CN111476173B - Power distribution network voltage sag source identification method based on BAS-SVM - Google Patents
Power distribution network voltage sag source identification method based on BAS-SVM Download PDFInfo
- Publication number
- CN111476173B CN111476173B CN202010273237.2A CN202010273237A CN111476173B CN 111476173 B CN111476173 B CN 111476173B CN 202010273237 A CN202010273237 A CN 202010273237A CN 111476173 B CN111476173 B CN 111476173B
- Authority
- CN
- China
- Prior art keywords
- svm
- bas
- voltage sag
- follows
- calculation formula
- 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 26
- 238000012360 testing method Methods 0.000 claims abstract description 20
- 230000009466 transformation Effects 0.000 claims abstract description 18
- 238000012549 training Methods 0.000 claims abstract description 17
- 239000011159 matrix material Substances 0.000 claims abstract description 7
- 238000012795 verification Methods 0.000 claims abstract description 3
- 238000004364 calculation method Methods 0.000 claims description 30
- 238000005070 sampling Methods 0.000 claims description 14
- 230000006698 induction Effects 0.000 claims description 10
- 230000035772 mutation Effects 0.000 claims description 10
- 241000254173 Coleoptera Species 0.000 claims description 9
- 239000013598 vector Substances 0.000 claims description 9
- 238000004422 calculation algorithm Methods 0.000 claims description 7
- 238000010606 normalization Methods 0.000 claims description 7
- 241001481710 Cerambycidae Species 0.000 claims description 5
- 239000000284 extract Substances 0.000 claims description 4
- 238000012545 processing Methods 0.000 claims description 4
- 230000000630 rising effect Effects 0.000 claims 1
- 238000004088 simulation Methods 0.000 abstract description 7
- 230000000694 effects Effects 0.000 abstract description 4
- 238000012706 support-vector machine Methods 0.000 description 20
- 230000008569 process Effects 0.000 description 4
- 238000000926 separation method Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 2
- 238000013145 classification model Methods 0.000 description 2
- 238000002790 cross-validation Methods 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 238000010845 search algorithm Methods 0.000 description 2
- 230000006978 adaptation Effects 0.000 description 1
- 230000001174 ascending effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000002431 foraging effect Effects 0.000 description 1
- 230000002068 genetic effect Effects 0.000 description 1
- 230000002401 inhibitory effect Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000006855 networking Effects 0.000 description 1
- 239000002245 particle Substances 0.000 description 1
- 238000005192 partition Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000011426 transformation method Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Biophysics (AREA)
- Molecular Biology (AREA)
- Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Power Engineering (AREA)
- Signal Processing (AREA)
- General Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention discloses a power distribution network voltage sag source identification method based on a BAS-SVM, which comprises the steps of obtaining relevant curves of various voltage sag signals by improving a result modulus matrix of S transformation; extracting a plurality of groups of characteristic index data of various voltage sag signals by applying improved S transformation; optimizing penalty factors and kernel function parameters of the SVM through the BAS to construct a BAS-SVM classifier; normalizing the extracted characteristic index data; the training set and the testing set are divided by adopting a 5-time cross verification method, the training set is used as a training sample of the BAS-SVM classifier, and the testing set is used as a testing sample of the BAS-SVM classifier to test, so that 100% accurate identification of a power distribution network voltage sag source is realized. Simulation results show that compared with CV-SVM, GA-SVM, PSO-SVM and other methods, the method can effectively improve the identification accuracy of different voltage sag sources and has better classification effect.
Description
Technical Field
The invention belongs to the technical field of power quality detection of a power distribution network, and particularly relates to a power distribution network voltage sag source identification method based on a BAS-SVM.
Background
Voltage sag is one of the most serious problems in the power quality, and the Institute of Electrical and Electronics Engineers (IEEE) defines voltage sag as the instantaneous reduction of the effective value of the supply voltage to 10% -90% of the rated value at the system frequency, with a duration of typically 0.5-30 power frequency cycles. In recent years, with the continuous popularization and development of automation and networking, the proportion of digital processors and power electronic components in power systems is increasing. These modern loads are more sensitive to voltage sags than traditional loads that are affected only by a power outage. Short-circuit faults, induction motor starting and switching events of transformers in a power transmission and distribution system are three main reasons for causing voltage sag. Therefore, the method has great significance in accurately classifying and identifying different voltage sag sources, and is a precondition for inhibiting and relieving the voltage sag.
