GENETIC ALGORITHM FOR ACTIVE NOISE CONTROL

This paper presents a genetic algorithm for an active noise control (ANC) system. The acoustic noise in aircraft cabins is not only harmful to the human ear but also will impede oral communication in such environments. The methods for suppressing acoustic noise using passive sound absorbers are effective for high frequency noise. However, at low frequencies it is not so as the acoustic wavelengths become large compared to the thickness of a typical acoustic absorber. For these reasons, a number of practically important acoustic noise problems are dominated by low frequency contributions. Hence generally these problems cannot be solved using passive methods as they are very expensive in terms of weight and bulk. This has necessitated the exploration of alternative methods of noise control. Active Noise Control (ANC) is one such alternative. The conventional ANC system often implements the filtered-x least mean square (FXLMS) algorithm to update the coefficients of FIR filters because of its simplicity but it requires identifying secondary path which increases the computational complexity while implementing multichannel ANC system and also it converges to local minima. In this paper, the FXLMS algorithm is replaced with genetic algorithm because it does not require identifying secondary path for ANC system and also it prevents local minima problem.