CN113468704B - Intermittent arc fault detection method and related device - Google Patents
Intermittent arc fault detection method and related device Download PDFInfo
- Publication number
- CN113468704B CN113468704B CN202110838493.6A CN202110838493A CN113468704B CN 113468704 B CN113468704 B CN 113468704B CN 202110838493 A CN202110838493 A CN 202110838493A CN 113468704 B CN113468704 B CN 113468704B
- Authority
- CN
- China
- Prior art keywords
- intermittent arc
- layer
- fault detection
- waveform
- convolution layer
- 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
- 238000001514 detection method Methods 0.000 title claims abstract description 127
- 238000012360 testing method Methods 0.000 claims abstract description 61
- 238000007781 pre-processing Methods 0.000 claims abstract description 19
- 238000012549 training Methods 0.000 claims description 36
- 238000000034 method Methods 0.000 claims description 35
- 239000012634 fragment Substances 0.000 claims description 23
- 238000013507 mapping Methods 0.000 claims description 16
- 238000013527 convolutional neural network Methods 0.000 claims description 15
- 239000011159 matrix material Substances 0.000 claims description 13
- 230000008569 process Effects 0.000 claims description 7
- 230000011218 segmentation Effects 0.000 claims description 5
- 230000006870 function Effects 0.000 description 6
- 238000010586 diagram Methods 0.000 description 5
- 238000004088 simulation Methods 0.000 description 5
- 238000005070 sampling Methods 0.000 description 4
- 230000004913 activation Effects 0.000 description 3
- 230000008878 coupling Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 230000006872 improvement Effects 0.000 description 3
- FGUUSXIOTUKUDN-IBGZPJMESA-N C1(=CC=CC=C1)N1C2=C(NC([C@H](C1)NC=1OC(=NN=1)C1=CC=CC=C1)=O)C=CC=C2 Chemical compound C1(=CC=CC=C1)N1C2=C(NC([C@H](C1)NC=1OC(=NN=1)C1=CC=CC=C1)=O)C=CC=C2 FGUUSXIOTUKUDN-IBGZPJMESA-N 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 230000035772 mutation Effects 0.000 description 2
- 238000010606 normalization Methods 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 230000008034 disappearance Effects 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/10—Geometric CAD
- G06F30/18—Network design, e.g. design based on topological or interconnect aspects of utility systems, piping, heating ventilation air conditioning [HVAC] or cabling
-
- 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
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- 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
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
- Y04S10/52—Outage or fault management, e.g. fault detection or location
Abstract
The application discloses an intermittent arc fault detection method and a related device, wherein a zero sequence current waveform to be detected is divided into a plurality of waveform segments, and preprocessing is performed to obtain a plurality of test samples; inputting each test sample into an intermittent arc fault detection model to obtain a fault detection result; sequentially determining whether intermittent arc faults occur in the current test sample according to fault detection results of the test samples, if not, setting a slave factor SF=SF+1, and if so, setting a master factor PF=PF+1 and setting SF=0; when PF is larger than a first preset threshold value and SF is larger than a second preset threshold value, judging that short-time intermittent arc faults occur in the power distribution network; when the PF is larger than a third preset threshold value, judging that the power distribution network has a long-time intermittent arc fault; when the PF is smaller than the fourth preset threshold value and the SF is larger than the second preset threshold value, the fact that the intermittent arc faults do not occur in the power distribution network is judged, and the technical problem that the detection accuracy is low in an existing intermittent arc fault detection method is solved.
Description
Technical Field
The application relates to the technical field of power distribution networks, in particular to an intermittent arc fault detection method and a related device.
Background
In the ground fault event of the resonant ground system, the single-phase ground fault has high occurrence probability and accounts for the vast majority of the total faults. In order to ensure the safety, stability and reliability of the power distribution network, timely detection and effective investigation are necessary before the single-phase earth fault actually occurs. The majority of single phase earth faults develop by a gradual transition from an initial intermittent arc fault (Intermittent Arc Faults, IAF) to a stable arc earth fault and finally to a single phase permanent earth fault. Therefore, the intermittent arc faults are effectively detected, so that potential safety hazards can be found as early as possible, fault equipment with potential safety hazards or external environments affecting the safe operation of the feeder line can be eliminated in time before more serious faults occur, and the purpose of protecting the stable operation of the power distribution network is achieved.
The existing intermittent arc fault detection method generally adopts a time-frequency analysis method to extract fault waveform characteristics, then the fault waveform characteristics are input into a classifier for classification, the matching of the extracted characteristics and the classifier is not optimal, and the existing classifier is not strong in interference resistance in the detection of actually measured waveforms, so that the detection accuracy is not high.
Disclosure of Invention
The application provides an intermittent arc fault detection method and a related device, which are used for solving the technical problem that the detection precision is not high in the existing intermittent arc fault detection method.
In view of this, a first aspect of the present application provides an intermittent arc fault detection method, including:
after the zero sequence current waveform to be detected of the power distribution network is obtained, dividing the zero sequence current waveform to be detected into a plurality of waveform segments according to a time sequence, and preprocessing the waveform segments to obtain a plurality of test samples;
sequentially inputting each test sample into an intermittent arc fault detection model for fault detection to obtain a fault detection result of each test sample;
sequentially determining whether intermittent arc faults occur in the current test sample according to fault detection results of the test samples, if not, setting a slave factor sf=sf+1, if so, setting a master factor pf=pf+1, and setting sf=0;
when the main factor PF is larger than a first preset threshold value and the auxiliary factor SF is larger than a second preset threshold value, judging that the power distribution network has short-time intermittent arc faults;
when the main factor PF is larger than a third preset threshold value, judging that the power distribution network has a long-time intermittent arc fault;
and when the master factor PF is smaller than a fourth preset threshold value and the slave factor SF is larger than the second preset threshold value, judging that the power distribution network has no intermittent arc faults.
