基于盲源分离和机器学习的光伏并网逆变器故障诊断

文章摘要

基于盲源分离和机器学习的光伏并网逆变器故障诊断

Open-Circuit Fault Diagnosis of Grid-Tie Inverters Based on Blind Source Separation and Machine Learning

投稿时间：2023-11-05 修订日期：2023-11-05

DOI：

中文关键词: 光伏并网逆变器，单管开路故障，快速独立成分分析，线性判别分析

英文关键词: Grid-Tie Inverters,Single Power Electronic Device Open-Circuit Fault, FastICA, LDA

基金项目:基于fastICA（Fast Independent Component Analysis快速独立分量分析）算法的光伏并网逆变器开路故障诊断技术研究

作者	单位	邮编
张磊^*	安徽水利水电职业技术学院	230041

摘要点击次数: 920

全文下载次数: 0

中文摘要:

针对光伏并网逆变器中的开关管发生单管开路不易察觉，存在运行隐患的问题，提出一种基于快速独立成分分析和线性判别分析相结合的诊断方法。首先，采用快速独立成分分析算法对三相输出电流进行分析，结果中仅包含正弦半波和正弦波成分时，可判定逆变器发生单管开路故障；其次，提取旋转电流Id在时域和频域下的特征；最后，以旋转电流Id特征为输入、逆变器工作状态编码为输出进行线性判别分析，以此定位开路的开关管。Matlab仿真验证表明，所提方法开路故障诊断准确率可达98.9%以上，具有较高的应用价值。

英文摘要:

A diagnosis method based on fast independent component analysis and linear discriminant analysis is proposed to solve the problem of hidden trouble in the grid-tie inverters. Firstly, the fast independent component analysis algorithm is used to analyze the three-phase output current, and the single power electronic device open-circuit fault can be judged when the results include only sinusoidal half-wave and sinusoidal components. Then, the characteristics of rotating current Id in time domain and frequency domain are extractedfinally. Finaly, the open-circuit switch is located by linear discriminant analysis the rotating current Id as input and the working state of the inverter as output. MATLAB simulation shows that the accuracy of open circuit fault diagnosis can reach more than 98.9% , and the proposed method has high application value.

View Fulltext 查看/发表评论下载PDF阅读器

关闭