CN109116196A - A kind of power cable fault discharging sound intelligent identification Method - Google Patents
A kind of power cable fault discharging sound intelligent identification Method Download PDFInfo
- Publication number
- CN109116196A CN109116196A CN201810737476.1A CN201810737476A CN109116196A CN 109116196 A CN109116196 A CN 109116196A CN 201810737476 A CN201810737476 A CN 201810737476A CN 109116196 A CN109116196 A CN 109116196A
- Authority
- CN
- China
- Prior art keywords
- cable fault
- sound
- data
- voice signal
- power cable
- 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.)
- Granted
Links
- 238000007599 discharging Methods 0.000 title claims abstract description 65
- 238000000034 method Methods 0.000 title claims abstract description 41
- 230000005236 sound signal Effects 0.000 claims abstract description 43
- 238000012706 support-vector machine Methods 0.000 claims abstract description 19
- 238000006243 chemical reaction Methods 0.000 claims abstract description 5
- 238000012545 processing Methods 0.000 claims description 12
- 239000000284 extract Substances 0.000 claims description 11
- 238000012549 training Methods 0.000 claims description 9
- 238000004364 calculation method Methods 0.000 claims description 6
- 239000011159 matrix material Substances 0.000 claims description 3
- 239000004744 fabric Substances 0.000 claims description 2
- 238000000605 extraction Methods 0.000 abstract description 3
- 238000010606 normalization Methods 0.000 description 7
- 238000001514 detection method Methods 0.000 description 5
- 239000000463 material Substances 0.000 description 4
- 239000000523 sample Substances 0.000 description 3
- 238000012360 testing method Methods 0.000 description 3
- 238000003745 diagnosis Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 235000013399 edible fruits Nutrition 0.000 description 2
- 230000005672 electromagnetic field Effects 0.000 description 2
- 230000004807 localization Effects 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 230000004888 barrier function Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000009413 insulation Methods 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000007634 remodeling Methods 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
- 238000010998 test method Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/12—Testing dielectric strength or breakdown voltage ; Testing or monitoring effectiveness or level of insulation, e.g. of a cable or of an apparatus, for example using partial discharge measurements; Electrostatic testing
- G01R31/1209—Testing dielectric strength or breakdown voltage ; Testing or monitoring effectiveness or level of insulation, e.g. of a cable or of an apparatus, for example using partial discharge measurements; Electrostatic testing using acoustic measurements
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/12—Testing dielectric strength or breakdown voltage ; Testing or monitoring effectiveness or level of insulation, e.g. of a cable or of an apparatus, for example using partial discharge measurements; Electrostatic testing
- G01R31/1227—Testing dielectric strength or breakdown voltage ; Testing or monitoring effectiveness or level of insulation, e.g. of a cable or of an apparatus, for example using partial discharge measurements; Electrostatic testing of components, parts or materials
- G01R31/1263—Testing dielectric strength or breakdown voltage ; Testing or monitoring effectiveness or level of insulation, e.g. of a cable or of an apparatus, for example using partial discharge measurements; Electrostatic testing of components, parts or materials of solid or fluid materials, e.g. insulation films, bulk material; of semiconductors or LV electronic components or parts; of cable, line or wire insulation
- G01R31/1272—Testing dielectric strength or breakdown voltage ; Testing or monitoring effectiveness or level of insulation, e.g. of a cable or of an apparatus, for example using partial discharge measurements; Electrostatic testing of components, parts or materials of solid or fluid materials, e.g. insulation films, bulk material; of semiconductors or LV electronic components or parts; of cable, line or wire insulation of cable, line or wire insulation, e.g. using partial discharge measurements
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
- G10L25/48—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 specially adapted for particular use
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Acoustics & Sound (AREA)
- Computational Linguistics (AREA)
- Signal Processing (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Human Computer Interaction (AREA)
- Multimedia (AREA)
- Locating Faults (AREA)
- Testing Relating To Insulation (AREA)
Abstract
A kind of power cable fault discharging sound intelligent identification Method, belongs to power cable fault field of detecting.It is characterized by comprising following steps: step 1, sampled voice signal simultaneously carries out analog-to-digital conversion；Step 2, data prediction and feature extraction；Step 3, the feature vector feeding support vector machines of the voice signal extracted is identified and obtains judging result；Step 4, present sound signals execute step 5~step 7 if it is cable fault discharging sound, sequence, if it is non-cable fault discharge sound, execute step 7；Step 5, the data of present sound signals are saved；Step 6, the calculating of related coefficient is carried out；Step 7, judging result and sound signal waveform are shown.In this power cable fault discharging sound intelligent identification Method, based on support vector machines and can automatic identification power cable fault discharge sound, get rid of for a long time rely on tester personal experience restriction, greatly promote power cable fault fixed point working efficiency.