The related characteristic index extraction of the voltage sag signals is a key factor affecting the classification and identification of voltage sag sources. The defect of standard S conversion can be effectively overcome by improving the S conversion, and the characteristic index of the voltage sag signal can be extracted more accurately. The Support Vector Machine (SVM) can effectively identify different types of voltage sags, but the penalty factors and the values of kernel function parameters of the support vector machine have great influence on classification effects. The longhorn beetle whisker search algorithm (BAS) is an intelligent algorithm for simulating the foraging process of the longhorn beetles, and can effectively solve the problem of optimizing multiple target parameters. Based on the current research of the problems at home and abroad, the invention further analyzes and researches a method for improving the classification and identification accuracy of the voltage sag sources by using improved S transformation and BAS, mainly extracts characteristic indexes of different voltage sag types accurately based on the improved S transformation, optimizes SVM, penalty factors and kernel function parameters by using BAS, and constructs a BAS-SVM classifier, thereby realizing accurate identification of the voltage sag sources of the power distribution network.
Disclosure of Invention
Aiming at the defects of the prior art, the invention provides a power distribution network voltage sag source identification method based on a BAS-SVM, which utilizes improved S transformation to extract a relevant amplitude curve and 16 characteristic indexes of a voltage sag signal, obtains optimal values of SVM punishment factors and kernel function parameters through the BAS, constructs a BAS-SVM classifier, carries out normalization processing on index data, adopts 5 times of cross validation to divide a training set and a testing set, and inputs the training set and the testing set into the BAS-SVM classifier to realize identification of different voltage sag sources of the power distribution network.
In order to achieve the technical purpose, the invention adopts the following technical scheme:
a power distribution network voltage sag source identification method based on a BAS-SVM comprises the following steps:
step 1, obtaining a fundamental frequency amplitude curve and a frequency amplitude envelope curve of various voltage sag signals by improving a result mode matrix of S transformation;
step 2, extracting a plurality of groups of characteristic index data of various voltage sag signals by applying improved S transformation;
step 3, normalizing a plurality of groups of characteristic index data extracted from each voltage sag signal;
step 4, optimizing penalty factors and kernel function parameters of the SVM through the BAS, and constructing a BAS-SVM classifier;
and 5, dividing the normalized plurality of sets of characteristic index data in the step 5 into a training set and a testing set by adopting a 5-time cross verification method, wherein the training set is used as a training sample of the BAS-SVM classifier, and the testing set is used as a testing sample of the BAS-SVM classifier for testing, so that 100% accurate identification of a power distribution network voltage sag source is realized.
In order to optimize the technical scheme, the specific measures adopted further comprise:
the voltage sag signal in the step 1 includes 5 kinds of voltage sag information caused by single-phase short-circuit fault, two-phase short-circuit fault, three-phase short-circuit fault, induction motor starting and transformer operation.
The characteristic indexes in the step 2 include mean value, standard deviation, RMS value, dip depth, energy, kurtosis, number of abrupt points, shannon entropy, logarithmic energy entropy, slope of down fundamental frequency amplitude, slope of up fundamental frequency amplitude, second harmonic content, skewness, waveform coefficient, crest factor and dip time ratio.
The number of the mutation points is determined by 11 times of fundamental frequency row vectors in a modular matrix of a result of improved S transformation, and the number of the mutation points is uniformly set to be 0 when the transformer is put into operation;
the second harmonic content is as follows:
wherein: u (U) 2 For the frequency value at the 32 nd sampling point in the frequency amplitude envelope curve, U 1 The frequency value at the 17 th sampling point in the frequency amplitude envelope curve is obtained;
the sag time ratio is as follows:
wherein: t (T) 1 For duration of dip, T 2 Is the time for the sag to be stable;
the dip depth is as follows:
MF＝U sag /U ref
in U ref And U sag Representing the effective values before and at the time of voltage sag, respectively;
the mean value calculation formula is as follows:
the standard deviation calculation formula is as follows:
the RMS value calculation formula is as follows:
the energy calculation formula is as follows:
the kurtosis calculation formula is as follows:
the shannon entropy calculation formula is as follows:
the logarithmic energy entropy calculation formula is as follows:
the deflection calculation formula is as follows:
the calculation formula of the waveform coefficient is as follows:
the crest factor calculation formula is as follows:
where xi is the amplitude of a certain sampling point in the disturbance sample vector x, N is the number of sampling points contained in the disturbance sample, and P is A, B, C three phases.