Optionally, the training process of the intermittent arc fault detection model is as follows:
acquiring a zero sequence current waveform to be trained, wherein the zero sequence current waveform to be trained comprises a waveform with intermittent arc faults and a waveform without intermittent arc faults;
dividing the zero sequence current waveform to be trained into a plurality of waveform fragments with the same size, and preprocessing the waveform fragments to obtain a plurality of training samples;
and training a preset convolutional neural network through the training sample to obtain an intermittent arc fault detection model.
Optionally, preprocessing the waveform segment to obtain a plurality of training samples or test samples, including:
mapping each waveform segment into a data sequence with a preset length based on a bicubic difference method;
normalizing each data sequence to obtain each normalized data sequence;
and converting each normalized data sequence into a matrix form according to a time sequence to obtain training samples or test samples corresponding to each waveform segment.
Optionally, the preset convolutional neural network is composed of an input layer, 5 separable convolutional layers, 4 connecting layers, 1 convolutional layer, 1 full-connection module and an output layer, wherein the full-connection module is composed of a plurality of full-connection layers in series;
the output ends of the input layers are respectively connected with the input ends of the first separable convolution layer, the second separable convolution layer and the third separable convolution layer;
the first connecting layer is connected with the output ends of the first separable convolution layer and the second separable convolution layer, and the second connecting layer is connected with the output ends of the second separable convolution layer and the third separable convolution layer;
the input end of the fourth separable convolution layer is connected with the output end of the first connecting layer, and the output end of the fourth separable convolution layer is connected with the input end of the convolution layer through the third connecting layer;
the input end of the fifth separable convolution layer is connected with the output end of the second connecting layer;
the input end of the fourth connecting layer is respectively connected with the output ends of the convolution layer and the fifth separable convolution layer, and the output end is connected with the output layer through the full-connecting layer module.
Optionally, the training samples include a fault sample and a non-fault sample, and the waveform segments corresponding to the fault sample are waveforms of one cycle before and after the fault mutation time in the zero-sequence current waveform to be trained.
A second aspect of the present application provides an intermittent arc fault detection device, comprising:
the device comprises a segmentation unit, a detection unit and a detection unit, wherein the segmentation unit is used for segmenting the zero sequence current waveform to be detected into a plurality of waveform fragments according to a time sequence after the zero sequence current waveform to be detected of the power distribution network is obtained, and preprocessing the waveform fragments to obtain a plurality of test samples;
the fault detection unit is used for sequentially inputting each test sample into the intermittent arc fault detection model to carry out fault detection, so as to obtain a fault detection result of each test sample;
the setting unit is used for sequentially determining whether intermittent arc faults occur in the current test sample according to the fault detection result of each test sample, if not, setting a slave factor SF=SF+1, if so, setting a master factor PF=PF+1, and setting SF=0;
the judging unit is used for judging that the power distribution network has short-time intermittent arc faults when the main factor PF is larger than a first preset threshold value and the auxiliary factor SF is larger than a second preset threshold value;
when the main factor PF is larger than a third preset threshold value, judging that the power distribution network has a long-time intermittent arc fault;
and when the master factor PF is smaller than a fourth preset threshold value and the slave factor SF is larger than the second preset threshold value, judging that the power distribution network has no intermittent arc faults.
Optionally, the method further comprises: training unit for:
acquiring a zero sequence current waveform to be trained, wherein the zero sequence current waveform to be trained comprises a waveform with intermittent arc faults and a waveform without intermittent arc faults;
dividing the zero sequence current waveform to be trained into a plurality of waveform fragments with the same size, and preprocessing the waveform fragments to obtain a plurality of training samples;
and training a preset convolutional neural network through the training sample to obtain an intermittent arc fault detection model.
Optionally, the preset convolutional neural network is composed of an input layer, 5 separable convolutional layers, 4 connecting layers, 1 convolutional layer, 1 full-connection module and an output layer, wherein the full-connection module is composed of a plurality of full-connection layers in series;
the output ends of the input layers are respectively connected with the input ends of the first separable convolution layer, the second separable convolution layer and the third separable convolution layer;
the first connecting layer is connected with the output ends of the first separable convolution layer and the second separable convolution layer, and the second connecting layer is connected with the output ends of the second separable convolution layer and the third separable convolution layer;
the input end of the fourth separable convolution layer is connected with the output end of the first connecting layer, and the output end of the fourth separable convolution layer is connected with the input end of the convolution layer through the third connecting layer;
the input end of the fifth separable convolution layer is connected with the output end of the second connecting layer;
the input end of the fourth connecting layer is respectively connected with the output ends of the convolution layer and the fifth separable convolution layer, and the output end is connected with the output layer through the full-connecting layer module.
A third aspect of the present application provides an intermittent arc fault detection apparatus, the apparatus comprising a processor and a memory;
the memory is used for storing program codes and transmitting the program codes to the processor;
the processor is configured to execute the intermittent arc fault detection method according to any one of the first aspects according to instructions in the program code.
A fourth aspect of the present application provides a computer readable storage medium for storing program code for performing the intermittent arc fault detection method of any one of the first aspects.
