CN111507417A - Fault transient protection method for power transmission line - Google Patents
Fault transient protection method for power transmission line Download PDFInfo
- Publication number
- CN111507417A CN111507417A CN202010314750.1A CN202010314750A CN111507417A CN 111507417 A CN111507417 A CN 111507417A CN 202010314750 A CN202010314750 A CN 202010314750A CN 111507417 A CN111507417 A CN 111507417A
- Authority
- CN
- China
- Prior art keywords
- fault
- waveform
- similarity
- transient
- protection
- 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.)
- Pending
Links
- 230000001052 transient effect Effects 0.000 title claims abstract description 107
- 238000000034 method Methods 0.000 title claims abstract description 71
- 230000005540 biological transmission Effects 0.000 title claims abstract description 45
- 238000001228 spectrum Methods 0.000 claims abstract description 52
- 241000544061 Cuculus canorus Species 0.000 claims abstract description 41
- 238000010845 search algorithm Methods 0.000 claims abstract description 17
- 238000004364 calculation method Methods 0.000 claims description 22
- 238000000354 decomposition reaction Methods 0.000 claims description 10
- 238000001514 detection method Methods 0.000 claims description 6
- 230000008859 change Effects 0.000 claims description 5
- 238000012360 testing method Methods 0.000 claims description 3
- 238000004422 calculation algorithm Methods 0.000 abstract description 37
- 238000004458 analytical method Methods 0.000 abstract description 14
- 230000003595 spectral effect Effects 0.000 abstract description 10
- 238000012216 screening Methods 0.000 abstract description 3
- 238000004088 simulation Methods 0.000 description 30
- 230000009471 action Effects 0.000 description 12
- 230000008569 process Effects 0.000 description 11
- 238000005070 sampling Methods 0.000 description 11
- 230000007704 transition Effects 0.000 description 10
- 235000005770 birds nest Nutrition 0.000 description 6
- 235000005765 wild carrot Nutrition 0.000 description 6
- 239000000284 extract Substances 0.000 description 3
- 238000005457 optimization Methods 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 230000009466 transformation Effects 0.000 description 3
- 238000010219 correlation analysis Methods 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 239000011159 matrix material Substances 0.000 description 2
- 230000010355 oscillation Effects 0.000 description 2
- VYMDGNCVAMGZFE-UHFFFAOYSA-N phenylbutazonum Chemical compound O=C1C(CCCC)C(=O)N(C=2C=CC=CC=2)N1C1=CC=CC=C1 VYMDGNCVAMGZFE-UHFFFAOYSA-N 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 241000272177 Cuculiformes Species 0.000 description 1
- 101100136092 Drosophila melanogaster peng gene Proteins 0.000 description 1
- 208000032365 Electromagnetic interference Diseases 0.000 description 1
- 230000002776 aggregation Effects 0.000 description 1
- 238000004220 aggregation Methods 0.000 description 1
- 230000004888 barrier function Effects 0.000 description 1
- 230000000052 comparative effect Effects 0.000 description 1
- 239000002131 composite material Substances 0.000 description 1
- 238000011217 control strategy Methods 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 238000012850 discrimination method Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000002955 isolation Methods 0.000 description 1
- 230000035772 mutation Effects 0.000 description 1
- 230000000717 retained effect Effects 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- 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
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2113/00—Details relating to the application field
- G06F2113/04—Power grid distribution networks
-
- 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
- G06F2218/16—Classification; Matching by matching signal segments
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Life Sciences & Earth Sciences (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Medical Informatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Locating Faults (AREA)
Abstract
In order to promote the practicability of the transient protection principle based on waveform uniqueness, the invention provides a transmission line transient protection method combining empirical mode decomposition-Virgerwell distribution and a Cuckoo algorithm. The method comprises the steps of decomposing fault transient signals by using EMD to obtain single-component signals of different frequency bands; and screening real transient signal components. And then, respectively solving the WVD for the true components, and linearly superposing the WVD of all the true components to obtain a time-frequency-amplitude energy spectral density graph of the fault transient signal. And finally, continuously iterating and matching the energy spectrum density graph and the energy spectrum density graph in the fault sample library by using a cuckoo search algorithm to calculate the similarity, and judging the faults inside and outside the area according to the similarity so as to realize the rapid and reliable protection of the transmission line faults. The EMD-WVD method can effectively improve the accuracy of fault transient waveform analysis, the cuckoo search algorithm can effectively improve the waveform matching speed, and the combination of the two is expected to promote the practicability of transient protection.
Description
Technical Field
The invention relates to a fault transient protection method for a power transmission line.
Background
With the rapid development of ultrahigh voltage power grids in China, ultrahigh voltage power transmission lines put higher requirements on relay protection reliability and quick action. Transient protection constructed based on transient quantity generated during fault has ultrahigh protection performance, and becomes a research hotspot in the field of relay protection at home and abroad.
Since the 70 s in the 20 th century, many scholars introduced wavelet transformation, Hilbert-Huang and other mathematical tools into transmission line transient protection, and transient protection technology developed vigorously. Among them, there is a document that performs super high speed direction protection and pilot protection using an amplitude comparison formula and a polarity comparison formula of the relative entropy of the S-transform energy. Or, the travelling wave distance protection is carried out on the power transmission line by utilizing wavelet transformation, but the method is difficult to select the scale, and the reliability is not high. There is a document that uses hilbert-yellow transform as a detection tool to sequentially extract transient ac components of each frequency to realize traveling wave distance protection. The method accurately displays various characteristic information in the time-frequency domain range, and has good operability; however, problems such as end-point effect, screening termination criteria and mode aliasing occur, so that the protection reliability is not high.
Most of the above documents use single time domain or single frequency domain fault feature information for protection, which may result in incomplete extraction of local fault feature information, and further affect protection reliability. In order to overcome the defect of single fault characteristics, documents propose a direct current transmission line pilot protection scheme based on traveling wave waveform correlation analysis according to the difference of the characteristics of the waveforms inside and outside the area. There is a permanent fault discrimination method based on waveform correlation of a least square method, which is proposed in literature by analyzing fault characteristics of a power transmission line with a shunt reactor under different fault properties of a single-phase earth fault. In order to solve the problems that a control strategy of the new energy grid connection is affected by amplitude, frequency, phase angle and the like, a main protection principle suitable for a new energy station sending-out line is provided in literature. According to the method, a waveform unique principle is utilized to synthesize time-frequency characteristic information to perform power transmission line fault protection, but the fault transient signal is matched with the waveform in a sample library, the similarity of the matched waveform needs to be calculated one by one, the calculation is complex, the required time is long, and the quick action of protection is greatly influenced.
Therefore, it is necessary to design a transient protection method for power transmission line fault.
Disclosure of Invention
The technical problem to be solved by the invention is to provide a power transmission line fault transient protection method which is easy to implement and has high fault positioning speed.
The technical solution of the invention is as follows:
a fault transient protection method for a power transmission line utilizes EMD (empirical mode decomposition, EMD, combined with empirical mode decomposition) to decompose fault transient signals to obtain single component signals of different frequency bands; similarity calculation is carried out on each single-component signal and the original transient signal by adopting a waveform correlation coefficient, and real transient signal components are screened;
then, respectively solving WVD (Wigner ville distribution, WVD) for the real transient signal components, and linearly superposing the WVD of all the real transient signal components to obtain a time-frequency-amplitude energy spectrum density graph of the fault transient signal;
and finally, continuously iterating and matching the energy spectrum density graph and the energy spectrum density graph in the fault sample library by using a Cuckoo Search algorithm (CS) to calculate the similarity, and judging the faults inside and outside the region according to the similarity so as to realize the rapid and reliable protection of the transmission line fault (namely rapidly identifying the fault and starting the protection).
The method comprises the following steps:
step 1: establishing a sample library;
the samples in the sample library are samples for simulating various fault conditions by utilizing the built power grid fault signal testing system or actual field detection data samples;
all patterns and signals can be expressed by matrix in nature. The existing algorithms for similarity matching are also used for extracting feature quantities and constructing a matrix for matching. The energy spectrum density graph is an external expression form of a fault signal, and the energy spectrum density graph is matched based on a waveform correlation principle and adopts a waveform to move a final recognition result.
Step 2: determining the waveform position x of the fault point of the transmission linebest (t)And a maximum number of iterations N;
and step 3: setting an initial probability parameter P, randomly dividing a line into a plurality of intervals according to the length, and setting the intervals as initial setting points of the waveform of a sample library;
and 4, step 4: for fault point position waveform xbest (t)Calculating, namely replacing the waveforms which do not meet the criterion with the waveforms which meet the criterion according to the waveform position updating criterion to obtain the local optimal waveform position to form a new generation of waveform set point;
the criterion is as follows: if the similarity value Q > P, the waveform position xi (t+1)Change occurs, otherwise the waveform position is not changed, and the result is still marked as xi (t+1)(ii) a The waveform is iterated continuously according to the algorithm, a waveform meeting the conditions is found, if the waveform is found, the original waveform is replaced by a new one, and the positions of all the waveforms are changed continuously. Essentially, the algorithm is a continuous iterative optimization process. The name is still called x after the change regardless of the position changei (t+1). Corresponding to xi (t+1)Is a dependent variable, if a new waveform meets the requirement, the original waveform is discarded, and the new waveform is still marked as xi (t+1))；
Q is actually the similarity calculation of two waveforms, and Q can be calculated by the following calculation formula:
wherein, αs、βs、ξsCharacteristic values representing the waveforms to be searched for in the sample library, αt、βt、ξtThe eigenvalues representing the eigenvector space, m, n and i being α respectivelys、βs、ξsThe number of (2).
And 5: determining a search result;
new n sample database waveforms and fault point waveform xbest (t)Similarity calculation is carried out, similarity waveform set points which accord with criteria are found out, and the similarity is recorded as k;
according to the formula k > ksetJudging by a protection criterion, and if the similarity k meets the condition, taking the position of the similarity k point as a search fault point; if not, returning to the step 4 to continue iterating to the maximum number N, and finding out the position of the similarity k as a final search result.
The EMD-WVD is mainly used for improving the accuracy of transient signals, and is directly input into an algorithm as a signal in the steps.
The calculation formula of the flight step length is as follows:
wherein x isi (t)Indicating the location of the ith sample bin setpoint waveform at the t-th iteration,
α represents step control quantity for point-to-point multiplication, L (lambda) represents waveform random search path, and column dimension distribution is obeyed.
α is mainly determined by the distance between a certain point waveform and the fault point waveform position in the sample library, the distance is α long and α short, however, α has randomness, and the value range of α is (0-1).
The cuckoo algorithm performs iterative analysis on the waveform, and the related waveform is the waveform processed by EMD-WVD.
The sample of the step 1 is a transient signal energy spectrum density map processed by EMD-WVD, and a cuckoo algorithm is embodied from the step 2 to the step 5; the innovation points of the invention are as follows: (1) the detection precision of the waveform is improved by using EMI) -WVD, and (2) the search efficiency is improved by using the cuckoo algorithm, namely, the waveform matching time is improved by using the cuckoo algorithm based on the waveform unique principle, namely, the quick action of relay protection is improved.
Has the advantages that:
in order to promote the practicability of the transient protection principle based on waveform uniqueness, the invention provides a transmission line transient protection method combining Empirical Mode Decomposition (EMD) -Wigner Ville Distribution (WVD) and a cuckoo algorithm. The method comprises the steps of decomposing fault transient signals by using EMD to obtain single-component signals of different frequency bands; and respectively carrying out similarity calculation on each single-component signal and the original transient signal by adopting a waveform correlation coefficient, and screening real transient signal components. And then, respectively solving the WVD for the true components, and linearly superposing the WVD of all the true components to obtain a time-frequency-amplitude energy spectral density graph of the fault transient signal. And finally, continuously iterating and matching the energy spectrum density graph and the energy spectrum density graph in the fault sample library by using a Cuckoo Search algorithm (CS) to calculate the similarity, and judging the internal and external faults of the area according to the similarity, thereby realizing the rapid and reliable protection of the transmission line faults. Simulation results show that the EMD-WVD method can effectively improve the accuracy of fault transient waveform analysis, the cuckoo search algorithm can effectively improve the waveform matching speed, and the combination of the EMD-WVD method and the cuckoo search algorithm is expected to promote the practicability of transient protection.