The normalization processing in the step 3 is performed on a plurality of sets of characteristic index data extracted from each voltage sag signal, and the normalization formula is as follows:
wherein: x is original data, X min X is the minimum value in the original data max Is the maximum value in the original data.
In the step 4, the penalty factor and the kernel function parameter of the SVM are optimized by the BAS, the optimizing result is that the penalty factor is 0.3798, and the kernel function parameter is 5.8570.
The building of the BAS-SVM classifier in the step 4 comprises the following steps:
(1) Constructing a random vector representing the direction of the longhorn beetles, and defining a space dimension k;
(2) Setting step factors as follows:
δ t ＝δ t-1 *eta
wherein: eta is a number near 1 between intervals [0,1], taking 0.95;
(3) Taking the identification accuracy of the SVM as an adaptability function;
(4) Initializing the space position of the longhorn beetles as an initial solution set of the BAS algorithm and storing the initial solution set in X best In (a) and (b);
(5) Calculating initial fitness function value of the longicorn at initial position according to fitness function, and storing in f best The fitness function is the identification accuracy of the SVM;
(6) Iteratively updating the positions of the left long antenna and the right long antenna of the longhorns, respectively solving the fitness function values of the left whisker and the right whisker when the longhorns are at the current positions, and updating X if the current function value is better than the initial fitness function value best And f best ；
(7) Stopping iteration if the iteration times are reached, turning to the step (8), otherwise, returning to the step (6) to continue iteration;
(8) And obtaining the optimal value of the fitness function, the SVM penalty factor and the optimal value of the kernel function parameter.
The invention has the following beneficial effects:
compared with CV-SVM, GA-SVM, PSO-SVM and other methods, the invention can more effectively improve the identification accuracy of different voltage sag sources and has better classification effect.
Drawings
Fig. 1 is a voltage sag model diagram caused by a short-circuit fault.
Fig. 2 is a graph of a voltage sag model caused by the startup of an induction motor.
Fig. 3 is a diagram of a voltage sag model caused by transformer operation.
Fig. 4 is a plot of fundamental frequency amplitude caused by a short circuit fault.
Fig. 5 is a plot of the amplitude of the fundamental frequency caused by the start of an induction motor.
Fig. 6 is a plot of the magnitude of the fundamental frequency caused by the operation of the transformer.
Fig. 7 is a frequency magnitude envelope resulting from a short circuit fault.
Fig. 8 is a frequency amplitude envelope resulting from the start of an induction motor.
Fig. 9 is a frequency amplitude envelope resulting from transformer operation.
Fig. 10 shows the number of abrupt points of a short-circuit fault.
Fig. 11 shows the number of mutation points for starting the induction motor.
Fig. 12 is a process of constructing a BAS-SVM classifier.
Detailed Description
Embodiments of the present invention are described in further detail below with reference to the accompanying drawings.
The invention discloses a method for identifying a voltage sag source of a power distribution network based on a BAS-SVM, which comprises the following steps:
step 1, obtaining a fundamental frequency amplitude curve and a frequency amplitude envelope curve of 5 voltage sag signals by improving a result mode matrix of S transformation, as shown in figures 4-9;
step 2, extracting 100 groups of characteristic index data of mean value, standard deviation, RMS value, sag depth, energy, kurtosis, mutation point number, shannon entropy, logarithmic energy entropy, fundamental frequency amplitude descending slope, fundamental frequency amplitude ascending slope, second harmonic content, skewness, waveform coefficient, crest factor and sag time ratio of 5 voltage sag signals by using improved S transformation;
namely, for 5 different voltage sag types, 100 groups of characteristic index data are extracted from each type, and the total is 500 groups of characteristic index data.