From the above technical scheme, the application has the following advantages:
the application provides an intermittent arc fault detection method, which comprises the following steps: after the zero sequence current waveform to be detected of the power distribution network is obtained, dividing the zero sequence current waveform to be detected into a plurality of waveform segments according to a time sequence, and preprocessing the waveform segments to obtain a plurality of test samples; sequentially inputting each test sample into an intermittent arc fault detection model for fault detection to obtain a fault detection result of each test sample; sequentially determining whether intermittent arc faults occur in the current test sample according to fault detection results of the test samples, if not, setting a slave factor SF=SF+1, if so, setting a master factor PF=PF+1, and setting SF=0; when the main factor PF is larger than a first preset threshold value and the auxiliary factor SF is larger than a second preset threshold value, judging that the power distribution network has short-time intermittent arc faults; when the main factor PF is larger than a third preset threshold value, judging that the power distribution network has a long-time intermittent arc fault; and when the master factor PF is smaller than the fourth preset threshold value and the slave factor SF is larger than the second preset threshold value, judging that the power distribution network has no intermittent arc faults.
In the method, fault detection is carried out through an intermittent arc fault detection model, and as the convolutional neural network has strong self-learning capability, the optimal characteristics and fault detection can be extracted in a self-adaptive manner, and the improvement of fault detection results is facilitated; in addition, the fact that the actual intermittent arc faults have random and intermittent properties and cannot be perfectly simulated by a simulation model is considered, and each test sample is further processed based on a master-slave factor method, so that the anti-interference performance and the accuracy of a detection result are improved, and the technical problem that the detection precision of the existing intermittent arc fault detection method is low is solved.
Drawings
In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described below, it being obvious that the drawings in the following description are only some embodiments of the present application, and that other drawings may be obtained according to these drawings without inventive faculty for a person skilled in the art.
Fig. 1 is a schematic flow chart of an intermittent arc fault detection method according to an embodiment of the present application;
FIG. 2 is a schematic diagram of an intermittent arc fault detection model according to an embodiment of the present disclosure;
FIG. 3 is a schematic diagram of a fault decision process based on master-slave factors according to an embodiment of the present application;
fig. 4 is a schematic structural diagram of an intermittent arc fault detection device according to an embodiment of the present application.
Detailed Description
In order to make the present application solution better understood by those skilled in the art, the following description will clearly and completely describe the technical solution in the embodiments of the present application with reference to the accompanying drawings in the embodiments of the present application, and it is apparent that the described embodiments are only some embodiments of the present application, not all embodiments. All other embodiments, which can be made by one of ordinary skill in the art without undue burden from the present disclosure, are within the scope of the present disclosure.
For ease of understanding, referring to fig. 1, one embodiment of an intermittent arc fault detection method provided herein includes:
and 101, after the zero sequence current waveform to be detected of the power distribution network is obtained, dividing the zero sequence current waveform to be detected into a plurality of waveform fragments according to a time sequence, and preprocessing the waveform fragments to obtain a plurality of test samples.
The complete zero sequence current waveform to be detected of the power distribution network can be obtained through the wave recording device, and the sampling step f is based s Dividing the zero sequence current waveform to be detected according to the time sequenceFor a plurality of waveform segments, one waveform segment corresponds to one detection window, and the segmented zero sequence current waveform to be detected consists of a plurality of continuous detection windows.
And then preprocessing each waveform segment to obtain a plurality of test samples. Specifically, mapping each waveform segment into a data sequence with a preset length based on a bicubic difference method; normalizing each data sequence to obtain each normalized data sequence; and converting each normalized data sequence into a matrix form according to the time sequence to obtain a test sample corresponding to each waveform segment.
In the embodiment of the application, the waveform segments are mapped into the data sequence with the length of 1024 by preferably using a bicubic difference method, and the numerical range of the ordinate is unchanged; and then normalizing each data sequence to normalize the ordinate numerical value of the data sequence to a scalar between 0 and 1, wherein the normalization formula is as follows:
wherein x 'is' i Is the normalized value, x, of the ith sample point in the kth waveform segment Win (k) i For the value of the ith sampling point in the kth waveform segment Win (k), max and min are respectively the maximum operator and the minimum operator.
The original data waveform is changed, each detection window waveform is equally divided into 32 segments, and 32 rows of 32 data points are arranged in the time direction, and thus the test sample format is a square matrix of 32×32 (1024=32×32).
Step 102, sequentially inputting each test sample into an intermittent arc fault detection model for fault detection, and obtaining a fault detection result of each test sample.
Further, the training process of the intermittent arc fault detection model is as follows:
acquiring a zero sequence current waveform to be trained, wherein the zero sequence current waveform to be trained comprises a waveform with intermittent arc faults and a waveform without intermittent arc faults; dividing the zero sequence current waveform to be trained into a plurality of waveform fragments with the same size, and preprocessing the waveform fragments to obtain a plurality of training samples; and training a preset convolutional neural network through a training sample to obtain an intermittent arc fault detection model.
Further, preprocessing the waveform segment to obtain a plurality of training samples or test samples, including:
mapping each waveform segment into a data sequence with a preset length based on a bicubic difference method; normalizing each data sequence to obtain each normalized data sequence; and converting each normalized data sequence into a matrix form according to the time sequence to obtain training samples corresponding to each waveform segment.
