摘要
针对传动系监测信号的特点,分别采用最小波分析、倒谱分析以及小波包分析,探讨车辆传动系统监测信号的预处理,以去除信号趋势项和去除冗余信息,消除信号噪声,提取特征频率.仿真计算结果证明,小波阈值去噪法可同时去噪与去除趋势项,利用倒谱分析法能够有效提取载波频率与调制频率,利用小波包分析可去除冗余信息,突出信号特征.
According to the features of the state monitoring data of vehicle transmission, a pretreatment procedure is proposed for the de-trend, de-noise, redundancy condensation, extraction of the feature frequency, where the wavelet analysis, wave-packet analysis and cepstrum analysis are integrated. The simulation results show that wavelet threshold de-noising method can eliminate noise and trend, cepstrum analysis is able to extract carrier frequency and frequency modulation, and wave-packet analysis is going to achieve redundancy condensation and stand out the features of the monitoring data.
出处
《车辆与动力技术》
2006年第2期43-46,共4页
Vehicle & Power Technology
基金
武器装备预研基金项目(514571201)
关键词
数据挖掘
特征提取
传动装置
data mining
feature extraction
transmission