Description
Technical field
A kind of power cable fault discharging sound intelligent identification Method, belongs to power cable fault field of detecting.
Background technique
Submarine transmission line of the power cable in Urban Underground power grid, the inside power supply circuit of industrial and mining enterprises and crossing river and sea
It is widely used in road.Cable once breaks down, and can produce to enterprise and cause loss of outage, bring not to resident living
Just, thus cable break down after need to find failure as early as possible and repaired.
Cable fault is searched to generally require by three fault diagnosis, fault localization and position determination of fault steps.Fault diagnosis
It is the insulation resistance value that the connectivity of each phase of cable, failure phase are checked with the tools such as multimeter and equipment, it is therefore an objective to distinguish fault
Matter selects suitable test method for subsequent step；Fault localization is to measure cable between fault point and test point with instrument
Length, it is therefore an objective to substantially determine the region where cable fault, reduce the range of trouble shoot；Position determination of fault is detected with instrument
The intensity of fault-signal or arrival time, it is therefore an objective to move closer to and finally confirm location of fault.
At present in type cable failure positioning link, event is mainly searched by the method for detection cable fault electric discharge sound both at home and abroad
Barrier point.There are two types of implementations for this method: sound detection harmony magnetic-synchro method.
The working principle of sound detection is to receive Method of Cable Trouble Point electric discharge sound, sound letter using the sounding probe for being placed in ground
Number host is transmitted to by probe, by the processing such as filtering, amplify, then by host will treated that voice signal is transmitted to listens to ear
Machine listens to for tester and identifies that tester compares the strong and weak variation of different location fault discharge sound by analysis, judges event
Hinder distance or the position of point.
The working principle of sound magnetic-synchro method is to receive cable fault using the sound magnetic-synchro detection probe for being placed in ground is synchronous
The voice signal and electromagnetic field signal that point electric discharge generates, sound and electromagnetic field signal are transmitted to host by popping one's head in, and host is to two kinds
Signal such as is filtered, amplifies at the processing, is then sent to voice signal and listens to earphone, while by sound and field waveform and two
Difference (delay of sound magnetic) between kind signal arrival time is displayed on the screen.The sound and master that the comprehensive earphone of tester listens to
The information that machine is shown is analyzed, and is identified fault discharge sound, is judged distance or the position of fault point.
Sound detection harmony magnetic-synchro method still has following shortcoming in terms of fault discharge voice recognition at present: (1)
Fault discharge voice recognition fully relies on tester personal experience, and the insufficient tester of experience is difficult to accurately identify fault discharge sound
Sound.(2) since test site is there are a large amount of ambient noise and interference, even if experienced tester often be difficult to standard
Really distinguish and identify fault discharge sound.(3) training time of manual identified fault discharge sound is long, at high cost, and artificial
Identify that the experience of fault discharge sound and technical ability are detached from site environment and are difficult to teach and inherit.
Summary of the invention
The technical problem to be solved by the present invention is overcome the deficiencies of the prior art and provide it is a kind of based on support vector machines and
Can automatic identification power cable fault discharge sound, get rid of for a long time rely on tester personal experience restriction, mention significantly
Rise the power cable fault discharging sound intelligent identification Method of power cable fault fixed point working efficiency.
The technical solution adopted by the present invention to solve the technical problems is: the power cable fault discharging sound intelligent recognition side
Method, characterized by the following steps:
Step 1, start sampled voice signal in cable fault discharging time and carry out analog-to-digital conversion to obtain initial data x0
(i)；
Step 2, DC processing is carried out to digitized voice signal, obtains data x1(i), then to DC processing
Data be normalized, obtain data x2(i), it finally extracts and obtains the feature vector of voice signal；
Step 3, the feature vector feeding support vector machines of the voice signal extracted is identified and obtains judgement knot
Fruit；
Step 4, whether present sound signals are cable fault discharging sound, and if it is cable fault discharging sound, sequence is executed
Step 5~step 7 executes step 7 if it is non-cable fault discharge sound；
Step 5, the data of present sound signals are saved；
Step 6, the acquisition for carrying out second of voice signal is gone forward side by side the calculating of Correlation series；
Step 7, judging result and sound signal waveform are shown.
