CN113222036B - Automatic defect identification method and device for high-voltage cable grounding system - Google Patents
Automatic defect identification method and device for high-voltage cable grounding system Download PDFInfo
- Publication number
- CN113222036B CN113222036B CN202110559003.9A CN202110559003A CN113222036B CN 113222036 B CN113222036 B CN 113222036B CN 202110559003 A CN202110559003 A CN 202110559003A CN 113222036 B CN113222036 B CN 113222036B
- Authority
- CN
- China
- Prior art keywords
- grounding system
- voltage cable
- model
- defect
- sheath
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 230000007547 defect Effects 0.000 title claims abstract description 101
- 238000000034 method Methods 0.000 title claims abstract description 50
- 238000012549 training Methods 0.000 claims abstract description 41
- 238000012360 testing method Methods 0.000 claims abstract description 36
- 230000008569 process Effects 0.000 claims abstract description 11
- 238000010276 construction Methods 0.000 claims abstract description 5
- 239000013598 vector Substances 0.000 claims description 43
- 230000006870 function Effects 0.000 claims description 32
- 238000007477 logistic regression Methods 0.000 claims description 14
- 238000010606 normalization Methods 0.000 claims description 14
- 238000012417 linear regression Methods 0.000 claims description 13
- 238000012545 processing Methods 0.000 claims description 11
- 230000002159 abnormal effect Effects 0.000 claims description 5
- 230000000007 visual effect Effects 0.000 claims description 2
- 238000003745 diagnosis Methods 0.000 abstract description 9
- 230000001419 dependent effect Effects 0.000 abstract description 2
- 239000010410 layer Substances 0.000 description 13
- 230000005540 biological transmission Effects 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 238000012423 maintenance Methods 0.000 description 4
- 239000002184 metal Substances 0.000 description 4
- 230000008878 coupling Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 238000007689 inspection Methods 0.000 description 3
- 230000015556 catabolic process Effects 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 230000001360 synchronised effect Effects 0.000 description 2
- 239000004593 Epoxy Substances 0.000 description 1
- 206010063385 Intellectualisation Diseases 0.000 description 1
- 238000012512 characterization method Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000012067 mathematical method Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 239000011241 protective layer Substances 0.000 description 1
- 230000001012 protector Effects 0.000 description 1
- 238000011084 recovery Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000017105 transposition Effects 0.000 description 1
- 238000012800 visualization Methods 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/08—Locating faults in cables, transmission lines, or networks
- G01R31/081—Locating faults in cables, transmission lines, or networks according to type of conductors
- G01R31/083—Locating faults in cables, transmission lines, or networks according to type of conductors in cables, e.g. underground
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/08—Locating faults in cables, transmission lines, or networks
- G01R31/088—Aspects of digital computing
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/50—Testing of electric apparatus, lines, cables or components for short-circuits, continuity, leakage current or incorrect line connections
- G01R31/52—Testing for short-circuits, leakage current or ground faults
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Business, Economics & Management (AREA)
- Evolutionary Biology (AREA)
- Economics (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Public Health (AREA)
- Probability & Statistics with Applications (AREA)
- Water Supply & Treatment (AREA)
- General Health & Medical Sciences (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Mathematical Physics (AREA)
- Testing Relating To Insulation (AREA)
Abstract
The application discloses a defect automatic identification method and device of a high-voltage cable grounding system, comprising the following steps: collecting a plurality of cable line sheath current construction sample data; dividing sample data into a training set and a testing set; adopting a training set to train a constructed automatic defect identification model of the high-voltage cable grounding system; inputting a test sample in the test set into a trained model, comparing a state number output by the model with an actual state number of the test sample, and judging the performance of the automatic defect identification model; acquiring the current of the cable line protection layer to be tested; and inputting the cable line protection layer current to be tested into the trained defect automatic identification model to obtain a state number identification result corresponding to the cable line protection layer current to be tested. The defect identification diagnosis process of the existing high-voltage cable grounding system is highly dependent on experience of patrol personnel, and the technical problem of automatic identification of defects of the cable grounding system is not achieved.
Description
Technical Field
The application relates to the technical field of defect identification of high-voltage cable grounding systems, in particular to an automatic defect identification method and device for a high-voltage cable grounding system.
Background
The high-voltage power cable not only meets the high-capacity power transmission requirement of the urban power grid, but also avoids the damage to the original urban appearance, and is widely applied to the urban power transmission system. Although the probability of the occurrence of the faults of the cable line is far lower than that of the traditional urban overhead transmission line, the positioning difficulty is high and the power supply recovery time is long after the faults occur because the power cable is laid in the channel, so that the social and economic effects are often greatly influenced.
The grounding system is an important component of the cabling and provides a diversion channel in the event of a cable failure. Engineering experience shows that the high-voltage cable line often has the phenomenon of increasing the sheath current before failure occurs. The state that the high-voltage cable is abnormal but does not develop to force the line to stop is the golden period for the operation and maintenance personnel to check and eliminate the hidden trouble. The sheath current can reflect the running state of the cable to a certain extent, and can be conveniently obtained under the condition that the line is not powered off, so that the sheath current becomes one of the state quantities focused by operation and maintenance personnel in the inspection process.
