CN114264915A - Power distribution network cable joint operation condition assessment early warning device and method - Google Patents
Power distribution network cable joint operation condition assessment early warning device and method Download PDFInfo
- Publication number
- CN114264915A CN114264915A CN202111518571.0A CN202111518571A CN114264915A CN 114264915 A CN114264915 A CN 114264915A CN 202111518571 A CN202111518571 A CN 202111518571A CN 114264915 A CN114264915 A CN 114264915A
- Authority
- CN
- China
- Prior art keywords
- sample
- ultraviolet
- cable joint
- output
- layer
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 71
- 238000012545 processing Methods 0.000 claims abstract description 44
- 238000011156 evaluation Methods 0.000 claims abstract description 25
- 238000012360 testing method Methods 0.000 claims abstract description 22
- 238000006243 chemical reaction Methods 0.000 claims abstract description 10
- 238000013210 evaluation model Methods 0.000 claims abstract description 4
- 238000002360 preparation method Methods 0.000 claims abstract description 4
- 238000012795 verification Methods 0.000 claims abstract description 4
- 230000006870 function Effects 0.000 claims description 39
- 210000002569 neuron Anatomy 0.000 claims description 24
- 230000008569 process Effects 0.000 claims description 21
- 238000004422 calculation algorithm Methods 0.000 claims description 19
- 238000012549 training Methods 0.000 claims description 18
- 239000011521 glass Substances 0.000 claims description 13
- 230000004913 activation Effects 0.000 claims description 11
- 238000013528 artificial neural network Methods 0.000 claims description 10
- 238000004364 calculation method Methods 0.000 claims description 10
- 238000012423 maintenance Methods 0.000 claims description 9
- 230000011218 segmentation Effects 0.000 claims description 8
- 230000005540 biological transmission Effects 0.000 claims description 7
- 230000003044 adaptive effect Effects 0.000 claims description 6
- 230000005284 excitation Effects 0.000 claims description 6
- 238000011478 gradient descent method Methods 0.000 claims description 6
- 238000010606 normalization Methods 0.000 claims description 6
- 238000003384 imaging method Methods 0.000 claims description 5
- 230000007613 environmental effect Effects 0.000 claims description 4
- 238000012544 monitoring process Methods 0.000 claims description 4
- 230000003287 optical effect Effects 0.000 claims description 3
- 239000002243 precursor Substances 0.000 claims description 3
- 239000011541 reaction mixture Substances 0.000 claims description 3
- 238000002834 transmittance Methods 0.000 claims description 3
- 238000012821 model calculation Methods 0.000 claims description 2
- 238000009413 insulation Methods 0.000 abstract description 2
- 238000001514 detection method Methods 0.000 description 6
- 238000003745 diagnosis Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000004927 fusion Effects 0.000 description 2
- 239000012212 insulator Substances 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 239000003595 mist Substances 0.000 description 2
- 241000566145 Otus Species 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000007664 blowing Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000009422 external insulation Methods 0.000 description 1
- 238000010438 heat treatment Methods 0.000 description 1
- 238000003331 infrared imaging Methods 0.000 description 1
- YOBAEOGBNPPUQV-UHFFFAOYSA-N iron;trihydrate Chemical compound O.O.O.[Fe].[Fe] YOBAEOGBNPPUQV-UHFFFAOYSA-N 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000010998 test method Methods 0.000 description 1
- 238000001931 thermography Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
- 238000000825 ultraviolet detection Methods 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Images
Abstract
The invention discloses a device and a method for evaluating and early warning the operation condition of a cable joint of a power distribution network, and belongs to the technical field of high voltage and insulation. The device comprises: the system comprises a phytotron, an ultraviolet imager, a thermal infrared imager and an information fuzzy reasoning system; a hygrometer and a sample are arranged in the artificial climate chamber; a video recorder, a format conversion module, an image acquisition module and an ultraviolet image processing module are sequentially arranged between the ultraviolet imager and the information fuzzy inference system; an infrared image processing module is arranged between the thermal infrared imager and the information fuzzy inference system. The method comprises the steps of sample preparation and test, information acquisition and processing, sample running state evaluation model establishment and verification, and finally establishment of a cable joint running state evaluation early warning database and application of the cable joint running state evaluation early warning database. The method can evaluate and early warn the running state of the cable joint, and has good popularization prospect.
Description
Technical Field
The invention belongs to the technical field of high voltage and insulation, and particularly relates to a device and a method for evaluating and early warning the operation condition of a cable joint of a power distribution network.
Background
Underground cables have become the main electric energy transmission mode for distribution network power supply due to the advantages of small occupied space and small influence of natural environment. However, in the service process, due to the influence of factors such as a channel structure, a cable trench environment and the like, the service performance of the cable is reduced, and the cable can be aged, partially discharged and the like.
In the actual operation process, the severe environment characteristics of the cable trench are main parameters influencing key parts of the underground cable, the influences of different temperature and humidity environments on the service state of the cable are greatly different, and the partial discharge characteristics are often shown, so that the temperature and humidity environments of the cable trench are closely related to the service characteristics of the cable, particularly, cracks are more easily generated near a cable joint, and finally, the phenomena of water branches, partial discharge and the like are easily generated on the cable joint under the combined action of the temperature and the humidity of the environment, so that the cable trench is a key position where the cable is easily broken down.
At present, the method for evaluating the external insulation state of the cable joint generally adopts methods such as a partial discharge test, an infrared temperature test, an ultraviolet current pulse test and the like to test the influence of influencing factors such as humidity, temperature or ultraviolet detection distance and the like on the failure of the cable joint. However, these methods have single limitations, cannot be completely suitable for the field environment to perform simple, fast and effective state detection on the cable connector, are prone to false detection and missed detection, are not beneficial to fault judgment and prediction, and are difficult to meet the requirements for evaluation of the actual conditions.
In the prior art, ultraviolet image diagnosis and infrared image diagnosis are widely applied to equipment such as insulators and high-voltage transformers. However, the infrared diagnosis only judges the operation state of the electrical equipment by heating, the detection mode is single, and the operation state of the cable joint of the power distribution network is difficult to evaluate from multiple angles, so that the information identification rate is low. And other information fusion technologies cannot be adapted to the evaluation and early warning of the operation condition of the cable joint due to respective defects, for example, the infrared and ultraviolet imaging fusion technology based on the radial neural network cannot be adapted due to the fact that the prediction precision and the training speed of the RBF neural network are insufficient, and the uncertainty of information identification is increased.