Preferably, feature vector described in step 2 includes:
According to data x3(i) it extracts voice signal short-time energy and is distributed pulse width feature z1:
Meeting conditionUnder the premise of, definition: wlast=max [1,2 ...,
N-D+1], wfirst=min [1,2 ..., n-D+1], then z1Are as follows: z1=wlast-wfirst
According to data x3(i) it extracts voice signal short-time energy and is distributed pulse depth-width ratio feature z2:
According to data x3(i) it extracts voice signal short-time energy and is distributed pulse position feature z3:
According to data x4(i) voice signal short-time zero-crossing rate feature z is extracted4:
Meeting conditionUnder the premise of, define rlast=max [1,2 ..., n-
D], rfirst=min [1,2 ..., n-D], then z4Are as follows:
Wherein: x3(i) for according to the data x2(i) the sound signal data short-time energy point that the step-length being calculated is D
Cloth；x4(i) for according to data x3(i) step-length being calculated is the sound signal data short-time zero-crossing rate of D.
Preferably, the data x that the sound signal data short-time energy that the step-length is D is distributed3(i) are as follows:
Wherein: D indicates step-length.
Preferably, the step-length is the data x of the sound signal data short-time zero-crossing rate of D4(i) are as follows:
Wherein: sign is sign function, and th indicates to be greater than zero threshold value, i ∈ [1,2 ..., n-D].
Preferably, the calculation formula of related coefficient described in step 6 are as follows:
Wherein: xk(i), xk+1(i), i ∈ [1,2 ..., n] respectively indicates the sound signal data acquired twice in succession.
Preferably, the training process of support vector machines described in step 3 includes the following steps:
Step 3-1 respectively prepares M cable fault discharging sound signal data and non-cable fault discharge acoustical signal data；
Step 3-2 respectively carries out M cable fault discharging sound signal data and non-cable fault discharge acoustical signal data
Processing, obtains eigenmatrix zij, i ∈ [1,2 ..., 2M], j ∈ [1,2,3,4],
Wherein, i=[1,2 ..., M] is cable fault discharging sound signal characteristic vector, i=[M+1, M+2 ..., 2M] right and wrong
Cable fault discharging sound signal characteristic vector；
Step 3-3, the column vector s of assignment 2M row indicate cable fault discharging sound wherein the 1st is assigned a value of 1 to M row, the
M+1 is assigned a value of 0 expression non-cable fault discharge sound to 2M row；
Step 3-4, by matrix zijWith column vector s input support vector machines training function, linear kernel is selected to be trained.
Compared with prior art, the present invention has the beneficial effects that
In this power cable fault discharging sound intelligent identification Method, based on support vector machines and being capable of automatic identification electric power
Cable fault electric discharge sound, gets rid of the restriction for relying on tester personal experience for a long time, it is fixed to greatly promote power cable fault
The working efficiency of point.
Due to power cable fault discharging sound intelligent identification Method can automatic identification power cable fault discharge sound, because
This greatly reduces the influence of a large amount of ambient noises and interference existing for scene.
It overcomes carry out sound magnetic-synchro method in the prior art to need to carry out the drawbacks of training for a long time, greatly reduces enterprise
Training cost and cost of labor.
Detailed description of the invention
Fig. 1 is power cable fault discharging sound intelligent identification Method flow chart.
Fig. 2 is 1 cable fault discharging sound original waveform figure of power cable fault discharging sound intelligent identification Method example.
Fig. 3 is that 1 cable fault discharging sound of power cable fault discharging sound intelligent identification Method example removes DC waveform figure.
Fig. 4 is 1 cable fault discharging sound normalization waveform figure of power cable fault discharging sound intelligent identification Method example.
Fig. 5 is that the 1 cable fault discharging sound short-time energy of power cable fault discharging sound intelligent identification Method example is distributed wave
Shape figure.
Fig. 6 is 1 cable fault discharging sound short-time zero-crossing rate waveform of power cable fault discharging sound intelligent identification Method example
Figure.
Fig. 7 is 1 cable fault discharging sound secondary acquisition waveform diagram of power cable fault discharging sound intelligent identification Method example.
Fig. 8 is 2 non-cable fault discharge sound original waveform figure of power cable fault discharging sound intelligent identification Method example.
Fig. 9 is that 2 non-cable fault discharge sound of power cable fault discharging sound intelligent identification Method example removes DC waveform figure.
