CN103018640A - Method for testing electricity discharge intensity of corona on surface of high-voltage insulator - Google Patents
Method for testing electricity discharge intensity of corona on surface of high-voltage insulator Download PDFInfo
- Publication number
- CN103018640A CN103018640A CN2012104900861A CN201210490086A CN103018640A CN 103018640 A CN103018640 A CN 103018640A CN 2012104900861 A CN2012104900861 A CN 2012104900861A CN 201210490086 A CN201210490086 A CN 201210490086A CN 103018640 A CN103018640 A CN 103018640A
- Authority
- CN
- China
- Prior art keywords
- discharge
- image
- gain
- electricity discharge
- insulator
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Landscapes
- Testing Relating To Insulation (AREA)
Abstract
The invention relates to a method for testing the electricity discharge intensity of the corona on the surface of a high-voltage insulator, comprising the following steps: collecting the electricity discharge video signal of the corona on the composite insulator at different electricity discharge intensity by adopting a solar-blind ultraviolet imager under the condition of different instrument gains and observation distances, segmenting an electricity discharge spot area by adopting the video analyzing and digital-image processing algorithm; obtaining the related data of the electricity discharge spot area, apparent electricity discharge amount, observation distance and instrument gains; establishing an electricity discharge intensity prediction model by adopting the regression algorithm through a least-square support vector machine accordingly; and finally testing the electricity discharge intensity of the corona on the surface of the high-voltage insulator by adopting the electricity discharge intensity prediction model. The method can be used for detecting the apparent electricity discharge amount of the corona on the high-voltage insulator without contacting the high-voltage insulator, has the advantage of operation safety and high detection efficiency and provides a convenient condition for knowing the operation state of equipment timely, evaluating the electricity discharge hazard and the development stage accurately and realizing the flashover early warning.
Description
Technical field
The present invention relates to a kind of can in the situation that not the connect insulator body detect the method for its corona discharge intensity, belong to technical field of measurement and test.
Background technology
The surface corona of high-tension insulator can be accelerated the aging speed of full skirt and sheath, the intensity of analyzing the insulator surface corona discharge helps to understand the running status of equipment, harmfulness and the developing stage of accurate evaluation discharge also are simultaneously the prerequisites that realizes the flashover early warning.
At present, domestic and international detection method to insulator arc-over intensity mainly contains following several:
One, pulse current method.Described method is used at present more extensive, but belongs to the contact measurement method, exists certain not enough in engineering is used: 1. need the on-the-spot current sensor of installing, and need do certain change to the structure of insulator, labour intensity is large and have certain danger; 2. the multiple high-pressure side of being born in of discharging, current sensor then is installed on low-pressure side, insulator chain length can reach several meters even more than ten meter, and on high-tension side discharge signal propagates into and has very large decay in the process of low-pressure side, causes sensor to be difficult to detect weak discharge; 3. transformer station and electric transmission line isolator quantity are very huge, if every a string insulation all installation detecting system will cause the problem that cost is too high, the installation and maintenance workload is excessive; 4. pulse current method is difficult to navigate to concrete discharge position.
Two, UV pulse method.What described method and imaging method detected all is the ultraviolet signal that discharge gives off, but it is output as corresponding pulse signal, can not carry out imaging to discharge, and described method is difficult to the orientation discharge position at the scene, when detector does not have over against discharge position, can there be very large detection error.
Three, ultraviolet imagery method.The ultraviolet imagery method had obtained applying progressively in recent years in electric system, but the Applicative time of ultraviolet imagery method in electric system is also shorter, and its present Research is as follows:
1) quantizes at present the ultraviolet imagery testing result and adopted so-called " photon number " parameter, there is following deficiency in this parameter: have complicated nonlinear relationship between the at first gain of photon number and ultraviolet imager, the distance, be difficult to obtain preferably law curve, be not easy to estimate strength of discharge; Secondly photon number is slow to the response speed of discharge, is not easy to reflect the dynamic process of discharge.
2) relevant research is in the relation between its ultraviolet parameter of research and the discharge capacity under fixed range and the fixed gain, when in engineering reality, detecting the discharge of insulator surface, its observed range and instrument gain are unfixing, and distance and gain can cause very large impact to testing result, cause carrying out quantitative analysis to strength of discharge.
Summary of the invention
The object of the invention is to the drawback for prior art, a kind of high pressure insulator surface corona discharge strength test method is provided, in the situation that connect insulator body not detects its corona discharge intensity accurately and efficiently.
Problem of the present invention realizes with following technical proposals:
A kind of high pressure insulator surface corona discharge strength test method, corona discharge vision signal when described method utilizes day blind ultraviolet imager to gather composite insulator in different strength of discharge under different instrument gain and observed range, then adopt video analysis and Digital Image Processing algorithm to be partitioned into the discharging light spot region, obtain the discharge facula area, Apparent discharge magnitude, this related data of observed range and instrument gain, adopt on this basis the least square method supporting vector machine regression algorithm to set up the discharge capacity Model To Describe Strength of Blended, utilize at last this model that high pressure insulator surface corona discharge intensity is tested, concrete operations may further comprise the steps:
A. apply the 50kV power frequency high voltage to insulator, utilize the discharge capacity of instrument for measuring partial discharge's test insulator, keep this voltage constant, select observed range to be respectively 4.0m, 8.0m, 16.0m, 24.5m, 29.6m, 33m, 41.5m and 50m, under above-mentioned each observed range, under being respectively 50%, 60%, 70% and 80% situation, the gain of ultraviolet imager records successively the corona discharge vision signal of composite insulator;
B. will be applied to that power frequency high voltage on the insulator increases 5kV successively until 120kV, the measuring process of repeating step a under each electrical voltage point;
C. the corona discharge vision signal to record at every turn, adopt video analysis and Digital Image Processing algorithm to be partitioned into the discharging light spot region, obtain the discharge facula area, and then this related data of facula area, Apparent discharge magnitude, observed range and instrument gain that obtains discharging;
D. the difference by the ultraviolet imager gain is divided into four groups with above-mentioned data, successively every group of the data least square method supporting vector machine regression algorithm is set up the discharge capacity forecast model, obtains and four four discharge capacity forecast models that gain is corresponding;
E. the gain of selecting ultraviolet imager is 50% or 60% or 70% or 80% to record the corona discharge vision signal of tested composite insulator, adopt video analysis and Digital Image Processing algorithm to be partitioned into the discharging light spot region, obtain the discharge facula area, measure simultaneously observed range, utilize this corresponding discharge capacity forecast model that gains to determine the corona discharge amount of tested composite insulator.
