Bean, Jacob Jon2019-10-182019-10-182018-04-25vt_gsexam:14591http://hdl.handle.net/10919/94626This dissertation is concerned with the active control of local sound fields, as applied to an active headrest system. Using loudspeakers and microphones, an active headrest is capable of attenuating ambient noise and providing a comfortable acoustic environment for an occupant. A finite element (FE) model of an active headrest is built and analyzed such that the expected noise reduction levels could be quantified for various geometries as well as primary sound field conditions. Both plane wave and diffuse primary sound fields are considered and it is shown that the performance deteriorates for diffuse sound fields. It is then demonstrated that virtual sensing can greatly improve the spatial extent of the quiet zones as well as the attenuation levels. A prototype of the active headrest was constructed, with characteristics similar to those of the FE model, and tested in both anechoic and reverberant sound fields. Multichannel feedforward and feedback control architectures are implemented in real-time and it is shown that adaptive feedback systems are capable of attenuating band-limited disturbances. The spatial attenuation pattern surrounding the head is also measured by shifting the head to various positions and measuring the attenuation at the ears. Two virtual sensing techniques are compared in both feedback and feedforward architectures. The virtual microphone arrangement, which assumes that the primary sound field is equivalent at the physical and virtual locations, results in the best performance when used in a feedback system attenuating broadband disturbances. The remote microphone technique, which accounts for the transfer response between the physical and virtual locations, offers the best performance for tonal primary sound fields. In broadband sound fields, a causal relationship rarely exists between the physical and virtual microphones, resulting in poor performance.ETDIn CopyrightActive headrestactive noise controladaptive controlhybrid controlfinite element modelingDissertation