主成分分析在提取周期性信号中的应用  

APPLICATION OF PRINCIPAL COMPONENT ANALYSIS IN EXTRACTING PERIODIC SIGNALS

作　　者：邵颖丽 王瑞莲 SHAO Yingli;WANG Ruilian(College of Statistics and Mathematics,Inner Mongolia University of Finance and Economics,Hohhot 010071,China)

机构地区：[1]内蒙古财经大学统计与数学学院,呼和浩特010071

出　　处：《内蒙古农业大学学报（自然科学版）》2020年第3期80-83,共4页Journal of Inner Mongolia Agricultural University(Natural Science Edition)

基　　金：国家自然科学基金项目,基于公平的交通流量分配模型及拥挤收费策略研究(71661024).

摘　　要：主成分分析是一种重要的统计方法,本文通过对主成分分析法的原理,数学模型以及计算过程的分析,提供了利用主成分提取周期性信号的理论依据。针对单缸柴油机排气噪声主要由以点火频率为基频的1种周期噪声的特点,对用声传感器采集的排气噪声信号进行了主成分分析,提取周期信号,并利用短时傅里叶变换,从频域提取信号特征,为排气消声器结构参数的选取提供理论依据。Principal Component Analysis(PCA)is an important statistical method.This paper provides a theoretical basis for ex-tracting periodic signals by PCA,by analyzing the principle,mathematical model and calculation process of PCA.In view of the char-acteristic that the exhaust noise of a single cylinder diesel engine mainly consists of a periodic noise with the ignition frequency as the fundamental frequency,the author analyzes the exhaust noise signal collected with the acoustic sensor by means of PCA,extracts the periodic signal,uses the short time Fourier transform,and extracts the signal features from frequency domain,hoping that he can pro-vide a theoretical basis for the selection of structural parameters of exhaust muffler.

关 键 词：主成分分析 信号分离 周期信号 

分 类 号：O213.9[理学—概率论与数理统计]