电力系统及其自动化学报2019，Vol.31Issue(3)：133-138,6.DOI:10.3969/j.issn.1003-8930.2019.03.020
综合RapidMiner与改进粒子群极限学习机算法的变压器故障诊断
Transformer Fault Diagnosis with the Combination of RapidMiner-modified Particle Swarm Optimization-extreme Learning Machine Algorithm
魏金萧 1周步祥 1唐浩 2张百甫 1杨常1
作者信息
- 1. 四川大学电气信息学院,成都 610065
- 2. 四川电力设计咨询有限责任公司,成都 610094
- 折叠
摘要
Abstract
To solve the problems such as the lack of encoding when the three-ratio method is used in transformer fault diagnosis and the weak anti-interference capabilities of various artificial intelligence methods,a transformer fault diag?nosis method is put forward by combing RapidMiner and the RapidMiner-modified particle swarm optimization-extreme learning machine(RM-MPSO-ELM). First,using RapidMiner tools and combining the sample data of transformers, the most relevant input variables of fault type are selected. Then,in light of the difficulty in the parameter selection us?ing ELM,an MPSO algorithm is used to optimize the parameters. Finally,ELM is used to identify the potential faults of the transformer,and the diagnostic performance is compared among IEC three-ratio method,support vector machine (SVM)method and different combinations of ELM algorithm. Results show that the proposed method has a higher diag?nosis accuracy.关键词
油浸式变压器/油中溶解气体分析/特征值选择/极限学习机/故障诊断Key words
oil immersed transformer/dissolved gas-in-oil analysis(DGA)/feature selection/extreme learning ma⁃chine(ELM)/fault diagnosis分类
信息技术与安全科学引用本文复制引用
魏金萧,周步祥,唐浩,张百甫,杨常..综合RapidMiner与改进粒子群极限学习机算法的变压器故障诊断[J].电力系统及其自动化学报,2019,31(3):133-138,6.