Above-mentioned high pressure insulator surface corona discharge strength test method adopts video analysis and Digital Image Processing algorithm to be partitioned into the discharging light spot region, and the concrete grammar that obtains the discharge facula area adopts following steps:
1. from the composite insulator corona discharge vision signal of blind ultraviolet imager collection of day, catch the ultraviolet frame of video, obtain a width of cloth color digital image, and this coloured image is converted to gray level image;
2. interested region of discharge in the image is intercepted, then adopting the dynamic threshold segmentation algorithm is bianry image with greyscale image transitions, realizes cutting apart of spot area;
3. adopt the erosion filter algorithm of binary mathematical morphology to remove the outer noise spot of region of discharge, adopt again expansion algorithm that image is processed, recover shape and the size of ultraviolet image;
4. gray-scale value is the number of 1 pixel in the statistical picture, and characterizes the discharge facula area with this
Size:
In the formula,
Be the bianry image matrix behind the morphologic filtering,
With
The position of the pixel of presentation video in matrix,
Expression is carried out summation operation to the image array gray-scale value.
Above-mentioned high pressure insulator surface corona discharge strength test method be to improve the sample data amount, under each strength of discharge under the gain of each ultraviolet imager the discharge facula area-the observed range relation curve (namely
Curve) observed range
, utilize fitting function
Carry out linear interpolation processing, the interpolation step-length is 0.1m, in the formula,
With
Be corresponding constant factor.
The present invention is under the prerequisite that does not affect on-the-spot discharge examination precision, utilize the input vector number of the method minimizing insulator corona discharge amount forecast model of artificial regulation ultraviolet imager yield value, effectively simplify the forecast model structure, improved corona discharge amount predetermined speed.Described method can in the situation that not the connect insulator body detect its corona discharge intensity, have handling safety, detection efficiency advantages of higher, in time understand the running status of equipment for the staff, the harmfulness of accurate evaluation discharge and developing stage and realize the flashover early warning condition of providing convenience.
Description of drawings
The invention will be further described below in conjunction with accompanying drawing.
Fig. 1 is the corona discharge spectral distribution graph;
Fig. 2 is the basic functional principle figure of day blind ultraviolet imager;
Fig. 3 is the imaging schematic diagram of day blind ultraviolet imager ultraviolet passage;
Fig. 4 is that image is processed and the calculation of parameter process flow diagram;
Fig. 5 is the elementary diagram of pilot system;
Fig. 6 be facula area and discharge capacity and voltage curve (
=8m,
=70%);
Fig. 7 is facula area and discharge capacity relation curve;
Fig. 8 be facula area and gain relationship curve (
=16m);
Fig. 9 is gain
Facula area is with the change curve of distance in the time of=70%;
Figure 10 is gain
Facula area is with the change curve of distance in the time of=50%;
Figure 11 is gain
Facula area is with the change curve of distance in the time of=60%;
Figure 12 is gain
Facula area is with the change curve of distance in the time of=80%;
Figure 13 is LS-SVM discharge capacity forecast model;
Figure 14 is predicted value and the raw data actual value of data.
Each symbol inventory is in the literary composition:
Be facula area,
Be the bianry image behind the morphologic filtering,
Be observed range,
With
Be corresponding constant factor,
Expression is carried out summation operation to the image array gray-scale value,
Be gain,
Be Apparent discharge magnitude.
Embodiment
1, the principle of work of day blind ultraviolet imager
1.1 the concept of discharge spectrum distribution characteristics and " day blind area "
Follow the transfer of energy and release in the process of discharge and give off light signal, the discharge of insulator surface is mainly the discharge of corona form, and its spectral distribution as shown in Figure 1.
As shown in Figure 1, the wavelength of optical signal that gives off during corona discharge mainly is distributed in 280-400nm ultraviolet light wave band, and the fraction wavelength is between 230-280nm.Earth surface also can be subject to solar radiation, but solar radiation spectrum is through effects such as Atmospheric Absorption and scatterings, UV-C(200-280nm) ultraviolet signal of wave band is almost all absorbed by the ozone in the atmosphere, the range of wavelengths that therefore will be lower than 280nm is called the sun " blind area ", and the ultraviolet signal of therefore surveying non-solar-blind band can be avoided the interference of sunshine.
1.2 the principle of work of day blind ultraviolet imager
Ultraviolet signal to discharge day blind area carries out the fresh approach that imaging is at present detection discharge, be referred to as a day blind ultraviolet imagery method, described method can directly obtain the image in Discharge illuminating zone, and the position that navigates to discharge has been avoided the interference of extraneous sunshine signal simultaneously.Day, blind ultraviolet became instrument to adopt two light spectrum image-forming principles, and basic structure as shown in Figure 2.
Imaging system of the present invention mainly is made of visual light imaging passage, ultraviolet imagery passage and image displaying part branch.There is beam splitter optics porch at instrument, and the light signal of incident is divided into two-way, wherein a road enter visible channel, another road then enters the ultraviolet passage.The structure of visible channel is identical with ordinary digital camera, mainly comprises camera lens and CCD(photoelectrical coupler, realizes light signal is converted to the electric signal components and parts).The structure of ultraviolet passage is relatively complicated, and its structure mainly is made of ultraviolet lens, day blind optical filter and ultraviolet imaging enhancer (comprising photocathode, microchannel plate (MCP), video screen, optical taper, CCD) as shown in Figure 3.Instrument internal is added to the ultraviolet channel image and then shows by display screen on the visible channel image.