Based on the problems in the background art, a device and a method specially for evaluating and early warning the operation condition of a cable joint of a power distribution network are needed to overcome the defects of high single information identification uncertainty and low accuracy rate in the evaluation methods of equipment such as a simple transfer insulator and a high-voltage transformer, and the operation state of the cable joint is comprehensively evaluated.
Disclosure of Invention
The invention aims to provide a device and a method for evaluating and early warning the operation condition of a cable joint of a power distribution network, and aims to solve the problems that the evaluation and early warning result is high in uncertainty and low in accuracy rate due to the lack of a comprehensive method for evaluating and early warning the operation condition of the cable joint.
In order to solve the problems, the technical scheme of the invention is as follows:
a distribution network cable joint operation condition assessment early warning device includes: the system comprises a phytotron, an ultraviolet imager, a thermal infrared imager and an information fuzzy reasoning system; a hygrometer and a sample are arranged in the artificial climate chamber, ultraviolet glass is arranged on one side of the artificial climate chamber, and a humidifier is arranged on the other side of the artificial climate chamber through a pipeline; a video recorder, a format conversion module, an image acquisition module and an ultraviolet image processing module are sequentially arranged between the ultraviolet imager and the information fuzzy inference system; an infrared image processing module is arranged between the thermal infrared imager and the information fuzzy inference system.
Furthermore, the artificial climate chamber is provided with 2 fans, the fans are respectively arranged at the opposite angles of the artificial climate chamber, and the pitch angle of the fans is set to be 45 degrees.
Furthermore, the light transmittance of the ultraviolet glass in the ultraviolet light signal of 240-280nm wave band can reach more than 95%.
Further, the observation distance of the ultraviolet imager from the sample through the ultraviolet glass is 7m, the observation angle is 20 degrees, and the gain is 70%; the observation distance of the infrared thermal imager from the sample through the ultraviolet glass is 2 m.
The early warning method of the early warning device comprises the following steps:
step A, sample preparation and testing;
sequentially and respectively installing samples with different operation state grades in an artificial climate chamber for electrifying, starting a humidifier and adjusting environmental parameters;
placing an ultraviolet imager 7m behind a sample in a phytotron, adjusting the elevation angle to 20 degrees for shooting and imaging, and shooting when the sample is pressurized and heated and tends to be stable;
placing a thermal infrared imager 2m behind a sample in an artificial climate chamber, and carrying out infrared shooting on the surface of the sample, wherein the shooting frequency is once every 5min, and 5 pictures are shot in each state grade;
meanwhile, the humidity is tested in real time through a humidity tester;
b, information acquisition and processing;
1. collecting and processing information of an ultraviolet imager;
shooting a sample by an ultraviolet imager, and storing an ultraviolet video by a video recorder;
then, the video obtained by the video recorder is converted into an AVI format through VTA software in a format conversion module, so that the video is convenient to store and process;
then, the obtained AVI format video passes through Corel video Audio System software in an image acquisition module to acquire ultraviolet images with specified frame number, wherein the specified frame number is 25 frames per second, and the ultraviolet images are in BMP format;
then inputting the obtained image into MATLAB in an ultraviolet image processing module, and according to the full white pixel pointsThe gray value (255) of the ultraviolet light source and the gray value (0) of the full black pixel point are subjected to gray processing, and finally, the presented white light spot is a discharge area in the ultraviolet image; continuously carrying out binarization on the image after the gray level processing by using an OTSU (optical transmission unit) adaptive threshold segmentation method, wherein the gray level value of pixels in the area of a white discharge area is between 0 and 1, and if the discharge of the sample is stronger, the gray level value of the sample is larger; the gray value of the pixel in the black area is 0, the sum of the gray values of the pixel in the white area represents the discharge intensity, the area of the white light spot is continuously calculated, and the area S of the white light spot of 250 images in each group is calculated according to 10S as a groupnAnd taking the 3 statistics as ultraviolet discharge characteristics;
2. collecting and processing information of the thermal infrared imager;
shooting the surface of the sample by using a thermal infrared imager, wherein the obtained infrared image is in a BMP format, and the obtained temperature data is in an Excel format;
inputting the infrared image in the data into MATLAB software in an infrared image processing module, performing gray processing and OTSU adaptive threshold segmentation, extracting the data of the sample, reading the temperature rise (difference between temperature and ambient temperature) of the corresponding position in a temperature data file according to the area coordinates of the sample surface, and defining the temperature rise data of the sample surface as TnCalculating the mean value, mode value and variance of the temperature rise of the surface area of the sample, and taking the 3 statistics as temperature characteristics;
3. acquiring information of a humidity tester;
humidity in artificial climate chamber detected by collection hygrometer is taken as humidity characteristic value Un；
Step C, establishing a sample running state evaluation model, verifying the model, and finally establishing a database for evaluating and early warning the running state of the cable joint;
and B, testing the product obtained in the step B for multiple times: data S of the first sample1、T1、U1Data S of the second sample2、T2、U2Data S of the n-th samplen、Tn、UnInputting the collected information into a fuzzy inference system for training, analyzing and processing;
an input-output structural model is built by adopting a command function anfis in MATLAB software, and the anfis function can initialize a fuzzy inference system and automatically generate a fuzzy rule; each rule may be expressed in an if-then statement, where if is a precondition and then is a "run state"; and the linear combination of the input variable and the constant term can obtain the output of the input variable, and the expression is as follows:
in the formula, yiIs the output of the ith rule; x is an output variable; a is a constant coefficient;
for example: if "very low temperature rise" + "very low discharge" + "very low humidity" the "cable joint remains operational; other rules are similar;
training three fourths of sample data as training data input in advance, wherein the sample data comprises: s1、T1，S2、T2，……,Sn、TnAnd the mean, mode and variance of (1) and the corresponding U1,U2……Un(ii) a Inputting the remaining one fourth of sample data into a trained ANFIS neural network for test verification to obtain the running state of the cable joint judged by ANFIS decision, and comparing the judgment result with the real running state to verify the feasibility of the method;
according to the basic structure of ANFIS, as shown in fig. 3:
the first layer is a membership function layer, and selects a proper membership function (generally, a Gaussian membership function is adopted to be more accurate) for fuzzifying input data aiming at data characteristics;
the second layer is a rule reasoning layer, specified neurons receive input information through corresponding fuzzification neurons, a normal form function of fuzzy T can be selected optionally, and rule excitation intensity is obtained through calculation;
the third layer is a normalization layer, the neuron input from the rule layer is all input into the neuron of the layer, and the normalized activation strength of the rule is determined through calculation;
the fourth layer is an inverse fuzzy layer, and the normalization neurons are connected to each corresponding neuron in the layer and receive initial input characteristic information at the same time;
the fifth layer is an output layer, the sum of the output of each defuzzification neuron is calculated, and according to the running state of a cable joint, the running state of a fuzzy output variable is divided into 4 fuzzy subsets, namely 'maintenance running', 'live-line maintenance', 'power failure maintenance' and 'equipment replacement';
d, applying a database for evaluating and early warning the running condition of the cable joint;
the sample needing to be evaluated and early-warned is tested, information is collected and processed according to the method in the step A and the step B, and corresponding S is obtainedi、TiMean, mode and variance of and UiInputting the value into the ANFIS model verified and perfected in the step C, and obtaining a corresponding output quantity through model calculation and evaluation;
according to the output preset value of the fifth layer in the step C, evaluating the running states of the sample, such as ultraviolet discharge intensity and the like, so as to obtain the running state grade of the tested sample, and realize intelligent monitoring, evaluation and early warning of the multi-source information of the running state of the cable joint of the power distribution network;
warp above Si、TiMean, mode and variance of and UiThe corresponding output quantities are simultaneously recorded in the database formed in step C.
