Distributed control of sound and vibration

The control of sound and vibration in large systems with many actuators and sensors has conventionally been achieved using fully-coupled, centralised, control strategies. The complexity of such a fully coupled controller rises rapidly with the number of transducers, however, and its stability is vulnerable to individual failures. Distributed control systems, of which a decentralised controller, having only local loops, is an extreme example, are not so interconnected, and may not be so susceptible to individual failures. It is possible to design decentralised systems for the active control of sound in which the individual control loops do not significantly interact, but then only local, rather than global control is generally achieved. For active vibration control, however, decentralised controllers can be made stable, with the appropriate choices of actuators and sensors, despite strong interaction between individual control loops, and significant global performance can then be achieved. Examples are considered in active vibration isolation and active vibroacoustic control. Finally, the distributed nature of the active mechanism of hearing in the inner ear is briefly described. Although the mechanisms associated with the individual control loops are poorly understood, it is clear that they are tuned very close to instability in order to achieve the greatest sensitivity.