摘要
为了从齿轮故障信号中提取出包含故障信号的特征频率,提出了基于EEMD自适应形态学解调方法。首先采用EEMD(集合经验模式分解)进行降噪,将原始信号与不同的白噪声叠加组成目标信号,然后将目标信号分解为有限个IMF分量,选取主要信息求和重构,再用形态学滤波器提取故障信号的特征频率。针对形态学结构元素尺寸的选择问题,利用遗传算法来优化形态学结构元素,自适应寻求最优解。通过数字仿真试验和齿轮故障模拟实验,并与EMD(经验模式分解)、SVD(奇异值分解)方法进行了比较,结果表明该算法要优于其他两种方法,能够清晰地提取出故障信号的各种频率特征。
In order to extract the characteristic frequencies from gear fault signals containing fuh information, the adaptive morphology demodulation method was proposed based on EEMD. EEMD （ensemble empirical mode decomposition） was used to reduce noise firstly, an original signal with superposition of different white noises formed target signals, and then the target signal was decomposed into a finite number of IMF components, the IMFs containing fault information were chosen and summed to reconstruct a signal. The morphologic filter was used to extract the characteristic frequencies containing fault information from the reconstructed signal. Aiming at the problem of morphologic structural element size selection, the genetic algoritms was used to adoptirely optimize the structural elements of morphology. Through numerical simulations and gear fault simulation tests, The results showed that the proposed method is superior to characteristic frequencies of gear faults. the proposed method was compared with EMD and SVD. the other two, it can be used to clearly extract various
出处
《振动与冲击》
EI
CSCD
北大核心
2014年第18期145-148,共4页
Journal of Vibration and Shock
基金
国家自然科学基金青年基金资助项目(51105284)
国家自然科学基金项目(51475339)
关键词
形态学
特征频率
EEMD
结构元素
遗传算法
morphology
characteristic frequency
EEMD
structural elements
genetic algorithm