The state of the current high-voltage cable grounding system is judged mainly by comparing the acquired sheath current amplitude with relevant standards by operation and maintenance personnel. The method has the advantages that human factors are large, no connection is established between the current characteristics of the protective layer and the defects of the cable, and the identification accuracy of related standards is insufficient to meet the requirements of daily operation and maintenance inspection. The sheath current diagnosis scheme aiming at the cable line is urgently needed to realize intellectualization.
Disclosure of Invention
The application provides a defect automatic identification method and device for a high-voltage cable grounding system, which solve the problems that the defect identification and diagnosis process of the existing high-voltage cable grounding system highly depends on experience of inspection personnel, and the automatic identification of the defects of the cable grounding system is not realized yet.
In view of this, the first aspect of the present application provides a method for automatically identifying defects of a high-voltage cable grounding system, the method comprising:
collecting a plurality of cable line sheath current construction sample data;
dividing the sample data into a training set and a testing set;
training the constructed automatic defect identification model of the high-voltage cable grounding system by adopting the training set;
inputting the test sample in the test set into a trained high-voltage cable grounding system defect automatic identification model, comparing a state number output by the high-voltage cable grounding system defect automatic identification model with an actual state number of the test sample, and judging the performance of the high-voltage cable grounding system defect automatic identification model;
acquiring the current of the cable line protection layer to be tested;
inputting the cable line protection layer current to be tested into the trained automatic identification model of the high-voltage cable grounding system defect, and obtaining a state number identification result corresponding to the cable line protection layer current to be tested.
Optionally, the collecting a plurality of cable line sheath current configuration sample data includes:
acquiring sheath currents of a plurality of sheath segments of a grounding system, acquiring amplitude ratio and phase angle difference between any two sheath currents, and forming a sheath current feature vector by the acquired amplitude ratio and phase angle difference;
and adding the cable line grounding system state number corresponding to the sheath current of the grounding system into the sheath current characteristic vector to form sample data.
Optionally, the automatic identification model for the defects of the high-voltage cable grounding system includes:
a model input interface for providing an input port for the sample data;
a polynomial feature constructor for extending the input sample data into a high-dimensional feature vector;
the data normalization processor is used for carrying out scaling processing on the high-dimensional feature vector, so that each feature value in the processed high-dimensional feature vector is subjected to standard normal distribution;
the logistic regression classifier is used for calculating the high-dimensional feature vector after the scaling treatment according to a preset first formula to obtain an output result of a state number;
the multi-classification result decision device is used for taking the state number with the largest occurrence number in the output results of the multiple logistic regression classifiers as a final output result;
and the model output interface is used for being connected with external display equipment to provide a visual result of the final output result.
Optionally, the calculating the scaled high-dimensional feature vector according to a preset first formula to obtain an output result of the state number includes:
multiplying each element of the high-dimensional feature vector by a weight, and representing the output value of the linear regression model in the form of accumulated summation as follows:
converting the output result of the linear regression model into an interval 0-1 to be expressed in the form of probability:
when the calculated probability is greater than or equal to 0.5, predicting that the grounding system is in a normal state, outputting 1 by the model, and outputting the state number corresponding to the training sample; if the calculated probability is smaller than 0.5, the model predicts that the state of the grounding system is abnormal, and the output value is 0; wherein y is w The output value of the linear regression model; n is the characteristic number of the grounding current sample after normalization treatment; x is x i The i element in the eigenvector is the i element in the eigenvector of the ground current sample after normalization processing; w (w) i The weight corresponding to the ith element; y is h The value is output for the Sigmoid function.
Optionally, the training the constructed automatic defect identification model of the high-voltage cable grounding system by using the training set includes:
the difference between the class predicted by the characteristics of the training set and the real class of the sample is represented by a loss function L, and when the loss function L takes the minimum value, the automatic identification model of the high-voltage cable grounding system defect completes parameter adjustment;
the loss function is:
wherein m is the number of samples, y (j) State number corresponding to the jth sample in the sample data set, y h (j) The value is output for the Sigmoid function.
A second aspect of the present application provides an automatic defect identification device for a high voltage cable grounding system, the device comprising:
the sample data acquisition unit is used for acquiring a plurality of cable line sheath currents to construct sample data;
the dividing unit is used for dividing the sample data into a training set and a testing set;
the training unit is used for training the constructed automatic defect identification model of the high-voltage cable grounding system by adopting the training set;
the comparison unit is used for inputting the test samples in the test set into the trained high-voltage cable grounding system defect automatic identification model, comparing the state number output by the high-voltage cable grounding system defect automatic identification model with the actual state number of the test samples, and judging the performance of the high-voltage cable grounding system defect automatic identification model;
the sample acquisition unit to be detected is used for acquiring the sheath current of the cable circuit to be detected;
the identification unit is used for inputting the cable line protection layer current to be tested into the trained automatic identification model of the high-voltage cable grounding system defect, and obtaining a state number identification result corresponding to the cable line protection layer current to be tested.