Further, in step C, the first layer adopts a gaussian membership function as:
in the formula (1), c and σ are parameters of a gaussian function, which are also called precursor parameters.
Further, the output of each node of the second layer in step C represents the rule excitation strength:
in the formula (2), OiIs the output of the second layer; w is aiAn activation function corresponding to the ith rule for the input signal;
and
respectively, of the second layer.
Further, the normalized activation intensity determined in the third layer in step C is:
in the formula (3), the reaction mixture is,
to normalize the activation intensity.
Further, in step C, the fourth layer neurons output:
in the formula (4), YiIs a function value; xiIs an input variable; { pi,ri,qiThe conclusion parameter, also called the post-element parameter, is determined by the training of anfis.
Further, in the fifth layer in the step C, the final output can be obtained by calculating the rule:
in the formula (5), Y is the total system output, YiThe output of the ith rule is shown;
after the membership function of each input variable and each output variable is determined and the fuzzy rule is established, the corresponding ANFIS system structure can be determined, then a large amount of data obtained by testing cable joints is taken as sample data, and a hybrid algorithm combining a BP algorithm and a least square method is applied to carry out system training until errors meet requirements;
the front-piece and back-piece parameters of the system can be identified by combining a gradient descent method and a least square method in the BP algorithm, and for the hybrid algorithm, the learning process of each period comprises forward transmission and backward propagation; in the forward learning process, fixing the front part parameters and identifying the back part parameters by using a least square method; if there are n groups of learning samples, the kth omic learning sample in the n groups is obtained by the formula (5):
let matrices a and S be:
S＝[p1 q1 r1 p2 q2 r2]T (8)
the total output Y of the system after the n group learning samples are input into the system can be expressed as
Y＝[Y1 Y2 … Yk … Yn]＝A|S (9)
The optimal estimation value S of the back-piece parameter vector under the minimum mean square error meaning can be obtained by following the least square method*
S*＝(ATA)-1ATY (10)
On the contrary, in the process of reverse learning, the error calculation is carried out on the back-piece parameters obtained in the last step, the BP algorithm in the feedforward neural network is adopted, the error is reversely transmitted from the output end to the input end, and finally the gradient descent method is utilized to correct the front-piece parameters;
after the ANFIS system is determined, according to the related membership function and the fuzzy rule, a group of input information is calculated by the ANFIS system to obtain the corresponding output; that is, when S, T, U in the input amount changes, the corresponding state output J changes with S, T, U; and the operation state of the cable joint in the power distribution network can be judged according to the finally obtained output value J.
The invention has the following beneficial effects:
(1) the device utilizes the environmental characteristics of the cable trench in the actual operation process simulated by the artificial climate chamber, heats and discharges the cable joint in the pressurization process, and simultaneously utilizes the ultraviolet imager to detect the discharge intensity; the ultraviolet imaging instrument shoots 240-280nm band ultraviolet signals when a sample runs, the ultraviolet signals are stored through an external video recorder for analyzing a discharge dynamic process, the video signals are transmitted to a computer, after being converted through VTA software in a format conversion module, Corel video System software in an image acquisition module is used for converting to obtain ultraviolet images with specific frame numbers, and then the image signals are processed through MATLAB software in an ultraviolet image processing module to obtain the number of white light spot pixels (the white light spot area is a discharge area of a cable sample), so that the discharge strength can be represented. And extracting the mode, the mean value and the variance of the area of the light spot as characteristic parameters, and finally entering an information fuzzy inference system to train the discharge information of each position of the sample.
The thermal infrared imager can monitor a sample to obtain an infrared image and temperature data, transmits image information to MATLAB software of an infrared image processing module in a computer to obtain the mode, the mean value and the variance of the temperature rise of the surface area of the sample as characteristic parameters, and finally enters an information fuzzy reasoning system to train the temperature rise information of each part of the sample.
The hygrometer can extract the humidity in the artificial climate chamber when the sample operates, measures the humidity value of the sample as a characteristic parameter in the same way, finally inputs the characteristic parameter into the information fuzzy reasoning system, combines the two groups of characteristic information to train, process and verify, and finally forms an operation state evaluation database for evaluating the cable joint state in the actual operation process of early warning.
