The effect of head-tracking resolution on the stability and performance of a local active noise control headrest system.

Abstract

Incorporating head-tracking techniques into local active noise control headrest systems enables the plant model used in the controller to be updated dynamically as the user moves their head. This reduces the mismatch between the plant model and the physical plant responses from the secondary sources to the users' ears, which increases the achievable noise reduction when head movement occurs. In practice, since the plant models for different head positions must be identified during a calibration procedure, it is necessary to limit the head-tracking resolution to constrain the complexity of this procedure. This leads to errors between the physical and modelled plant responses as the user's head moves, which impacts the control system's stability and performance. However, the relationship between the control system behaviour and the tracking accuracy is not well understood. This paper investigates the impact of head-tracking resolution, considering translational and rotational movements, on the stability and performance of an active headrest. Assuming the error signals at the user's ears are available for adaptive control, it is shown that the system has an upper-frequency limit beyond which controller instability occurs, and this frequency is influenced by the tracking resolution, the initial head position, and the type of head movement.