In this embodiment of the present application, the training samples include a fault sample and a non-fault sample, the waveform segment corresponding to the fault sample is a waveform of one cycle before and after the fault mutation time in the zero sequence current waveform to be trained, that is, in this embodiment of the present application, the length unit of each waveform segment is two power frequency cycles, and the sampling frequency may be 10kHz. The complete waveform data stream of intermittent arc faults is divided into a plurality of waveform fragments taking two cycles as units, one waveform fragment corresponds to one detection window, the sequence of the detection window is mapped into a data sequence with the length of 1024 by using a bicubic interpolation method, and the numerical range of the ordinate is unchanged. The minimum and maximum normalization method then normalizes the ordinate data values to a scalar between 0 and 1. Finally, the original data waveform is changed, each detection window waveform is equally divided into 32 segments, and 32 rows of 32 data points are arranged in the time direction, and therefore, the training sample format is a square matrix of 32×32 (1024=32×32).
Further, the preset convolutional neural network in the embodiment of the present application is composed of an Input layer (Input), 5 separable convolutional layers (sep_conv), 4 connection layers (connection), 1 convolutional layer (Conv), 1 full connection module and an Output layer (Output), the full connection module is composed of a plurality of full connection layers (Dense) connected in series, and specific network structure and Input/Output characteristic diagram size parameters of each layer can refer to fig. 2;
the output end of the input layer is respectively connected with the input ends of the first separable convolution layer, the second separable convolution layer and the third separable convolution layer;
the first connecting layer is connected with the output ends of the first separable convolution layer and the second separable convolution layer, and the second connecting layer is connected with the output ends of the second separable convolution layer and the third separable convolution layer;
the input end of the fourth separable convolution layer is connected with the output end of the first connecting layer, and the output end of the fourth separable convolution layer is connected with the input end of the convolution layer through the third connecting layer;
the input end of the fifth separable convolution layer is connected with the output end of the second connecting layer;
the input end of the fourth connecting layer is respectively connected with the output ends of the fifth separable convolution layers of the convolution layers, and the output end is connected with the output layer through the full connecting layer module.
The intermittent arc fault detection model in the application has a 10-layer network structure, wherein layers 2 to 11 are innovative structures combining an acceptance architecture and a Dense block, the structure is used for replacing a manually designed feature extraction method, and layers 12 to 15 are full-connection layers and mainly play a role of a classifier.
Layer 1 is the input layer, and the input test sample format is a square matrix of 32×32; layers 2-4 are separable convolution layers, divided into two phases: first, a single 3 x 3 convolution kernel is used to perform feature extraction while keeping the output feature mapping matrix size unchanged using a padding technique. Secondly, 36 feature mapping matrixes are generated by using 36 one-dimensional convolution kernels; the 5 th layer and the 6 th layer are all connecting layers, the 5 th layer combines the feature mapping matrixes output by the 2 nd layer and the 4 th layer into 108 feature mapping matrix sets in series, and the 6 th layer combines the feature mapping matrixes output by the 3 rd layer and the 4 th layer into 72 feature mapping matrix sets in series; the 7 th-8 th layers are separable convolution layers, and 36 feature mapping matrixes with the output features unchanged are generated; the layer 9 combines the feature mapping matrix output by the layer 3 and the layer 7 in series with the layer 6; layer 10 is a conventional convolutional layer; the 11 th layer is a connecting layer, and the feature mapping matrixes output by the 8 th layer and the 10 th layer are combined in series to form 72 feature mapping matrix sets; the layers 12 to 15 are full-connection layers, the layer 12 expands all feature mapping matrixes into 1-dimensional tensors, the activation function of the layer 13 is 'ReLU', the full-connection layer of the layer 14 adopts dropout, the rejection rate is preferably set to 25%, and in order to reduce the possibility of overfitting, the activation function of the layer 15 is 'Softmax', and the loss function corresponding to the activation function of the layer 15 is a cross entropy function.
And sequentially inputting each test sample into an intermittent arc fault detection model according to the time direction to perform fault detection to obtain a fault detection result of each test sample, wherein the intermittent arc fault detection model outputs 0 to indicate that no intermittent arc fault occurs in the current test sample, and the intermittent arc fault detection model outputs 1 to indicate that intermittent arc fault occurs in the current test sample.
The neural network architecture provided by the embodiment of the application has stronger self-learning and associative storage capacity, can quickly find an optimal solution, has stronger robustness and better adaptability, and has high accuracy in both simulation and actual measurement samples; according to the embodiment of the application, the advantages of the acceptance structure and the advantages of the Dense block are combined, the width of the network is increased, the front and the back of the network layer are associated, and a network structure capable of avoiding overfitting and avoiding gradient disappearance is formed; moreover, the intermittent arc faults can be detected in real time without a fault triggering algorithm.
Step 103, determining whether an intermittent arc fault occurs in the current test sample according to the fault detection result of each test sample, if not, setting a slave factor sf=sf+1, if so, setting a master factor pf=pf+1, and setting sf=0.
In the embodiment of the application, the fact that the actual intermittent arc faults are random and intermittent and cannot be perfectly simulated by a simulation model is considered, and in order to improve the anti-interference performance and accuracy of a detection result, the detection of the single-node intermittent arc faults is realized, and a master-slave factor method is provided. Referring to fig. 3, fig. 3 (a) is a schematic diagram of an intermittent process of intermittent arc faults, fig. b is an ideal master factor variation graph, and fig. c is an ideal slave factor variation graph. The complete zero sequence current waveform (dataStream) according to the sampling step f s Sequentially inputting the detection results into a detection window in a time direction, wherein one detection reference corresponds to one test sample, fault detection results (0 or 1) of all the test samples can be obtained through the steps, whether intermittent arc faults occur in the current test sample or not is sequentially determined according to the fault detection results of all the test samples, if not, a slave factor SF=SF+1 is set, if yes, a master factor PF=PF+1 is set, and SF=0 is set. Wherein the primary factor PF and the secondary factor SF have an initial value of 0.
