CN115983265A - Relay protection defect setting and grading method based on convolutional neural network - Google Patents

Relay protection defect setting and grading method based on convolutional neural network Download PDF

Info

Publication number
CN115983265A
CN115983265A CN202310013026.9A CN202310013026A CN115983265A CN 115983265 A CN115983265 A CN 115983265A CN 202310013026 A CN202310013026 A CN 202310013026A CN 115983265 A CN115983265 A CN 115983265A
Authority
CN
China
Prior art keywords
defect
neural network
relay protection
text
convolutional neural
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
Application number
CN202310013026.9A
Other languages
Chinese (zh)
Inventor
郑少明
董鹏
杨心平
杜鹃
刘丹
宿洪智
陶畅
王书鸿
薛安成
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
North China Grid Co Ltd
Original Assignee
North China Grid Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by North China Grid Co Ltd filed Critical North China Grid Co Ltd
Priority to CN202310013026.9A priority Critical patent/CN115983265A/en
Publication of CN115983265A publication Critical patent/CN115983265A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • YGENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y04INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
    • Y04SSYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
    • Y04S10/00Systems supporting electrical power generation, transmission or distribution
    • Y04S10/50Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications

Landscapes

  • Image Analysis (AREA)

Abstract

The invention relates to the technical field of relay protection device defect analysis, and provides a relay protection device defect grading method based on a convolutional neural network. The method comprises the following steps: the method comprises the steps of preprocessing data based on a certain regional power grid relay protection operation defect record to obtain a defect fixed-level data set, extracting features of a one-dimensional convolution layer opposite quantization text matrix through Markov assumption, transmitting the obtained features into a fully-connected neural network to obtain an identification vectorization text prediction result, and performing preset times of model iteration to convergence through a forward propagation gradient to obtain a target convolution neural network model; and inputting the test set into a target convolutional neural network model based on the selected prediction parameters to obtain a defect classification recognition result. The invention improves the classification accuracy of the grading prediction result of the defect recording text.

