Ben Cazzolato

Danielle Moreau

Dick Peterson

Virtual sensors for active noise control

Ben S. Cazzolato, Danielle Moreau and Dick Peterson

Active noise control systems aim to reduce noise by creating anti-noise using controller driven loudspeakers. Early work in the field of active noise control focussed on achieving global control where the objective was to reduce the noise throughout the entire acoustic domain of interest, for instance a car interior. Global control does not always attenuate the noise to an acceptable level in practical problems, especially in modally dense enclosures and in sound fields which have a large spatial variance. In this case, local control might be the only feasible solution. Unfortunately, local control generally achieves the greatest noise reductions at the error sensor locations, which might not always be the location where the maximum attenuation is required. Furthermore, the zone of quiet around the error sensors tends to be very small. Two methods to overcome the problems encountered in local active noise control systems are currently under investigation. The first method aims to extend the zone of quiet by using either an energy density control strategy or a spatial control strategy. The second method, called virtual sensing, aims to move the zone of quiet away from the physical microphones to the desired location of maximum attenuation. Both methods have the potential to improve the performance of a local active noise control system. This project is focussed on developing new and improved sensing and control methodologies for local active noise control systems by combining both methods.

The figure below illustrates the concept. For more detailed information on the technique, please see the following publications page on energy density and virtual sensing.

Figure 1: Comparison of local active noise control (a) at a physical sensor and (b) at a virtual sensor.