104, when the main factor PF is larger than a first preset threshold value and the secondary factor SF is larger than a second preset threshold value, judging that the power distribution network has short-time intermittent arc faults; when the main factor PF is larger than a third preset threshold value, judging that the power distribution network has a long-time intermittent arc fault; and when the master factor PF is smaller than the fourth preset threshold value and the slave factor SF is larger than the second preset threshold value, judging that the power distribution network has no intermittent arc faults.
The first preset threshold in the embodiment of the present application is preferably η 1 ×f s The second preset threshold is preferably β×f s The third preset threshold is preferably η 2 ×f s The fourth preset threshold is preferably η 3 ×f s 。η 1 The threshold coefficient indicating whether the main factor is judged to be a short-time intermittent fault or not can take a value of 0.2-0.3; η (eta) 2 The threshold coefficient indicating whether the main factor is judged to be a long-time intermittent fault or not can take the value of 1-2; η (eta) 3 The threshold coefficient indicating whether the main factor judges intermittent faults occur or not can take a value of 0.05; beta represents a threshold coefficient for judging whether intermittent arc faults occur from factors, and can take a value of 0.1.
When the main factor PF>η 1 ×f s And from factor SF>β×f s When the power distribution network is in short-time intermittent arc faults, the power distribution network is judged to be in short-time intermittent arc faults;
when the main factor PF>η 2 ×f s When the power distribution network is in the intermittent arc fault for a long time, the power distribution network is judged to have the intermittent arc fault for a long time;
when the main factor PF<η 3 ×f s And from factor SF>β×f s When the power distribution network is judged to not generate intermittenceArc fault and initializing the master factor PF and the slave factor SF to 0, and returning to the initial state.
The intermittent arc fault detection model is combined with the master-slave factor method, so that a plurality of fragment information under the same fault waveform can be analyzed, the accidental of judgment errors is greatly reduced by combining the master-slave factor method with the intermittent arc fault detection model, and the detection accuracy of intermittent arc faults is improved.
In the embodiment of the application, the intermittent arc fault detection model is used for fault detection, and the convolutional neural network has strong self-learning capability, so that the optimal characteristics and fault detection can be extracted in a self-adaptive manner, and the fault detection result can be improved; in addition, the fact that the actual intermittent arc faults have random and intermittent properties and cannot be perfectly simulated by a simulation model is considered, and each test sample is further processed based on a master-slave factor method, so that the anti-interference performance and the accuracy of a detection result are improved, and the technical problem that the detection precision of the existing intermittent arc fault detection method is low is solved.
The above is an embodiment of an intermittent arc fault detection method provided in the present application, and the following is an embodiment of an intermittent arc fault detection device provided in the present application.
Referring to fig. 4, an intermittent arc fault detection apparatus provided in an embodiment of the present application includes:
the segmentation unit is used for segmenting the zero sequence current waveform to be detected into a plurality of waveform fragments according to the time sequence after the zero sequence current waveform to be detected of the power distribution network is obtained, and preprocessing the waveform fragments to obtain a plurality of test samples;
the fault detection unit is used for sequentially inputting each test sample into the intermittent arc fault detection model to carry out fault detection, so as to obtain a fault detection result of each test sample;
the setting unit is used for sequentially determining whether intermittent arc faults occur in the current test sample according to the fault detection results of the test samples, if not, setting a slave factor sf=sf+1, if so, setting a master factor pf=pf+1, and setting sf=0;
the judging unit is used for judging that the power distribution network has short-time intermittent arc faults when the main factor PF is larger than a first preset threshold value and the auxiliary factor SF is larger than a second preset threshold value;
when the main factor PF is larger than a third preset threshold value, judging that the power distribution network has a long-time intermittent arc fault;
and when the master factor PF is smaller than the fourth preset threshold value and the slave factor SF is larger than the second preset threshold value, judging that the power distribution network has no intermittent arc faults.
As a further improvement, further comprising: training unit for:
acquiring a zero sequence current waveform to be trained, wherein the zero sequence current waveform to be trained comprises a waveform with intermittent arc faults and a waveform without intermittent arc faults;
dividing the zero sequence current waveform to be trained into a plurality of waveform fragments with the same size, and preprocessing the waveform fragments to obtain a plurality of training samples;
and training a preset convolutional neural network through a training sample to obtain an intermittent arc fault detection model.
As a further improvement, the preset convolutional neural network consists of an input layer, 5 separable convolutional layers, 4 connecting layers, 1 convolutional layer, 1 full-connection module and an output layer, wherein the full-connection module consists of a plurality of full-connection layers which are connected in series;
the output end of the input layer is respectively connected with the input ends of the first separable convolution layer, the second separable convolution layer and the third separable convolution layer;
the first connecting layer is connected with the output ends of the first separable convolution layer and the second separable convolution layer, and the second connecting layer is connected with the output ends of the second separable convolution layer and the third separable convolution layer;
the input end of the fourth separable convolution layer is connected with the output end of the first connecting layer, and the output end of the fourth separable convolution layer is connected with the input end of the convolution layer through the third connecting layer;
the input end of the fifth separable convolution layer is connected with the output end of the second connecting layer;
the input end of the fourth connecting layer is respectively connected with the output ends of the fifth separable convolution layers of the convolution layers, and the output end is connected with the output layer through the full connecting layer module.
