基于DSP的实时降噪系统设计与实现  

Design and implementation of real-time noise control systems based on DSP

作　　者：张大勇[1] 凌强[1] 计炜梁 张俊斌 徐酉亮 熊远波 吴艺英 

机构地区：[1]中国科学技术大学自动化系,安徽合肥230027 [2]白城兵器实验中心轻武器部,吉林白城137001

出　　处：《微型机与应用》2016年第14期89-92,95,共5页Microcomputer & Its Applications

基　　金：国家自然科学基金(61273112)

摘　　要：针对周期噪声与窄带噪声,基于DSP硬件平台,进行主动降噪系统设计与实现。主要工作为次级通道辨识、对算法定点化处理和基于DSP的实时降噪实现。所提出的定点化方法大大提高了定点DSP的运算速度,保证了降噪的实时性。其中降噪程序采用自适应算法中的前馈滤波-X-最小均方算法,DSP采用德州仪器公司的16位定点DSP TMS320VC5509A,使用TLV320AIC23B作为音频处理芯片,取得了明显降噪效果。In order to actively cancel periodic and narrow-band noise, this paper proposes designs and implements an active noise control system based on a DSP hardware platform. Our main tasks include the identification of the secondary acoustic channel, the fixed - point realization of algorithms and the real-time noise cancellation. Due to the implemented fixed-pointed realization, the computational speed of the noise cancel- lation algorithm is greatly enhanced, which guarantees the real-time processing of noise cancellation. Our system takes the feedforward filtered- X-least mean square （FXLMS） algorithm, the TMS320VC5509A by TI as the MCU, and the TLV320AIC23B to process audio signals. The ex- perimental results demonstrate the effectiveness of the system.

关 键 词：主动降噪 滤波-X-最小均方算法 定点化 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]