Optionally, the sample data obtaining unit is specifically configured to obtain sheath currents of a plurality of sheath segments of the grounding system, obtain an amplitude ratio and a phase angle difference between any two sheath currents, and form a sheath current feature vector by the obtained plurality of amplitude ratios and phase angle differences;
and adding the cable line grounding system state number corresponding to the sheath current of the grounding system into the sheath current characteristic vector to form sample data.
Optionally, the training unit is specifically configured to represent a difference between a class predicted by the training set feature and a real class of the sample by using a loss function L, and when the loss function L takes a minimum value, the automatic identification model of the high-voltage cable grounding system defect completes parameter adjustment;
the loss function is:
wherein m is the number of samples, y (j) State number corresponding to the jth sample in the sample data set, y h (j) The value is output for the Sigmoid function.
A third aspect of the present application provides an automatic defect identification device for a high voltage cable grounding system, the device comprising a processor and a memory:
the memory is used for storing program codes and transmitting the program codes to the processor;
the processor is configured to execute the steps of the method for automatically identifying defects of the high-voltage cable grounding system according to the first aspect according to the instructions in the program code.
A fourth aspect of the present application provides a computer readable storage medium for storing program code for performing the method of the first aspect described above.
From the above technical scheme, the application has the following advantages:
in the embodiment of the application, a defect automatic identification method of a high-voltage cable grounding system is provided, and a plurality of cable line protection layer current construction sample data are collected; dividing sample data into a training set and a testing set; adopting a training set to train a constructed automatic defect identification model of the high-voltage cable grounding system; inputting a test sample in the test set into a trained high-voltage cable grounding system defect automatic identification model, comparing a state number output by the high-voltage cable grounding system defect automatic identification model with an actual state number of the test sample, and judging the performance of the high-voltage cable grounding system defect automatic identification model; acquiring the current of the cable line protection layer to be tested; and inputting the cable line protection layer current to be tested into the trained defect automatic identification model to obtain a state number identification result corresponding to the cable line protection layer current to be tested.
According to the method, the current of a plurality of small sections in the high-voltage cable grounding system is extracted, namely the defects of the high-voltage cable grounding system are classified in multiple sections, meanwhile, the obtained multiple sections of sheath currents are extracted according to a preset mode, and the constructed current characteristic vectors and corresponding state numbers are input into the constructed automatic identification model to obtain the automatic identification model of the defects of the high-voltage cable grounding system, so that the defects of the high-voltage cable grounding system are automatically identified. Compared with the existing diagnosis method, the method can effectively reduce the dependence of the diagnosis result on the manual experience, and simultaneously improve the efficiency and accuracy of the state diagnosis of the cable grounding system.
Drawings
FIG. 1 is a method flow diagram of one embodiment of a method for automatically identifying defects in a high voltage cable grounding system according to the present application;
FIG. 2 is a device block diagram of one embodiment of an automatic defect identification device for a high voltage cable grounding system of the present application;
FIG. 3 is a schematic diagram illustrating current collection of the sheath after cable line division in an embodiment of the present application;
fig. 4 is a schematic structural diagram of an automatic defect identification model of a high-voltage cable grounding system according to an embodiment of the present application.
Detailed Description
In order to make the present application solution better understood by those skilled in the art, the following description will clearly and completely describe the technical solution in the embodiments of the present application with reference to the accompanying drawings in the embodiments of the present application, and it is apparent that the described embodiments are only some embodiments of the present application, not all embodiments. All other embodiments, which can be made by one of ordinary skill in the art without undue burden from the present disclosure, are within the scope of the present disclosure.
Fig. 1 is a flowchart of a method for automatically identifying defects of a high-voltage cable grounding system according to an embodiment of the present application, as shown in fig. 1, where fig. 1 includes:
101. collecting a plurality of cable line sheath current construction sample data;
it should be noted that, the present application may collect the sheath current of a plurality of high voltage cable grounding systems to form sample data for training the subsequent model.
In a specific embodiment, the collection of sample data comprises:
1011. acquiring sheath currents of a plurality of sheath segments of a grounding system, acquiring amplitude ratio and phase angle difference between any two sheath currents, and forming a sheath current feature vector by the acquired amplitude ratio and phase angle difference;
it should be noted that, this application can divide into a plurality of cross-connection ground main sections (each main section corresponds one sample data) with high-voltage cable ground system, this application can divide into a plurality of minor segments with each main section in order to discern the trouble in different regions in the main section, the sheath electric current of every minor segment of acquisition (the sheath electric current of every minor segment is directly related with the trouble that this minor segment region appears), consequently, the sheath electric current of every sheath minor segment after can acquire the division to calculate amplitude ratio and phase angle difference between any two minor segment sheath electric currents, constitute sheath electric current feature vector by a plurality of amplitude ratios and phase angle differences that obtain.