(2) The method is different from single test methods such as partial discharge test, infrared temperature test, ultraviolet current pulse test and the like, obtains characteristic data which can represent the running state of the cable joint, such as discharge capacity, temperature rise, humidity and the like of the cable joint through ultraviolet and infrared imaging and a hygrometer, establishes an anfis data prediction model on the basis of the characteristic data, and can evaluate and early warn the cable joint through intelligent monitoring of multi-source information after establishing the model. And carrying out a large amount of detection on samples in different running states in the early stage of model establishment and collecting the obtained characteristic information of discharge capacity, temperature rise, humidity and the like. And then establishing a model to train the obtained characteristic information, and finally establishing a cable joint running state evaluation database. And then, the method can be applied reversely, parameters such as ultraviolet discharge, infrared temperature and humidity of the cable joint are extracted to be used as the input of a prediction model, and the obtained output can represent the running state of the cable joint. Therefore, the method can evaluate and early warn the running state of the cable joint, and has good popularization prospect.
Drawings
FIG. 1 is a detection flow chart of a power distribution network cable joint operation condition evaluation early warning;
FIG. 2 is a feasibility analysis diagram of a power distribution network cable joint operation condition evaluation early warning method;
fig. 3 is a diagram of an ANFIS model network structure used in the power distribution network cable joint operation condition assessment and early warning method in step C.
The reference numbers are as follows: 1-a humidifier; 2-artificial climate chamber; 3-a fan; 4-a hygrometer; 5-ultraviolet glass; 6-ultraviolet imager; 7-infrared thermal imaging system; 8-sample; 9-a video recorder; 10-a format conversion module; 11-an image acquisition module; 12-an ultraviolet image processing module; 13-an infrared image processing module; 14-information fuzzy inference system.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the technical solutions of the embodiments of the present invention will be described clearly and completely with reference to the accompanying drawings of the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all embodiments of the present invention. All other embodiments, which can be obtained by a person skilled in the art without any inventive step based on the embodiments of the present invention, are within the scope of the present invention.
Thus, the following detailed description of the embodiments of the present invention, presented in the figures, is not intended to limit the scope of the invention, as claimed, but is merely representative of selected embodiments of the invention.
Examples
As shown in fig. 1, an evaluation and early warning device for the operation condition of a cable joint of a power distribution network comprises: the system comprises a phytotron 2, an ultraviolet imager 6, a thermal infrared imager 7 and an information fuzzy inference system 14.
2 fans 3, a hygrometer 4 and a sample 8 are arranged in the artificial climate chamber 2, ultraviolet glass 5 is arranged on one side of the artificial climate chamber 2, and the light transmittance of the ultraviolet glass 5 in the 240-280nm waveband ultraviolet light signal can reach more than 95%; the other side of the artificial climate chamber 2 is provided with a humidifier 1 through a pipeline. The fans 3 are respectively arranged at the opposite corners of the artificial climate chamber, the pitch angle is set to be 45 degrees, and the fans are used for blowing away the fog of the humidifier 1 and are uniformly distributed in the artificial climate chamber 2. The hygrometer 4 is connected to an information fuzzy inference system 14. The sample 8 may be electrically pressurized.
A video recorder 9, a format conversion module 10, an image acquisition module 11 and an ultraviolet image processing module 12 are sequentially arranged between the ultraviolet imager 6 and the information fuzzy inference system 14. The observation distance of the ultraviolet imager 6 from the sample 8 through the ultraviolet glass 5 is 7m, the observation angle is 20 degrees, and the gain is 70%.
An infrared image processing module 13 is arranged between the thermal infrared imager 7 and the information fuzzy inference system 14. The infrared thermal imager 7 transmits the ultraviolet glass 5 and has an observation distance of 2m from the sample.
In the implementation:
the climatic chamber 2 is about 2m in diameter and about 3m high.
The humidifier adopts an industrial ultrasonic cold mist humidifier 1, the maximum mist output amount can reach 12kg/h, and the humidifier can be manually controlled.
The hygrometer 4 is a digital hygrometer.
The ultraviolet imager 6 is of a CoroCAM 504 type in south Africa and can shoot ultraviolet signals in the wave band of 240-280 nm.
The video recorder 9 is of a Sony VDR-MC3 type and can record video of the shooting of the ultraviolet imager 7.
The format conversion module 10 adopts VTA software, and can change the video format into AVI format, which is convenient for storage.
The image acquisition module 11 can convert the picture into BMP format by using the Corel video system software.
The ultraviolet image processing module 12 employs MATLAB (gray scale processing, OTSU segmentation).
The thermal infrared imager 7 is of a Fldke Ti-32 type, and can obtain an infrared image in a BMP format and a temperature data file in an Excel format respectively through shooting. And the picture display is more beautiful when the picture display is adjusted to be iron red.
The infrared image processing module 13 also employs MATLAB (gray scale processing, OTUS segmentation).
The information fuzzy inference model in the information fuzzy inference system 14 is an ANFIS model built by MATLAB software.
Sample 8 was classified into four classes according to actual operating conditions: i, II, III and IV grades.
The early warning method of the early warning device for evaluating the operation condition of the cable joint of the power distribution network is specifically divided into the following steps as shown in fig. 1-2:
step A, sample preparation and testing;
and sequentially and respectively installing the samples 8 with different running state grades in the artificial climate chamber 2, electrifying, starting the humidifier 1, and adjusting the environmental parameters.
The ultraviolet imager 6 is arranged at a position 7m behind the sample 8 in the climatic chamber 2, the elevation angle is adjusted to 20 degrees to prepare shooting imaging, and shooting is carried out when the sample 8 is pressurized and heated to be stable.
And placing the thermal infrared imager 7 2m behind the sample 8 in the artificial climate chamber 2, and carrying out infrared shooting on the surface of the sample, wherein the shooting frequency is once every 5min, and 5 pictures are shot in each state grade.
And simultaneously the humidity is tested in real time by the humidity tester 4.
B, information acquisition and processing;
1. collecting and processing information of the ultraviolet imager 6;
the ultraviolet imager 6 shoots a sample 8, and an ultraviolet video is stored through the video recorder 9;
then, the video obtained by the video recorder 9 is converted into an AVI format through VTA software in the format conversion module 10, so that the video is convenient to store and process;
then, the obtained AVI format video passes through Corel video System software in an image acquisition module 11 to acquire ultraviolet images with specified frame number, wherein the specified frame number is 25 frames per second, and the ultraviolet images are in BMP format;
then, the obtained image is input into MATLAB in the ultraviolet image processing module 12, and gray processing is performed according to the gray value (255) of the full white pixel point and the gray value (0) of the full black pixel point, and finally, the displayed white light spot is the discharge area in the ultraviolet image. Continuously carrying out binarization on the image after the gray level processing by using an OTSU (optical transmission unit) adaptive threshold segmentation method, wherein the gray level value of pixels in the area of a white discharge area is between 0 and 1, and if the discharge of the sample is stronger, the gray level value of the sample is larger; the gray value of the pixel in the black area is 0, the sum of the gray values of the pixel in the white area represents the discharge intensity, the area of the white light spot is continuously calculated, and the area S of the white light spot of 250 images in each group is calculated according to 10S as a groupnAnd the 3 statistics are taken as the ultraviolet discharge characteristics.