Figure 10 is 2 non-cable fault discharge sound normalization waveform of power cable fault discharging sound intelligent identification Method example
Figure.
Figure 11 is 2 non-cable fault discharge sound short-time energy of power cable fault discharging sound intelligent identification Method example distribution
Waveform diagram.
Figure 12 is 2 non-cable fault discharge sound short-time zero-crossing rate wave of power cable fault discharging sound intelligent identification Method example
Shape figure.
Figure 13 is 2 non-cable fault discharge sound secondary acquisition waveform of power cable fault discharging sound intelligent identification Method example
Figure.
Specific embodiment
Fig. 1~13 are highly preferred embodiment of the present invention, and 1~13 the present invention will be further described with reference to the accompanying drawing.
As shown in Figure 1, a kind of power cable fault discharging sound intelligent identification Method, includes the following steps:
Step 1, start sampled voice signal in cable fault discharging time and carry out analog-to-digital conversion；
During power cable fault fixed test, Method of Cable Trouble Point breakdown moment cable can completely generate pulsed magnetic field
Signal is acquired, one section of sound of a length of T when recording since the fault discharge moment using signal triggering sound signal data
Signal data, and analog-to-digital conversion is carried out to the sound signal data.
Step 2, the pretreatment such as direct current, normalization and feature extraction is carried out to digitized voice signal.
Sound signal data acquisition in most cases uses unipolarity A/D converter, obtains data and first has to direct current
Amount processing, for the initial data x of N A/D converters output0(i), i ∈ [1,2 ..., n] goes DC quantity treated data
x1(i) are as follows:
x1(i)=x0(i)-2N-1,i∈[1,2,…,n]
Then to going DC quantity treated data x1(i) it is normalized, obtains normalized result x2(i):
Then to normalization data x2(i) it is handled, the sound signal data short-time energy that material calculation is D is distributed x3
(i):
To normalization data x3(i) it is handled, material calculation is the sound signal data short-time zero-crossing rate x of D4(i):
Wherein, sign is sign function, and th is greater than zero threshold value.
Feature extraction finally is carried out to normalized data, is specifically included:
(1) according to data x3(i) it extracts voice signal short-time energy and is distributed pulse width feature z1:
Meeting conditionUnder the premise of, definition: wlast=max [1,2 ...,
N-D+1], wfirst=min [1,2 ..., n-D+1], then z1Are as follows: z1=wlast-wfirst
(2) according to data x3(i) it extracts voice signal short-time energy and is distributed pulse depth-width ratio feature z2:
(3) according to data x3(i) it extracts voice signal short-time energy and is distributed pulse position feature z3:
(4) according to data x4(i) voice signal short-time zero-crossing rate feature z is extracted4:
Meeting conditionUnder the premise of, define rlast=max [1,2 ..., n-
D], rfirst=min [1,2 ..., n-D], then z4Are as follows:
Above-mentioned feature z1、z2、z3、z4Constitute voice signal property vector z={ z1、z2、z3、z4}。
Step 3, the feature vector feeding support vector machines of the voice signal extracted is identified and obtains judgement knot
Fruit；
The sound characteristic information extracted in step 2 is sent into support vector machines, by support vector machines to sound spy
Reference breath is judged, judges whether present sound signals belong to cable fault discharging sound.
It before carrying out the identification of sound characteristic information using support vector machines, needs to be trained support vector machines, have
Body training process includes the following steps:
Step 3-1 respectively prepares M cable fault discharging sound signal data according to the requirement of aforementioned sound signal data acquisition
With non-cable fault discharge acoustical signal data.
Step 3-2, according to aforementioned data pretreatment and requirements for extracting features, respectively to M cable fault electric discharge acoustical signal
Data and non-cable fault discharge acoustical signal data are handled, and eigenmatrix z is obtainedij, i ∈ [1,2 ..., 2M], j ∈ [1,2,
3,4].Wherein, i=[1,2 ..., M] is cable fault discharging sound signal characteristic vector, and i=[M+1, M+2 ..., 2M] is non-electrical
Cable fault discharge acoustical signal feature vector.
Step 3-3, the column vector s of assignment 2M row, wherein the 1st to M row is assigned a value of 1 (expression is cable fault electric discharge
Sound), M+1 to 2M row is assigned a value of 0 (expression is not cable fault discharging sound).
Step 3-4, by matrix zijWith column vector s input support vector machines training function, linear kernel is selected to be trained.