Description

Relay protection defect setting and grading method based on convolutional neural network
Technical Field
The invention relates to the technical field of relay protection device defect analysis, in particular to a relay protection device defect grading method based on a convolutional neural network.
Background
In the long-term operation process of the relay protection equipment, a large amount of defect text data are recorded and accumulated through means such as routing inspection, tests and the like. After the text data is stored in the system, the text data is usually used as record data, and a large amount of valuable information contained in the record data cannot be discovered, so that the record data is key data for classifying the defect levels of the equipment.
However, a large amount of classification work of the defect levels of the equipment needs to be finished manually, so that the efficiency is low, the workload is large, and the embarrassment situation that accurate judgment is difficult is often caused for certain sub-health defects with strong ambiguity, and therefore, the classification accuracy is affected.
Disclosure of Invention
In view of this, the invention provides a relay protection defect setting and grading method based on a convolutional neural network, so as to solve the problems that in the prior art, the defect grade classification work needs to be completed manually, the efficiency is low, and the classification accuracy is poor.
The invention provides a relay protection defect grading method based on a convolutional neural network, which comprises the following steps:
s1, preprocessing data based on a certain regional power grid relay protection operation defect record to obtain a defect grading data set, wherein the defect grading data set comprises a training set, a testing set and a verification set;
s2, performing feature extraction on a text matrix subjected to opposite quantization by adopting a one-dimensional convolutional layer through Markov assumption, transmitting the obtained features into a fully-connected neural network to obtain a text prediction result subjected to identification vectorization, and performing model iteration of preset times to convergence through a forward propagation gradient to obtain a target convolutional neural network model;
and S3, inputting the test set into the target convolutional neural network model based on the selected prediction parameters to obtain a defect grading identification result.
Further, the S1 includes:
s11, based on a relay protection operation defect record of a certain regional power grid, removing stop words and irrelevant words by adopting a relay protection defect dictionary, performing vectorization processing to form a vectorized text matrix, and dividing the vectorized text matrix into three defect grades of critical, serious and general;
s12, performing word segmentation processing on the critical, serious and general three defect levels respectively by using a method of combining the jieba word segmentation with a professional dictionary, and dividing the three defect levels into the training set, the testing set and the verification set according to the following ratio of 6.
Further, the training of the model in S2 includes: batch _ size =128, and preset number = 10000.
Further, the S2 includes:
s21, constructing 3 one-dimensional convolution kernels with the length of 2,3,4 according to a Markov hypothesis;
s22, performing convolution operation on the vector text matrix by using the constructed one-dimensional convolution kernel, extracting a characteristic part containing a set amount of information entropy in a text, transmitting the characteristic part into the fully-connected neural network, calculating by using sigmoid as an activation function, and outputting probability distribution of the category to which the vectorized text belongs by using a softmax function to obtain a text prediction result for identifying vectorization;
and S23, based on the recognition vectorization text prediction result, performing model iteration of preset times to convergence by using a cross entropy loss function and through a forward propagation gradient to obtain the target convolutional neural network model.
Further, the probability distribution of the category to which the vectorized text belongs ranges between [0,1] and is 1.
Further, the cross entropy loss function includes the following calculation formula:
Figure BDA0004038894340000021
wherein M represents the number of categories, y ic Representing a symbolic function, taking 1 if the true class of sample i is equal to c, otherwise taking 0 ic Representing the probability that the observed sample i belongs to class c.
Further, the S3 includes:
s31, representing the prediction precision of the target convolutional neural network model by adopting the accuracy, the recall rate and the F1 score;
and S32, inputting the test set into the target convolutional neural network model to obtain a defect classification recognition result.
Further, the obtaining of the model evaluation index includes the following calculation formula:
Figure BDA0004038894340000031
Figure BDA0004038894340000032
Figure BDA0004038894340000033
wherein P represents accuracy, R represents recall, F represents F1 score, TP represents number of correctly ranked defect texts, FP represents number of incorrectly ranked defect texts, and FN represents number of undetected texts.
Compared with the prior art, the invention has the following beneficial effects:
1. the classification method adopts the softmax function to calculate, outputs the probability distribution of the category to which the vectorized text belongs, and improves the classification accuracy of the grading prediction result of the defect record text.
2. According to the method, the characteristic part containing the set amount of information entropy in the text is extracted, the loss function is obtained to be converged to the preset value, the calculation speed is high, and the situation that the text falls into the local optimal solution is avoided.
3. According to the method, a series of model evaluation index parameters of accuracy, recall rate and F1 score are selected, so that the prediction precision of the target convolutional neural network model is guaranteed, and the defects can be timely processed and reported.
Drawings
In order to more clearly illustrate the technical solution of the present invention, the drawings needed for the embodiment or the prior art description will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings without creative efforts.
Fig. 1 is a flowchart of a relay protection defect setting and grading method based on a convolutional neural network provided by the present invention.
Detailed Description
In the following description, for purposes of explanation and not limitation, specific details are set forth, such as particular system structures, techniques, etc. in order to provide a thorough understanding of the embodiments of the invention. It will be apparent, however, to one skilled in the art that the present invention may be practiced in other embodiments that depart from these specific details. In other instances, detailed descriptions of well-known systems, devices, circuits, and methods are omitted so as not to obscure the description of the present invention with unnecessary detail.
The relay protection defect setting and grading method based on the convolutional neural network provided by the invention will be described in detail below with reference to the accompanying drawings.
Fig. 1 is a flowchart of a relay protection defect setting and grading method based on a convolutional neural network provided by the invention.
As shown in fig. 1, the relay protection defect setting and grading method includes:
s1, preprocessing data based on a power grid relay protection operation defect record in a certain area to obtain a defect grading data set, wherein the defect grading data set comprises a training set, a testing set and a verification set;
