智能软降噪算法研究

AN INTELLIGENT SOFT ACOUSTIC NOISE REDUCTION ALGORITHM

针对噪声空间分布的复杂性、混沌性和未知性提出一种智能软降噪算法。该算法在建立"广义噪声"定义下典型"未知噪声信号群"描述的基础上,通过转换矩阵的分解运算,使得噪声源以极坐标来表达其空间位置,因此能够将来自空域不同位置及其发射方向角的噪声源进行分解与定位。同时利用"包含预知特定频率信号的未知噪声信号群",将其中有用"预知信号"在去除所有噪声干扰的情况下提取出来,达到高效镇噪效果。

In this paper,an intelligent soft acoustic noise reduction algorithm is proposed in light of the properties of complex,chaotic and unknown the noise spatial distribution has.Based on defining the description of typical ＂unknown noise signal groups＂ in the definition of ＂generalised noise＂,the algorithm employs decomposition operation of transformation matrix to make the noise sources express their spatial location in polar coordinate,therefore it is able to decompose and locate the noise sources emerged from different positions of spatial regions with various emission direction angles.At the same time,this algorithm makes use of the ＂unknown noise signal groups containing foreseen specific frequency signals＂,and extracts the useful ＂foreseen singles＂ in condition of removing all the noise interference to achieve the effect of effective cancelling out the noise signals.