In the embodiment of the application, the intermittent arc fault detection model is used for fault detection, and the convolutional neural network has strong self-learning capability, so that the optimal characteristics and fault detection can be extracted in a self-adaptive manner, and the fault detection result can be improved; in addition, the fact that the actual intermittent arc faults have random and intermittent properties and cannot be perfectly simulated by a simulation model is considered, and each test sample is further processed based on a master-slave factor method, so that the anti-interference performance and the accuracy of a detection result are improved, and the technical problem that the detection precision of the existing intermittent arc fault detection method is low is solved.
The embodiment of the application also provides intermittent arc fault detection equipment, which comprises a processor and a memory;
the memory is used for storing the program codes and transmitting the program codes to the processor;
the processor is configured to perform the intermittent arc fault detection method of the foregoing method embodiments according to instructions in the program code.
The embodiment of the application also provides a computer readable storage medium, and the computer readable storage medium is used for storing program codes, and the program codes are used for executing the intermittent arc fault detection method in the embodiment of the method.
It will be clear to those skilled in the art that, for convenience and brevity of description, specific working procedures of the apparatus and units described above may refer to corresponding procedures in the foregoing method embodiments, which are not described herein again.
The terms "first," "second," "third," "fourth," and the like in the description of the present application and in the above-described figures, if any, are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that embodiments of the present application described herein may be capable of operation in sequences other than those illustrated or described herein, for example. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
It should be understood that in this application, "at least one" means one or more, and "a plurality" means two or more. "and/or" for describing the association relationship of the association object, the representation may have three relationships, for example, "a and/or B" may represent: only a, only B and both a and B are present, wherein a, B may be singular or plural. The character "/" generally indicates that the context-dependent object is an "or" relationship. "at least one of" or the like means any combination of these items, including any combination of single item(s) or plural items(s). For example, at least one (one) of a, b or c may represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", wherein a, b, c may be single or plural.
In the several embodiments provided in this application, it should be understood that the disclosed apparatus and method may be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative, e.g., the division of the units is merely a logical function division, and there may be additional divisions when actually implemented, e.g., multiple units or components may be combined or integrated into another system, or some features may be omitted or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be an indirect coupling or communication connection via some interfaces, devices or units, which may be in electrical, mechanical or other form.
The units described as separate units may or may not be physically separate, and units shown as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
In addition, each functional unit in each embodiment of the present application may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units may be implemented in hardware or in software functional units.
The integrated units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present application may be embodied in essence or a part contributing to the prior art or all or part of the technical solution in the form of a software product stored in a storage medium, including several instructions to execute all or part of the steps of the methods described in the embodiments of the present application by a computer device (which may be a personal computer, a server, or a network device, etc.). And the aforementioned storage medium includes: u disk, mobile hard disk, read-Only Memory (ROM), random access Memory (Random Access Memory, RAM), magnetic disk or optical disk, etc.
The above embodiments are merely for illustrating the technical solution of the present application, and not for limiting the same; although the present application has been described in detail with reference to the foregoing embodiments, it should be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the corresponding technical solutions.
Claims (7)
1. A method of intermittent arc fault detection, comprising:
after the zero sequence current waveform to be detected of the power distribution network is obtained, dividing the zero sequence current waveform to be detected into a plurality of waveform segments according to a time sequence, and preprocessing the waveform segments to obtain a plurality of test samples;
sequentially inputting each test sample into an intermittent arc fault detection model for fault detection to obtain a fault detection result of each test sample;
sequentially determining whether intermittent arc faults occur in the current test sample according to fault detection results of the test samples, if not, setting a slave factor sf=sf+1, if so, setting a master factor pf=pf+1, and setting sf=0;
when the main factor PF is larger than a first preset threshold value and the auxiliary factor SF is larger than a second preset threshold value, judging that the power distribution network has short-time intermittent arc faults;
when the main factor PF is larger than a third preset threshold value, judging that the power distribution network has a long-time intermittent arc fault;
when the master factor PF is smaller than a fourth preset threshold value and the slave factor SF is larger than the second preset threshold value, judging that the power distribution network has no intermittent arc fault;
the training process of the intermittent arc fault detection model comprises the following steps:
acquiring a zero sequence current waveform to be trained, wherein the zero sequence current waveform to be trained comprises a waveform with intermittent arc faults and a waveform without intermittent arc faults;
dividing the zero sequence current waveform to be trained into a plurality of waveform fragments with the same size, and preprocessing the waveform fragments to obtain a plurality of training samples;
training a preset convolutional neural network through the training sample to obtain an intermittent arc fault detection model;
preprocessing the waveform segment to obtain a plurality of training samples or test samples, wherein the method comprises the following steps:
mapping each waveform segment into a data sequence with a preset length based on a bicubic difference method;
normalizing each data sequence to obtain each normalized data sequence;
and converting each normalized data sequence into a matrix form according to a time sequence to obtain training samples or test samples corresponding to each waveform segment.