the S1 comprises:
s11, based on relay protection operation defect records of a power grid in a certain area, adopting a relay protection defect dictionary to remove stop words and irrelevant words, and performing vectorization processing to form a vectorized text matrix, wherein the vectorized text matrix is divided into three defect levels of emergency, serious and general;
s12, performing word segmentation processing on the critical, serious and general three defect levels respectively by using a method of combining jieba word segmentation with a professional dictionary, and dividing the three defect levels into the training set, the testing set and the verification set according to the following steps of 2.
Defect grading dataset information is shown in Table 1:
data set Training set Test set Verification set Total up to
Urgency (Label 0) 972 301 290 1563
Severe (Label 1) 705 238 188 1131
In general (Label 2) 723 239 242 1204
Total up to 2400 778 720 3898
S2, performing feature extraction on a text matrix subjected to opposite quantization by adopting a one-dimensional convolutional layer through Markov assumption, transmitting the obtained features into a fully-connected neural network to obtain a text prediction result subjected to identification vectorization, and performing model iteration of preset times to convergence through a forward propagation gradient to obtain a target convolutional neural network model;
the training of the model in the S2 comprises the following steps: batch _ size =128, and preset number = 10000.
The S2 comprises the following steps:
s21, constructing 3 one-dimensional convolution kernels with the length of 2,3,4 according to a Markov hypothesis;
according to the Markov assumption: the probability of any word appearing is only related to one or a few words appearing before the word, and 3 one-dimensional convolution kernels with the length of 2,3,4 are constructed, namely the influence of 2,3,4 words nearby the word on the one-dimensional convolution kernels is considered.
S22, performing convolution operation on the vector text matrix by using the constructed one-dimensional convolution kernel, extracting a characteristic part containing a set amount of information entropy in a text, transmitting the characteristic part into the fully-connected neural network, calculating by using sigmoid as an activation function, and outputting probability distribution of the category to which the vectorized text belongs by using a softmax function to obtain a text prediction result for identifying vectorization;
specifically, a constructed one-dimensional convolution kernel is utilized to carry out convolution operation on a vector quantization text matrix, a characteristic part containing a set amount of information entropy in a text is extracted, the extracted characteristic part is input into a 5-layer fully-connected neural network, and sigmoid is adopted as an activation function, namely
Figure BDA0004038894340000061
The fully-connected neural network is a convolutional neural network model, and the convolutional neural network model comprises a plurality of one-dimensional convolutions to obtain the characteristic representation of the N-gram in the sentence.
Outputting probability distribution of the category to which the vectorized text belongs through the fully-connected neural network, and then calculating by utilizing a softmax function to convert the probability into the probability distribution with the range of [0,1] and 1, wherein the obtaining of the softmax function comprises the following calculation formula:
Figure BDA0004038894340000062
wherein Z is i Is the output value of the ith node, and C is the number of output nodes, i.e. the number of classified categories.
And S23, based on the recognition vectorization text prediction result, performing model iteration of preset times to convergence by using a cross entropy loss function and through a forward propagation gradient to obtain the target convolutional neural network model.
Calculating model gaps of the prediction classification and the actual classification by using a cross entropy loss function, wherein the obtaining of the cross entropy loss function comprises the following calculation formula:
Figure BDA0004038894340000071
where M is the number of classes, y ic For the sign function, if the true class of sample i is equal to c taken 1, otherwise 0 is taken ic Is the probability that the observation sample i belongs to class c.
And based on a random gradient descent method, taking the cross entropy loss function as a target function, continuously and randomly extracting samples to calculate gradient, and performing gradient descent according to a set step length until the cross entropy loss function converges to a set value to obtain the target convolutional neural network model. The invention adopts the random extraction of the gradient calculation by the random gradient descent method, is not easy to fall into the local optimal solution and has higher calculation speed.
And S3, inputting the test set into the target convolutional neural network model based on the selected prediction parameters to obtain a defect grading identification result.
The S3 comprises the following steps:
s31, representing the prediction precision of the target convolutional neural network model by adopting the accuracy, the recall rate and the F1 score;
and S32, inputting the test set into the target convolutional neural network model to obtain a defect classification recognition result.
And the target convolutional neural network model reflects the identification result of defect grading through the evaluation index of the target convolutional neural network model.
The obtaining of the target convolutional neural network model evaluation index comprises the following calculation formula:
Figure BDA0004038894340000072
Figure BDA0004038894340000073
Figure BDA0004038894340000074
wherein P represents accuracy, R represents recall, F represents F1 score, TP represents number of correctly ranked defect texts, FP represents number of incorrectly ranked defect texts, and FN represents number of undetected texts.
In order to comprehensively evaluate the accuracy of the training model, batch _ size =128 in the training process is set, after one round of training, each batch takes out the optimal model as the basic model of the next round of training, and the following results are obtained through 1 million times of training:
P=0.6833
R=0.7560
F=0.2472
according to the method, the probability distribution of the category to which the vectorized text belongs is output by adopting softmax function calculation, so that the classification accuracy of the recognition vectorized text prediction result is improved; by extracting the characteristic part containing the set amount of information entropy in the text, the loss function is obtained to be converged to a preset value, the calculation speed is high, and the situation of falling into a local optimal solution is avoided; a series of model evaluation parameters of accuracy, recall rate and F1 score are selected, so that the prediction precision of the target convolutional neural network model is guaranteed, and the defects can be timely processed and reported.
All the above optional technical solutions may be combined arbitrarily to form optional embodiments of the present application, and are not described herein again.
It should be understood that, the sequence numbers of the steps in the foregoing embodiments do not imply an execution sequence, and the execution sequence of each process should be determined by functions and internal logic of the process, and should not limit the implementation process of the embodiments of the present invention in any way.
The above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; such modifications and substitutions do not substantially depart from the spirit and scope of the embodiments of the present invention, and are intended to be included within the scope of the present invention.

Claims (8)

1. A relay protection defect setting and grading method based on a convolutional neural network is characterized by comprising the following steps:
s1, preprocessing data based on a certain regional power grid relay protection operation defect record to obtain a defect grading data set, wherein the defect grading data set comprises a training set, a testing set and a verification set;
s2, performing feature extraction on a text matrix subjected to opposite quantization by adopting a one-dimensional convolutional layer through Markov assumption, transmitting the obtained features into a fully-connected neural network to obtain a text prediction result subjected to identification vectorization, and performing model iteration of preset times to convergence through a forward propagation gradient to obtain a target convolutional neural network model;
and S3, inputting the test set into the target convolutional neural network model based on the selected prediction parameters to obtain a defect grading identification result.
2. The relay protection device defect grading method according to claim 1, wherein S1 comprises:
s11, based on a relay protection operation defect record of a certain regional power grid, removing stop words and irrelevant words by adopting a relay protection defect dictionary, performing vectorization processing to form a vectorized text matrix, and dividing the vectorized text matrix into three defect grades of critical, serious and general;
s12, performing word segmentation processing on the critical, serious and general three defect levels respectively by using a method of combining the jieba word segmentation with a professional dictionary, and dividing the three defect levels into the training set, the testing set and the verification set according to the following ratio of 6.
3. The relay protection device defect grading method according to claim 1, wherein the model training in S2 comprises: batch _ size =128, and the preset number = 10000.
4. The relay protection device defect grading method according to claim 1, wherein the S2 comprises:
s21, constructing 3 one-dimensional convolution kernels with the length of 2,3,4 according to the Markov assumption;
s22, performing convolution operation on the vector text matrix by using the constructed one-dimensional convolution kernel, extracting a characteristic part containing a set amount of information entropy in a text, transmitting the characteristic part into the fully-connected neural network, calculating by using sigmoid as an activation function, and outputting probability distribution of the category to which the vectorized text belongs by using a softmax function to obtain a text prediction result for identifying vectorization;
and S23, based on the recognition vectorization text prediction result, performing model iteration of preset times to convergence by using a cross entropy loss function and through a forward propagation gradient to obtain the target convolutional neural network model.
5. The relay protection device defect grading method according to claim 4, wherein the probability distribution of the category to which the vectorized text belongs is in the range of [0,1] and is 1.
6. The relay protection device defect grading method according to claim 4, wherein the cross entropy loss function comprises the following calculation formula:
Figure FDA0004038894330000021
where M denotes the number of categories, yi c Representing a symbolic function, taking 1 if the true class of sample i is equal to c, otherwise taking 0 c Representing the probability that the observed sample i belongs to the class c.
7. The relay protection device defect grading method according to claim 1, wherein the S3 comprises:
s31, representing the prediction precision of the target convolutional neural network model by adopting the accuracy, the recall rate and the F1 score;
and S32, inputting the test set into the target convolutional neural network model to obtain a defect classification recognition result.
8. The relay protection device defect grading method according to claim 7, wherein the obtaining of the model evaluation index comprises the following calculation formula:
Figure FDA0004038894330000022
Figure FDA0004038894330000023
Figure FDA0004038894330000024
wherein P represents accuracy, R represents recall, F represents F1 score, TP represents number of correctly ranked defect texts, FP represents number of incorrectly ranked defect texts, and FN represents number of undetected texts.
CN202310013026.9A 2023-01-05 2023-01-05 Relay protection defect setting and grading method based on convolutional neural network Pending CN115983265A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202310013026.9A CN115983265A (en) 2023-01-05 2023-01-05 Relay protection defect setting and grading method based on convolutional neural network