Step 4, whether present sound signals are cable fault discharging sound, and if it is cable fault discharging sound, sequence is executed
Step 5~step 7 executes step 7 if it is non-cable fault discharge sound.
Step 5, the data of present sound signals are saved.
Step 6, the acquisition for carrying out second of voice signal is gone forward side by side the calculating of Correlation series；
It repeats step 1 and step 2, carries out the acquisition of second of voice signal, and by second of voice signal number
It is executed after word and removes direct current, normalized, finally extracted and obtain the data of second of voice signal, and for successively continuous two
The sound signal data of secondary acquisition calculates sound signal data x twicek(i), xk+1(i), the related coefficient of i ∈ [1,2 ..., n]
C:
After the current related coefficient C of sound signal data twice is calculated, by related coefficient C and presetting phase relation
Number threshold value C0It is compared, whether is that cable fault discharging sound is reaffirmed to the voice signal collected for the first time.
Step 7, judging result and sound signal waveform are shown.
Separately below by a cable fault discharging sound example and the example of a non-cable fault discharge sound to above-mentioned
Step is described in detail.Example 1: when acoustic information is cable fault discharging sound:
Step 1, acquisition duration T is one end voice signal of 100ms, and uses unipolarity resolution, N for the A/D of 12bit
Converter obtains the initial data x for the cable fault discharging sound that one group of sampling number n is 8000(i), waveform is as shown in Figure 2.
Step 2, to the initial data x of cable fault discharge sound0(i) DC processing is carried out, after obtaining DC processing
Data x1(i), as shown in Figure 3.Then to data x1(i) it is normalized, obtains data x2(i), as shown in Figure 4.
Utilize normalization data x2(i), the sound signal data short-time energy that material calculation D is 50 is distributed x3(i) (such as Fig. 5
It is shown), calculate the short-time zero-crossing rate x that threshold value th is 0.54(i), (as shown in Figure 6).
Utilize short-time energy distributed data x3(i), it calculates short-time energy and is distributed pulse width feature z1=95, depth-width ratio is special
Levy z2=0.0024, position feature z3=121.Utilize short-time zero-crossing rate data x4(i) short-time zero-crossing rate feature z is calculated4=500.
z1、z2、z3、z4Constitutive characteristic vector z={ 95,0.0024,121,500 }.
Feature vector z={ 95,0.0024,121,500 } are sent into support vector machines, by support vector machines by step 3~5
The voice signal is obtained into after judging as the conclusion of cable fault discharging sound, and saves the data of present sound signals.
Step 6, the voice signal x of a length of 100ms when acquiring againk+1(i), waveform is as shown in fig. 7, the sound is believed
Number the data x of voice signal that collects of data and last timek(i) (i.e. x0(i), waveform is shown in Fig. 2) carry out related coefficient meter
It calculates, the related coefficient C=0.954 of two voice signals is calculated, it is final to confirm current sound information for cable fault electric discharge
Sound.
Step 7, the waveform of judging result and acoustic information that current sound information is cable fault discharging sound is carried out
Display.
Example 2: when acoustic information is non-cable fault discharge sound:
Step 1, acquisition duration T is one end voice signal of 100ms, and uses unipolarity resolution, N for the A/D of 12bit
Converter obtains the initial data x' for the cable fault discharging sound that one group of sampling number n is 8000(i), waveform such as Fig. 8 institute
Show.
Step 2, to the initial data x' of cable fault discharge sound0(i) DC processing is carried out, after obtaining DC processing
Data x'1(i), as shown in Figure 9.Then to data x'1(i) it is normalized, obtains data x'2(i), as shown in Figure 10.
Utilize normalization data x'2(i), the sound signal data short-time energy that material calculation D is 50 is distributed x'3(i) (such as
Shown in Figure 11), calculate the short-time zero-crossing rate x' that threshold value th is 0.54(i), (as shown in figure 12).
Utilize short-time energy distributed data x'3(i), it calculates short-time energy and is distributed pulse width feature z1'=615, depth-width ratio
Feature z'2=1.41 × 10-4, position feature z'3=426.Utilize short-time zero-crossing rate data x'4(i) short-time zero-crossing rate feature is calculated
z'4=1826.z1’、z’2、z’3、z’4Constitutive characteristic vector z'={ 615,1.41 × 10-4,426,1826}。
Step 3~4, by feature vector z'={ 615,1.41 × 10-4, 426,1826 } and it is sent into support vector machines, by supporting
Vector machine obtains the voice signal into after judging as the conclusion of non-cable fault discharge sound.