2. Collecting and processing information of the thermal infrared imager 7;
the thermal infrared imager 7 shoots the surface of the sample 8, the obtained infrared image is in a BMP format, and the obtained temperature data is in an Excel format;
inputting the infrared image in the data into MATLAB software in an infrared image processing module 13, performing gray processing and OTSU adaptive threshold segmentation, extracting data of the sample 8, reading the temperature rise (difference between temperature and ambient temperature) of the corresponding position in the temperature data file according to the area coordinates of the sample surface, and defining the temperature rise data of the sample surface as TnThe mean, mode and variance of the temperature rise in the surface area of the sample 8 were calculated as 3The statistics are characterized as temperatures.
3. Acquiring information of the humidity tester 4;
humidity in the artificial climate chamber 2 detected by the hygrometer 4 is collected as a humidity characteristic value Un。
Step C, establishing a sample running state evaluation model, verifying the model, and finally establishing a database for evaluating and early warning the running state of the cable joint;
and B, testing the product obtained in the step B for multiple times: data S of the first sample 81、T1、U1Data S of the second sample 82、T2、U2Data S of the n-th sample 8n、Tn、UnAfter the collection, the information is input into the fuzzy inference system 14 for training and analyzing.
Namely, an input-output structure model is built by adopting a command function anfis in MATLAB software, and the anfis function can initialize the fuzzy inference system and automatically generate fuzzy rules. Each rule may be expressed in an if-then statement, where if is a precondition and then is a "run state". And the linear combination of the input variable and the constant term can obtain the output of the input variable, and the expression is as follows:
in the formula, yiIs the output of the ith rule; x is an output variable; a is a constant coefficient.
For example: if "very low temperature rise" + "very low discharge" + "very low humidity" the "cable joint remains operational; other rules are similar.
Training three fourths of sample data as training data input in advance, wherein the sample data comprises: s1、T1，S2、T2，……,Sn、TnAnd the mean, mode and variance of (1) and the corresponding U1,U2……Un. Inputting the remaining one fourth of sample data into the trained ANFIS neural network for test verification to obtainAnd judging the running state of the cable joint by ANFIS decision, comparing the judgment result with the real running state, and verifying the feasibility of the method.
According to the basic structure of ANFIS, as shown in fig. 3:
the first layer is a membership function layer. And (3) selecting a proper membership function (generally, a Gaussian membership function is more accurate) for data characteristics to fuzzify the input data.
And the Gaussian function membership function is:
in the formula (1), c and σ are parameters of a gaussian function, which are also called precursor parameters.
The second layer is a rule reasoning layer, the specified neurons receive input information from corresponding fuzzification neurons, a normal form function of fuzzy T can be selected optionally, and rule excitation strength is obtained through calculation, namely the output of each node represents the rule excitation strength:
in the formula (2), OiIs the output of the second layer; w is aiAn activation function corresponding to the ith rule for the input signal;
and
respectively, of the second layer.
The third layer is a normalization layer, the neuron inputs from the rule layer are all input into the neurons of the layer, and the normalized activation strength of the rule is determined by calculation:
in the formula (3), the reaction mixture is,
to normalize the activation intensity.
The fourth layer is an inverse fuzzy layer, the normalization neurons are connected to each corresponding neuron in the layer and receive initial input characteristic information at the same time, and the neurons in the layer output:
in the formula (4), YiIs a function value; xiIs an input variable; { pi,ri,qiThe conclusion parameter, also called the post-element parameter, is determined by the training of anfis.
The fifth layer is an output layer, the sum of the output of each defuzzification neuron is calculated, and the fuzzy output variable 'running state' is divided into 4 fuzzy subsets according to the running state of the cable joint, namely 'maintenance running', 'live-line maintenance', 'power failure maintenance' and 'equipment replacement'. The final output can be obtained by calculating the rule:
in the formula (5), Y is the total system output, YiIndicating the output of the ith rule.
After the membership function of each input variable and each output variable is determined and the fuzzy rule is established, the corresponding ANFIS system structure can be determined, then a large amount of data obtained by testing cable joints is used as sample data, and a hybrid algorithm combining a BP algorithm and a least square method is applied to carry out system training until errors meet requirements.
The front-piece and back-piece parameters of the system can be identified by combining a gradient descent method and a least square method in the BP algorithm, and the learning process of each period comprises forward transmission and backward propagation for the hybrid algorithm. In the forward learning process, the front piece parameters are fixed, and the back piece parameters are identified by using a least square method. If there are n groups of learning samples, the kth omic learning sample in the n groups is obtained by the formula (5):
let matrices a and S be:
S＝[p1 q1 r1 p2 q2 r2]T (8)
the total output Y of the system after the n group learning samples are input into the system can be expressed as
Y＝[Y1 Y2 … Yk … Yn]＝AS (9)
The optimal estimation value S of the back-piece parameter vector under the minimum mean square error meaning can be obtained by following the least square method*
S*＝(ATA)-1ATY (10)
On the contrary, in the process of reverse learning, the error calculation is carried out on the back-piece parameters obtained in the last step, the BP algorithm in the feedforward neural network is adopted, the error is reversely transmitted from the output end to the input end, and finally the gradient descent method is used for correcting the front-piece parameters.
After the ANFIS system is determined, according to the related membership function and the fuzzy rule, a group of input information is calculated by the ANFIS system to obtain the corresponding output; that is, when S, T, U changes in the input amount, the corresponding state output J changes with S, T, U. And the operation state of the cable joint in the power distribution network can be judged according to the finally obtained output value J.
In the specific implementation:
in the training process, the error of the model is reduced by adjusting the parameter values of all membership functions, so that the calculated data in the system is closer to the actual numerical value.