The greatest innovation point of the invention is as follows: based on the waveform unique principle, the research continues. The method aims to improve the accuracy of the waveform by utilizing EMD-WVD, and the iterative analysis is carried out by utilizing the cuckoo algorithm so as to solve the problem that the protection speed is possibly influenced due to overlong matching time caused by overlarge samples of the sample library constructed by the existing literature.
The waveform energy spectrum density graph is an external expression form of the fault signal, compared with a single frequency domain or a single time domain, the energy spectrum density graph integrates fault information of time-frequency-amplitude, obtained information is more comprehensive, an expression form is more three-dimensional and visual, simulation is enabled to be close to an actual site to the maximum extent, and protection is enabled to be more reliable. This is the most important reason for the present invention to use waveform spectral density maps.
The invention provides a transmission line transient protection method combining an EMD-WVD algorithm and a cuckoo algorithm on the basis of accurately analyzing a transient signal by using the EMD-WVD algorithm. The method utilizes EMD-WVD to accurately analyze the transient signal, removes the influence of pseudo signals such as interference and the like, and improves the authenticity of the transient signal; and then, performing waveform matching by adopting a cuckoo algorithm, and quickly calculating the waveform similarity. A protection scheme is provided by combining an EMD-WVD algorithm and a cuckoo algorithm, and a protection criterion is constructed according to the similarity.
Drawings
FIG. 1 is a simulation model of a 500kV transmission line;
FIG. 2 is a transient spike signal;
FIG. 3 is a graph of energy spectral density of a single-phase ground fault processed by EMD-WVD;
figure 4 is a graph of the energy spectral density of a Wigner Ville distribution without EMD treatment.
Detailed Description
The invention will be described in further detail below with reference to the following figures and specific examples:
example 1:
EMD-WVD analysis of 1 transient signals
In the aspect of analyzing multi-component and non-stationary signals, the WVD extracts the time and the frequency of the signals respectively, so that the defect of mutual restriction of the time and the frequency is avoided, and the WVD is a powerful signal analysis tool. But the cross term interference thereof brings great obstruction to the time frequency distribution. In order to solve the problem of cross terms and more accurately analyze the waveform characteristic information of fault transient signals, the invention adopts an EMD-WVD method to extract the characteristic quantity of the signals and obtain the energy spectrum density chart of the signals, and the specific implementation process is as follows:
x (t) is defined as an energy-limited transient real signal, which is decomposed by EMD, i.e.:
the IMF (intrinsic mode action) is a finite number of eigenmode functions.
In equation (1), x (t) is decomposed into n IMF single component signals, where k true components α existiN-k pseudo components zk+1The spurious component can interfere with the WVD of the signal. Therefore, the false component needs to be removed, and the true component is retained.
In order to eliminate the pseudo component, the invention utilizes the waveform correlation coefficient to respectively calculate the similarity between the single component signals of different frequency bands generated by EMD decomposition and the original transient state signal, and the waveform correlation coefficient is shown as the formula (2):
in the formula (2), R (x, IMF)i) Representing the correlation coefficient of the original transient signal with the EMD decomposed single component signal. R (x, IMF)i) Is between-1 and 1, the sign indicating the direction of correlation. When R (x, IMF)i) The closer to 1 the absolute value of (A) is, the more similar the waveforms of the two signals are, the greater the correlation is; when R (x, IMF)i) The absolute value of (a) is close to 0, which means that the similarity difference between the two waveforms is large and the correlation is small. The invention is calculated to know that the values of the true components are all larger than 0.3, and the value of the false component is far smaller than 0.3, so that the invention leads the correlation coefficient R (x, IMF)i) The threshold value of (2) is set to 0.3, and the spurious component is removed according to the magnitude of the correlation coefficient.
After eliminating the pseudo components, solving WVD for each single component signal to obtain an energy spectrum density graph, wherein WVD is shown as a formula (3). The single-component energy spectral density map obtained at this time has almost no cross-term interference. Linearly superposing the energy spectrum density graphs obtained by the WVD of each component to obtain an energy spectrum density graph of the transient signal, wherein the energy spectrum density graph is shown in a formula (4):
the transient signal processed by EMD-WVD can inhibit the interference of cross terms to a great extent, so that the waveform characteristic is more obvious, the precision of the energy spectrum density graph of the transient signal is improved, and the protection is more reliable. In the following, aiming at the problems of complex calculation and long time consumption of the protection method based on the existing waveform unique principle, the cuckoo search algorithm is introduced in the invention. The energy spectrum density graph of the detected transient signal is subjected to waveform similarity calculation by using a cuckoo algorithm and the energy spectrum density graph in the sample library, so that the calculation speed of similarity matching can be effectively increased.
2 cuckoo search algorithm principle
The cuckoo algorithm (CS) is an emerging heuristic algorithm proposed in recent years and is used for effectively solving the optimization problem. Cuckoos represent solutions with egg nests, and new solutions are continuously used for replacing non-optimal solutions in the iteration process.
The existing protection method based on the waveform unique principle mostly simulates a large number of fault points, matches fault transient signals with waveforms in a sample library, and needs to calculate the similarity of the matched waveforms one by one. In order to solve the problem, the matching process of the energy spectrum density waveform processed by EMD-WVD and the waveform in the sample library is converted into the cuckoo search algorithm optimizing process, so that the calculated amount is greatly reduced. The process is as follows:
the similarity matching process follows a column-dimensional flight, as shown in equation (5):
L(s)～|s|-1-λ(5)
during flight, the fault point waveform has certain probability P (P ∈ [0, 1]) found by the waveform in the sample library, if found, the sample library discards the waveform and randomly extracts a new waveform as a set point, if not found, the fault point waveform replaces the waveform of the set point, and the position is updated as formula (6).
