摘要
内燃机广泛应用于工程、动力等领域,然而内燃机因燃烧和机械运动引起的冲击与振动导致其减振降噪一直是研究的热点,而如何准确识别振源则是减振的前提。本文针对振源盲分离时观测信号不少于源信号数目要求不易满足的问题,利用经验小波变换(Empirical wavelet transform,EWT)结合快速独立成分分析(Fast independent component analysis,FastICA)实现对内燃机振源信号的识别。首先使用时域同步平均法对内燃机缸盖的振动信号进行预处理,然后进行经验小波变换,之后再利用皮尔逊相关系数选择有效经验模态分量作为快速独立成分分析(FastICA)的输入,最终分离结果表明:该方法可以有效地从内燃机缸盖振动信号中识别出燃烧信号和气阀机构开启时的气体冲击信号。
The internal combustion engine is widely used in engineering,power and other fields.However,due to combustion impact,mechanic vibration and the complexity of its structure,the vibration and noise reduction of internal combustion engines has always been a focus of research,and especially how to accurately identify the vibration source is the premise of vibration reduction.For the problem that the number of observed signals is not less than the number of source signals is not easy to meet in the blind separation of vibration sources,in this paper,the empirical wavelet transform(EWT)and fast independent component analysis(FastICA)are used to identify vibration sources of internal combustion engines.Firstly,the time domain synchronous average technique is used to preprocess vibration signal of the internal combustion engines cylinder head,and then experience the wavelet transform,next the Pearson correlation coefficient is used to choose effective empirical mode component as fast independent component analysis(FastICA)input,and finally the separation results show that this method can effectively identify the combustion gas signal and the valve body opens gas combustion impact signal from the internal combustion engines cylinder head vibration signal.
作者
史嘉伟
伍星
刘韬
杨启超
SHI Jiawei;WU Xing;LIU Tao;YANG Qichao(Faculty of Mechanical and Electrical Engineering,Kunming University of Science and Technology,Kunming 650500,China)
出处
《机械科学与技术》
CSCD
北大核心
2021年第5期741-748,共8页
Mechanical Science and Technology for Aerospace Engineering
基金
国家自然科学基金项目(51875272,51675251)
云南省重点科研项目(2017FA028)。
关键词
内燃机
时域同步平均
经验小波变换
快速独立成分分析
连续小波变换
internal combustion engines
time-domain synchronous average
empirical wavelet transform
fast independent component analysis
continuous wavelet transform