Step 3, normalizing 100 groups of the 16 characteristic index data extracted from each voltage sag signal;
step 4, optimizing penalty factors and kernel function parameters of the SVM through the BAS, and constructing a BAS-SVM classifier;
the SVM is a bi-classification model linear classifier defined in feature space with maximum separation distance, the basic idea of which is to solve for the separation hyperplane that not only correctly partitions the training dataset, but also has the maximum geometric separation. The BAS is an intelligent algorithm similar to a genetic algorithm, a particle swarm algorithm and the like, which is proposed in 2017, and can be applied to the optimizing process of multiple objective functions, and the fitness function is defined as the identification accuracy of the SVM.
The steps for constructing the BAS-SVM classifier are shown in fig. 12, and include:
(1) A random vector is constructed to represent the longhorn beetle whisker orientation and define the spatial dimension k.
(2) Setting a step size factor. The step size can be divided into a fixed step size and a variable step size, in order to ensure that the searching process does not fall into local optimum, the step size is set as the variable step size, and the initial step size is set as large as possible as follows:
δ t ＝δ t-1 *eta
wherein: eta is a number near 1 between intervals [0,1], taking 0.95.
(3) And taking the identification accuracy of the SVM as a fitness function.
(4) Initializing the space position of the longhorn beetles as an initial solution set of the BAS algorithm and storing the initial solution set in X best Is a kind of medium.
(5) Calculating initial fitness function value of the longicorn at initial position according to fitness function, and storing in f best The fitness function is the recognition accuracy of the SVM.
(6) Iteratively updating the positions of the left long antenna and the right long antenna of the longicorn, and respectively solving the fitness function values of the left whisker and the right whisker when the longicorn is at the current position. If the current function value is better than the initial fitness function value, updating X best And f best 。
(7) Stopping iteration if the iteration times are reached, turning to the step (8), otherwise, returning to the step (6) to continue iteration.
(8) And obtaining the optimal value of the fitness function, the SVM penalty factor and the optimal value of the kernel function parameter.
And 5, dividing the 100 groups of characteristic index data normalized in the step 5 into a training set and a testing set (50 groups of training set and testing set respectively) by adopting a 5-time cross validation method, wherein the training set is used as a training sample of the BAS-SVM classifier, and the testing set is used as a testing sample of the BAS-SVM classifier to test, so that 100% accurate identification of a power distribution network voltage sag source is realized, as shown in a table 1.
In an embodiment, the voltage sag signal in step 1 includes 5 kinds of voltage sag information caused by a single-phase short-circuit fault, a two-phase short-circuit fault, a three-phase short-circuit fault, an induction motor start-up and a transformer operation.
In the embodiment, the number of the mutation points is determined by 11 times of fundamental frequency row vectors in a mode matrix of a result of improved S transformation, and the number of the mutation points cannot be accurately extracted due to a large number of harmonic components when the transformer is put into operation, and the mutation points are uniformly set to 0, as shown in fig. 10 and 11.
The second harmonic content is as follows:
wherein: u (U) 2 For the frequency value at the 32 nd sampling point in the frequency amplitude envelope curve, U 1 The frequency value at the 17 th sampling point in the frequency amplitude envelope curve is obtained;
the sag time ratio is as follows:
wherein: t (T) 1 For duration of dip, T 2 Is the time for the sag to be stable;
the dip depth is as follows:
MF＝U sag /U ref
in U ref And U sag Representing the effective values before and at the time of voltage sag, respectively;
the mean value calculation formula is as follows:
the standard deviation calculation formula is as follows:
the RMS value calculation formula is as follows:
the energy calculation formula is as follows:
the kurtosis calculation formula is as follows:
the shannon entropy calculation formula is as follows:
the logarithmic energy entropy calculation formula is as follows:
the deflection calculation formula is as follows:
the calculation formula of the waveform coefficient is as follows:
the crest factor calculation formula is as follows:
where xi is the amplitude of a certain sampling point in the disturbance sample vector x, N is the number of sampling points contained in the disturbance sample, and P is A, B, C three phases.
In the embodiment, the normalization processing in step 3 extracts 100 sets of the above 16 feature index data from each voltage sag signal, where the normalization formula is as follows:
wherein: x is original data, X min X is the minimum value in the original data max Is the maximum value in the original data.
In the embodiment, in the step 4, the penalty factor and the kernel function parameter of the SVM are optimized by the BAS, and the optimizing result is that the penalty factor is 0.3798 and the kernel function parameter is 5.8570.
