摘要
小波包分析方法是一种能有效地进行时 -频定位和微弱信号提取的工具 .但是小波滤波器组的频域特性和隔点采样会造成频谱混叠 ,导致分频结果不正确 .改进的小波包分频算法根据小波包混频的原因 ,结合 FFT分析进行处理 ,较好地消除了混频现象 .仿真研究表明 。
The wavelet packet analysis algorithm (WPAA) is a powerful tool in time\|frequency localization and weak signal detection. However, the frequency characters of quadrature mirror filters (QMFs) and dyadic\|down sampling cause serious frequency aliasing in signal decomposition, which would result in spurious signal components. A modified WPAA combined with FFT transform technique is presented in this paper, which treats the coarse signal and detail signal accordingly to satisfy the Shannon sampling theory, and frequency aliasing is successfully avoid. The proposed algorithm is then used to analyze the vibration signals of a gear case system to detect possible incipient faults. Simulation results demonstrate the improved WPAA is superior to the common WPAA in weak fault detection and forecast.
出处
《浙江大学学报(工学版)》
EI
CAS
CSCD
北大核心
2001年第3期307-311,共5页
Journal of Zhejiang University:Engineering Science
基金
国家自然科学 (青年 )基金资助项目 (2 0 0 76 0 40 )
关键词
时一频分析
小波包分解
故障检测
time\|frequency analysis
wavelet packet decomposition
fault detection