基于最优阈值小波包分析的转子振动信号降噪研究  

Denoising of Rotor Vibration Signals Based on the Optimal Threshold Wavelet Packet Analysis

作　　者：赵凯[1] 李本威[1] 徐圣田 

机构地区：[1]海军航空工程学院飞行器工程系 [2]92635部队

出　　处：《燃气涡轮试验与研究》2014年第3期22-26,共5页Gas Turbine Experiment and Research

基　　金：国家自然科学基金青年基金(61102167)

摘　　要：针对采集转子振动信号时信号中往往夹杂着噪声干扰的问题,提出了一种自寻优阈值的小波包降噪方法。该方法可对不同振动信号自寻优地设置阈值,且阈值的选取与所研究的信号有关而与信号长度无关;并根据采样定理计算不同转子转速对应的小波包分解的最低层数。对多种信号进行的降噪研究表明,该方法可有效降低转子振动信号中的噪声干扰。Since the rotor vibration signal acquisition often mingled with the noise interference, a self-opti-mized threshold wavelet packet denoising method was proposed. Self-optimized threshold, which was linked to the signal itself rather than the length of signal, was set for different vibration signals. And the low-est layer of wavelet packet was decomposed according to the rotor speed. The research results show that, this method can reduce the noise disturbance in rotor vibration signal effectively, and has certain applica-tion value.

关 键 词：航空发动机 转子振动 最优阈值 小波包分析 信号降噪 阈值函数 采样定理 

分 类 号：V235.13[航空宇航科学与技术—航空宇航推进理论与工程]