The ultimate principle that this instrument detects discharge and orientation discharge position is as follows:
When discharge occurs in the high-tension apparatus surface, light signal (the light that discharges and self give off that the light signal that the high-tension apparatus body sends (light in reflect ambient light source) and region of discharge give off, the ultraviolet signal that comprises part) enters simultaneously ultraviolet imager, the beam splitter of ultraviolet imager inside is divided into two ways of optical signals with incident optical signal, wherein a road directly enter visible channel, incident optical signal images in the CCD surface behind lens reflection, this image also is the image that apparatus body becomes.
Another road light signal then enters into the ultraviolet light passage, and this passage has special " day is blind " optical filter, only allows the ultraviolet signal of 240-280nm pass through, because the non-solar-blind band light signal strength that the high-tension apparatus surface corona gives off is
The order of magnitude, very faint and belong to the invisible radiation image, for being embodied as picture, the ultraviolet passage has adopted the method as conversion and image intensifying to be embodied as picture: utilize first the ultraviolet light photo negative electrode that ultraviolet image is converted to electronic image, then through MCP electronic image is gained and amplify, the electron stream of MCP output bombards at a high speed on the video screen at its rear portion, and electronic image is converted to again visible images, then gathers imaging through CCD.
The ultraviolet passage only carries out imaging to the Discharge illuminating zone, for position that can orientation discharge, has adopted Image Fusion ultraviolet image to be added on the visible images in its instrument internal, thereby has demonstrated discharge position.
2. the extraction of ultraviolet imagery image quantization parameter and calculating
According to its discharge hot spot (high field intensity zone ionization and luminous during discharge in the ultraviolet image, in ultraviolet image, be shown as the zone of white, this project is defined as the discharge hot spot with it) characteristic of the significant change with strength of discharge, proposed a kind of new discharge quantization method at this: by ultraviolet video or image are processed the relevant image parameter of extraction for detection of result's quantification, its image is processed and parameter is carried calculation flow chart as shown in Figure 4.
Ultraviolet imager can for video, be write first video playback and analysis software continuous or random catching ultraviolet frame of video from the ultraviolet video to testing result with video or image format storage, and each frame is a width of cloth color digital image; For image, then directly from the storage card of ultraviolet imager, read.The image that the picture frame that obtains from video or ultraviolet imager are taken is coloured image, for ease of subsequent treatment, this project intercepts interested region of discharge in the image first, coloured image is converted to gray level image and then carries out image segmentation, region of discharge is some white hot spots in ultraviolet image, its gray-scale value is 1 or near 1, and the gray-scale value of background image is that bianry image realized the cutting apart of spot area thereby this project has adopted the dynamic threshold segmentation algorithm with above-mentioned greyscale image transitions generally far below 1.
In above-mentioned bianry image, white image also has the noise spot of part sometimes except region of discharge, the pixel number that comprises according to the noise region image is generally much smaller than the characteristic of region of discharge, the present invention has adopted the erosion filter algorithm of binary mathematical morphology (Mathematical Morphology), but merely image being carried out erosion operation can cause image area to reduce, calculate the band error for follow-up facula area, invention has adopted again expansion algorithm that image is processed for this reason, like this in filtering also kept preferably shape and the size of ultraviolet image in the noise in image.
Discharge stronger, hot spot is larger, thereby can utilize the size of image spot to characterize its discharge power.Bianry image form with matrix in computing machine is preserved, the corresponding gray-scale value of each pixel, the gray-scale value of white point is 1, the gray-scale value of black color dots is 0, therefore be that the number of 1 pixel can characterize its spot area size by adding up gray-scale value in its zone, the present invention is defined as " facula area "
, its calculating formula is as follows:
In the formula,
Expression is carried out summation operation to the image array gray-scale value,
Be the bianry image behind the morphologic filtering.According to above-mentioned definition as can be known, facula area is actually the number of discharging light spot region pixel, thus this project to define its unit at this be pixel (Pixel).
3. the structure of pilot system and test method
3.1 testing equipment and wiring
The pilot system wiring of this project build as shown in Figure 5.Test specimen is the 110kV composite insulator, model is FXBW-110/100, its high-pressure side gold utensil is owing to suffering thunderbolt to have a burn point, easily form the corona point that the position is fixed and strength of discharge is more stable at this point after the pressurization, be convenient to study the relation between facula area and discharge capacity, observed range and the instrument gain.Discharge capacity adopts instrument for measuring partial discharge to measure, model is that LDS-6(Germany produces, the minimum detectable Apparent discharge magnitude＜1pC), detecting impedance directly is serially connected in the low-pressure side ground loop of composite insulator, detection impedance bandwidth is 100kHz-1000kHz, utilize an electric capacity to consist of coupling capacitance for the high-voltage capacitor of 0.015uF, produce corona discharge for preventing high-voltage conducting wires, being socketed diameter outside wire is the soft aluminium corrugated tube of 10cm.The ultraviolet imager model is South Africa CoroCAM504, for the ease of analyzing the dynamic process of discharge, outputting video signal is stored by external video record equipment, and except the gain that changes instrument, other parameters of instrument adopt default setting in the process of test.
3.2 test procedure and method
Test procedure and method are as follows:
1) at first studies relation between discharge capacity and voltage and the facula area.The discharge capacity that utilizes the demarcation signal source to inject 1000pC at the insulator two ends is played a game and is put the demarcation that detection system is carried out discharge capacity, then apply power frequency high voltage to insulator, studied respectively 50kV, 55kV, 60kV, 65kV until the 120kV discharge capacity under totally 15 electrical voltage points and the variation characteristic of ultraviolet image.
2) facula area under the different strength of discharges of research
With distance
And gain
Variation characteristic.Increase voltage to 50kV, keep this voltage constant, select observed range to be respectively 4.0m, 8.0m, 16.0m, 24.5m, 29.6m, 33m, 41.5m and 50m, under above-mentioned each observed range, under being respectively 50%, 60%, 70% and 80% situation, the gain of ultraviolet imager records the ultraviolet vision signal successively.
3) be that 55kV, 60kV, 65kV are until repeat whole 2 under each electrical voltage point of 120kV at voltage successively) measuring process.
