Accuracy and Performance Evaluation in the Genetic Optimization of Nonlinear Systems for Active Noise Control

This paper investigates the performance of genetic optimization in a nonlinear system for active noise control based on Volterra filters. While standard Filtered-X algorithms may converge to local minima, genetic algorithms (GAs) may handle this problem efficiently. In addition, this class of algorithms does not require the identification of the secondary paths. This is a key advantage of the proposed approach. Computer simulations show that a simple GA is able to find satisfactory solutions even in the presence of nonlinearities in the secondary path. The results are more accurate than using the linear techniques and the nonlinear systems based on classical LMS algorithms.