Step 7, by current sound information be non-cable fault discharge sound judging result and acoustic information waveform into
Row display.
The conclusion that voice signal is non-cable fault discharge sound can also be verified by step 6:
The voice signal x ' of a length of 100ms when acquiring againk+1(i), waveform is as shown in figure 13, by the voice signal
The data x ' for the voice signal that data and last time collectk(i) (i.e. x '0(i), waveform is shown in Fig. 8) carry out related coefficient meter
It calculates, related coefficient C '=0.025 of two voice signals is calculated, it is final to confirm that current sound information is that non-cable failure is put
Electroacoustic.
The above described is only a preferred embodiment of the present invention, being not that the invention has other forms of limitations, appoint
What those skilled in the art changed or be modified as possibly also with the technology contents of the disclosure above equivalent variations etc.
Imitate embodiment.But without departing from the technical solutions of the present invention, according to the technical essence of the invention to above embodiments institute
Any simple modification, equivalent variations and the remodeling made, still fall within the protection scope of technical solution of the present invention.
Claims (6)
1. a kind of power cable fault discharging sound intelligent identification Method, characterized by the following steps:
Step 1, start sampled voice signal in cable fault discharging time and carry out analog-to-digital conversion to obtain initial data x0(i)；
Step 2, DC processing is carried out to digitized voice signal, obtains data x1(i), then to the number of DC processing
According to being normalized, data x is obtained2(i), it finally extracts and obtains the feature vector of voice signal；
Step 3, the feature vector feeding support vector machines of the voice signal extracted is identified and obtains judging result；
Step 4, whether voice signal is cable fault discharging sound before judging, if it is cable fault discharging sound, sequence executes step
Rapid 5~step 7 executes step 7 if it is non-cable fault discharge sound；
Step 5, the data of present sound signals are saved；
Step 6, the acquisition for carrying out second of voice signal is gone forward side by side the calculating of Correlation series；
Step 7, judging result and sound signal waveform are shown.
2. power cable fault discharging sound intelligent identification Method according to claim 1, it is characterised in that: institute in step 2
The feature vector stated includes:
According to data x3(i) it extracts voice signal short-time energy and is distributed pulse width feature z1:
Meeting conditionUnder the premise of i ∈ [1,2 ..., n-D+1], definition: wlast=max [1,2 ..., n-D+
1], wfirst=min [1,2 ..., n-D+1], then z1Are as follows: z1=wlast-wfirst
According to data x3(i) it extracts voice signal short-time energy and is distributed pulse depth-width ratio feature z2:
According to data x3(i) it extracts voice signal short-time energy and is distributed pulse position feature z3:
According to data x4(i) voice signal short-time zero-crossing rate feature z is extracted4:
Meeting conditionUnder the premise of i ∈ [1,2 ..., n-D], r is definedlast=max [1,2 ..., n-D],
rfirst=min [1,2 ..., n-D], then z4Are as follows:
Wherein: data x3(i) for according to the data x2(i) the sound signal data short-time energy point that the step-length being calculated is D
Cloth；Data x4(i) for according to data x3(i) step-length being calculated is the sound signal data short-time zero-crossing rate of D.
3. power cable fault discharging sound intelligent identification Method according to claim 2, it is characterised in that: the step-length
The data x being distributed for the sound signal data short-time energy of D3(i) are as follows:
Wherein: D indicates step-length.
4. power cable fault discharging sound intelligent identification Method according to claim 2, it is characterised in that: the step-length
For the data x of the sound signal data short-time zero-crossing rate of D4(i) are as follows:
Wherein: sign is sign function, and th indicates to be greater than zero threshold value, i ∈ [1,2 ..., n-D].
5. power cable fault discharging sound intelligent identification Method according to claim 1, it is characterised in that: institute in step 6
The calculation formula for the related coefficient stated are as follows:
Wherein: xk(i), xk+1(i), i ∈ [1,2 ..., n] respectively indicates the sound signal data acquired twice in succession.
