摘要
针对航空发动机机匣观测信号为多个振动源的混合信号的问题,提出了一种基于航空发动机振动信号特征的盲分离算法,能够从混合的观测信号中确定振动源的个数以及提取发动机内部各个振动源的振动信息。算法的核心是基于航空发动机等旋转机械转轴故障振动信号的频谱特点,其故障信号的振动谱一般包含基频、谐波成分和次谐波成分,如不对中、碰磨、裂纹等。算法用到的主要工具是连续小波变换和时间同步平稳法,主要步骤为:首先通过连续小波变换将航空发动机不同观测通道上的观测信号分解,根据谱峰值分析确定主要振动源及对应的基频;然后通过时间同步平稳法,分别从每个观测通道上提取各个振动源的谐波和次谐波成分,从每个观测通道上提取出了源信号;对于同一源信号,从每个观测通道上都能提取出一个基本映像,最后通过对比每个映像信号的二范数,确定每个源信号的最优估计。通过数值模拟信号和实测航空发动机加速度振动信号对算法进行了验证,结果表明,在对发动机转速有一定先验知识的情况下,所提算法能够估计出发动机内部的主要振动源的个数,并能提取源信号的主要成分。结合旋转机械振动频谱特点等先验知识,说明了所提算法的正确性和实用性。
To solve the difficulty that the observed signals on aircraft engine case are generally a mixture of multiple vibration sources,a blind source separation algorithm is proposed based on the characteristics of aircraft engine vibration sources,where the spectral characteristics of the rotating shaft vibration signal of aircraft engine are aimed at.The vibration spectrum of shaft fault signals of rotating machinery usually includes fundamental frequency,harmonic components and sub-harmonic components,corresponding to such as misalignment,rubbing and crack.The observed signals from different channels are decomposed by continuous wavelet transform,the main vibration sources and the corresponding fundamental frequencies are determined according to the spectral peak analysis,and the harmonics and subharmonic components of each vibrationsource are extracted from each observed signal with time synchronous averaging method,thus an image of each source can be extracted from each observation channel,and several image signals are obtained for the same source.The optimal estimation is determined by comparing the 2-norm of image signals of each source.The proposed algorithm is validated by numerical simulation signal and measured aerodynamic acceleration signal from aircraft engine case.The results show that the algorithm can estimate the number of main vibration sources inside the engine and extract the main components of each source signal in the case of a certain priori knowledge of the engine speed.
作者
杨广振
荆建平
明阳
颜逸飞
陈昌敏
YANG Guangzhen;JING Jianping;MING Yang;YAN Yifei;CHEN Changmin(School of Mechanical and Power Engineering, Shanghai Jiaotong University, Shanghai 200240, China;AECC Commercial Aircraft Engine Co. , Ltd. , Shanghai 200240, China)
出处
《西安交通大学学报》
EI
CAS
CSCD
北大核心
2017年第6期20-27,共8页
Journal of Xi'an Jiaotong University
基金
国家自然科学基金资助项目(10772114
11672179)
关键词
航空发动机
盲分离
信号特征
振动信号
aircraft engine
blind source separation
signal feature
vibration signal