When the training work is finished and the error meets the requirement, the model can directly calculate the input characteristic quantity to automatically obtain the corresponding output quantity, and therefore the prediction estimation of a certain object is realized.
The most classical machine learning algorithm of ANFIS is a mixed algorithm formed by a least square method and a BP algorithm, each front piece measurement parameter in the neural network is counted through the BP algorithm, and the center and the width of each front piece measurement parameter of the neural network are adjusted by the least square method, so that the modeling process is more accurate.
After the ANFIS model is established, different input characteristic parameters are input into the model under the constraint of the previously solved membership function and fuzzy rule, and different output quantities can be obtained.
Assuming that the experiment obtained 200 data, three quarters of the data were trained and the remaining data were used for testing. Setting the network error as 0.05 and the iteration times as 100, verifying the established model, and finally establishing a database for evaluating and early warning the operation condition of the cable joint.
D, applying a database for evaluating and early warning the running condition of the cable joint;
the sample needing to be evaluated and early-warned is tested, information is collected and processed according to the method in the step A and the step B, and corresponding S is obtainedi、TiMean, mode and variance of and UiAnd inputting the value into the ANFIS model verified and perfected in the step C, and calculating and evaluating through the model to obtain the corresponding output quantity.
And C, evaluating the running states of the samples such as ultraviolet discharge intensity and the like according to the output preset value of the fifth layer in the step C to obtain the running state grade of the tested samples, and realizing intelligent monitoring, evaluation and early warning of the multi-source information of the running state of the cable joint of the power distribution network.
Warp above Si、TiMean, mode and variance of and UiThe corresponding output quantities are simultaneously recorded in the database formed in step C.
Claims (10)
1. The utility model provides a distribution network cable joint running condition aassessment early warning device which characterized in that: the method comprises the following steps: the system comprises a phytotron (2), an ultraviolet imager (6), a thermal infrared imager (7) and an information fuzzy reasoning system (14); a hygrometer (4) and a sample (8) are arranged in the artificial climate chamber (2), ultraviolet glass (5) is arranged on one side of the artificial climate chamber (2), and a humidifier (1) is arranged on the other side of the artificial climate chamber (2) through a pipeline; a video recorder (9), a format conversion module (10), an image acquisition module (11) and an ultraviolet image processing module (12) are sequentially arranged between the ultraviolet imager (6) and the information fuzzy inference system (14); an infrared image processing module (13) is arranged between the thermal infrared imager (7) and the information fuzzy reasoning system (14).
2. The power distribution network cable joint operation condition evaluation and early warning device of claim 1, wherein: the artificial climate chamber (2) is provided with 2 fans (3), the fans (3) are respectively arranged at opposite angles of the artificial climate chamber, and the pitch angle of the fans (3) is set to be 45 degrees.
3. The power distribution network cable joint operation condition evaluation and early warning device of claim 1, wherein: the light transmittance of the ultraviolet glass (5) in the ultraviolet light signal with the wave band of 240-280nm can reach more than 95 percent.
4. The power distribution network cable joint operation condition evaluation and early warning device of claim 1, wherein: the observation distance between the ultraviolet imager (6) and the sample (8) after the ultraviolet glass (5) penetrates through the ultraviolet imager is 7m, the observation angle is 20 degrees, and the gain is 70 percent; the infrared thermal imager (7) penetrates through the ultraviolet glass (5) and has an observation distance of 2m from the sample (8).
5. An early warning method of an early warning apparatus as claimed in any one of claims 1 to 4, wherein: the method comprises the following steps:
step A, sample preparation and testing;
sequentially and respectively installing samples 8 with different operation state grades in an artificial climate chamber 2 for electrifying, starting a humidifier 1, and adjusting environmental parameters;
placing an ultraviolet imager 6 at a position 7m behind a sample 8 in the artificial climate chamber 2, adjusting the elevation angle to be a 20-degree angle for shooting imaging, and shooting when the sample 8 is pressurized and heated and tends to be stable;
placing the thermal infrared imager 7 2m behind the sample 8 in the artificial climate chamber 2, and carrying out infrared shooting on the surface of the sample, wherein the shooting frequency is once every 5min, and 5 pictures are shot in each state grade;
meanwhile, the humidity is tested in real time by a humidity tester 4;
b, information acquisition and processing;
1. collecting and processing information of the ultraviolet imager 6;
the ultraviolet imager 6 shoots a sample 8, and an ultraviolet video is stored through the video recorder 9;
then, the video obtained by the video recorder 9 is converted into an AVI format through VTA software in the format conversion module 10, so that the video is convenient to store and process;
then, the obtained AVI format video passes through Corel video System software in an image acquisition module 11 to acquire ultraviolet images with specified frame number, wherein the specified frame number is 25 frames per second, and the ultraviolet images are in BMP format;
inputting the obtained image into MATLAB in the ultraviolet image processing module 12, and carrying out gray processing according to the gray value (255) of the full white pixel point and the gray value (0) of the full black pixel point, wherein the displayed white light spot is a discharge area in the ultraviolet image; continuously carrying out binarization on the image after the gray level processing by using an OTSU (optical transmission unit) adaptive threshold segmentation method, wherein the gray level value of pixels in the area of a white discharge area is between 0 and 1, and if the discharge of the sample is stronger, the gray level value of the sample is larger; the gray value of the pixel in the black area is 0, the sum of the gray values of the pixel in the white area represents the discharge intensity, the area of the white light spot is continuously calculated, and the area S of the white light spot of 250 images in each group is calculated according to 10S as a groupnAnd taking the 3 statistics as ultraviolet discharge characteristics;
2. collecting and processing information of the thermal infrared imager 7;
the thermal infrared imager 7 shoots the surface of the sample 8, the obtained infrared image is in a BMP format, and the obtained temperature data is in an Excel format;