The method of the invention is verified by simulation as follows:
and establishing simulation models of 5 voltage dips caused by single-phase short-circuit faults, two-phase short-circuit faults, three-phase short-circuit faults, induction motor starting and transformer operation based on MATLAB/Simulink, and performing simulation experiments, wherein the simulation models are shown in figures 1-3.
The power supply voltage grade of the power system is 11kv, the capacity is 30 MV.A, the transformation ratio of the transformer is 11kv/0.4kv, the sampling frequency is set to 1600Hz during simulation, the sampling point number is set to 512 points, and the fundamental frequency takes the power frequency of 50Hz. For a short-circuit fault, changing the starting and stopping time, the fault position and the line load of the fault; for induction motor starting, varying the capacity of the motor and the starting time of the motor; for transformer operation, the capacity, connection mode, switching time and line load of the transformer are changed. 100 groups of sample data of 5 voltage dip types are obtained through simulation, wherein 50 groups of data of each voltage dip type are used as training samples of the BAS-SVM classifier, and the other 50 groups of data are used as test samples.
From the data in table 1, it can be seen that: compared with standard S transformation, the improved S transformation can extract the characteristic indexes of different voltage sag signals more accurately, and is more beneficial to classifying and identifying the signals by a classifier; based on the improved S transformation, 16 characteristic indexes are extracted, and a longhorn beetle whisker search algorithm (BAS) optimizes penalty factors and kernel function parameters of a Support Vector Machine (SVM), so that compared with other traditional classification models, the formed BAS-SVM classifier can improve the recognition accuracy of different voltage sag categories and has a good recognition effect.
Table 1 recognition accuracy of different classifiers based on standard S-transform and improved S-transform
| Transformation method | SVM | CV-SVM | GA-SVM | PSO-SVM | BAS-SVM |
| Standard S transform | 83.6％ | 83.6％ | 76.8％ | 83.6％ | 83.6％ |
| Improved S-transform | 82.8％ | 92.8％ | 94.8％ | 94.8％ | 100％ |
The above is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above examples, and all technical solutions belonging to the concept of the present invention belong to the protection scope of the present invention. It should be noted that modifications and adaptations to the invention without departing from the principles thereof are intended to be within the scope of the invention as set forth in the following claims.
Claims (5)
1. The utility model provides a distribution network voltage sag source identification method based on BAS-SVM, which is characterized by comprising the following steps:
step 1, obtaining a fundamental frequency amplitude curve and a frequency amplitude envelope curve of various voltage sag signals by improving a result mode matrix of S transformation;
the voltage sag signals comprise 5 kinds of voltage sag information caused by single-phase short circuit faults, two-phase short circuit faults, three-phase short circuit faults, induction motor starting and transformer operation;
step 2, extracting a plurality of groups of characteristic index data of various voltage sag signals by applying improved S transformation;
step 3, normalizing a plurality of groups of characteristic index data extracted from each voltage sag signal;
step 4, optimizing penalty factors and kernel function parameters of the SVM through the BAS, and constructing a BAS-SVM classifier;
step 5, dividing the normalized plurality of sets of characteristic index data in the step 3 into a training set and a testing set by adopting a 5-time cross verification method, wherein the training set is used as a training sample of the BAS-SVM classifier, and the testing set is used as a testing sample of the BAS-SVM classifier for testing, so that 100% accurate identification of a power distribution network voltage sag source is realized;
and step 2, the characteristic indexes comprise a mean value, a standard deviation, an RMS value, a dip depth, energy, kurtosis, the number of mutation points, shannon entropy, logarithmic energy entropy, a base frequency amplitude falling slope, a base frequency amplitude rising slope, a second harmonic content, skewness, a waveform coefficient, a crest coefficient and a dip time ratio.