2. The intermittent arc fault detection method according to claim 1, wherein the preset convolutional neural network is composed of an input layer, 5 separable convolutional layers, 4 connecting layers, 1 convolutional layer, 1 full-connection module and an output layer, and the full-connection module is composed of a plurality of full-connection layers connected in series;
the output ends of the input layers are respectively connected with the input ends of the first separable convolution layer, the second separable convolution layer and the third separable convolution layer;
the first connecting layer is connected with the output ends of the first separable convolution layer and the second separable convolution layer, and the second connecting layer is connected with the output ends of the second separable convolution layer and the third separable convolution layer;
the input end of the fourth separable convolution layer is connected with the output end of the first connecting layer, and the output end of the fourth separable convolution layer is connected with the input end of the convolution layer through the third connecting layer;
the input end of the fifth separable convolution layer is connected with the output end of the second connecting layer;
the input end of the fourth connecting layer is respectively connected with the output ends of the convolution layer and the fifth separable convolution layer, and the output end is connected with the output layer through the full-connecting layer module.
3. The intermittent arc fault detection method according to claim 1, wherein the training samples comprise fault samples and non-fault samples, and the waveform segments corresponding to the fault samples are waveforms of one cycle before and after a fault abrupt change time in the zero-sequence current waveform to be trained.
4. An intermittent arc fault detection apparatus, comprising:
the device comprises a segmentation unit, a detection unit and a detection unit, wherein the segmentation unit is used for segmenting the zero sequence current waveform to be detected into a plurality of waveform fragments according to a time sequence after the zero sequence current waveform to be detected of the power distribution network is obtained, and preprocessing the waveform fragments to obtain a plurality of test samples;
the fault detection unit is used for sequentially inputting each test sample into the intermittent arc fault detection model to carry out fault detection, so as to obtain a fault detection result of each test sample;
the setting unit is used for sequentially determining whether intermittent arc faults occur in the current test sample according to the fault detection result of each test sample, if not, setting a slave factor SF=SF+1, if so, setting a master factor PF=PF+1, and setting SF=0;
the judging unit is used for judging that the power distribution network has short-time intermittent arc faults when the main factor PF is larger than a first preset threshold value and the auxiliary factor SF is larger than a second preset threshold value;
when the main factor PF is larger than a third preset threshold value, judging that the power distribution network has a long-time intermittent arc fault;
when the master factor PF is smaller than a fourth preset threshold value and the slave factor SF is larger than the second preset threshold value, judging that the power distribution network has no intermittent arc fault;
training unit for:
acquiring a zero sequence current waveform to be trained, wherein the zero sequence current waveform to be trained comprises a waveform with intermittent arc faults and a waveform without intermittent arc faults;
dividing the zero sequence current waveform to be trained into a plurality of waveform fragments with the same size, and preprocessing the waveform fragments to obtain a plurality of training samples;
training a preset convolutional neural network through the training sample to obtain an intermittent arc fault detection model;
preprocessing the waveform segment to obtain a plurality of training samples or test samples, wherein the method comprises the following steps:
mapping each waveform segment into a data sequence with a preset length based on a bicubic difference method;
normalizing each data sequence to obtain each normalized data sequence;
and converting each normalized data sequence into a matrix form according to a time sequence to obtain training samples or test samples corresponding to each waveform segment.
5. The intermittent arc fault detection device according to claim 4, wherein the preset convolutional neural network is composed of an input layer, 5 separable convolutional layers, 4 connecting layers, 1 convolutional layer, 1 full-connection module and an output layer, and the full-connection module is composed of a plurality of full-connection layers connected in series;
the output ends of the input layers are respectively connected with the input ends of the first separable convolution layer, the second separable convolution layer and the third separable convolution layer;
the first connecting layer is connected with the output ends of the first separable convolution layer and the second separable convolution layer, and the second connecting layer is connected with the output ends of the second separable convolution layer and the third separable convolution layer;
the input end of the fourth separable convolution layer is connected with the output end of the first connecting layer, and the output end of the fourth separable convolution layer is connected with the input end of the convolution layer through the third connecting layer;
the input end of the fifth separable convolution layer is connected with the output end of the second connecting layer;
the input end of the fourth connecting layer is respectively connected with the output ends of the convolution layer and the fifth separable convolution layer, and the output end is connected with the output layer through the full-connecting layer module.
6. An intermittent arc fault detection apparatus, the apparatus comprising a processor and a memory;
the memory is used for storing program codes and transmitting the program codes to the processor;
the processor is configured to perform the intermittent arc fault detection method of any of claims 1-3 according to instructions in the program code.
