Comparative Study of Different Adaptive Control Strategies in Noise Cancellation Applications

Abstract

In this paper, the system identification and noise cancellation has been done and further the adaptive control algorithms like LMS(Least mean square),NLMS(normalized least mean square),NLMF(normalized least mean forth) and RLS(recursive least square) filters are compared. System identification identifies an unknown system given an input and output. It is used in active vibration and noise control applications. The LMS algorithm has lowest computations involved than all other ones. RLS is a computationally complex filter algorithm but it works more efficiently. In all of these filter algorithms, the weight coefficient is continuously updated until the convergence is reached. These algorithms are implemented and are compared by using parameters such as MSE (mean square error), PSNR (peak signal to noise ratio), convergence, complexity and accuracy.