In the formula (6), xi (t)Indicating the location of the ith sample bin setpoint waveform at the t-th iteration,
for point-to-point multiplication, α denotes a step control quantity, L (λ) is a waveform random search path, and the column dimension distribution is obeyed.
The waveform (iterative) position updating criterion is that in the process of searching for matching, a similarity value Q is obtained by calculating the similarity, and Q is between 0 and 1. If Q > P, the waveform position xi (t+1)Change occurs, otherwise it is not changed, the result is still marked as xi (t+1)。
In the traditional cuckoo algorithm, step generation is random and lacks adaptivity. In order to solve the problem, quantum codes are reversely generated according to binary codes to solve the problem of step randomness, but the calculation is complex, and the overall convergence speed is seriously influenced. Literature [ schhaoran, zhankhon, lie, penchenhui ] transformer fault diagnosis based on buguo bird algorithm and support vector machine [ J]Protection and control of electric power systems 2015, 43 (08): 8-13.XUE Haoran, ZHANG Keheng, L I Bin, PENG Chenhui. Fault diagnostics of transporter based on the cuckoo search and vector machine J]Power System Protection and Control，2015，43(08)：8-13]The step size is controlled by equation (7), where stepmaxAnd stepminMaximum and minimum step sizes are represented, respectively; n isbestThe optimal state of the current bird nest position is represented; n isiIndicating the ith bird nest position; dmaxIndicating the maximum distance between the optimal position and other bird nests. However, the method also has the problem of long calculation time, and is not suitable for transient protection of the power transmission line.
Based on this, on the premise of ensuring reliability, the invention intends to adopt the document [ Yang Xinsthee, DebS].International Journal ofMathematical Modeling and Numerical，2010，1(4)：330-343.]On the basis of equation (6), dot-multiplied by a coefficient xi (t)-xbest (t)And is used for controlling the flight step length as shown in formula (8).
The coefficient is the current position x of a certain point waveform in the sample libraryi (t)And fault point waveform xbest (t)The difference in absolute value of. From the equation (8), when the waveform x is at a certain point in the sample banki (t)And fault point waveform position xbest (t)The farther away, xi (t)-xbest (t)The greater the absolute value of (a) is,
the larger the search step size, the smaller the search step size. Transient protection combining EMD-WVD and cuckoo algorithm
3.1 analysis of the differences in similarity between the inside and outside of the region
The waveform characteristic difference is obvious when faults exist inside and outside the area. The method utilizes the similarity of the transient signal waveform energy spectrum density graph to represent the difference of the internal and external faults.
As shown in fig. 1, a 500kV transmission line is provided with fault points at intervals of 1km, transient signal waveform spectrum density maps generated under different fault conditions are simulated at the fault points, actual field detection data samples are collected, and a sample library is established. F is arranged at positions 5km, 45km and 99km away from the M end of the bus2Simulating faults in the line area; f is arranged at 105km and 130km away from the M end3Simulating a forward out-of-range fault; f is arranged at the positions 30km and 50km away from the M end of the bus1Simulating a reverse out-of-range fault. Accurately analyzing the transient signals by using EMD-WVD to obtain an energy spectrum density graph without cross-item interference; waveform generation using cuckoo algorithmAnd matching and quickly calculating the waveform similarity. The results are shown in Table 1.
TABLE 1 similarity of energy spectral density plots for different fault locations
Table 1Energy spectral density map similarity under different faultlocations
As can be seen from table 1, when the power transmission line has an intra-area fault, the similarity between the energy spectrum density map of the fault transient signal and the energy spectrum density map in the sample library is high, and the numerical value is close to 1; when an out-of-area fault occurs, the energy spectrum density graph of the transient signal of the fault has a larger difference with the energy spectrum density graph in the sample library, and the similarity values are all less than 0.3, so that the similarity difference of the energy spectrum density graph of the waveform of the transient signal is obvious when the out-of-area fault occurs in the power transmission line.
3.2 protection scheme
The difference of transient state signal waveform energy spectrum density graphs of internal and external faults is obvious as known from section 3.1. Therefore, the transient protection of the power transmission line can be carried out by judging the faults inside and outside the area by utilizing the similarity of the transient signal waveform energy spectrum density diagram, and the protection criterion is as follows:
k＞kset(9)
wherein k issetExpressed as a protection action value. Considering the influence of factors such as sample interval, sampling frequency, power grid parameters, data windows, bird nest number of the cuckoo algorithm and the like on the protection reliability and the difference of the similarity of waveforms inside and outside the region, the invention sets the action value to be 0.9.
According to the formula (9), when the similarity of the transient signal waveform spectrum density graph is greater than the protection action value, judging that the fault is in the region; otherwise, the mobile terminal is judged to be out-of-range fault. Meanwhile, on the basis of accurately analyzing the transient signal by EMD-WVD to obtain a waveform energy spectrum density graph, the method adopts the cuckoo algorithm to carry out waveform matching and quickly calculate the waveform similarity, and the protection scheme is as follows.
(1) Establishing a sample library: the sample library comprises samples for simulating various fault conditions by using the built power grid fault signal testing system and actual field detection data samples.
(2) Determining the waveform position x of the fault point of the transmission linebest (t)And a maximum number of iterations N.
(3) And setting an initial probability parameter P, randomly dividing the samples in the sample library into n intervals according to the length of the power transmission line, and setting the intervals as initial waveform setting points of the sample library.
(4) Position waveform x of fault pointbest (t)Calculating according to the method in chapter 2, and replacing the waveforms meeting the criterion with the waveforms not meeting the criterion according to the waveform position updating criterion in chapter 2 to obtain the local optimal waveform position to form a new generation waveform set point.
(5) New n sample database waveforms and fault point waveform xbest (t)Similarity calculation is carried out, similarity waveform set points which accord with criteria are found out, and the similarity is recorded as k; judging according to the protection criterion of the formula (9), and if the similarity k meets the condition, taking the position of the similarity k point as a search fault point; and if not, returning to the step (4) to continue iterating to the maximum number N, finding out the position of the similarity k, and setting the position as a final search result.
