Detail Active Noise Control Inside 3d Enclosure Using Recurrent Neural Networks   Oleh:  Bambang, Riyanto ; Wardana, Bayu Jenis: Article from Article Dalam koleksi: Final Program and Book of Abstracts: The 4th Asian Control Conference, September 25-27, 2002 (Sep. 2002), page 2180-2185. Topik: Neural Networks; Noise Control; Enclosure Fulltext: AC021457.PDF (270.74KB) Isi artikel This paper presents application of recurrent neural networks to modeling and control of acoustic noise and implement the neural control scheme and its learning in real-time using digital signal processor. A method is proposed for learning algorithm of recurrent neural networks suited for active noise control application. Realtime experiments using digital signal processor are carried out to reject acoustic noise within 3D enclosure. The dynamic nonlinear mapping capability of recurrent neural networks, is employed to model secondary electrical and acoustic path in ANC and to implement adaptive nonlinear controller. The proposed control scheme consists of two stages: first, modeling secondary path of ANC using recurrent neural networks, and secondly utilizing this model neural networks, in conjunction with controller neural networks, to generate antinoise signal. Opini Anda Klik untuk menuliskan opini Anda tentang koleksi ini! Kembali

Copyright © 2006, 2007 Unika Atma Jaya, all rights reserved