4. test findings and analysis
4.1 the relation between discharge capacity and voltage and the facula area
Progressively increase voltage for insulator, when voltage reached approximately 40kV, ultraviolet imager begins to observe gold utensil burn point place, high-pressure side faint discharge, and hot spot is very little, mostly be discrete point-like, the Apparent discharge magnitude mean value that instrument for measuring partial discharge detects is 133pC.After voltage reached 50kV, the facula area of ultraviolet imager obviously increased, and discharge capacity mean value is 230pC.Continuing increases voltage, and facula area and discharge capacity be corresponding increase also, and the facula area average is 8157Pixel after voltage reaches 120kV, and Apparent discharge magnitude has then reached 4013pC.Based on test figure, Fig. 6 has provided the relation curve between facula area and discharge capacity and the voltage.
As shown in Figure 6, progressively increase along with voltage, its facula area and discharge capacity be corresponding increase also, but have certain non-linearly, discharge capacity begins to increase slower, when voltage has reached approximately 75kV, gathering way of discharge capacity increases to some extent, after voltage had reached 105kV, the speed that discharge capacity increases descended again to some extent, but the hot spot face is always apparent in view with the voltage increase.Based on above-mentioned test figure, Fig. 7 has provided the relation curve of discharge capacity and facula area.
As can be seen from Figure 7 discharge capacity corresponding increase with the increase of facula area utilizes the facula area parameter can characterize preferably the variation of its discharge capacity.
4.2 gain is on the impact of facula area
The needs that detect in order to adapt to different strength of discharges need be regulated the gain of ultraviolet imager when carrying out discharge examination, this project has been studied the relation between facula area under the different strength of discharges and gain.
Under same observed range and approximately uniform strength of discharge, along with the gain increase of ultraviolet imager, its facula area also significantly increases.Based on test figure, strength of discharge be respectively between facula area in 418pC (60kV), 980pC (80kV), 2475pC (100kV) and 4013pC (120kV) situation and the gain relation as shown in Figure 8.
Observe match formula that this project of variation tendency of its data point proposes shown in (2):
In the formula,
Be instrument gain,
With
Be corresponding constant factor.Fitting function expression formula and goodness of fit coefficient
As shown in table 1.
Table 1 facula area and gain fitting function expression formula
| Discharge capacity/pC | The fitting function expression formula | The goodness of fit
|
| 418 | 0.9866 | |
| 980 |
|
0.9780 |
| 2475 |
|
0.9938 |
| 4013 |
|
0.9889 |
Coefficient
Value generally between 0.06-0.08 the time its goodness of fit coefficient close to 1(
Span between 0 to 1,
Value illustrates that more near 1 regression curve is better to the fitting degree of observed reading), also be to be similar to according to referring to the function rule between facula area and the gain to change, in the gain margin of 50%-80%, every increase by 10% that generally gains, facula area approximately increases 1.5-2 doubly.
4.3 observed range is on the impact of facula area
Along with observed range increases, the image size of ultraviolet imagery passage progressively reduces, thereby the facula area in the ultraviolet image also reduces corresponding.Obviously reduce along with distance increases facula area, according to test figure, instrument gain is 70%, and facula area and observed range Relations Among were as shown in Figure 9 when strength of discharge was respectively 418pC, 980pC, 2475pC and 4013pC.
The fitting formula that this project of data variation characteristic proposes in the analysis chart is shown in (3):
Fitting function expression formula and goodness of fit coefficient
As shown in table 2.
Table 2 facula area and distance relation fitting function expression formula
| Discharge capacity/pC | The fitting function expression formula | The goodness of fit
|
| 418 |
|
0.9894 |
| 980 |
|
0.9958 |
| 2475 |
|
0.9910 |
| 4013 | 0.9707 |
The analysis showed that it
Its goodness of fit coefficient also is that facula area changes according to power function with distance close to 1 when generally changing between 1.7-2.0.
4.4 the lower facula area of other gain and distance relation
Fig. 9 is that the gain at ultraviolet imager is fixed as resulting related data in 70% situation, and gain also has obvious impact to facula area, for characterizing better facula area under different gains and the different strength of discharges along with the variation characteristic of observed range, provided gain at this and be respectively resulting related data in 50%, 60% and 80% situation, such as Figure 10, Figure 11, shown in Figure 12.
5. the quantification of insulator corona discharge intensity
Facula area is relevant with discharge capacity, observed range and instrument gain, therefore based on the ultraviolet imagery method discharge is quantized to be actually according to facula area
, observed range
And gain
Estimate discharge capacity
A process, this problem is a multidimensional nonlinear mapping problems
:
（4）
And based on the above-mentioned experimental study of this project as can be known,
With
,
With
Relation between three parameters is complicated, is difficult to provide its concrete function expression.The LS-SVM regression algorithm has good non-linear mapping capability, can approach multidimensional nonlinear characteristic between the input and output by corresponding study and training, has utilized LS-SVM to set up above-mentioned discharge based on this this project and has quantized forecast model.
Directly utilize
,
With
Predict the corona discharge amount, the LS-SVM model of foundation has three input vectors, will cause the forecast model complicated structure.The research in earlier stage of this project finds that gain is set to the needs that 50%, 60%, 70% and 80% 4 fixing value can satisfy on-the-spot discharge examination substantially, therefore this project is set up respectively separately independently forecast model to above-mentioned four kinds of gains, thereby the simplified model structure, the forecast model structure of foundation as shown in figure 13.
In Figure 13, it is the separately forecast model independently of setting up for 50%, 60%, 70% and 80% time that LS-SVM50, LS-SVM60, LS-SVM70 and LS-SVM80 represent respectively gain, and the input vector of each forecast model only has two like this: observed range
With the facula area under this distance
, output is the Apparent discharge magnitude for predicting then
When high-volume predicting, get final product according to the corresponding forecast model of the gain selection of instrument.
Specify the method for building up of its model as 70% data take gain in this project.
1) the training sample data determines.Each bar curve represents under certain strength of discharge facula area with the variation characteristic of distance in Fig. 9, correspondingly can obtain distance according to each curve
, the facula area under this distance
With correspondence
Data, this be expressed as (
) form, as being the curve of 4013pC for discharge capacity, its data are respectively (4,22137,4013), (8,8157,4013), (16,2022,4013), (24.5,917,4013), (29.6,632,4013), (33,426,4013), (41.5,282,4013).Also can set up above-mentioned similar data set based on the corresponding curve of other discharge capacity equally.
Because the observed range in the test has only been got the several discrete values of 4.0m, 8.0m, 16.0m, 24.5m, 29.6m, 33m and 41.5m, the sample data that 7 distance values above only utilizing obtain is very few, for this this project under each strength of discharge
The distance of relation curve has been carried out linear interpolation processing, and the interpolation step-length is 0.1m, obtains altogether 1140 groups of sample datas after the interpolation.
2) training of regression model and test
Support vector machine need to select suitable kernel function that data-mapping is arrived more high-dimensional feature space, then carries out linear regression at high-dimensional feature space, and this project is selected the gaussian radial basis function kernel function at this.Evenly half in the sample drawn data is training data, and second half is check data.The training sample input model is trained, utilize genetic algorithm to regular parameter
With
Parameter is carried out optimizing, and the average error value of training sample is 1.3%, and the average error of verification sample is 1.5%, illustrates that this model has good precision of prediction and generalization ability.Raw data and checking data are input to this prognoses system, and the result of its raw data and predicted data as shown in figure 14.
As shown in Figure 14, to sample and non-sample input data, its predicted data overlaps substantially with real data, also is that the error of predicted value and actual value is less, and system has good precision of prediction.Utilize identical method can set up LS-SVM discharge capacity forecast model under other three kinds of gains.
Claims (3)
1. high pressure insulator surface corona discharge strength test method, it is characterized in that, corona discharge vision signal when described method utilizes day blind ultraviolet imager to gather composite insulator in different strength of discharge under different instrument gain and observed range, then adopt video analysis and Digital Image Processing algorithm to be partitioned into the discharging light spot region, obtain the discharge facula area, Apparent discharge magnitude, this related data of observed range and instrument gain, adopt on this basis the least square method supporting vector machine regression algorithm to set up the discharge capacity Model To Describe Strength of Blended, utilize at last this model that high pressure insulator surface corona discharge intensity is tested; Concrete operations may further comprise the steps:
A. apply the 50kV power frequency high voltage to insulator, utilize the discharge capacity of instrument for measuring partial discharge's test insulator, keep this voltage constant, select observed range to be respectively 4.0m, 8.0m, 16.0m, 24.5m, 29.6m, 33m, 41.5m and 50m, under above-mentioned each observed range, under being respectively 50%, 60%, 70% and 80% situation, the gain of ultraviolet imager records successively the corona discharge vision signal of composite insulator;
B. will be applied to that power frequency high voltage on the insulator increases 5kV successively until 120kV, the measuring process of repeating step a under each electrical voltage point;
C. the corona discharge vision signal to record at every turn, adopt video analysis and Digital Image Processing algorithm to be partitioned into the discharging light spot region, obtain the discharge facula area, and then this related data of facula area, Apparent discharge magnitude, observed range and instrument gain that obtains discharging;
D. the difference by the ultraviolet imager gain is divided into four groups with above-mentioned data, successively every group of the data least square method supporting vector machine regression algorithm is set up the discharge capacity forecast model, obtains and four four discharge capacity forecast models that gain is corresponding;
E. the gain of selecting ultraviolet imager is 50% or 60% or 70% or 80% to record the corona discharge vision signal of tested composite insulator, adopt video analysis and Digital Image Processing algorithm to be partitioned into the discharging light spot region, obtain the discharge facula area, measure simultaneously observed range, utilize this corresponding discharge capacity forecast model that gains to determine the corona discharge amount of tested composite insulator.
2. a kind of high pressure insulator surface corona discharge strength test method according to claim 1 is characterized in that, adopts video analysis and Digital Image Processing algorithm to be partitioned into the discharging light spot region, obtains the discharge facula area, specifically according to the following steps operation:
1. from the composite insulator corona discharge vision signal of blind ultraviolet imager collection of day, catch the ultraviolet frame of video, obtain a width of cloth color digital image, and this coloured image is converted to gray level image;
2. interested region of discharge in the image is intercepted, then adopting the dynamic threshold segmentation algorithm is bianry image with greyscale image transitions, realizes cutting apart of spot area;
3. adopt the erosion filter algorithm of binary mathematical morphology to remove the outer noise spot of region of discharge, adopt again expansion algorithm that image is processed, recover shape and the size of ultraviolet image;
4. gray-scale value is the number of 1 pixel in the statistical picture, and characterizes the discharge facula area with this
Size:
In the formula,
Be the bianry image matrix behind the morphologic filtering,
With
The position of the pixel of presentation video in matrix,
Expression is carried out summation operation to the image array gray-scale value.
3. a kind of high pressure insulator surface corona discharge strength test method according to claim 1 and 2 is characterized in that, to the observed range of the discharge facula area under each strength of discharge under the gain of each ultraviolet imager-observed range relation curve
, utilize fitting function
Carry out linear interpolation processing, the interpolation step-length is 0.1m, in the formula,
With
Be corresponding constant factor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2012104900861A CN103018640A (en) | 2012-11-27 | 2012-11-27 | Method for testing electricity discharge intensity of corona on surface of high-voltage insulator |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2012104900861A CN103018640A (en) | 2012-11-27 | 2012-11-27 | Method for testing electricity discharge intensity of corona on surface of high-voltage insulator |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103018640A true CN103018640A (en) | 2013-04-03 |
Family
ID=47967467
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2012104900861A Pending CN103018640A (en) | 2012-11-27 | 2012-11-27 | Method for testing electricity discharge intensity of corona on surface of high-voltage insulator |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103018640A (en) |
Cited By (34)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103323720A (en) * | 2013-07-04 | 2013-09-25 | 华北电力大学（保定） | Method for estimating conductor corona loss under high-altitude wind-blown sand conditions |
| CN103454556A (en) * | 2013-08-09 | 2013-12-18 | 国家电网公司 | Tour inspection device with 3D scanning function and detection method thereof |
| CN103543394A (en) * | 2013-10-27 | 2014-01-29 | 华北电力大学（保定） | Discharge ultraviolet imaging quantization parameter extraction method of high-voltage electric equipment |
| CN103698676A (en) * | 2014-01-10 | 2014-04-02 | 深圳供电局有限公司 | Method and system for evaluating corona discharge of power transmission and transformation equipment |
| CN103744005A (en) * | 2013-12-31 | 2014-04-23 | 国网青海省电力公司电力科学研究院 | Discharging strength conversion method based on ultraviolet imaging method |
| CN103954897A (en) * | 2014-05-20 | 2014-07-30 | 电子科技大学 | Intelligent power grid high-voltage insulation damage monitoring system and method based on ultraviolet imaging |
| CN104035007A (en) * | 2014-06-10 | 2014-09-10 | 华北电力大学（保定） | Short-air-gap breakdown prewarning method of high-voltage electrical equipment |
| CN104280670A (en) * | 2014-09-26 | 2015-01-14 | 北京环境特性研究所 | Corona detection method based on solar-blind ultraviolet imager |
| CN105158059A (en) * | 2015-09-14 | 2015-12-16 | 吉林大学 | Test method for strength of car body embedded part of composite material |
| CN105259527A (en) * | 2015-10-09 | 2016-01-20 | 国网新疆电力公司检修公司 | Test method for calibrating detection results of different ultraviolet imaging instruments |
| CN105334440A (en) * | 2015-11-26 | 2016-02-17 | 云南电网有限责任公司电力科学研究院 | Partial discharge detecting system and method |
| CN105372562A (en) * | 2015-10-21 | 2016-03-02 | 国网新疆电力公司检修公司 | Method and system for detecting ultraviolet discharge degree by means of set indexes |
| CN105372564A (en) * | 2015-11-23 | 2016-03-02 | 北京环境特性研究所 | Ultraviolet imaging corona detection method and device |
| CN106291269A (en) * | 2016-07-19 | 2017-01-04 | 国家电网公司 | The determination method and apparatus of high voltage electric equipment strength of discharge |
| CN106940886A (en) * | 2017-03-08 | 2017-07-11 | 华北电力大学（保定） | A kind of electrical equipment electric discharge ultraviolet imagery quantization parameter extracting method based on gray scale |
| CN106969833A (en) * | 2017-03-23 | 2017-07-21 | 华中科技大学 | A kind of ultraviolet detector system of big visual field and its application |
| CN107505546A (en) * | 2017-08-25 | 2017-12-22 | 国家电网公司 | A kind of method that corona discharge is monitored using ultraviolet imager |
| CN108008259A (en) * | 2017-11-14 | 2018-05-08 | 国网江西省电力有限公司电力科学研究院 | Based on infrared, the integrated detection method of Uv and visible light image co-registration and device |
| CN108399486A (en) * | 2018-01-22 | 2018-08-14 | 国网福建省电力有限公司泉州供电公司 | A kind of electrical equipment fault detection method based on least square support vector machines |
| CN109975670A (en) * | 2019-04-04 | 2019-07-05 | 华北电力大学 | A kind of method of optics and electric field signal Conjoint Analysis the air gap discharge process |
| CN110146791A (en) * | 2019-05-07 | 2019-08-20 | 国网山东省电力公司电力科学研究院 | A kind of corona detection method based on image procossing |
| CN110161374A (en) * | 2019-06-18 | 2019-08-23 | 苏州微纳激光光子技术有限公司 | A kind of day blind ultraviolet imagery background Xanthophyll cycle optical system and method |
| CN111352006A (en) * | 2020-03-27 | 2020-06-30 | 国网甘肃省电力公司电力科学研究院 | External insulation equipment real-time discharge intensity quantification and evaluation system based on ultraviolet imaging |
| CN111443264A (en) * | 2020-04-30 | 2020-07-24 | 国网江苏省电力有限公司电力科学研究院 | Corona ultraviolet imaging equipment-based corona discharge intensity identification method and system and storage medium |
| CN112285504A (en) * | 2020-10-27 | 2021-01-29 | 华北电力大学 | Multispectral solar-blind narrow-band ultraviolet imager and method for detecting different discharge states by using same |
| CN113189391A (en) * | 2021-04-23 | 2021-07-30 | 中国人民解放军陆军工程大学 | Optical remote measuring method for ground flashback current |
| CN113238131A (en) * | 2021-07-13 | 2021-08-10 | 北方夜视科技(南京)研究院有限公司 | Insulator corona discharge state evaluation method based on deep learning |
| CN113781438A (en) * | 2021-09-12 | 2021-12-10 | 华北电力大学 | Method for judging corona initial voltage by using spot area based on discharge ultraviolet image |
| CN114002566A (en) * | 2021-11-09 | 2022-02-01 | 国网山东省电力公司电力科学研究院 | Partial discharge detection device and method based on multi-light fusion |
| CN114034997A (en) * | 2021-11-10 | 2022-02-11 | 国网江苏省电力有限公司检修分公司 | Insulator degradation degree prediction method and system based on multiple parameters |
| CN115389882A (en) * | 2022-08-26 | 2022-11-25 | 中国南方电网有限责任公司超高压输电公司广州局 | Corona discharge state evaluation method and device, computer equipment and storage medium |
| CN117611784A (en) * | 2023-11-30 | 2024-02-27 | 华北电力大学 | Cooperative control method and device for multispectral high-timeliness discharge detection |
| CN117806161A (en) * | 2023-11-22 | 2024-04-02 | 国网湖北省电力有限公司超高压公司 | Ultraviolet device self-adaptive gain control method, storage medium and electronic device |
| CN118425710A (en) * | 2024-07-04 | 2024-08-02 | 中国南方电网有限责任公司超高压输电公司电力科研院 | Ultraviolet imaging device for insulator fault detection and insulator fault detection system |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101551435A (en) * | 2009-05-15 | 2009-10-07 | 湖北省电力试验研究院 | Method for detecting corona discharging and device thereof |
| CN101726693A (en) * | 2009-11-26 | 2010-06-09 | 绍兴电力局 | Method for seeking discharge regions of power devices on ultraviolet images |
| CN102175953A (en) * | 2011-02-28 | 2011-09-07 | 长沙理工大学 | Discharge online monitoring and fault positioning method for power transmission line insulator |
| CN102288884A (en) * | 2011-09-13 | 2011-12-21 | 华北电力大学（保定） | External insulation discharging detecting method based on ultraviolet light spots |
-
2012
- 2012-11-27 CN CN2012104900861A patent/CN103018640A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101551435A (en) * | 2009-05-15 | 2009-10-07 | 湖北省电力试验研究院 | Method for detecting corona discharging and device thereof |
| CN101726693A (en) * | 2009-11-26 | 2010-06-09 | 绍兴电力局 | Method for seeking discharge regions of power devices on ultraviolet images |
| CN102175953A (en) * | 2011-02-28 | 2011-09-07 | 长沙理工大学 | Discharge online monitoring and fault positioning method for power transmission line insulator |
| CN102288884A (en) * | 2011-09-13 | 2011-12-21 | 华北电力大学（保定） | External insulation discharging detecting method based on ultraviolet light spots |
Non-Patent Citations (2)
| Title |
|---|
| 律方成 等: "基于紫外成像图像信息的绝缘子表面放电量化方法", 《电工技术学报》 * |
| 王胜辉: "基于紫外成像的污秽悬式绝缘子放电检测及评估", 《中国博士学位论文全文数据库》 * |
Cited By (45)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103323720B (en) * | 2013-07-04 | 2016-03-09 | 华北电力大学（保定） | Conductor corona loss estimating method under High aititude dust storm condition |
| CN103323720A (en) * | 2013-07-04 | 2013-09-25 | 华北电力大学（保定） | Method for estimating conductor corona loss under high-altitude wind-blown sand conditions |
| CN103454556A (en) * | 2013-08-09 | 2013-12-18 | 国家电网公司 | Tour inspection device with 3D scanning function and detection method thereof |
| CN103454556B (en) * | 2013-08-09 | 2016-01-20 | 国家电网公司 | A kind of inspection device and detection method thereof with 3D scan function |
| CN103543394A (en) * | 2013-10-27 | 2014-01-29 | 华北电力大学（保定） | Discharge ultraviolet imaging quantization parameter extraction method of high-voltage electric equipment |
| CN103543394B (en) * | 2013-10-27 | 2016-02-03 | 华北电力大学（保定） | A kind of high voltage electric equipment electric discharge ultraviolet imagery quantization parameter extracting method |
| CN103744005A (en) * | 2013-12-31 | 2014-04-23 | 国网青海省电力公司电力科学研究院 | Discharging strength conversion method based on ultraviolet imaging method |
| CN103698676A (en) * | 2014-01-10 | 2014-04-02 | 深圳供电局有限公司 | Method and system for evaluating corona discharge of power transmission and transformation equipment |
| CN103954897A (en) * | 2014-05-20 | 2014-07-30 | 电子科技大学 | Intelligent power grid high-voltage insulation damage monitoring system and method based on ultraviolet imaging |
| CN104035007A (en) * | 2014-06-10 | 2014-09-10 | 华北电力大学（保定） | Short-air-gap breakdown prewarning method of high-voltage electrical equipment |
| CN104280670A (en) * | 2014-09-26 | 2015-01-14 | 北京环境特性研究所 | Corona detection method based on solar-blind ultraviolet imager |
| CN104280670B (en) * | 2014-09-26 | 2017-03-01 | 北京环境特性研究所 | A kind of corona detection method based on day blind ultraviolet imager |
| CN105158059A (en) * | 2015-09-14 | 2015-12-16 | 吉林大学 | Test method for strength of car body embedded part of composite material |
| CN105158059B (en) * | 2015-09-14 | 2017-09-01 | 吉林大学 | A kind of composite-material body built-in fitting intensity experiment method |
| CN105259527B (en) * | 2015-10-09 | 2018-08-17 | 国网新疆电力公司检修公司 | A kind of test method that different ultraviolet imager testing results are calibrated |
| CN105259527A (en) * | 2015-10-09 | 2016-01-20 | 国网新疆电力公司检修公司 | Test method for calibrating detection results of different ultraviolet imaging instruments |
| CN105372562A (en) * | 2015-10-21 | 2016-03-02 | 国网新疆电力公司检修公司 | Method and system for detecting ultraviolet discharge degree by means of set indexes |
| CN105372564A (en) * | 2015-11-23 | 2016-03-02 | 北京环境特性研究所 | Ultraviolet imaging corona detection method and device |
| CN105334440A (en) * | 2015-11-26 | 2016-02-17 | 云南电网有限责任公司电力科学研究院 | Partial discharge detecting system and method |
| CN106291269A (en) * | 2016-07-19 | 2017-01-04 | 国家电网公司 | The determination method and apparatus of high voltage electric equipment strength of discharge |
| CN106940886A (en) * | 2017-03-08 | 2017-07-11 | 华北电力大学（保定） | A kind of electrical equipment electric discharge ultraviolet imagery quantization parameter extracting method based on gray scale |
| CN106940886B (en) * | 2017-03-08 | 2019-11-22 | 贵州众创巨电力科技有限公司 | A kind of electrical equipment electric discharge ultraviolet imagery quantization parameter extracting method based on gray scale |
| CN106969833A (en) * | 2017-03-23 | 2017-07-21 | 华中科技大学 | A kind of ultraviolet detector system of big visual field and its application |
| CN107505546A (en) * | 2017-08-25 | 2017-12-22 | 国家电网公司 | A kind of method that corona discharge is monitored using ultraviolet imager |
| CN108008259A (en) * | 2017-11-14 | 2018-05-08 | 国网江西省电力有限公司电力科学研究院 | Based on infrared, the integrated detection method of Uv and visible light image co-registration and device |
| CN108399486A (en) * | 2018-01-22 | 2018-08-14 | 国网福建省电力有限公司泉州供电公司 | A kind of electrical equipment fault detection method based on least square support vector machines |
| CN109975670A (en) * | 2019-04-04 | 2019-07-05 | 华北电力大学 | A kind of method of optics and electric field signal Conjoint Analysis the air gap discharge process |
| CN110146791A (en) * | 2019-05-07 | 2019-08-20 | 国网山东省电力公司电力科学研究院 | A kind of corona detection method based on image procossing |
| WO2020224458A1 (en) * | 2019-05-07 | 2020-11-12 | 国网山东省电力公司电力科学研究院 | Method for detecting corona discharge employing image processing |
| CN110161374A (en) * | 2019-06-18 | 2019-08-23 | 苏州微纳激光光子技术有限公司 | A kind of day blind ultraviolet imagery background Xanthophyll cycle optical system and method |
| CN111352006A (en) * | 2020-03-27 | 2020-06-30 | 国网甘肃省电力公司电力科学研究院 | External insulation equipment real-time discharge intensity quantification and evaluation system based on ultraviolet imaging |
| CN111443264A (en) * | 2020-04-30 | 2020-07-24 | 国网江苏省电力有限公司电力科学研究院 | Corona ultraviolet imaging equipment-based corona discharge intensity identification method and system and storage medium |
| CN112285504B (en) * | 2020-10-27 | 2023-11-17 | 华北电力大学 | Multispectral solar blind narrow-band ultraviolet imager and method for detecting different discharge states thereof |
| CN112285504A (en) * | 2020-10-27 | 2021-01-29 | 华北电力大学 | Multispectral solar-blind narrow-band ultraviolet imager and method for detecting different discharge states by using same |
| CN113189391A (en) * | 2021-04-23 | 2021-07-30 | 中国人民解放军陆军工程大学 | Optical remote measuring method for ground flashback current |
| CN113238131A (en) * | 2021-07-13 | 2021-08-10 | 北方夜视科技(南京)研究院有限公司 | Insulator corona discharge state evaluation method based on deep learning |
| CN113781438A (en) * | 2021-09-12 | 2021-12-10 | 华北电力大学 | Method for judging corona initial voltage by using spot area based on discharge ultraviolet image |
| CN114002566A (en) * | 2021-11-09 | 2022-02-01 | 国网山东省电力公司电力科学研究院 | Partial discharge detection device and method based on multi-light fusion |
| CN114034997A (en) * | 2021-11-10 | 2022-02-11 | 国网江苏省电力有限公司检修分公司 | Insulator degradation degree prediction method and system based on multiple parameters |
| CN115389882A (en) * | 2022-08-26 | 2022-11-25 | 中国南方电网有限责任公司超高压输电公司广州局 | Corona discharge state evaluation method and device, computer equipment and storage medium |
| CN115389882B (en) * | 2022-08-26 | 2024-05-28 | 中国南方电网有限责任公司超高压输电公司广州局 | Corona discharge state evaluation method, corona discharge state evaluation device, computer device, and storage medium |
| CN117806161A (en) * | 2023-11-22 | 2024-04-02 | 国网湖北省电力有限公司超高压公司 | Ultraviolet device self-adaptive gain control method, storage medium and electronic device |
| CN117611784A (en) * | 2023-11-30 | 2024-02-27 | 华北电力大学 | Cooperative control method and device for multispectral high-timeliness discharge detection |
| CN118425710A (en) * | 2024-07-04 | 2024-08-02 | 中国南方电网有限责任公司超高压输电公司电力科研院 | Ultraviolet imaging device for insulator fault detection and insulator fault detection system |
| CN118425710B (en) * | 2024-07-04 | 2024-09-24 | 中国南方电网有限责任公司超高压输电公司电力科研院 | Ultraviolet imaging device for insulator fault detection and insulator fault detection system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103018640A (en) | Method for testing electricity discharge intensity of corona on surface of high-voltage insulator | |
| CN101551435B (en) | Method for detecting corona discharging and device thereof | |
| CN100483144C (en) | Corona detecting method and system for insulator remote surface on optical amplifier | |
| Wang et al. | Attachment process in rocket‐triggered lightning strokes | |
| CN112285504B (en) | Multispectral solar blind narrow-band ultraviolet imager and method for detecting different discharge states thereof | |
| CN102230952B (en) | Corona detection method based on ultraviolet photons | |
| CN105447471A (en) | Infrared detection based device gas leakage identification method and apparatus | |
| CN103412246B (en) | Based on the Automatic detection method of abnormal discharge of power equipment of ultraviolet video | |
| CN111308283A (en) | Multifunctional switch cabinet working state sensor and early warning method | |
| CN105301542A (en) | Full-solar-blind ultraviolet imaging quantified detection method based on photon counting technology | |
| CN103744005A (en) | Discharging strength conversion method based on ultraviolet imaging method | |
| CN104280670A (en) | Corona detection method based on solar-blind ultraviolet imager | |
| CN101718836B (en) | Electrical insulator fault detector and detecting method thereof | |
| CN101458300B (en) | Circuit discharging detecting system | |
| CN113253069B (en) | Partial discharge risk assessment method based on multispectral intensity | |
| CN112130316A (en) | Multi-channel multi-spectral-band optical filter structure and application and method thereof | |
| CN111443264A (en) | Corona ultraviolet imaging equipment-based corona discharge intensity identification method and system and storage medium | |
| CN113160095A (en) | Infrared detection signal pseudo-color processing method, device and system and storage medium | |
| CN201355386Y (en) | Circuit discharge detecting device | |
| CN113238131B (en) | Insulator corona discharge state evaluation method based on deep learning | |
| CN211785898U (en) | Ultraviolet warning thermal infrared imager | |
| Koopmans | Current status of the LOFAR EoR key science project | |
| CN103616621B (en) | A kind of vacuum discharge detection device of electronic gun | |
| Li et al. | Machine vision-based network monitoring system for solar-blind ultraviolet signal | |
| CN108844637B (en) | InGaAs FPAs assembly performance evaluation device based on image quality and application thereof |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20130403 |