• CN:11-2187/TH
  • ISSN:0577-6686

›› 2002, Vol. 38 ›› Issue (2): 100-103.

• 论文 • 上一篇    下一篇

基于神经网络的动反馈管道有源消声系统声学通道辨识

吴斌;费仁元;周大森;杨立勇   

  1. 北京工业大学机电学院
  • 发布日期:2002-02-15

IDENTIFICATION FOR ACOUSTICAL PATHS OF THE ACTIVE NOISE CONTROL SYSTEM IN DUCTS USING MOTIONAL FEEDBACK LOUDSPEAKER BASED ON NEURAL NETWORKS

Wu Bin;Fei Renyuan;Zhou Dasen;Yang Liyong   

  1. Beijing Polytechnic University
  • Published:2002-02-15

摘要: 介绍了一种采用动反馈式扬声器的管道有源自适应噪声控制系统,系统中动反馈扬声器既作为有源消声系统的执行元件,也作为检测控制效果的声传感器。分析了系统中存在的各声学通道,并进行了基于自适应线性神经网络的辨识。该方法实现简便,对工作环境无干扰,各声学通首传递函数估计精度高。

关键词: 动反馈扬声器, 神经网络辨识, 有源消声

Abstract: An active adaptive noise control system in ducts is proposed,which adopts a motional feedback loudspeaker not only as an noise controlling actuator but also as a detecting sensor. By analysing each acoustical path, identification based adaptive linear neural networks is made. This kind of identifing method can be achieved expediently, and it would’t bring disturbance to working environment. The estimated result of each acoustical path agrees well with its real acoustical character, respectively.

Key words: Active noise control, Motional feedback loudspeaker, Neural networks identification

中图分类号: