Relay Protection Devices Defect Text Data Mining Technology And Application

Relay protection device is the first line of defense of the power system.Its reliability is crucial,and the accurate and efficient achieves defect processing has attracted much attention.However,with the expansion of the power grid,the number of relay protection devices has increased sharply,and the per capita equipment carrying capacity of operation and maintenance personnel is severely overloaded.On the other hand,the operation and maintenance site has accumulated a large number of defective text records,which contains rich experience in defecting processing,and is yet to be excavated.At the same time,the vigorous development of natural language processing technology provides technical support for excavating defective text records and serving the operation and maintenance management of relay protection devices.In view of this,this article is based on the shortcoming data record data recorded by a regional power grid relay protection device,and the following tasks have been carried out:(1)Combined with various simple and integrated classification algorithms,a defective model based on professional dictionary and text directionalization is established,and the level method based on the Bagging model is determined.Specifically,the defective level process and application scenarios are first introduced.Secondly,based on professional dictionaries and Jieba function packages as text preprocessing,combined with word frequency-inverted file frequency(TF-idf)assignment method to achieve the discharge of the defective text of the relay protection device;Thirdly,combined with the Python Tool Library,the defective models of integrated algorithms such as logical regression and decision trees and Bagging,XGBOOST and other integrated algorithms are built respectively.;Eventually,the Bagging Classifier was used to build a defective level method.(2)In response to the text data of the defect,the knowledge map construction method of the defective field of relay protection devices was proposed,and the visualization of the knowledge map was completed based on the NEO4J platform.Specifically,first of all,the knowledge map technology and Neo4J technology are introduced in concept;second,the defect text is divided into two types:structural characteristics and non-alternative data based on structural characteristics.For structural data,according to the determined data attributes,according to the determined data attributes The label directly forms the knowledge map according to the establishment logic;for non-structured data,the structured data of the data is completed through information,and then the knowledge map is constructed.The combination of dependent clause analysis technology;finally,a visual knowledge map is formed based on the PY2NEO development framework and the Neo4J web page version platform.(3)The preliminary application of three aspects of the knowledge map of relay protection devices in texture,analysis of defect characteristics,speculation and disappearance of deficit reasons,and deficiency reasons.In terms of improvement of text quality,based on the standard format of non-structured data,information correction was realized using word constraints or structured data.In terms of defect characteristics,according to the search conditions,nodes and attribute matching methods are adopted to traverse the search structured data knowledge diagram to realize the online query of defect type and the performance analysis of the equipment of various manufacturers.In terms of intelligent diagnosis and online discharge,the knowledge diagram of nodes and attribute matching traversing non-structured data is used to achieve the determination of deficiencies to speculate and eliminate shortage of shortcomings.