The process for obtaining the sheath current characteristic vector specifically comprises the following steps:
because the long-distance cable transmission line generally adopts a cross-connection grounding mode. When the circuit is designed and planned, the high-voltage cable grounding system is generally divided into a plurality of cross-connection main sections according to the length of the circuit, and the grounding system of each main section is composed of 9 small sections (A1, A2 … C3) of the cable metal sheath, 12 intermediate connectors (JA 0, JB0 … JC 3), 2 direct grounding boxes (GB 1, GB 2) at the head and the tail and 2 cross-connection boxes (CB 1, CB 2) at the middle, as shown in FIG. 3. The two ends of each small section of the cable metal sheath are connected to the cross-connecting box or the direct grounding box in the joint through connecting wires, the transposition connection of the cable metal sheath is realized in the cross-connecting box, and the grounding connection of the cable metal sheath is realized in the direct grounding box.
Common defects of the grounding system of the cross interconnection high-voltage cable include open circuit of the sheath loop, water inlet of the cross interconnection box, breakdown of a joint epoxy prefabricated member, damage of an outer sheath, breakdown of a protector and the like, and the common defects can be divided into the large categories of open circuit of the sheath loop, newly added branches among the sheath loops and newly added pairs of ground supports 3 of the sheath loop according to circuit topology, as shown in table 1. The 3 major grounding system defects can be further divided into 27 minor categories according to the position of the defects, and the 27 minor categories are marked as '0' to '3Q' and are numbered in total.
Table 1 classification of high voltage cable grounding system status
After the defect is divided, in order to obtain the current states of different positions in the main section of the cable grounding system, the application can select the small section A1, B1 and C1 end sheath current I as shown in FIG. 3 m1 、I m2 、I m3 With the head-end sheath current I of A3, B3 and C3 small sections m4 、I m5 、I m6 Is a characteristic state quantity. I is as follows m1 As a basic value, a 3-path synchronous collector is adopted to collect I m1 And I m2 Amplitude ratio x of (2) 1 And phase angle difference x 5 Acquisition I m1 And I m3 Amplitude ratio x of (2) 2 And phase angle difference x 6 . I is as follows m4 As a basic value, a 3-path synchronous collector is adopted to collect I m4 And I m5 Amplitude ratio x of (2) 3 And phase angle difference x 7 Acquisition I m4 And I m6 Amplitude ratio x of (2) 4 And phase angle difference x 8 . Thus forming the 8-dimensional eigenvector [ x ] of the sheath current 1 、x 2 、x 3 、x 4 、x 5 、x 6 、x 7 、x 8 ]。
1012. And adding the cable line grounding system state number corresponding to the sheath current of the grounding system into the sheath current characteristic vector to form sample data.
Note that, the state number of the cable line grounding system to which the sample data collected by each main section belongs is denoted as y, and the state number corresponds to the fault type of the main section. The sample data set is composed of a multidimensional vector representing the current characteristics of the sheath and a state number representing the grounding system. Specifically, if the sheath current feature vector is an 8-dimensional vector, each sample data is 1 9-dimensional array [ x ] 1 、x 2 、x 3 、x 4 、x 5 、x 6 、x 7 、x 8 、y]The collected m samples form an mx9-dimensional sample dataset.
102. Dividing sample data into a training set and a testing set;
it should be noted that, the present application may randomly divide the sample data set into two major classes, namely, a training set and a testing set according to a ratio of 4:1.
103. Adopting a training set to train a constructed automatic defect identification model of the high-voltage cable grounding system;
the automatic identification model for the defects of the high-voltage cable grounding system is composed of 6 parts and sequentially comprises a model input interface, a polynomial characteristic constructor, a data normalization processor, a logistic regression classifier, a multi-classification result decision maker and a model output interface.
In a specific embodiment, the high voltage cable grounding system defect automatic identification model is shown in fig. 4. The application can take functions in the scikit-learn database as examples to construct an automatic defect identification model, which comprises the following steps:
(1) Model input interface
The model input interface is used to provide an input port for feature vectors in the upper sample set.
(2) Polynomial feature constructor
The application can use a polynomial feature constructor to extend the 8-maintenance-layer current feature vector into a higher-order high-dimensional feature vector. According to the method and the device, the complexity of the model is improved in a mode of constructing polynomial characteristics, so that the model can mine more implicit relations between samples.
(3) Data normalization processor
The data normalization processor can perform scaling processing on the data output by the polynomial feature constructor, and each feature value of the processed feature vector is subjected to standard normal distribution so as to prevent the influence of larger values of certain types of feature values on other feature values with smaller values due to different physical quantities, different measurement units and the like.
