DEVELOPMENT AND EVALUATION OF A WIRELESS ACTIVE NOISE CONTROL SYSTEM IN AN ENCLOSED ENVIRONMENT

Abstract

Active noise control (ANC) is a method used to reduce acoustical vibrations mainly within the low-frequency range. This thesis aims to implement a wireless active noise control system in an enclosed environment, evaluate relevant system performance characteristics and compare it with a wired version of an active noise control system. In doing so, a method to build the wireless ANC system is outlined. The ANC system used the Filtered X-Least Mean Square algorithm (FxLMS) as a standard method to control the noise. In addition, online system identification is implemented to track any changes in the secondary path while a user is moving around. A single-frequency tonal noise source, secondary loudspeakers, and microphones are used for testing and evaluation of the complete wireless system. It is found that the wireless system works well when validation of a single-channel system is carried out in an air handling duct test rig. Similar noise reduction levels were obtained for the wireless systems when a user wears an in-ear error microphone; it is found that the system performs well for stationary cases. When the user walks at an average pace, the system can diverge if offline system identification is used. However, if online system identification is used, the system performs well.