太赫兹科学与电子信息学报
太赫玆科學與電子信息學報
태혁자과학여전자신식학보
Information and Electronic Engineering
2014年
3期
457-460,470
,共5页
级联五电平逆变电路%载波相移脉冲宽度调制%BP神经网络%故障诊断
級聯五電平逆變電路%載波相移脈遲寬度調製%BP神經網絡%故障診斷
급련오전평역변전로%재파상이맥충관도조제%BP신경망락%고장진단
cascaded five-level inverter%phase-shifted Pulse Width Modulation%Back Propagation neural network%fault diagnosis
因含有大量的开关器件，多电平逆变器难以用基于模型的方法进行故障诊断。针对这一问题，提出了一种基于神经网络的故障识别和分类方法。采用载波相移脉冲宽度调制(PWM)策略搭建级联五电平逆变电路，对逆变器的输出电压信号进行快速傅里叶变换，获取其频谱并以此作为特征信息。利用反向传播算法(BP)神经网络对输出电压模式进行分类。Matlab仿真结果表明，本文设计的 BP神经网络有效地实现了对逆变器的故障诊断。
因含有大量的開關器件，多電平逆變器難以用基于模型的方法進行故障診斷。針對這一問題，提齣瞭一種基于神經網絡的故障識彆和分類方法。採用載波相移脈遲寬度調製(PWM)策略搭建級聯五電平逆變電路，對逆變器的輸齣電壓信號進行快速傅裏葉變換，穫取其頻譜併以此作為特徵信息。利用反嚮傳播算法(BP)神經網絡對輸齣電壓模式進行分類。Matlab倣真結果錶明，本文設計的 BP神經網絡有效地實現瞭對逆變器的故障診斷。
인함유대량적개관기건，다전평역변기난이용기우모형적방법진행고장진단。침대저일문제，제출료일충기우신경망락적고장식별화분류방법。채용재파상이맥충관도조제(PWM)책략탑건급련오전평역변전로，대역변기적수출전압신호진행쾌속부리협변환，획취기빈보병이차작위특정신식。이용반향전파산법(BP)신경망락대수출전압모식진행분류。Matlab방진결과표명，본문설계적 BP신경망락유효지실현료대역변기적고장진단。
It is difficult to diagnose a multilevel inverter using mathematical models because of the large amount of switching devices. Therefore, a neural network-based method is applied to identify and classify various faults. In this paper, a cascaded five-level inverter modulated by the strategy of the carrier phase-shifted Pulse Width Modulation(PWM) is developed. By utilizing Fast Fourier Transform technique, the spectrum of the output voltage is obtained and is taken as the feature information. A Back Propagation(BP) neural network is applied to the fault diagnosis of the inverter,and Matlab simulation results show that the BP neural network can diagnose the faults of the inverter successfully.