(4) Logistic regression classifier
The logistic regression classifier consists of a linear regression model and Sigmoid functions. The linear regression model is used for representing the degree of the normal state of the grounding system in the form of accumulated summation by multiplying each element of the current characteristic vector by a weight, and specifically comprises the following steps:
the Sigmoid function functions to transform the result of the linear regression model characterization into the interval (0, 1) to be presented in the form of probability. When the calculated probability is greater than or equal to 0.5, the model predicts that the grounding system is in a normal state, and the model outputs 1. If the calculated probability is smaller than 0.5, the model predicts that the state of the grounding system is abnormal, the output value is 0, and the method specifically comprises the following steps:
wherein: y is w The output value of the linear regression model; n is the characteristic number of the grounding current sample after normalization treatment; x is x i The i element in the eigenvector is the i element in the eigenvector of the ground current sample after normalization processing; w (w) i The weight corresponding to the i-th element. y is h Output value for Sigmoid function, y 0 And outputting values for the classification logistic regression model.
Since the present model involves 27 defect classes from "0" to "3Q", one logistic regression classifier can be constructed between every two state numbers (classes), and then 351 logistic regression classifiers can be constructed in total.
(5) Multi-classification result decision maker
For the results generated by the 351 logistic regression classifiers, the multi-classification result decision maker selects the state number y with the most occurrence in the 351 results as the final output result.
(6) Model output interface
For connection to an external display device, providing a visualization of the output results.
The specific training process of the automatic defect identification model of the high-voltage cable grounding system comprises the following steps of;
after model creation, the model parameters need to be adjusted according to the sample set. For a logistic regression model, the parameters of the model are weights corresponding to the feature quantities. The difference between the model predicted by the sample set features and the sample true category is represented by a loss function L, and the loss function is as follows:
wherein m is the number of samples, y (j) State number corresponding to the jth sample in the sample data set, y h (j) The value is output for the Sigmoid function. When the loss function L takes a minimum value, the model completes the parameter adjustment. The parameter value of the model when the L takes the minimum value can be obtained by utilizing the mathematical methods such as Newton method and the like. In order to prevent the model from being over fitted, an L2 regular term is selected to correct the loss function during parameter adjustment.
104. Inputting a test sample in the test set into a trained high-voltage cable grounding system defect automatic identification model, comparing a state number output by the high-voltage cable grounding system defect automatic identification model with an actual state number of the test sample, and judging the performance of the high-voltage cable grounding system defect automatic identification model;
it should be noted that, the present application may adopt a test sample in a test set to input to a trained automatic identification model of the high-voltage cable grounding system defect, compare a state number output by the automatic identification model of the high-voltage cable grounding system defect with an actual state number of the test sample, and determine performance of the automatic identification model of the high-voltage cable grounding system defect.
105. Acquiring the current of the cable line protection layer to be tested;
it should be noted that, for the trained automatic defect identification model of the high-voltage cable grounding system, the application can adopt current data collected on site or current data of a high-voltage cable grounding system collected remotely, and input the current data into the trained automatic defect identification model of the high-voltage cable grounding system to output the defect type, so as to realize automatic identification of the defect type of the high-voltage cable grounding system.
106. And inputting the cable line sheath current to be tested into a trained automatic identification model of the high-voltage cable grounding system defect, and obtaining a state number identification result corresponding to the cable line sheath current to be tested.
According to the method, the current of a plurality of small sections in the high-voltage cable grounding system is extracted, namely the defects of the high-voltage cable grounding system are classified in multiple sections, meanwhile, the obtained multiple sections of sheath currents are extracted according to a preset mode, and the constructed current characteristic vectors and corresponding state numbers are input into the constructed automatic identification model to obtain the automatic identification model of the defects of the high-voltage cable grounding system, so that the defects of the high-voltage cable grounding system are automatically identified. Compared with the existing diagnosis method, the method can effectively reduce the dependence of the diagnosis result on the manual experience, and simultaneously improve the efficiency and accuracy of the state diagnosis of the cable grounding system.
The application also provides an embodiment of a defect automatic identification device of a high-voltage cable grounding system, as shown in fig. 2, fig. 2 includes:
a sample data acquisition unit 201, configured to acquire a plurality of cable-line sheath currents to construct sample data;
a dividing unit 202 for dividing the sample data into a training set and a test set;
the training unit 203 is configured to train the constructed automatic defect identification model of the high-voltage cable grounding system by adopting a training set;
the comparison unit 204 is configured to input a test sample in the test set to the trained automatic identification model of the high-voltage cable grounding system defect, compare a state number output by the automatic identification model of the high-voltage cable grounding system defect with an actual state number of the test sample, and determine performance of the automatic identification model of the high-voltage cable grounding system defect;
the sample to be tested obtaining unit 205 is configured to obtain a sheath current of the cable circuit to be tested;
the identifying unit 206 is configured to input the current of the cable protection layer to be tested into the trained automatic defect identifying model of the high-voltage cable grounding system, and obtain a status number identifying result corresponding to the current of the cable protection layer to be tested.