6. power cable fault discharging sound intelligent identification Method according to claim 1, it is characterised in that: institute in step 3
The training process for the support vector machines stated includes the following steps:
Step 3-1 respectively prepares M cable fault discharging sound signal data and non-cable fault discharge acoustical signal data；
Step 3-2, respectively to M cable fault discharging sound signal data and non-cable fault discharge acoustical signal data at
Reason, obtains eigenmatrix zij, i ∈ [1,2 ..., 2M], j ∈ [1,2,3,4],
Wherein, i=[1,2 ..., M] is cable fault discharging sound signal characteristic vector, and i=[M+1, M+2 ..., 2M] is non-cable
Fault discharge acoustical signal feature vector；
Step 3-3, the column vector s of assignment 2M row indicate cable fault discharging sound, M+1 wherein the 1st is assigned a value of 1 to M row
0 expression non-cable fault discharge sound is assigned a value of to 2M row；
Step 3-4, by matrix zijWith column vector s input support vector machines training function, linear kernel is selected to be trained.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810737476.1A CN109116196B (en) | 2018-07-06 | 2018-07-06 | Intelligent power cable fault discharge sound identification method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810737476.1A CN109116196B (en) | 2018-07-06 | 2018-07-06 | Intelligent power cable fault discharge sound identification method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109116196A true CN109116196A (en) | 2019-01-01 |
| CN109116196B CN109116196B (en) | 2020-09-25 |
Family
ID=64823074
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810737476.1A Active CN109116196B (en) | 2018-07-06 | 2018-07-06 | Intelligent power cable fault discharge sound identification method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109116196B (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112013425A (en) * | 2019-05-30 | 2020-12-01 | 宁波方太厨具有限公司 | Electric quantity detection method and detection system for gas cooker |
| CN113327629A (en) * | 2021-05-06 | 2021-08-31 | 上海交通大学 | Power equipment sound diagnosis method and system |
| CN113466616A (en) * | 2021-06-22 | 2021-10-01 | 海南电网有限责任公司乐东供电局 | Method and device for quickly positioning cable fault point |
| CN113759208A (en) * | 2021-06-02 | 2021-12-07 | 青岛鼎信通讯股份有限公司 | Abnormal waveform identification method based on fault indicator |
| CN115762529A (en) * | 2022-10-17 | 2023-03-07 | 国网青海省电力公司海北供电公司 | Method for preventing cable from being broken outside by using voice recognition perception algorithm |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1821795A (en) * | 2006-03-29 | 2006-08-23 | 保定天威集团有限公司 | Method and its device for detecting discharge fault position of high voltage electric equipment |
| CN102426835A (en) * | 2011-08-30 | 2012-04-25 | 华南理工大学 | Method for identifying local discharge signals of switchboard based on support vector machine model |
| CN102628917A (en) * | 2012-04-25 | 2012-08-08 | 广州供电局有限公司 | Partial discharge recognition method and system |
| KR101410733B1 (en) * | 2013-01-10 | 2014-06-24 | 한국전기연구원 | Partial discharge measurement device using reference MIC and method which can guide fault position |
| US20140216157A1 (en) * | 2013-02-05 | 2014-08-07 | Tzu-Feng Wang-Lee | Method for Determining Arc Discharge as well as Failure of Mechanical Vibration and Abnormal Displacement in Power Equipment Using Acoustic Energy Feature |
| CN104409073A (en) * | 2014-11-04 | 2015-03-11 | 贵阳供电局 | Substation equipment sound and voice identification method |
| CN106782505A (en) * | 2017-02-21 | 2017-05-31 | 南京工程学院 | A kind of method based on electric discharge voice recognition high-tension switch cabinet state |
-
2018
- 2018-07-06 CN CN201810737476.1A patent/CN109116196B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1821795A (en) * | 2006-03-29 | 2006-08-23 | 保定天威集团有限公司 | Method and its device for detecting discharge fault position of high voltage electric equipment |
| CN102426835A (en) * | 2011-08-30 | 2012-04-25 | 华南理工大学 | Method for identifying local discharge signals of switchboard based on support vector machine model |
| CN102628917A (en) * | 2012-04-25 | 2012-08-08 | 广州供电局有限公司 | Partial discharge recognition method and system |
| KR101410733B1 (en) * | 2013-01-10 | 2014-06-24 | 한국전기연구원 | Partial discharge measurement device using reference MIC and method which can guide fault position |
| US20140216157A1 (en) * | 2013-02-05 | 2014-08-07 | Tzu-Feng Wang-Lee | Method for Determining Arc Discharge as well as Failure of Mechanical Vibration and Abnormal Displacement in Power Equipment Using Acoustic Energy Feature |
