01-16 | 2020-06-02 | 华南理工大学 | Voltage sag disturbance event discovery and positioning method, device, equipment and medium |
| CN112731049A (en) * | 2020-12-08 | 2021-04-30 | 深圳供电局有限公司 | Clock synchronization abnormity monitoring method and device and computer readable storage medium |
Non-Patent Citations (1)
| Title |
|---|
| 基于实测数据的陕西电网电压暂降特征分布分析;胡长青;曹爱民;黄研利;王建波;刘长荣;唐正聪;贺瀚青;周冬;;智慧电力(第09期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113886183A (en) | 2022-01-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wen et al. | A new ensemble residual convolutional neural network for remaining useful life estimation | |
| CN113886183B (en) | Method for measuring and calculating occurrence time of voltage sag event | |
| CN105279365B (en) | For the method for the sample for learning abnormality detection | |
| ElNozahy et al. | A probabilistic load modelling approach using clustering algorithms | |
| CN103135009B (en) | Electric appliance detection method and system based on user feedback information | |
| CN109978229A (en) | The method that the full battery core multi-point temperature of a kind of pair of power battery pack and tie point temperature carry out thermal runaway prediction | |
| CN109472088B (en) | Shale gas-conditioned production well production pressure dynamic prediction method | |
| CN108074015B (en) | Ultra-short-term prediction method and system for wind power | |
| CN109492748B (en) | Method for establishing medium-and-long-term load prediction model of power system based on convolutional neural network | |
| CN112434390B (en) | PCA-LSTM bearing residual life prediction method based on multi-layer grid search | |
| CN113723010A (en) | Bridge damage early warning method based on LSTM temperature-displacement correlation model | |
| CN115828140A (en) | Neighborhood mutual information and random forest fusion fault detection method, system and application | |
| CN109066651B (en) | Method for calculating limit transmission power of wind power-load scene | |
| CN113657678A (en) | Power grid power data prediction method based on information freshness | |
| Cheng et al. | A novel fuzzy time series forecasting method based on fuzzy logical relationships and similarity measures | |
| CN115456306A (en) | Bus load prediction method, system, equipment and storage medium | |
| CN112305441A (en) | Power battery health state assessment method under integrated clustering | |
| CN112287980A (en) | Power battery screening method based on typical feature vector | |
| CN115481726A (en) | Industrial robot complete machine health assessment method and system | |
| Van Tinh et al. | A new hybrid fuzzy time series forecasting model based on combining fuzzy c-means clustering and particle swam optimization | |
| CN116776245A (en) | Three-phase inverter equipment fault diagnosis method based on machine learning | |
| CN117034762A (en) | Composite model lithium battery life prediction method based on multi-algorithm weighted sum | |
| CN111612227A (en) | Load prediction method based on K-means clustering and bat optimization neural network | |
| CN113240904B (en) | Traffic flow prediction method based on feature fusion | |
| Oprea et al. | Electricity load profile calculation using self-organizing maps |
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 |