In a specific embodiment, the sample data obtaining unit 201 is specifically configured to obtain sheath currents of a plurality of sheath segments of the grounding system, obtain an amplitude ratio and a phase angle difference between any two sheath currents, and form a sheath current feature vector from the obtained plurality of amplitude ratios and phase angle differences; and adding the cable line grounding system state number corresponding to the sheath current of the grounding system into the sheath current characteristic vector to form sample data.
In a specific embodiment, the training unit 203 is specifically configured to represent a difference between a class predicted by a feature of the training set and a true class of the sample by using a loss function L, and when the loss function L takes a minimum value, the automatic identification model of the defect of the high-voltage cable grounding system completes parameter adjustment;
the loss function is:
wherein m is the number of samples, y (j) State number corresponding to the jth sample in the sample data set, y h (j) The value is output for the Sigmoid function.
It will be clear to those skilled in the art that, for convenience and brevity of description, specific working procedures of the above-described systems, apparatuses and units may refer to corresponding procedures in the foregoing method embodiments, which are not repeated herein.
The terms "first," "second," "third," "fourth," and the like in the description of the present application and in the above-described figures, are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that embodiments of the present application described herein may be implemented in sequences other than those illustrated or otherwise described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
It should be understood that in this application, "at least one" means one or more, and "a plurality" means two or more. "and/or" for describing the association relationship of the association object, the representation may have three relationships, for example, "a and/or B" may represent: only a, only B and both a and B are present, wherein a, B may be singular or plural. The character "/" generally indicates that the context-dependent object is an "or" relationship. "at least one of" or the like means any combination of these items, including any combination of single item(s) or plural items(s). For example, at least one (one) of a, b or c may represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", wherein a, b, c may be single or plural.
In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods may be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative, e.g., the division of the units is merely a logical function division, and there may be additional divisions when actually implemented, e.g., multiple units or components may be combined or integrated into another system, or some features may be omitted or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be an indirect coupling or communication connection via some interfaces, devices or units, which may be in electrical, mechanical or other form.
The units described as separate units may or may not be physically separate, and units shown as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
In addition, each functional unit in each embodiment of the present application may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units may be implemented in hardware or in software functional units.
The integrated units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present application may be embodied in essence or a part contributing to the prior art or all or part of the technical solution in the form of a software product stored in a storage medium, including several instructions to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in the embodiments of the present application. And the aforementioned storage medium includes: u disk, mobile hard disk, read-Only Memory (ROM), random access Memory (RandomAccess Memory, RAM), magnetic disk or optical disk, etc.
The above embodiments are merely for illustrating the technical solution of the present application, and not for limiting the same; although the present application has been described in detail with reference to the foregoing embodiments, it should be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the corresponding technical solutions.
Claims (7)
1. The automatic defect identification method for the high-voltage cable grounding system is characterized by comprising the following steps of:
collecting a plurality of cable line sheath current construction sample data;
dividing the sample data into a training set and a testing set;
training the constructed automatic defect identification model of the high-voltage cable grounding system by adopting the training set;
inputting the test sample in the test set into a trained high-voltage cable grounding system defect automatic identification model, comparing a state number output by the high-voltage cable grounding system defect automatic identification model with an actual state number of the test sample, and judging the performance of the high-voltage cable grounding system defect automatic identification model;
acquiring the current of the cable line protection layer to be tested;
inputting the cable line protection layer current to be tested into the trained automatic identification model of the high-voltage cable grounding system defect, and obtaining a state number identification result corresponding to the cable line protection layer current to be tested;
wherein, gather a plurality of cable line sheath electric currents and construct sample data, include:
acquiring sheath currents of a plurality of sheath segments of a grounding system, acquiring amplitude ratio and phase angle difference between any two sheath currents, and forming a sheath current feature vector by the acquired amplitude ratio and phase angle difference;
adding a cable line grounding system state number corresponding to the sheath current of the grounding system into the sheath current characteristic vector to form sample data;
the automatic defect identification model of the high-voltage cable grounding system comprises the following steps:
a polynomial feature constructor for extending the input sample data into a high-dimensional feature vector;
the data normalization processor is used for carrying out scaling processing on the high-dimensional feature vector, so that each feature value in the processed high-dimensional feature vector is subjected to standard normal distribution;
the logistic regression classifier is used for calculating the high-dimensional feature vector after the scaling treatment according to a preset first formula to obtain an output result of a state number;
the calculating the high-dimensional feature vector after the scaling treatment according to a preset first formula to obtain an output result of a state number comprises the following steps:
multiplying each element of the high-dimensional feature vector by a weight, and representing the output value of the linear regression model in the form of accumulated summation as follows:
;
converting the output result of the linear regression model into an interval 0-1 to be expressed in the form of probability:
;
when the calculated probability is greater than or equal to 0.5, predicting that the grounding system is in a normal state, outputting 1 by the model, and outputting the state number corresponding to the training sample; if the calculated probability is smaller than 0.5, the model predicts that the state of the grounding system is abnormal, and the output value is 0; in the method, in the process of the invention,y w the output value of the linear regression model;nthe characteristic number of the grounding current sample after normalization processing;x i the first characteristic vector of the ground current sample after normalization processingiAn element;w i is the firstiThe weight corresponding to each element;y h the value is output for the Sigmoid function.