3.3 cuckoo Algorithm tachykinetic analysis
When the number of the bird nests is more, the number of times of iteration required each time is more; the longer the flight step, the faster the search but the lower the precision; the P probability is too high to converge to the optimal solution. The reliability and the quick action of the transient protection of the power transmission line are comprehensively considered, and the analysis shows that the requirements of the invention are met when the number N of the bird nests is 20, the P is 0.5 and the maximum iteration number N is about 40. In order to verify the speed-action performance of the cuckoo algorithm, as shown in a power transmission line model shown in fig. 1, every 1km is provided with a spacing point, a single-phase earth fault occurs at a position 50km in a set area, and an initial phase angle theta is equal to10 0Distance of failure dM40km, transition resistance Rg100 Ω, noise SRN10dB, the sampling frequency is 1MHZ and the time window size is 2 ms. The invention selects 200 waveforms under the above conditions as sample library waveforms, inputs the fault point transient signal waveforms into the cuckoo search algorithm, and directly normalizes the matchingMatching algorithm, Hausdorff template matching algorithm, and comparison, the results are shown in Table 2.
TABLE 2 comparison of cuckoo search algorithm with several matching algorithms
Table 2Comparison of cuckoo search algorithm and several matchingalgorithms
As shown in table 2, the three matching algorithms all satisfy the criterion of equation (9), and an intra-area fault occurs. Although the Hausdorff template matching algorithm is higher in similarity than the cuckoo search algorithm, the average calculation time is 9.528 ms. And the average calculation time of the cuckoo search algorithm is only 3.472ms under the condition of ensuring the reliability. The result shows that the matching calculation time is greatly improved by using the cuckoo search algorithm.
4 simulation analysis
4.1 EMD-WVD simulation analysis of transient signals
The invention utilizes EMD method to decompose the fault signal, screen the real component and solve WVD for the real component. And performing linear superposition on all the WVD to obtain a WVD time-frequency-amplitude energy spectral density graph without cross term interference. Simulation analysis of transient signals is shown below.
Fig. 1 shows a model of a 500kV high-voltage transmission line. A. M, N, B is a transmission line l1、l2、l3The bus end of (a). Transmission line l1、l2The lengths of the two pieces of the composite material are 60km, 100km and 40km respectively. The protection device is arranged at the M end of the bus and the line l2Is set as protected3And (4) protecting the circuit. f. of1Indicating a reverse out-of-range fault, f2Indicates an in-zone fault, f3Indicating a forward out-of-range fault.
The invention utilizes EMD-WVD method to make fault signal have single-phase earth fault in 50km place in power transmission line zone shown in figure 1, initial phase angle theta is 100Distance of failure dM40km, transition resistance Rg100 Ω, noise SRNThe simulation was performed under 10dB condition, and the results are shown in fig. 2.
In fig. 2, the transient signal is obviously subjected to a characteristic quantity mutation at the time of 0.216ms, which indicates that the WVD can accurately detect the transient signal. In order to further verify the feasibility of the method, the invention performs EMD processing on the transient signal to obtain 6 transient signal components, and the correlation coefficient of the transient signal components is shown in Table 3.
TABLE 3 correlation analysis
Table 3Correlation Analvsis
As can be seen from table 3, correlation coefficients of the four signal components IMF1, IMF2, IMF3 and IMF4 are all greater than 0.3, and are true components; the IMF5 and IMF6 signal components have small correlation coefficients, which are far less than 0.3, and are pseudo components. Removing IMF5 and IMF6 component signals, respectively obtaining Wigner Ville distribution of IMF1, IMF2, IMF3 and IMF4, and superposing to obtain a transient waveform energy spectrum density diagram, as shown in FIG. 3.
In fig. 3, the energy spectrum density map processed by EMD-WVD reflects the characteristic information of time, frequency, amplitude, etc. of the transient signal at the fault point, and the transient protection method based on the waveform unique principle can be realized by its high resolution and good time-frequency-amplitude aggregation.
Fig. 4 is an energy spectral density plot of the original fault transient signal directly using WVD processing. Comparing fig. 3 and fig. 4, it can be seen that the energy spectrum density map without EMD processing shows cross interference terms. The interference of the cross terms causes the area of the broadband to have stronger energy distribution, thereby greatly influencing the reliability of protection.
Comparing fig. 3 and fig. 4, it can be known that the method of the present invention can effectively suppress the cross terms of the signals in the time-frequency domain range, obtain an accurate fault transient signal WVD energy spectrum density map, and effectively improve the accuracy of the transient signal waveform. The waveform energy spectrum density graph of the transient signal processed by EMD-WVD contains multi-scale panoramic information of the transient signal in time-frequency-amplitude, more intuitively shows the characteristics of the transient signal and creates conditions for a transient protection method based on a waveform unique principle.
4.2 EMD-WVD simulation analysis of transient signals
4.2.1 different fault locations
In the simulation model of fig. 1, different fault positions of the power transmission line are simulated. F is arranged at the positions 0.1km, 0.5km, 40km, 50km, 50.5km, 60km, 99km, 99.5km and 99.9km away from the M end of the bus2Simulating faults in the line area; f is arranged at positions 100.1km, 100.5km, 120km, 130km, 135km and 139.9km away from the M end3Simulating a forward out-of-range fault; f is arranged at the positions 0.1km, 0.5km, 30km, 30.5km, 50km and 59.9km away from the end M of the bus1Simulating a reverse out-of-range fault. The method comprises the steps of collecting energy spectrum density graphs of transient signal waveforms of different fault positions by using an EMI-WVD (electro-magnetic interference-WVD) method, obtaining transient waveform characteristic information, performing characteristic quantity optimization matching in a sample library by using a cuckoo algorithm, calculating the similarity of a global optimal energy spectrum density graph, judging a fault area, and obtaining a simulation result as shown in a table 4.
Table 4 simulation results for different fault locations (θ 60 °, R100 Ω)
Table 4Simu]ation results under different fault locations(θ＝60°，R＝100Ω)
As can be seen from Table 4, in the power transmission line l2When the head end, the middle end and the tail end of the energy spectrum density graph have faults, the characteristic quantity matching correlation coefficients of the energy spectrum density graph obtained by the calculation result are all larger than the set threshold value, and the protection can reliably act; and in l2When a fault occurs at the position, close to the local bus, of the back-end line and the opposite-end line, the calculated energy spectrum density graph characteristic quantity matching correlation coefficient is far smaller than a threshold value, and the protection device does not act.
