电子测量与仪器学报

彭丽维,张彼德,孔令瑜.级联H桥七电平逆变器故障的SKSNN-LPP特征提取方法[J].电子测量与仪器学报,2019,33(2):110-116

级联H桥七电平逆变器故障的SKSNN-LPP特征提取方法

Fault feature extraction method for cascaded H-bridge seven-level inverter based on SKSNN-LPP

DOI：

英文关键词:cascaded H-bridge seven-level inverter supervised kernel shared nearest neighbor feature extraction locality preserving projection

基金项目:四川省科技厅应用基础研究项目(2017JY0204)、四川省电力电子节能技术与装备重点实验室项目(SZJJ2015064)、西华大学研究生创新基金(ycjj2017063)资助项目

作者	单位
彭丽维	1.西华大学电气与电子信息学院
张彼德	1.西华大学电气与电子信息学院
孔令瑜	1.西华大学电气与电子信息学院

Author	Institution
Peng Liwei	1.School of Electric Engineering and Electronic Information, Xihua University
Zhang Bide	1.School of Electric Engineering and Electronic Information, Xihua University
Kong Lingyu	1.School of Electric Engineering and Electronic Information, Xihua University

摘要点击次数: 121

全文下载次数: 1

中文摘要:

针对级联H桥七电平逆变器结构复杂，故障特征属性相互交叉，相似故障类间区分度低的问题，提出一种监督核共享近邻(SKSNN)相似性度量方法，并运用于局部保持投影(LPP)算法中，形成一种新的基于监督核共享近邻的局部保持投影(SKSNN LPP)特征提取算法，用于提取七电平逆变器IGBT开路故障低维敏感特征。首先，采集各故障状态下逆变器交流侧三相电流原始信号数据；其次，利用SKSNN LPP算法提取嵌入于原始数据中的低维敏感特征；然后，将提取的特征作为支持向量机(SVM)的输入建立故障诊断模型；最后，与传统信号处理及统计分析方法进行了仿真对比分析。结果表明，所提出的特征提取方法能有效减低相似故障类误诊率，诊断精度高达964%，明显高于其他方法。

英文摘要:

Seven level inverter has a complex structure and several fault attributes cross each other, thus reducing the discrimination among similar fault classes. Given this, a supervised kernel shared nearest neighbor (SKSNN) algorithm was proposed and applied to locality preserving projection (LPP), which formed a new feature extraction algorithm for The IGBT open circuit fault feature extraction of cascade seven level inverter. Firstly, the three phase current of the AC side was collected as the original signal corresponding to each fault status. On that basis, the low dimensional sensitive features embedded in raw data would be extracted by the SKSNN LPP algorithm. Then, the extracted fault features were taken as the input of support vector machine （SVM） to establish fault diagnosis model. Finally, through the comparative analysis of the diagnostic effects, it can be shown that the proposed method is superior to traditional signal processing and statistical analysis methods, which can effectively reduce the misdiagnosis rate of similar fault categories and can achieve 964% diagnostic accuracy.

查看全文查看/发表评论下载PDF阅读器