Open AccessDistribution network reliability investment effectiveness evaluation based on defect data mining
Author(s) -
Liu Keyan,
Liu Yangtao,
Hu Lijuan,
Du Songhuai,
Su Juan
Publication year - 2017
Publication title -
the journal of engineering
Language(s) - English
Resource type - Journals
ISSN - 2051-3305
DOI - 10.1049/joe.2017.0692
Subject(s) - reliability (semiconductor) , data mining , computer science , reliability engineering , classifier (uml) , grid , investment (military) , artificial intelligence , power (physics) , engineering , mathematics , physics , quantum mechanics , politics , law , political science , geometry
Distribution network investment benefit assessment affects the future power grid planning directly. Reasonable distribution network investment benefit evaluation plays an important role in real power grid construction. In this study, an improved term frequency‐inverse document frequency algorithm is proposed aimed at the reliability investment and reliability index of the distribution network based on the existing reliability defect text data. Especially the irrelevance between planned outage data and the investment data of reliability enhancement measures during the construction of the distribution network. The algorithm extends the key words of the classifier by synonyms and weighs the weight of the short text. It firstly solves the problem that many information fields in the professional field are neglected in the training set and then solves the influence of the feature length on the expression of the category theme. After the improved algorithm to deal with the defect text, the ‘power outage content basic thesaurus’ is formed. Through the semantic mapping relation, the semantic mapping table of keyword and reliability investment project is established, and the relationship between reliability investment and reliability of each distribution network is obtained. Finally, the model of the reliability investment of the distribution network considering the pre‐arranged power outage is established. Finally, taking the reliability data of a region as an example, the investment benefit is evaluated and the validity and practicability of the model are verified.