Active control of repetitive transient noise

This paper presents the possibilities of active noise control techniques to reduce repetitive transient noise. Although there is a significant difference in nature between periodic and transient noise, up till now no specific research on active noise control of transient noise was reported. The presented research focuses on the development of control algorithms, dedicated for the control of repetitive transient noise. The effectiveness of the control algorithms is demonstrated for the reduction of impulse noise in a duct by a secondary loudspeaker. A linear time-invariant feedback controller is developed to drive this loudspeaker. The performance of this time-invariant controller is limited for several reasons: time variance of the system parameters, limited controller bandwidth, nonlinearities, etc. However, when the noise consists of successive impulsive sounds that exhibit a repetitive character, it is possible to extend the developed time-invariant feedback controller with a learning behaviour, based on the additional information about the repetitiveness. At each pulse, the control signal for the secondary loudspeaker can be adapted on the basis of the residual noise at the previous pulses, such that the radiated noise is reduced. This latter control technique, called iterative learning control, can significantly increase the performance of a feedback active noise control system.