2. The method for identifying the voltage sag sources of the power distribution network based on the BAS-SVM according to claim 1, wherein the number of the mutation points is determined by 11 times of fundamental frequency row vectors in a result module matrix of improved S transformation, and the number of the mutation points is uniformly set to 0 when the transformer is put into operation;
the second harmonic content is as follows:
wherein: u (U) 2 For the frequency value at the 32 nd sampling point in the frequency amplitude envelope curve, U 1 The frequency value at the 17 th sampling point in the frequency amplitude envelope curve is obtained;
the sag time ratio is as follows:
wherein: t (T) 1 For duration of dip, T 2 Is the time for the sag to be stable;
the dip depth is as follows:
MF＝U sag /U ref
in U ref And U sag Representing the effective values before and at the time of voltage sag, respectively;
the mean value calculation formula is as follows:
the standard deviation calculation formula is as follows:
the RMS value calculation formula is as follows:
the energy calculation formula is as follows:
the kurtosis calculation formula is as follows:
the shannon entropy calculation formula is as follows:
the logarithmic energy entropy calculation formula is as follows:
the deflection calculation formula is as follows:
the calculation formula of the waveform coefficient is as follows:
the crest factor calculation formula is as follows:
where xi is the amplitude of a certain sampling point in the disturbance sample vector x, N is the number of sampling points contained in the disturbance sample, and P is A, B, C three phases.
3. The method for identifying a voltage sag source of a power distribution network based on BAS-SVM according to claim 1, wherein the normalization processing in step 3 extracts a plurality of sets of characteristic index data of each voltage sag signal, and the normalization formula is as follows:
wherein: x is original data, X min X is the minimum value in the original data max Is the maximum value in the original data.
4. The method for identifying a voltage sag source of a power distribution network based on a BAS-SVM according to claim 1, wherein in the step 4, the penalty factor and the kernel function parameter of the SVM are optimized by the BAS, the optimizing result is that the penalty factor is 0.3798, and the kernel function parameter is 5.8570.
5. The method for identifying a voltage sag source of a power distribution network based on BAS-SVM according to claim 1, wherein the constructing a BAS-SVM classifier in step 4 includes:
(1) Constructing a random vector representing the direction of the longhorn beetles, and defining a space dimension k;
(2) Setting step factors as follows:
δ t ＝δ t-1 *eta
wherein: eta is a number near 1 between intervals [0,1], taking 0.95;
(3) Taking the identification accuracy of the SVM as an adaptability function;
(4) Initializing the space position of the longhorn beetles as an initial solution set of the BAS algorithm and storing the initial solution set in X best In (a) and (b);
(5) Calculating initial fitness function value of the longicorn at initial position according to fitness function, and storing in f best The fitness function is the identification accuracy of the SVM;
(6) Iteratively updating the positions of the left long antenna and the right long antenna of the longhorns, respectively solving the fitness function values of the left whisker and the right whisker when the longhorns are at the current positions, and updating X if the current function value is better than the initial fitness function value best And f best ；
(7) Stopping iteration if the iteration times are reached, turning to the step (8), otherwise, returning to the step (6) to continue iteration;
(8) And obtaining the optimal value of the fitness function, the SVM penalty factor and the optimal value of the kernel function parameter.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010273237.2A CN111476173B (en) | 2020-04-09 | 2020-04-09 | Power distribution network voltage sag source identification method based on BAS-SVM |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010273237.2A CN111476173B (en) | 2020-04-09 | 2020-04-09 | Power distribution network voltage sag source identification method based on BAS-SVM |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111476173A CN111476173A (en) | 2020-07-31 |
| CN111476173B true CN111476173B (en) | 2024-03-05 |
Family
ID=71751363
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010273237.2A Active CN111476173B (en) | 2020-04-09 | 2020-04-09 | Power distribution network voltage sag source identification method based on BAS-SVM |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111476173B (en) |
Families Citing this family (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112036318B (en) * | 2020-09-01 | 2023-10-20 | 海南电网有限责任公司电力科学研究院 | Voltage sag source homogeneous integration identification method |
| CN112380945B (en) * | 2020-11-06 | 2021-12-21 | 中国电力科学研究院有限公司 | Method and system for identifying potential hazards of power transmission channel environment based on satellite remote sensing image |