inputting the infrared image in the data into MATLAB software in an infrared image processing module 13, performing gray processing and OTSU adaptive threshold segmentation, extracting data of the sample 8, reading the temperature rise (difference between temperature and ambient temperature) of the corresponding position in the temperature data file according to the area coordinates of the sample surface, and defining the temperature rise data of the sample surface as TnCalculating the mean value, mode value and variance of the temperature rise of the surface area of the sample 8, and taking the 3 statistic values as temperature characteristics;
3. acquiring information of the humidity tester 4;
humidity in the artificial climate chamber 2 detected by the hygrometer 4 is collected as a humidity characteristic value Un；
Step C, establishing a sample running state evaluation model, verifying the model, and finally establishing a database for evaluating and early warning the running state of the cable joint;
and B, testing the product obtained in the step B for multiple times: data S of the first sample 81、T1、U1Data S of the second sample 82、T2、U2Data S of the n-th sample 8n、Tn、UnAfter the information is collected, the information is input into a fuzzy inference system 14 for training, analyzing and processing;
an input-output structural model is built by adopting a command function anfis in MATLAB software, and the anfis function can initialize a fuzzy inference system and automatically generate a fuzzy rule; each rule may be expressed in an if-then statement, where if is a precondition and then is a "run state"; and the linear combination of the input variable and the constant term can obtain the output of the input variable, and the expression is as follows:
in the formula, yiIs the output of the ith rule; x is an output variable; a is a constant coefficient；
For example: if "very low temperature rise" + "very low discharge" + "very low humidity" the "cable joint remains operational; other rules are similar;
training three fourths of sample data as training data input in advance, wherein the sample data comprises: s1、T1，S2、T2，……,Sn、TnAnd the mean, mode and variance of (1) and the corresponding U1,U2……Un(ii) a Inputting the remaining one fourth of sample data into a trained ANFIS neural network for test verification to obtain the running state of the cable joint judged by ANFIS decision, and comparing the judgment result with the real running state to verify the feasibility of the method;
according to the basic structure of ANFIS, as shown in fig. 3:
the first layer is a membership function layer, and selects a proper membership function (generally, a Gaussian membership function is adopted to be more accurate) for fuzzifying input data aiming at data characteristics;
the second layer is a rule reasoning layer, specified neurons receive input information through corresponding fuzzification neurons, a normal form function of fuzzy T can be selected optionally, and rule excitation intensity is obtained through calculation;
the third layer is a normalization layer, the neuron input from the rule layer is all input into the neuron of the layer, and the normalized activation strength of the rule is determined through calculation;
the fourth layer is an inverse fuzzy layer, and the normalization neurons are connected to each corresponding neuron in the layer and receive initial input characteristic information at the same time;
the fifth layer is an output layer, the sum of the output of each defuzzification neuron is calculated, and according to the running state of a cable joint, the running state of a fuzzy output variable is divided into 4 fuzzy subsets, namely 'maintenance running', 'live-line maintenance', 'power failure maintenance' and 'equipment replacement';
d, applying a database for evaluating and early warning the running condition of the cable joint;
sample basis steps needing evaluation and early warningA. The method in the step B carries out testing, information acquisition and processing to obtain corresponding Si、TiMean, mode and variance of and UiInputting the value into the ANFIS model verified and perfected in the step C, and obtaining a corresponding output quantity through model calculation and evaluation;
according to the output preset value of the fifth layer in the step C, evaluating the running states of the sample, such as ultraviolet discharge intensity and the like, so as to obtain the running state grade of the tested sample, and realize intelligent monitoring, evaluation and early warning of the multi-source information of the running state of the cable joint of the power distribution network;
warp above Si、TiMean, mode and variance of and UiThe corresponding output quantities are simultaneously recorded in the database formed in step C.
6. The power distribution network cable joint operation condition evaluation and early warning method of claim 5, wherein: in the step C, the first layer adopts a Gaussian membership function as follows:
in the formula (1), c and σ are parameters of a gaussian function, which are also called precursor parameters.
7. The power distribution network cable joint operation condition evaluation and early warning method of claim 5, wherein: the output of each node of the second layer in the step C represents the rule excitation strength:
in the formula (2), OiIs the output of the second layer; w is aiAn activation function corresponding to the ith rule for the input signal;
and
respectively, of the second layer.
8. The power distribution network cable joint operation condition evaluation and early warning method of claim 5, wherein: the normalized activation intensity determined in the third layer in step C is:
in the formula (3), the reaction mixture is,
to normalize the activation intensity.
9. The power distribution network cable joint operation condition evaluation and early warning method of claim 5, wherein: the fourth layer of neurons in step C output:
in the formula (4), YiIs a function value; xiIs an input variable; { pi,ri,qiThe conclusion parameter, also called the post-element parameter, is determined by the training of anfis.
10. The power distribution network cable joint operation condition evaluation and early warning method of claim 5, wherein: and in the fifth layer in the step C, the final output can be obtained through a calculation rule:
in the formula (5), Y is the total system output, YiThe output of the ith rule is shown;
after the membership function of each input variable and each output variable is determined and the fuzzy rule is established, the corresponding ANFIS system structure can be determined, then a large amount of data obtained by testing cable joints is taken as sample data, and a hybrid algorithm combining a BP algorithm and a least square method is applied to carry out system training until errors meet requirements;
the front-piece and back-piece parameters of the system can be identified by combining a gradient descent method and a least square method in the BP algorithm, and for the hybrid algorithm, the learning process of each period comprises forward transmission and backward propagation; in the forward learning process, fixing the front part parameters and identifying the back part parameters by using a least square method; if there are n groups of learning samples, the kth omic learning sample in the n groups is obtained by the formula (5):
let matrices a and S be:
S＝[p1 q1 r1 p2 q2 r2]T (8)
the total output Y of the system after the n group learning samples are input into the system can be expressed as
Y＝[Y1 Y2 … Yk … Yn]＝AS (9)
The optimal estimation value S of the back-piece parameter vector under the minimum mean square error meaning can be obtained by following the least square method*