7. A computer readable storage medium, characterized in that the computer readable storage medium is for storing a program code for performing the intermittent arc fault detection method of any one of claims 1-3.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110838493.6A CN113468704B (en) | 2021-07-23 | 2021-07-23 | Intermittent arc fault detection method and related device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110838493.6A CN113468704B (en) | 2021-07-23 | 2021-07-23 | Intermittent arc fault detection method and related device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113468704A CN113468704A (en) | 2021-10-01 |
| CN113468704B true CN113468704B (en) | 2024-02-02 |
Family
ID=77882139
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110838493.6A Active CN113468704B (en) | 2021-07-23 | 2021-07-23 | Intermittent arc fault detection method and related device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113468704B (en) |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103529316A (en) * | 2013-08-15 | 2014-01-22 | 国家电网公司 | Comprehensive detection method for high resistance ground faults of electric power system |
| CN104375061A (en) * | 2014-12-02 | 2015-02-25 | 国网上海市电力公司 | Intermittent grounding fault detection system of power distribution network |
| CN104597344A (en) * | 2015-01-08 | 2015-05-06 | 上海交通大学 | Fault arc online detecting method based on wavelet first-layer high-frequency component correlation |
| CN107871044A (en) * | 2017-11-13 | 2018-04-03 | 杭州电魂网络科技股份有限公司 | Course method to set up and device |
| CN110398663A (en) * | 2019-07-03 | 2019-11-01 | 东南大学 | A kind of flexible direct current electric network fault recognition methods based on convolutional neural networks |
| CN110501631A (en) * | 2019-08-19 | 2019-11-26 | 重庆大学 | A kind of online intermittent fault detection and diagnostic method |
| CN110717144A (en) * | 2019-08-19 | 2020-01-21 | 珠海格力电器股份有限公司 | Fault arc detection method, device, terminal and storage medium |
| CN111123048A (en) * | 2019-12-23 | 2020-05-08 | 温州大学 | Series fault arc detection device and method based on convolutional neural network |
| CN111707908A (en) * | 2020-07-29 | 2020-09-25 | 中国科学技术大学先进技术研究院 | Multi-load loop series fault arc detection method and device and storage medium |
| CN113049922A (en) * | 2020-04-22 | 2021-06-29 | 青岛鼎信通讯股份有限公司 | Fault arc signal detection method adopting convolutional neural network |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10503998B2 (en) * | 2016-11-07 | 2019-12-10 | Gracenote, Inc. | Recurrent deep neural network system for detecting overlays in images |
| EP3791196A4 (en) * | 2018-05-07 | 2022-01-05 | Inhand Networks Inc. | System for locating fault in power distribution network based on mixed mode wave recording |
-
2021
- 2021-07-23 CN CN202110838493.6A patent/CN113468704B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103529316A (en) * | 2013-08-15 | 2014-01-22 | 国家电网公司 | Comprehensive detection method for high resistance ground faults of electric power system |
| CN104375061A (en) * | 2014-12-02 | 2015-02-25 | 国网上海市电力公司 | Intermittent grounding fault detection system of power distribution network |
| CN104597344A (en) * | 2015-01-08 | 2015-05-06 | 上海交通大学 | Fault arc online detecting method based on wavelet first-layer high-frequency component correlation |
| CN107871044A (en) * | 2017-11-13 | 2018-04-03 | 杭州电魂网络科技股份有限公司 | Course method to set up and device |
| CN110398663A (en) * | 2019-07-03 | 2019-11-01 | 东南大学 | A kind of flexible direct current electric network fault recognition methods based on convolutional neural networks |
| CN110501631A (en) * | 2019-08-19 | 2019-11-26 | 重庆大学 | A kind of online intermittent fault detection and diagnostic method |
| CN110717144A (en) * | 2019-08-19 | 2020-01-21 | 珠海格力电器股份有限公司 | Fault arc detection method, device, terminal and storage medium |
| CN111123048A (en) * | 2019-12-23 | 2020-05-08 | 温州大学 | Series fault arc detection device and method based on convolutional neural network |
| CN113049922A (en) * | 2020-04-22 | 2021-06-29 | 青岛鼎信通讯股份有限公司 | Fault arc signal detection method adopting convolutional neural network |
| CN111707908A (en) * | 2020-07-29 | 2020-09-25 | 中国科学技术大学先进技术研究院 | Multi-load loop series fault arc detection method and device and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113468704A (en) | 2021-10-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wang et al. | ArcNet: Series AC arc fault detection based on raw current and convolutional neural network | |
| CN113049922B (en) | Fault arc signal detection method adopting convolutional neural network | |
| CN103558529B (en) | A kind of mode identification method of three-phase cartridge type supertension GIS partial discharge altogether | |
| CN108737406A (en) | A kind of detection method and system of abnormal flow data | |
| Aydin et al. | An approach for automated fault diagnosis based on a fuzzy decision tree and boundary analysis of a reconstructed phase space | |
| CN111475680A (en) | Method, device, equipment and storage medium for detecting abnormal high-density subgraph | |
| CN110141219A (en) | Myocardial infarction automatic testing method based on lead fusion deep neural network | |
| CN102016607A (en) | Method and apparatus for analyzing waveform signals of a power system | |
| CN109257383B (en) | BGP anomaly detection method and system | |
| CN113222036B (en) | Automatic defect identification method and device for high-voltage cable grounding system | |
| CN116524200A (en) | High-voltage circuit breaker fault diagnosis method based on image recognition | |
| CN102324007A (en) | Method for detecting abnormality based on data mining | |
| CN112381351A (en) | Power utilization behavior change detection method and system based on singular spectrum analysis | |
| CN116432112A (en) | Arc fault detection method based on wavelet packet transformation and residual convolution neural network | |
| CN113468704B (en) | Intermittent arc fault detection method and related device | |
| CN112651849B (en) | Method and system for identifying abnormal voltage monitoring data based on unbalanced data set | |
| CN115015683B (en) | Cable production performance test method, device, equipment and storage medium | |
| KR20200116425A (en) | RNN based partial discharge detection apparatus and method in gas-insulated switchgear | |
| CN116127447A (en) | Virtual power plant false data injection attack detection method, device, terminal and medium | |
| CN115564324A (en) | Electric power fingerprint identification method, device and equipment based on event detection | |
| CN115114956A (en) | Fault diagnosis method and system | |
| CN114943254A (en) | A-CNN-based optical fiber differential protection algorithm for output line of photovoltaic power station | |
| CN114186590A (en) | Power distribution network single-phase earth fault identification method based on wavelet and deep learning | |
| CN117290756B (en) | Power transmission line fault identification method and device based on federal learning and electronic equipment | |
| CN117272143A (en) | Power transmission line fault identification method and device based on gram angle field and residual error network |
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 |