| CN112529066B (en) * | 2020-12-07 | 2024-04-16 | 南京工程学院 | Voltage sag source identification method based on BAS-BP classifier model |
| CN113608011B (en) * | 2021-08-11 | 2024-01-30 | 国网江苏省电力有限公司 | Direct-current high-voltage measuring device with self-calibration function and operation method thereof |
| CN113688786B (en) * | 2021-09-10 | 2022-07-12 | 广东电网有限责任公司广州供电局 | PSO (particle swarm optimization) width learning-based voltage sag multiple disturbance source identification method |
| CN114118130B (en) * | 2021-09-28 | 2024-04-05 | 西安交通大学 | Transformer partial discharge mode identification method, system, medium and equipment |
| CN114626487B (en) * | 2022-05-16 | 2023-09-05 | 南昌工程学院 | Linear transformation relation checking method based on random forest classification algorithm |
| CN117743807B (en) * | 2024-02-19 | 2024-04-26 | 四川大学 | Multi-stage voltage sag cause identification method for power distribution network |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109026647A (en) * | 2018-08-14 | 2018-12-18 | 东华大学 | A kind of the Hydraulic pump fault detection method and system of BASA optimization GRBF-SVM |
| CN110954779A (en) * | 2019-11-29 | 2020-04-03 | 国网上海市电力公司 | Voltage sag source feature identification method based on S transformation and multidimensional fractal |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108414923A (en) * | 2018-02-05 | 2018-08-17 | 武汉大学 | A kind of analog-circuit fault diagnosis method based on the extraction of depth confidence network characterization |
-
2020
- 2020-04-09 CN CN202010273237.2A patent/CN111476173B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109026647A (en) * | 2018-08-14 | 2018-12-18 | 东华大学 | A kind of the Hydraulic pump fault detection method and system of BASA optimization GRBF-SVM |
| CN110954779A (en) * | 2019-11-29 | 2020-04-03 | 国网上海市电力公司 | Voltage sag source feature identification method based on S transformation and multidimensional fractal |
Non-Patent Citations (1)
| Title |
|---|
| 赵洛印等.基于GWO-SVM 的电压暂降扰动源识别.电测与仪表.2019,第56卷(第56期),全文. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111476173A (en) | 2020-07-31 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111476173B (en) | Power distribution network voltage sag source identification method based on BAS-SVM | |
| CN107979086B (en) | Voltage sag reason identification method based on EM algorithm and gradient lifting tree | |
| CN109975673B (en) | Method for identifying fault arc at direct current side of photovoltaic microgrid | |
| CN108345906B (en) | Non-invasive electrical appliance identification method based on Boost model | |
| CN110070048A (en) | Device type recognition methods and system based on double secondary K-means clusters | |
| CN109858564B (en) | Improved Adaboost-SVM model generation method suitable for wind power converter fault diagnosis | |
| CN102982347A (en) | Method for electric energy quality disturbance classification based on KL distance | |
| Guomin et al. | Deep learning‐based fault location of DC distribution networks | |
| CN113702767B (en) | Island direct-current microgrid fault diagnosis method based on wavelet sliding window energy | |
| Chen et al. | Reliable detection method of variable series arc fault in building integrated photovoltaic systems based on nonstationary time series analysis | |
| CN111398721A (en) | Power distribution network voltage sag source classification and identification method introducing adjustment factors | |
| Qaisar et al. | Time-domain identification of the power quality disturbances based on the event-driven processing | |
| CN112036318A (en) | Homogeneous integration identification method for voltage sag source | |
| Zhibin et al. | Research on household appliances recognition method based on data screening of deep learning | |
| CN116502149A (en) | Low-voltage power distribution network user-transformation relation identification method and system based on current characteristic conduction | |
| Li et al. | A training-free non-intrusive load monitoring approach for high-frequency measurements based on graph signal processing | |
| Mei et al. | Random forest based adaptive non-intrusive load identification | |
| Liu et al. | Classification and identification of electric shock current for safety operation in power distribution network | |
| CN116125199A (en) | Low-voltage distribution network line fault early warning method and system | |
| CN106503290B (en) | PLS-SVM-based general single-phase harmonic load modeling method | |
| Duan et al. | PD pattern recognition of XLPE cable based on parameter optimal support vector machine algorithm | |
| CN112529066B (en) | Voltage sag source identification method based on BAS-BP classifier model | |
| CN112949563B (en) | Non-invasive load identification method based on variable point detection and improved KNN algorithm | |
| Huang et al. | Fault diagnosis of short circuit in transmission line based on variational modal decomposition and whale algorithm | |
| CN116628454B (en) | Low-voltage distribution network topology identification method and device based on intelligent switch |
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 |