gallo09-COLLOQUE

Active noise control: Study of the algorithm

Marco GALLO
LMS International

Active control has been proposed as a possible solution for noise reduction. The job done was focused on evaluate an adaptive feed-forward controller on engine sound quality ; the control strategy is based on the feed-forward filtered LMS algorithm (Fx-LMS), with some modifications in order to improve the convergence rate.
Fig.(a) Evolution of error in time for different disturbance frequencies for the Fx-LMS Fig.(b) Evolution of error in time for different frequencies for the NEX-LMS
First of all the convergence speed of Fx-LMS and NEX-LMS is studied. As it can be seen in Fig.(a) and Fig.(b), the NEX-LMS algorithm (Normalized Equalized x- LMS) extends the fast convergence to the whole frequency range. Then, simulation with real car data and latency time study were carried out, in order to choose hardware components of the system, like the board and the processor. Right now the job is focused on programming, debugging and optimizing board and processor.