摘要
传统的EMD处理故障特征时,因为分解的虚假模态分量影响而使得故障特征不够准确,需要将其去除。借鉴了较低采样频率(二倍Nyqvist频率以下)基于能量原理去除虚假模态分量的方法,将其改进应用于较大采样频率的汽轮机故障特征提取之中,实现对汽轮机故障信号去除经验模态分解过程的虚假分量。并用此改进EMD方法对汽轮机油膜震荡和动静碰磨的信号进行处理,最终得到更准确的汽轮机故障特征信息。
The traditional EMD with the impact of the decomposed false modal component makes the fault feature not accurate enough, so needs to be removed. This paper based on the method of the removing false modal component by the energy principle under the lower sampling frequency （twice Nyqvist frequency）, applies the improved EMD to the fault feature extraction of the steam turbine under the larger sampling frequency in order to achieve the empirical mode decomposition of the steam turbine fault signals removed false modal component. Finally, the improved EMD is used to deal with the oil film concussion of the steam turbine and Rubbing signals, and the accurate turbine fault feature information is ultimately obtained.
出处
《汽轮机技术》
北大核心
2013年第6期447-450,457,共5页
Turbine Technology
