<![CDATA[In a response to a growing demand for environments of 70dB or less noise levels, many industrial sectors have focused with some form of noise control system. Active noise control (ANC) has proven to be the most effective technology. This paper mainly investigates application of neural network on self-adaptation system in active noise control (ANC). An active silencing control system is made which adopts a motional feedback loudspeaker as not a noise controlling source but a detecting sensor. The working fundamentals and the characteristics of the motional feedback loudspeaker are analyzed in detail. By analyzing each acoustical path, identification based adaptive linear neural network is built. This kind of identifying method can be achieved conveniently. The estimated result of each sound channel matches well with its real sound character, respectively. DOI: http://dx.doi.org/10.11591/telkomnika.v10i7.1580]]>