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202310013026.9A CN115983265A (en) 2023-01-05 2023-01-05 Relay protection defect setting and grading method based on convolutional neural network

Publications (1)

Publication Number Publication Date
CN115983265A true CN115983265A (en) 2023-04-18

Family

ID=85966533

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202310013026.9A Pending CN115983265A (en) 2023-01-05 2023-01-05 Relay protection defect setting and grading method based on convolutional neural network

Country Status (1)

Country Link
CN (1) CN115983265A (en)

Similar Documents

Publication Publication Date Title
CN108073568B (en) Keyword extraction method and device
CN111914090B (en) Method and device for enterprise industry classification identification and characteristic pollutant identification
CN111881983B (en) Data processing method and device based on classification model, electronic equipment and medium
CN110188047B (en) Double-channel convolutional neural network-based repeated defect report detection method
CN106095928A (en) A kind of event type recognition methods and device
CN109165284A (en) A kind of financial field human-computer dialogue intension recognizing method based on big data
CN111767398A (en) Secondary equipment fault short text data classification method based on convolutional neural network
CN111767397A (en) Electric power system secondary equipment fault short text data classification method
CN108304567B (en) Method and system for identifying working condition mode and classifying data of high-voltage transformer
CN110097096B (en) Text classification method based on TF-IDF matrix and capsule network
CN112685324B (en) Method and system for generating test scheme
CN109800309A (en) Classroom Discourse genre classification methods and device
CN109063983A (en) A kind of natural calamity loss real time evaluating method based on social media data
CN116522334A (en) RTL-level hardware Trojan detection method based on graph neural network and storage medium
CN115146062A (en) Intelligent event analysis method and system fusing expert recommendation and text clustering
CN106815209B (en) Uygur agricultural technical term identification method
CN110825642A (en) Software code line-level defect detection method based on deep learning
CN112559741B (en) Nuclear power equipment defect record text classification method, system, medium and electronic equipment
CN117197538B (en) Bayesian convolution neural network structure apparent damage identification method based on Gaussian distribution weight sampling
CN111737993A (en) Method for extracting health state of equipment from fault defect text of power distribution network equipment
CN114077663A (en) Application log analysis method and device
CN115983265A (en) Relay protection defect setting and grading method based on convolutional neural network
CN107122394A (en) Abnormal deviation data examination method and device
CN111160756A (en) Scenic spot assessment method and model based on secondary artificial intelligence algorithm
CN114970601A (en) Power equipment partial discharge type identification method, equipment and storage 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