基于LMD的PWM整流电路故障特征提取新方法

Fault Feature Extraction of PWM Rectifier Based on LMD

脉冲宽度调制(PWM)整流电路结构日益复杂,对其可靠运行提出了更高的要求;对局域均值分解(LMD)用于PWM整流电路的故障特征提取进行研究,提出一种基于LMD和加权频带能量法的特征提取新方法;该方法通过逐步抽取调频调幅成分将故障信号在频域上展开,然后基于信号能量的频带分布特点,充分考虑各频带成分与故障的相关性,构造故障特征向量,实现特征提取;最后以PWM整流电路为例进行仿真,相电压380V,仿真时间0.5s,0.1s时注入故障;结果表明,该方法能有效地提取故障信号的特征,并降低特征向量的维数。

Performance improvement of Pulse Width Modulation （PWM） rectifier brings about a complicated structure, which makes the reliable operation face higher demand increasingly. Therefore, this paper proposes a new method of fault feature extraction based on LMD and weighted frequency range energy operator. This method unfolds the energy distribution of the fault signal in term of frequency through decomposing it into amplitude and frequency modulated signals. Then the feature vector can be constructed accord to the energy distribution and the correlation between each product function and the fault signal. Finally, the three-phase three-level 380V PWM rectifier is utilized as an example, fault injected on 0.1 s of 0.5 s simulation time. And the simulation results show that the proposed approach can catch the fault feature sensitively, while achieving dimension reduction of feature vector.