... 该模块根据语音状态检测器的输出来控制滑动窗自适应 滤波器的系数更新，采用归一化最小均方算法(Normalized Least Mean Squares, NLMS) [3-4] ，递推式为 ( ) ( ) ( 1) ( ) ( ) e n n W n W n n ...
基于2个网页-相关网页
域归一化最小均方算法 TDNLMS
仿真结果表明最小均方自适应算法、归一化最小均方算法、基于QR分解的平方根信息自适应滤波器和基于QR分解的最小二乘格型自适应滤波算法都能有效地滤除海浪噪声。
It can be drawn from the simulation of the sea adaptive filter adopting various algorithms mentioned that the LMS, NLMS, QRRLS and QRDLSL are all effective to denoise the sea noise.
提出一种改进的归一化最小均方(MNLMS)算法，并用该算法驱动FIR滤波器以实现对象模型及其逆的辨识。
A modified NLMS (MNLMS) algorithm was proposed and applied to drive a FIR filter to approximate both the model and the inverse of plants.
NLMS算法是回声消除器中最常用的算法之一，然而语音信号的强相关性使NLMS(归一化最小均方)算法的收敛速度变慢。
NLMS (normalized least mean square) algorithm is widely used in echo canceller. However, due to high correlation of speech signals, the performance of echo canceller based on NLMS is depressed.
应用推荐