2. The automatic defect recognition method for a high-voltage cable grounding system according to claim 1, wherein the automatic defect recognition model for a high-voltage cable grounding system further comprises:
a model input interface for providing an input port for the sample data;
the multi-classification result decision device is used for taking the state number with the largest occurrence number in the output results of the multiple logistic regression classifiers as a final output result;
and the model output interface is used for being connected with external display equipment to provide a visual result of the final output result.
3. The automatic defect identification method for the high-voltage cable grounding system according to claim 1, wherein the training the constructed automatic defect identification model for the high-voltage cable grounding system by using the training set comprises the following steps:
using a loss function to predict the difference between the class of the training set characteristics and the true class of the sampleLIndicating when the loss functionLWhen the minimum value is taken, the automatic identification model of the high-voltage cable grounding system defect completes parameter adjustment;
the loss function is:
;
in the method, in the process of the invention,min order to obtain the number of samples,y j() is the firstjThe corresponding state numbers of the individual samples in the sample dataset,y h j() the value is output for the Sigmoid function.
4. The utility model provides a defect automatic identification equipment of high voltage cable earth connection system which characterized in that includes:
the sample data acquisition unit is used for acquiring a plurality of cable line sheath currents to construct sample data;
the dividing unit is used for dividing the sample data into a training set and a testing set;
the training unit is used for training the constructed automatic defect identification model of the high-voltage cable grounding system by adopting the training set;
the comparison unit is used for inputting the test samples in the test set into the trained high-voltage cable grounding system defect automatic identification model, comparing the state number output by the high-voltage cable grounding system defect automatic identification model with the actual state number of the test samples, and judging the performance of the high-voltage cable grounding system defect automatic identification model;
the sample acquisition unit to be detected is used for acquiring the sheath current of the cable circuit to be detected;
the identification unit is used for inputting the cable line protection layer current to be tested into the trained automatic identification model of the high-voltage cable grounding system defect to obtain a state number identification result corresponding to the cable line protection layer current to be tested;
the sample data acquisition unit is specifically configured to:
acquiring sheath currents of a plurality of sheath segments of a grounding system, acquiring amplitude ratio and phase angle difference between any two sheath currents, and forming a sheath current feature vector by the acquired amplitude ratio and phase angle difference;
adding a cable line grounding system state number corresponding to the sheath current of the grounding system into the sheath current characteristic vector to form sample data;
the automatic defect identification model of the high-voltage cable grounding system comprises the following steps:
a polynomial feature constructor for extending the input sample data into a high-dimensional feature vector;
the data normalization processor is used for carrying out scaling processing on the high-dimensional feature vector, so that each feature value in the processed high-dimensional feature vector is subjected to standard normal distribution;
the logistic regression classifier is used for calculating the high-dimensional feature vector after the scaling treatment according to a preset first formula to obtain an output result of a state number;
wherein the logistic regression classifier is specifically configured to:
multiplying each element of the high-dimensional feature vector by a weight, and representing the output value of the linear regression model in the form of accumulated summation as follows:
;
converting the output result of the linear regression model into an interval 0-1 to be expressed in the form of probability:
;
when the calculated probability is greater than or equal to 0.5, predicting that the grounding system is in a normal state, outputting 1 by the model, and outputting the state number corresponding to the training sample; if the calculated probability is smaller than 0.5, the model predicts that the state of the grounding system is abnormal, and the output value is 0; in the method, in the process of the invention,y w the output value of the linear regression model;nthe characteristic number of the grounding current sample after normalization processing;x i the first characteristic vector of the ground current sample after normalization processingiAn element;w i is the firstiThe weight corresponding to each element;y h the value is output for the Sigmoid function.
5. The automatic defect recognition device of a high voltage cable grounding system according to claim 4, wherein the training unit is specifically configured to use a loss function for a gap between a class predicted by the training set feature and a sample true classLIndicating when the loss functionLWhen the minimum value is taken, the automatic identification model of the high-voltage cable grounding system defect completes parameter adjustment;
the loss function is:
;
in the method, in the process of the invention,min order to obtain the number of samples,y j() is the firstjThe corresponding state numbers of the individual samples in the sample dataset,y h j() the value is output for the Sigmoid function.
6. An automatic defect identification device for a high voltage cable grounding system, the device comprising a processor and a memory:
the memory is used for storing program codes and transmitting the program codes to the processor;
the processor is configured to execute the method for automatically identifying defects in a high voltage cable grounding system according to any one of claims 1-3 according to instructions in the program code.