S*＝(ATA)-1ATY (10)
On the contrary, in the process of reverse learning, the error calculation is carried out on the back-piece parameters obtained in the last step, the BP algorithm in the feedforward neural network is adopted, the error is reversely transmitted from the output end to the input end, and finally the gradient descent method is utilized to correct the front-piece parameters;
after the ANFIS system is determined, according to the related membership function and the fuzzy rule, a group of input information is calculated by the ANFIS system to obtain the corresponding output; that is, when S, T, U in the input amount changes, the corresponding state output J changes with S, T, U; and the operation state of the cable joint in the power distribution network can be judged according to the finally obtained output value J.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111518571.0A CN114264915A (en) | 2021-12-13 | 2021-12-13 | Power distribution network cable joint operation condition assessment early warning device and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111518571.0A CN114264915A (en) | 2021-12-13 | 2021-12-13 | Power distribution network cable joint operation condition assessment early warning device and method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114264915A true CN114264915A (en) | 2022-04-01 |
Family
ID=80827199
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111518571.0A Pending CN114264915A (en) | 2021-12-13 | 2021-12-13 | Power distribution network cable joint operation condition assessment early warning device and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114264915A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115684835A (en) * | 2022-10-28 | 2023-02-03 | 贵州电网有限责任公司信息中心 | Power distribution network fault diagnosis method, device, equipment and storage medium |
| CN116973671A (en) * | 2023-09-22 | 2023-10-31 | 江苏大圆电子科技有限公司 | Aging early warning method and system for cable |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101576600A (en) * | 2009-05-15 | 2009-11-11 | 重庆大学 | Self-walking underground cable failure detection intelligent instrument |
| CN201527462U (en) * | 2009-09-10 | 2010-07-14 | 国网电力科学研究院 | Compact type artificial climate box for ultraviolet detection |
| CN103091603A (en) * | 2013-01-14 | 2013-05-08 | 华北电力大学 | Breakdown intelligent classification and positioning method of electric transmission line |
| CN104155574A (en) * | 2014-07-31 | 2014-11-19 | 国网湖北省电力公司武汉供电公司 | Power distribution network fault classification method based on adaptive neuro-fuzzy inference system |
| CN105891708A (en) * | 2016-05-11 | 2016-08-24 | 四川大学 | ANFIS-based condition assessment for high-voltage circuit breaker operating mechanism |
| CN111833312A (en) * | 2020-06-22 | 2020-10-27 | 国网江苏省电力有限公司电力科学研究院 | Ultraviolet image diagnosis method and system for detecting discharge of fault insulator |
-
2021
- 2021-12-13 CN CN202111518571.0A patent/CN114264915A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101576600A (en) * | 2009-05-15 | 2009-11-11 | 重庆大学 | Self-walking underground cable failure detection intelligent instrument |
| CN201527462U (en) * | 2009-09-10 | 2010-07-14 | 国网电力科学研究院 | Compact type artificial climate box for ultraviolet detection |
| CN103091603A (en) * | 2013-01-14 | 2013-05-08 | 华北电力大学 | Breakdown intelligent classification and positioning method of electric transmission line |
| CN104155574A (en) * | 2014-07-31 | 2014-11-19 | 国网湖北省电力公司武汉供电公司 | Power distribution network fault classification method based on adaptive neuro-fuzzy inference system |
| CN105891708A (en) * | 2016-05-11 | 2016-08-24 | 四川大学 | ANFIS-based condition assessment for high-voltage circuit breaker operating mechanism |
| CN111833312A (en) * | 2020-06-22 | 2020-10-27 | 国网江苏省电力有限公司电力科学研究院 | Ultraviolet image diagnosis method and system for detecting discharge of fault insulator |
Non-Patent Citations (3)
| Title |
|---|
| 曹培, 徐鹏, 高凯等: "基于边缘计算的电缆接头运行状态智能传感与监测", 高压电器, vol. 56, no. 9, pages 26 - 32 * |
| 李和明;王胜辉;律方成;刘云鹏;陈磊;魏金祥;: "基于放电紫外成像参量的绝缘子污秽状态评估", 电工技术学报, no. 12, pages 237 - 240 * |
| 金立军;艾建勇;段伟: "基于多光源图像的接触网外绝缘健康状态评估", 高电压技术, vol. 42, no. 11, pages 161 - 169 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115684835A (en) * | 2022-10-28 | 2023-02-03 | 贵州电网有限责任公司信息中心 | Power distribution network fault diagnosis method, device, equipment and storage medium |
| CN116973671A (en) * | 2023-09-22 | 2023-10-31 | 江苏大圆电子科技有限公司 | Aging early warning method and system for cable |
| CN116973671B (en) * | 2023-09-22 | 2023-12-01 | 江苏大圆电子科技有限公司 | Aging early warning method and system for cable |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110598736B (en) | Power equipment infrared image fault positioning, identifying and predicting method | |
| CN111209434B (en) | Substation equipment inspection system and method based on multi-source heterogeneous data fusion | |
| CN111242123B (en) | Power equipment fault diagnosis method based on infrared image | |
| CN114264915A (en) | Power distribution network cable joint operation condition assessment early warning device and method | |
| CN110829417B (en) | Electric power system transient stability prediction method based on LSTM double-structure model | |
| CN110017901B (en) | Infrared spectrum diagnosis method for operating state of electrical equipment in power system | |
| CN110411580B (en) | Diagnosis method and diagnosis system for heating defects of electric equipment | |
| CN109284938A (en) | A kind of comprehensive estimation method and device of power cable line state | |
| CN113344475B (en) | Transformer bushing defect identification method and system based on sequence modal decomposition | |
| CN114359695A (en) | Insulator breakage identification method based on uncertainty estimation | |
| CN110766313A (en) | Cable tunnel comprehensive state evaluation method based on operation and maintenance system | |
| CN115131747A (en) | Knowledge distillation-based power transmission channel engineering vehicle target detection method and system | |
| CN111884236A (en) | Intelligent transient stability evaluation system for power grid | |
| CN116738355A (en) | Electric power multi-mode feature level fusion method considering data difference | |
| CN116338545A (en) | Method, system, equipment and medium for identifying metering error state of current transformer | |
| JP3234742B2 (en) | Method and apparatus for estimating remaining life of electronic component | |
| CN113468823B (en) | Optical module damage detection method and system based on machine learning | |
| CN112183745A (en) | High-voltage cable partial discharge mode identification method based on particle swarm algorithm and DBN | |
| CN111429028A (en) | Power transmission line icing disaster risk assessment method suitable for mountainous terrain | |
| Azari et al. | Intelligent fault detection in power distribution systems using thermos-grams by ensemble classifiers | |
| CN117491357B (en) | Quality monitoring method and system for paint | |
| CN112700425B (en) | Determination method for X-ray image quality of power equipment | |
| CN116091501B (en) | Method, device, equipment and medium for identifying partial discharge type of high-voltage electrical equipment | |
| CN117496223A (en) | Light insulator defect detection method and device based on deep learning | |
| CN116305705A (en) | Imaging system environment adaptability assessment method based on infrared simulation |
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 |