| CN104409073A (en) * | 2014-11-04 | 2015-03-11 | 贵阳供电局 | Substation equipment sound and voice identification method |
| CN106782505A (en) * | 2017-02-21 | 2017-05-31 | 南京工程学院 | A kind of method based on electric discharge voice recognition high-tension switch cabinet state |
Non-Patent Citations (2)
| Title |
|---|
| 周玲: "基于超声波信号的局部放电故障识别算法研究", 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》 * |
| 陈志坚 等: ""利用盲源分离技术提取地下电缆故障放电声的研究"", 《城市建设理论研究（电子版）》 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112013425A (en) * | 2019-05-30 | 2020-12-01 | 宁波方太厨具有限公司 | Electric quantity detection method and detection system for gas cooker |
| CN112013425B (en) * | 2019-05-30 | 2021-12-03 | 宁波方太厨具有限公司 | Electric quantity detection method and detection system for gas cooker |
| CN113327629A (en) * | 2021-05-06 | 2021-08-31 | 上海交通大学 | Power equipment sound diagnosis method and system |
| CN113759208A (en) * | 2021-06-02 | 2021-12-07 | 青岛鼎信通讯股份有限公司 | Abnormal waveform identification method based on fault indicator |
| CN113466616A (en) * | 2021-06-22 | 2021-10-01 | 海南电网有限责任公司乐东供电局 | Method and device for quickly positioning cable fault point |
| CN115762529A (en) * | 2022-10-17 | 2023-03-07 | 国网青海省电力公司海北供电公司 | Method for preventing cable from being broken outside by using voice recognition perception algorithm |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109116196B (en) | 2020-09-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109116196A (en) | A kind of power cable fault discharging sound intelligent identification Method | |
| CN104007374B (en) | Inference signal prevention method in converter transformer partial discharge detection experiment | |
| CN105548832B (en) | High voltage power cable fault recognition method | |
| CN106771922B (en) | A kind of high-tension electricity system of detecting partial discharge in equipment and Recognition of Partial Discharge | |
| Peng et al. | Rough set theory applied to pattern recognition of Partial Discharge in noise affected cable data | |
| CN107218518B (en) | A kind of detection method of the detection device for drain line blockage failure | |
| CN101802628A (en) | A method for detecting, identifying and locating partial discharges occurring in a discharge site along an electric apparatus | |
| CN102759686B (en) | Method for locating power cable faults | |
| CN101196547A (en) | Live testing apparatus for local discharge based on ultrasound assistance | |
| CN106228107B (en) | A kind of supersonic guide-wave broken rail monitoring method based on independent component analysis | |
| CN105606968B (en) | A kind of conductor corona electric discharge audible noise processing method | |
| CN103389446B (en) | Electrical positioning method for partial discharge of power transformer | |
| CN107037338B (en) | A kind of GIS oscillatory surge pressure test default kind identification method | |
| CN111007365A (en) | Ultrasonic partial discharge identification method and system based on neural network | |
| CN107367676A (en) | Partial discharge indicator based on audio intelligent identification | |
| CN103558532A (en) | Partial discharge on-line detection system of high-voltage crosslinked polyethylene power cable | |
| CN114325256A (en) | Power equipment partial discharge identification method, system, equipment and storage medium | |
| Le et al. | Joint causal and anti-causal segmentation and location of transitions in power disturbances | |
| CN110161376A (en) | A kind of traveling wave fault moment extraction algorithm | |
| CN108802586B (en) | A kind of recognition methods for transmission line of electricity corona discharge and paradoxical discharge | |
| CN106771928A (en) | A kind of online pick-up method of partial discharge pulse's initial time | |
| CN107192892B (en) | Thunder positioning system automatic trigger method based on lightning electromagnetic signal identification technology | |
| CN112560673A (en) | Thunder detection method and system based on image recognition | |
| CN106019042A (en) | Method of identifying high voltage cable inner insulation defect type based on specific chromatograms | |
| CN109884483A (en) | Insulating tube type busbar shelf depreciation acoustics on-line monitoring method and device |
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 | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20210727 Address after: No.16, Sanying Road, Zhangdian District, Zibo City, Shandong Province Patentee after: SHANDONG KEHUI POWER AUTOMATION Co.,Ltd. Patentee after: QINGDAO KEHUI ELECTRIC Co.,Ltd. Address before: No.16, Sanying Road, Zhangdian District, Zibo City, Shandong Province Patentee before: SHANDONG KEHUI POWER AUTOMATION Co.,Ltd. |