4.2.2 different initial phase angles of failure and different transition resistances
The simulation sets that the initial fault phase angles are respectively 20 degrees and the 60-degree transition resistances are respectively 50 omega and 100 omega, and the validity of the method is comprehensively verified under the four conditions. F is arranged at the positions 0.1km, 50km and 99.9km away from the end M of the bus2Cause in simulation areaA barrier; f is arranged at a position 100.1km away from the M end of the bus3Simulating a forward out-of-range fault; f is arranged at a position 0.1km away from the M end of the bus1Simulating a reverse out-of-range fault. And (3) performing characteristic quantity optimizing matching in the sample library by using a cuckoo algorithm, calculating the similarity of the global optimal energy spectrum density graph, and judging a fault area, wherein the simulation result is shown in a table 5.
TABLE 5 simulation results for different fault initial phase angles and different excess resistances
Table 5Simulation results for different initial phase angles anddifferent over-resistances
It can be known from table 5 that when the initial fault angles are respectively selected to be 20 ° and 60 °, and the transition resistances are respectively 50 Ω and 100 Ω, and when the internal and external faults occur at different positions, the faults can be accurately discriminated by using the protection method of the present invention, which indicates that the protection method is less affected by the initial fault angles and the transition resistances, and the protection is reliable.
4.2.3 different failure types
In the case of different fault types, in order to analyze and verify the effectiveness of the protection method, f is arranged at a fault point 50km away from the M end of the bus2Simulating an intra-area fault; f is arranged at a position 130km away from the M end of the bus at a fault point3Simulating a forward out-of-range fault; f is arranged at the position 40km away from the M end of the bus at the fault point1Simulating a reverse out-of-range fault. Different types of fault conditions (single-phase grounding, two-phase short circuit and three-phase short circuit) are set, and the initial fault phase angle is 600The transition resistance is 100 Ω. The simulation results are shown in table 6.
TABLE 6 simulation results under different fault types
Table 6Simulation results for different fault types
The results of the simulation experiments in Table 6 show that under different fault types, the transmission line l to be protected2When the internal fault occurs, the protection method can realize the rapid isolation protection of the line; when faults occur outside the forward area and the reverse area of the power transmission line, the method can also lock the protection device of the line, so that the protection device does not act. The simulation results can also be used to obtain that different fault types have no influence on the protection method of the invention.
4.2.4 Gaussian white noise of different intensities
The invention simulates Gaussian white noise simulation with different intensities at 50km in a region, wherein a circuit is set to be an A-phase single-term grounding fault, and the initial phase angle is 600The transition resistance is over 100 Ω, and the simulation results are shown in table 7.
TABLE 7 simulation results under white Gaussian noise of different intensities
Table 7Simulation results under different intensity Gaussian whitenoise
As can be seen from Table 7, with the enhancement of Gaussian white noise, the similarity k meets the set threshold value, and the protection action shows that the method can still accurately identify the internal and external faults under the condition of strong noise interference.
4.2.5 different sampling frequencies and different time windows
When transient signals are analyzed, the sampling frequency is too low, which may cause the time when the sudden change of the signals arrives to be missed, and the precision of the waveform is reduced, thereby affecting the reliability of protection; the choice of the size of the time window directly affects the integrity of the waveform. In order to further verify the reliability of the protection method, the method is used for simulating under the conditions of different sampling frequencies and different time windows.
Suppose that FIG. 1 showsA single-phase earth fault occurs at 50km in a transmission line area, and the initial phase angle theta of the fault is 600Transition resistance Rg100 Ω. The simulation results for different sampling frequencies and different time windows are shown in table 8.
TABLE 8 simulation results for different sampling frequencies and different time windows
Table 8Simulation results with different sampling frequencies anddifferent time windows
As can be seen from table 8, when the time window is 0.2ms, the similarity k is less than 0.9, and the out-of-range fault is determined according to equation (9), which may be because the accuracy of the waveform is affected by selecting the time window too small. And when the time window is greater than 0.5ms, the similarity k is greater than 0.9, and an intra-zone fault occurs. When the sampling frequency is 100KHZ and the time window is greater than 0.5ms, the similarity is greater than 0.9, and an intra-zone fault occurs. Simulation results show that the method still has reliable protection effect when the time window is 0.5ms and the sampling frequency is 100 KHZ.
4.3 simulation comparison with other protection schemes
In order to verify the superiority of the protection method, the invention is designed to carry out comparative simulation analysis with a distance protection scheme singly using frequency domain fault information and a natural frequency pilot protection scheme singly using time domain fault information.
The simulation line model is shown in figure 1, transient signals are analyzed by wavelet transformation, A-phase single-phase earth fault is set, and initial phase angle of the fault is 200The transition resistance was 100 Ω, and the white gaussian noise was 20dB, and the results are shown in table 9, where the simulation was performed at 0.1km in the region, 50km in the region, 130km outside the forward region, and 40km outside the reverse region, respectively.
TABLE 9 comparison of action results of different protection methods
Table 9Comparison of action results of different protection methods
As can be seen from table 9, under the above conditions, the distance protection is 0.1km in the region, and the time domain information of the initial traveling wave signal cannot be accurately obtained at a position 50km in the region, so that the protection is rejected; the pilot protection based on the natural frequency may have blind spots at 50km in the region, and the full-line protection cannot be realized.
In conclusion, the protection method provided by the invention comprehensively uses the transient signal time-frequency-amplitude characteristic information, overcomes the situation that the protection fails when a single signal is used, and improves the reliability of protection; the cuckoo algorithm is utilized to match the transient waveform energy spectrum density graph with the energy spectrum density graph in the sample library through continuous iteration, a method for protecting similarity is found out, and the waveform matching speed is effectively improved.
