基于神经网络的风机管道噪声有源控制的实验研究  

Experimental Study on Active Control of Noise in Fan Duct on the Basis of Neural Network

作　　者：张奇志[1] 曾成[1] 关碧华 尹建民 周雅莉 

机构地区：[1]北京机械工业学院计算机及自动化系,北京100085

出　　处：《电声技术》2002年第7期54-56,共3页Audio Engineering

基　　金：北京市教委资助科技项目(99KJ144);北京市优秀人才培养专项经费资助。

摘　　要：通过实验来考察空气流动对传声器测量信号的影响,给出了基于神经网络的风机管道噪声有源控制的实验结果。实验结果表明,风机管道中空气流动对传声器测量信号的影响很大。在实验中采用了几项保护措施,使得次级路径的辨识结果对风机管道中空气流动不再敏感。实验结果表明了基于神经网络的风机管道噪声有源控制系统的有效性。The objective of this paper is to show the results of the practical implementation of neural network controller on an active noise control system for a fan duct,and exams the in-fluence of airflow to the measurements with microphone.The experimental result shows that the measurements with microphone are highly sensitive to airflow in the fan duct.Several protective measures have been taken in our experiments so that then the identification of secondary path is insensitive to airflow in the fan duct.The usefulness and validity of the proposed neural network control scheme are examined in experiments.

关 键 词：神经网络 风机 管道噪声 有源控制 实验研究 

分 类 号：TH43[机械工程—机械制造及自动化] TB535[理学—物理]