A robust filtered-s LMS algorithm for nonlinear active noise control

• Published in: Applied acoustics Vol. 73; no. 8; pp. 836 - 841
• Main Authors: George, Nithin V., Panda, Ganapati
• Format: Journal Article
• Language: English
• Published: Kidlington Elsevier Ltd 01.08.2012; Elsevier
• Subjects: Acoustics; Active noise control; Exact sciences and technology; Filtered-s least mean square algorithm; Functional link artificial neural network; Fundamental areas of phenomenology (including applications); Impulsive noise; Noise: its effects and control; Physics; Robust algorithm; Filtered-s least mean square algorithm; Functional link artificial neural network; Impulsive noise; Robust algorithm; Active noise control; Performance evaluation; Filtering; Nonlinear analysis; Neural network; Algorithm; Noise control; Simulation; Active control
• ISSN: 0003-682X; 1872-910X
• DOI: 10.1016/j.apacoust.2012.02.005

► Active noise control system which is robust against disturbances. ► Based on a functional link artificial neural network. ► Does not require selection of thresholds. ► Simulation study reveals enhanced performance. The performance of a nonlinear active noise control (ANC) system based on the recently developed filtered-s least mean square (FsLMS) algorithm deteriorates when strong disturbances in the ANC system are acquired by the microphones. To surmount this shortcoming, a novel robust FsLMS (RFsLMS) algorithm is proposed for a functional link artificial neural network (FLANN) based ANC system. The new ANC system is least sensitive to such disturbances and does not call for any prior information on the noise characteristics. The results obtained from simulation study establish the effectiveness of this new ANC scheme.