5 final phrase
The invention provides a power transmission line transient fault protection method combining an EMD-WVD algorithm and a cuckoo algorithm based on a waveform unique principle, which mainly works as follows:
1) aiming at the problem of cross item interference in the WVD process, the transient signal is decomposed by using EMD, correlation coefficients are calculated to screen real transient signal components, all real component transient signals are superposed to obtain a WVD time-frequency-amplitude energy spectrum density graph without cross item interference, and the accuracy of the transient signal is improved.
2) And continuously iterating and matching the energy spectrum density graph of the fault transient signal and the energy spectrum density graph in the fault sample library by using a cuckoo search algorithm to calculate the similarity, and then judging the internal and external faults according to the similarity, so that the matching speed of the waveform similarity is effectively improved, and the internal and external faults of the transmission line area are quickly and reliably protected.
3) And simulating different fault positions, different initial phase angles, different transition resistances, different fault types, different intensity Gaussian noises, different sampling frequencies, different time windows and the like. Simulation results show that the protection method disclosed by the invention has higher reliability and quick action, and is expected to promote the practicability of transient protection.
In the invention, the EMD is used for decomposing the transient signal into the prior art, and the method comprises the following specific steps:
empirical mode decomposition
The main function of Empirical Mode Decomposition (EMD) is to decompose a multi-component signal into single-component signals of different frequency bands, and then process the single-component signals respectively, so as to avoid mutual interference of the single-component signals. The single component signal after the multi-component is decomposed is defined as the "Intrinsic modal function" (IMF). The IMF component should generally satisfy the following condition:
1. the zero and extreme points should generally be equal in number, but not equal to each other by at most one.
2. In a time interval required by a signal, the average value of the envelope with the maximum local range and the minimum local range is defined to be 0.
The key to IMF realization is to satisfy the above two conditions. From these two conditions, it follows that: IMF reveals the nature of the natural oscillation mode of the signal, and the defined zero-crossing point can only comprise one oscillation mode in one period, so that no other waveform is added in the decomposition process, and the problem of cross term interference generated by Weiganer-Weill distribution can be effectively avoided.
The steps of decomposing the transient signal by using the empirical mode decomposition method are as follows:
1. all extreme points of the signal x (t) are calculated.
2. In the time interval required by the signal, utilizing a cubic spline function to interpolate the maximum and minimum local range of the signal to obtain an upper envelope line and a lower envelope line, wherein the upper envelope line and the lower envelope line are respectively expressed by u (t) and l (t).
3. Order to
m1(t)＝[u(t)+l(t)]/2 (3.7)
In the formula, m1And (t) is the mean value of the upper envelope line and the lower envelope line.
4. Order to
h1(t)＝x(t)-m1(t) (3.8)
This is done by one iteration. Will be formula (3.8) h1And (t) verifying whether the IMF requirements can be met, and if not, starting calculation according to the step (5).
5. The function h is processed by utilizing the cubic spline function again1(t) obtaining the upper and lower envelope u from the maximum and minimum local ranges of the specific section11(t)、l11(t) substituting the envelope curve into the formula (3.7) to obtain a mean curve m11(t), and then substituting the mean curve into the formula (3.8) to obtain the formula (3.9):
h11(t)＝h1(t)-m11(t) (3.9)
6. h obtained from formula (3.9)11(t) verifying whether it satisfies two conditions of IMF; if the condition is met, continuing to operate in the step 7; if the condition can not be met, iteration is carried out according to the step 5 until two conditions of IMF are met.
7. When h is generated1k(t)＝h1(k-1)(t)-m1k(t) when the IMF condition is satisfied, let
c1(t)＝h1k(t) (3.10)
In equation (3.10), the first IMF component to be decomposed is c by iterative operation1(t)。
8. Order to
r1(t)＝x(t)-c1(t) (3.11)
As shown in equation (3.11), the first IMF component c is linearly subtracted from the signal x (t)1(t) obtaining a first residual component r1(t) of (d). And similarly, repeating the steps 1 to 7 continuously to obtain the IMF components and the residual components of the rest items respectively.
In the formula (3.12), the IMF component selected by empirical mode decomposition is c2(t)，…，cm-1(t)，cm(t) of (d). When the residual term rm(t) the function becomes a monotonic function or in a range of intervalsWhen only one extreme point exists, the method terminates the calculation. At this point, the signal x (t) is decomposed into m IMF components and a residual term rm(t), namely:
Claims (3)
1. a fault transient protection method for a power transmission line is characterized in that an EMD (empirical mode decomposition) is utilized to decompose fault transient signals to obtain single component signals of different frequency bands; similarity calculation is carried out on each single-component signal and the original transient signal by adopting a waveform correlation coefficient, and real transient signal components are screened;
then, respectively solving WVD (Wigner ville distribution, WVD) of the real transient signal components, and linearly superposing the WVD of all the real transient signal components to obtain a time-frequency-amplitude energy spectrum density graph of the fault transient signal;
and finally, continuously iterating and matching the energy spectrum density graph and the energy spectrum density graph in the fault sample library by using a Cuckoo Search algorithm (CS) to calculate the similarity, and judging the internal and external faults of the area according to the similarity, thereby realizing the rapid and reliable protection of the transmission line faults.
2. The transmission line fault transient protection method according to claim 1, characterized by comprising the following steps:
step 1: establishing a sample library;
the samples in the sample library are samples for simulating various fault conditions by utilizing the built power grid fault signal testing system or actual field detection data samples;
step 2: determining the waveform position x of the fault point of the transmission linebest (t)And a maximum number of iterations N;
and step 3: setting an initial probability parameter P, randomly dividing a line into a plurality of intervals according to the length, and setting the intervals as initial setting points of the waveform of a sample library;
and 4, step 4:for fault point position waveform xbest (t)Calculating, namely replacing the waveforms which do not meet the criterion with the waveforms which meet the criterion according to the waveform position updating criterion to obtain the local optimal waveform position to form a new generation of waveform set point;
the criterion is as follows: if the similarity value Q>P, waveform position xi (t+1)Change occurs, otherwise the waveform position is not changed, and the result is still marked as xi (t+1)；
And 5: determining a search result;
new n sample database waveforms and fault point waveform xbest (t)Similarity calculation is carried out, similarity waveform set points which accord with criteria are found out, and the similarity is recorded as k;
according to the formula k>ksetJudging by a protection criterion, and if the similarity k meets the condition, taking the position of the similarity k point as a search fault point; if not, returning to the step 4 to continue iterating to the maximum number N, and finding out the position of the similarity k as a final search result.
3. The transmission line fault transient protection method according to claim 2, wherein the calculation formula of the flight step length is as follows:
wherein x isi (t)Indicating the location of the ith sample bin setpoint waveform at the t-th iteration,
α represents step control quantity for point-to-point multiplication, L (lambda) represents waveform random search path, and column dimension distribution is obeyed.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010314750.1A CN111507417A (en) | 2020-04-21 | 2020-04-21 | Fault transient protection method for power transmission line |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010314750.1A CN111507417A (en) | 2020-04-21 | 2020-04-21 | Fault transient protection method for power transmission line |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111507417A true CN111507417A (en) | 2020-08-07 |
Family
ID=71876294
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010314750.1A Pending CN111507417A (en) | 2020-04-21 | 2020-04-21 | Fault transient protection method for power transmission line |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111507417A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112114229A (en) * | 2020-09-15 | 2020-12-22 | 广东电网有限责任公司 | Power distribution network fault positioning method, device and system |
| CN113471947A (en) * | 2021-07-27 | 2021-10-01 | 中国电力工程顾问集团西北电力设计院有限公司 | Active suppression method and system for very fast transient overvoltage of GIS |
| CN116184124A (en) * | 2023-04-26 | 2023-05-30 | 华东交通大学 | Power distribution network fault type identification method |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109581166A (en) * | 2019-01-15 | 2019-04-05 | 清华大学 | Transmission line lightning stroke or Fault Locating Method based on Sobel energy spectrum |
| CN110596539A (en) * | 2019-09-24 | 2019-12-20 | 长沙理工大学 | Power distribution network fault positioning method based on transient waveform correlation |
-
2020
- 2020-04-21 CN CN202010314750.1A patent/CN111507417A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109581166A (en) * | 2019-01-15 | 2019-04-05 | 清华大学 | Transmission line lightning stroke or Fault Locating Method based on Sobel energy spectrum |
| CN110596539A (en) * | 2019-09-24 | 2019-12-20 | 长沙理工大学 | Power distribution network fault positioning method based on transient waveform correlation |
Non-Patent Citations (4)
| Title |
|---|
| 徐长航等: "经验模态分解和魏格纳-维利分布在往复泵泵阀振动信号特征提取中的应用", 《中国石油大学学报(自然科学版)》 * |
| 李泽文等: "基于波形相关性分析的输电线路暂态保护方法", 《电力系统及其自动化学报》 * |
| 李荣雨等: ""自适应步长布谷鸟搜索算法"", 《计算机科学》 * |
| 蔡艳平等: "基于经验模态分解的WVD交叉项抑制法", 《计算机工程》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112114229A (en) * | 2020-09-15 | 2020-12-22 | 广东电网有限责任公司 | Power distribution network fault positioning method, device and system |
| CN113471947A (en) * | 2021-07-27 | 2021-10-01 | 中国电力工程顾问集团西北电力设计院有限公司 | Active suppression method and system for very fast transient overvoltage of GIS |
| CN113471947B (en) * | 2021-07-27 | 2023-06-13 | 中国电力工程顾问集团西北电力设计院有限公司 | Active suppression method and system for GIS (gas insulated switchgear) ultra-fast transient overvoltage |
| CN116184124A (en) * | 2023-04-26 | 2023-05-30 | 华东交通大学 | Power distribution network fault type identification method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111507417A (en) | Fault transient protection method for power transmission line | |
| CN107451557B (en) | Power transmission line short-circuit fault diagnosis method based on empirical wavelet transform and local energy | |
| CN107301381A (en) | Recognition Method of Radar Emitters based on deep learning and multi-task learning strategy | |
| CN103149514B (en) | Method for identifying multi-insulated defect mode in GIS (gas insulated switchgear) | |
| CN106841915B (en) | Power transmission line fault positioning method based on compressed sensing | |
| CN111985383A (en) | Transient electromagnetic signal noise separation and identification method based on improved variational modal decomposition | |
| CN109507304A (en) | A kind of defect inspection method based on ultrasonic inspection | |
| CN116361718A (en) | Single-phase earth fault line selection method based on improved variational modal decomposition and dispersion entropy | |
| CN110927527A (en) | Transient fault sensing method based on transient energy double channels and information fusion | |
| Omar et al. | Fault classification on transmission line using LSTM network | |
| Guowei et al. | Research on network intrusion detection method of power system based on random forest algorithm | |
| CN110905478B (en) | Well drilling data cleaning method based on box plot method and Markov's square distance method | |
| Xu et al. | Novel fault location for high permeability active distribution networks based on improved VMD and S-transform | |
| Mukherjee et al. | Classification and localization of transmission line faults using curve fitting technique with Principal component analysis features | |
| CN112327104A (en) | Fault detection and positioning method for power distribution network with distributed power supply | |
| CN108089100A (en) | The detection method of small current neutral grounding system arc light resistance earth fault | |
| CN109635428B (en) | GIS mechanical fault diagnosis method based on mechanical state signal analysis | |
| CN111340196A (en) | Countermeasure network data generation method and abnormal event detection method | |
| CN111428772A (en) | Photovoltaic system depth anomaly detection method based on k-nearest neighbor adaptive voting | |
| CN116610990A (en) | Method and device for identifying hidden danger of breaker based on characteristic space differentiation | |
| CN109632974A (en) | A kind of ultrasonic examination echo-signal separation method | |
| CN114689991A (en) | Method and system for determining fault time of high-voltage cable and related components | |
| CN115327325A (en) | Charging pile series arc fault detection method and device and electronic equipment | |
| CN114397569A (en) | Circuit breaker fault arc detection method based on VMD parameter optimization and sample entropy | |
| CN114091593A (en) | Network-level arc fault diagnosis method based on multi-scale feature fusion |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200807 |
|
| RJ01 | Rejection of invention patent application after publication |