基于改进希尔伯特-黄变换的发动机气门故障诊断  

A Study on Improved Hilbert-Huang Transform Diagnosis for Leakage of Engine Valves

作　　者：杨恭勇 周小龙[1] 梁秀霞 李家飞 

机构地区：[1]东北电力大学工程训练教学中心,吉林吉林132012 [2]河南信宇石油机械制造股份有限公司,河南濮阳457001

出　　处：《东北电力大学学报》2017年第3期66-72,共7页Journal of Northeast Electric Power University

摘　　要：由于发动机系统及工作环境等因素的影响,发动机气门故障信号往往呈现出非线性和非平稳性的特点。为此,提出一种基于改进希尔伯特-黄变换的故障诊断方法。以气门声音信号为研究对象,首先,采用快速独立分量分析法去除环境噪声因素对于信号诊断准确性的影响,对降噪后信号进行改进经验模态分解,得到表征信号特性的固有模态函数,并通过相关性分析法去除虚假分量,从而获得信号的希尔伯特谱和边际谱,最后,结合时域和频域特征进行故障诊断。通过仿真研究证实了本文所提方法的准确性,实际试验证明:希尔伯特谱和边际谱能够有效并准确地反映出故障信号的时频信息,为该类问题的解决提供一种切实有效的方法。Because of the influence of engine system and working environment, the leakage of engine valves signal is proved to be non-stationary and non-stationary. In view of this characteristic, a fault diagnosis method based on improved Hilbert-Huang transform is proposed. The engine cylinder knocking sound signals as the research object. Firstly, these signals are pretreated by using the fast independent component analysis method to eliminate ambient noise. Then, with improved empirical mode decomposition method, the intrinsic mode func- tions are obtained and sensitive intrinsic mode functions are selected by correlation analysis method. Finally, the Hilbert spectrum and marginal spectrum of the signals are obtained with Hilbert transform. By combining the features of time domain and frequency domain, the faults can be diagnosed. The simulation experiment shows the effectiveness of the proposed method. The actual test shows that the Hilbert spectrum and marginal spectrum can display the physical information of the fault signals effectively and accurately. It also provides an effective method for this problem.

关 键 词：希尔伯特-黄变换 相关系数 气门 故障诊断 

分 类 号：TK401[动力工程及工程热物理—动力机械及工程]