Optimal active noise control in large rooms using a “locally global” control strategy

Abstract

The feasibility of applying active noise control in large rooms, in which global control is very difficult, is investigated. Local control can only ensure sound attenuation near error-sensor positions. Considering that workers usually work only in certain regions of a workroom, a new “locally global” control strategy is proposed. The objective is to reduce the acoustic potential energy in the target region. Compared to local control, “locally global” control ensures overall noise reduction over the target region. Compared to global control, it allows the number of control channels to be significantly reduced. The placements of the control loudspeakers and error microphones must be optimized to ensure that, while the sum of the squared sound pressures at the error sensors is minimized, the potential energy in the target region is reduced. Room sound fields are modeled using the image-source method and point sources. Genetic algorithms are used to optimize the locations of the control loudspeakers and error microphones. Both numerical and experimental results are presented. The sensitivities of control performance to variations in the excitation frequency, the control-source positions, and the error-sensor locations are investigated.