Application of Gammatone Filter Bank to Active Noise Control Algorithm

With the aid of research achievement of human auditory system, a new active noise control (ANC) algorithm considering the human auditory subjectivity was proposed in this paper. By using Gammatone filter bank to simulate filtering and frequency decomposition function of human ear, the real-time response of human ear to noise was computed. The error filter in filtered-Error least-mean-square (FELMS) algorithm was designed on the basis of real-time computed response. The design principle was to retain the intense noise components and inhibit insensitive parts of human ear's perception. Computer simulations demonstrate that the algorithm proposed in this paper can not only simulate the human ear's perception to control different environmental noise adaptively, but also improve the noise reduction and convergent rate compared with the traditional filtered-X least-mean-square (FXLMS) algorithm.