Analysis and DSP implementation of an ANC system using a filtered-error neural network

Active Noise Control Using a Functional Link Artificial Neural Network with the Simultaneous Perturbation Learning Rule

“…The exponential increase of noise pollution and ineffectiveness of passive techniques for noise attenuation have led to the development of active noise control (ANC) system [1][2][3][4][5][6]. This technique has been successfully applied to heating, ventilating, and air conditioning systems [4,5], exhaust and motor noise [4,7], headsets [4,6], and airplanes [4].…”

“…To deal with these effects, different structures for nonlinear controllers have been recently proposed in the literature [3,4,[7][8][9][10][11][12][13][14][15][16][17]. In [3], a nonlinear controller based on a filtered-error BP neural network was proposed to offer computational advantage over the filtered-x BP neural network algorithm.…”

“…In [3], a nonlinear controller based on a filtered-error BP neural network was proposed to offer computational advantage over the filtered-x BP neural network algorithm. In [4], a nonlinear controller based on a normalized Gaussian radial basis neural network was proposed to compensate the nonminimum-phase secondary path transfer function and control the nonlinear noise process.…”

“…Observations of various structures and algorithms reported in [3,4,[7][8][9][10][11][12][13][14][15][16][17][18][19][20][21] reveal that the structure based on FLANN with the FSLMS algorithm is a preferable nonlinear structure for active control of nonlinear noise processes because of its interesting properties. But, in this control method, in order to update the weights of the FLANN, we need a gradient of the error function, namely, we must know the model of the secondary path [3].…”

“…But, in this control method, in order to update the weights of the FLANN, we need a gradient of the error function, namely, we must know the model of the secondary path [3].…”