7. A computer readable storage medium for storing program code for performing the method of automatic defect identification of a high voltage cable grounding system as in any of claims 1-3.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110559003.9A CN113222036B (en) | 2021-05-21 | 2021-05-21 | Automatic defect identification method and device for high-voltage cable grounding system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110559003.9A CN113222036B (en) | 2021-05-21 | 2021-05-21 | Automatic defect identification method and device for high-voltage cable grounding system |
Publications (2)
Publication Number | Publication Date |
---|---|
CN113222036A CN113222036A (en) | 2021-08-06 |
CN113222036B true CN113222036B (en) | 2024-03-22 |
Family
ID=77097814
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110559003.9A Active CN113222036B (en) | 2021-05-21 | 2021-05-21 | Automatic defect identification method and device for high-voltage cable grounding system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113222036B (en) |
Families Citing this family (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114035118B (en) * | 2021-11-30 | 2024-05-28 | 徐忠林 | Protective layer ground fault detection method, positioning method, detection system and positioning system |
CN114460490A (en) * | 2022-01-17 | 2022-05-10 | 武汉朕泰智能科技有限公司 | Method for diagnosing ground short circuit fault inside high-voltage cable sheath grounding box |
CN115656719B (en) * | 2022-10-26 | 2023-04-07 | 徐忠林 | High-voltage cable sheath defect online diagnosis method based on sheath current transaction |
CN116502051B (en) * | 2023-06-26 | 2023-09-26 | 广东电网有限责任公司珠海供电局 | Submarine cable local defect identification method and device |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2004003578A1 (en) * | 2002-06-26 | 2004-01-08 | Adaptive Regelsysteme Gesellschaft Mbh | Method for determination of a parameter of an electrical network |
CN111257690A (en) * | 2020-02-17 | 2020-06-09 | 广东电网有限责任公司 | Fault diagnosis and positioning method for cross-connection high-voltage cable sheath protector |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US9684026B2 (en) * | 2013-10-08 | 2017-06-20 | Rockwell Automation Technologies, Inc. | Apparatus and method for automatic ground fault location determination in high resistance grounded motor drive system |
-
2021
- 2021-05-21 CN CN202110559003.9A patent/CN113222036B/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2004003578A1 (en) * | 2002-06-26 | 2004-01-08 | Adaptive Regelsysteme Gesellschaft Mbh | Method for determination of a parameter of an electrical network |
CN111257690A (en) * | 2020-02-17 | 2020-06-09 | 广东电网有限责任公司 | Fault diagnosis and positioning method for cross-connection high-voltage cable sheath protector |
Also Published As
Publication number | Publication date |
---|---|
CN113222036A (en) | 2021-08-06 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN113222036B (en) | Automatic defect identification method and device for high-voltage cable grounding system | |
CN108107324B (en) | Power distribution network fault positioning method based on deep convolutional neural network | |
CN107909118B (en) | Power distribution network working condition wave recording classification method based on deep neural network | |
CN107832855B (en) | Line loss multi-source diagnosis method and system based on correlation analysis | |
Decanini et al. | Robust fault diagnosis in power distribution systems based on fuzzy ARTMAP neural network-aided evidence theory | |
CN109409444B (en) | Multivariate power grid fault type discrimination method based on prior probability | |
CN110609200A (en) | Power distribution network earth fault protection method based on fuzzy metric fusion criterion | |
CN110766313A (en) | Cable tunnel comprehensive state evaluation method based on operation and maintenance system | |
CN109936113A (en) | A kind of protection act intelligent diagnosing method and system based on random forests algorithm | |
CN107515339A (en) | A kind of Risk Identification Method and system based on DC current distribution situation | |
CN111654392A (en) | Low-voltage distribution network topology identification method and system based on mutual information | |
CN113657622A (en) | Method, device, terminal and storage medium for fusing multidimensional state data of electrical equipment | |
CN116298684A (en) | Fault studying, judging and positioning system for power distribution network | |
CN111999591B (en) | Method for identifying abnormal state of primary equipment of power distribution network | |
CN107462810A (en) | A kind of fault section location method suitable for active power distribution network | |
Benzerga et al. | Low-voltage network topology and impedance identification using smart meter measurements | |
CN109784777B (en) | Power grid equipment state evaluation method based on time sequence information fragment cloud similarity measurement | |
CN107632237A (en) | A kind of fault line selection method for single-phase-to-ground fault based on wavelet character amount correlation after failure | |
CN112595918A (en) | Low-voltage meter reading fault detection method and device | |
CN115902557A (en) | Switch cabinet fault diagnosis processing method and device and nonvolatile storage medium | |
US20150227639A1 (en) | System data compression system and method thereof | |
CN114062832A (en) | Method and system for identifying short-circuit fault type of power distribution network | |
CN112465358A (en) | Voltage quality classification method and device based on support vector machine | |
CN114662251A (en) | Power distribution network fault positioning method based on deep neural network | |
CN113655344B (en) | Power distribution network fault identification method, device, terminal and medium |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |