Modified-Error Adaptive Feedback Active Noise Control System Using Linear Prediction Filter

Nobuhiro MIYAZAKI  Yoshinobu KAJIKAWA  

IEICE TRANSACTIONS on Fundamentals of Electronics, Communications and Computer Sciences   Vol.E97-A   No.10   pp.2021-2032
Publication Date: 2014/10/01
Online ISSN: 1745-1337
DOI: 10.1587/transfun.E97.A.2021
Type of Manuscript: PAPER
Category: Engineering Acoustics
active noise control,  modified-error feedback ANC,  linear prediction filter,  MR noise,  head-mounted ANC system,  

Full Text: PDF>>
Buy this Article

In this paper, we propose a modified-error adaptive feedback active noise control (ANC) system using a linear prediction filter. The proposed ANC system is advantageous in terms of the rate of convergence, while maintaining stability, because it can reduce narrowband noise while suppressing disturbance, including wideband components. The estimation accuracy of the noise control filter in the conventional system is degraded because the disturbance corrupts the input signal to the noise control filter. A solution of this problem is to utilize a linear prediction filter. The linear prediction filter is utilized for the modified-error feedback ANC system to suppress the wideband disturbance because the linear prediction filter can separate narrowband and wideband noise. Suppressing wideband noise is important for the head-mounted ANC system we have already proposed for reducing the noise from a magnetic resonance imaging (MRI) device because the error microphones are located near the user's ears and the user's voice consequently corrupts the input signal to the noise control filter. Some simulation and experimental results obtained using a digital signal processor (DSP) demonstrate that the proposed feedback ANC system is superior to a conventional feedback ANC system in terms of the estimation accuracy and the rate of convergence of the noise control filter.