基于改进萤火虫算法优化ELM的电力电容器故障诊断  

Fault Diagnosis of Power Capacitor Based on ELM Optimized By Improved Firefly Algorithm

作　　者：张博皓 吕冰[1] ZHANG Bohao;L Bing(School of Automation and Electronic Engineering, Qingdao University of Science and Technology, Qingdao 266100)

机构地区：[1]青岛科技大学自动化与电子工程学院,青岛266100

出　　处：《微型电脑应用》2019年第5期107-110,共4页Microcomputer Applications

摘　　要：针对ELM分类预测的结果易受其初始输入权值和阀值的影响,提出了一种改进萤火虫算法优化ELM的电力电容器故障诊断模型。选择电力电容器故障诊断的准确率为适应度,通过IFA优化ELM的初始输入权值和阈值,实现电力电容器故障自适应诊断。研究结果表明,与其他算法比较可知,IFA_ELM可以有效提高电力电容器故障诊断的准确率和降低误判率,为电力电容器故障诊断提供新的方法和途径。The result of ELM classification prediction is easily influenced by the initial input weight and threshold value, a fault diagnosis model of power capacitor is proposed to improve the optimization of ELM algorithm. The accuracy of power capacitor fault diagnosis is selected as adaptability, and the initial input weight and threshold value of ELM are optimized by IFA to realize the self-adaptive diagnosis of power capacitor fault. The results show that the IFA_ELM can effectively improve the accuracy and reduce the misjudgment rate of power capacitor fault diagnosis, and it provides a new method and approach for power capacitor fault diagnosis.

关 键 词：萤火虫算法 极限学习机 电力电容器 遗传算法 粒子群算法 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]