CAJ | 학술논문

主要研究基于Wiener自适应滤波LMS算法与RLS算法在语音信号去噪中的应用.对语音信号与噪声在相关和不相关、平稳和不平稳这四种情况下,分别采用RLS和LMS算法对语音信号进行去噪分析.研究发现,不管信号与噪声处于上面四种情况的任何一种,RLS算法比较优越,而且在处理相关和非平稳的情况下优势更大;对于不相关和平稳的情况,LMS算法的效果较好.
주요연구기우Wiener자괄응려파LMS산법여RLS산법재어음신호거조중적응용.대어음신호여조성재상관화불상관、평은화불평은저사충정황하,분별채용RLS화LMS산법대어음신호진행거조분석.연구발현,불관신호여조성처우상면사충정황적임하일충,RLS산법비교우월,이차재처리상관화비평은적정황하우세경대;대우불상관화평은적정황,LMS산법적효과교호.
This paper studies speech signal denoising based on the Wiener adaptive filtering LMS algorithm and RLS algorithm.Voice signal and noise in the relevant and irrelevant, smooth and not smooth the four cases, denoising speech signal analysis is discussed. respectively RLS and LMS algorithm. The study found regardless of the signal and noise in any one of the above four cases,RLS algorithm is superior,and in dealing with the greater advantage of the associated and non-stationary case. Irrelevant and stationary case,the LMS algorithm is better.