Behera Santosh Kumar, Das Debi Prasad, Subudhi Bidyadhar
Process Modeling and Instrumentation Cell, CSIR-Institute of Minerals and Materials Technology, Bhubaneswar-751013, India.
Department of Electrical Engineering, National Institute of Technology, Rourkela, India.
J Acoust Soc Am. 2017 Aug;142(2):573. doi: 10.1121/1.4996125.
The active headrest is an important application of active noise control (ANC) where the occupant gets a quiet zone during use. However, its performance degrades due to head movement, which is an important issue. Unlike conventional active headrest algorithm, in the present paper, the two error microphones are attached to a head attachment such as a hat or band so that both error microphones can move along with the head, keeping the relative distance of ears and the microphones intact. The two secondary loudspeakers are kept near to the head, facing the ears. The performance of the active headrest was studied through real-time experimentation considering head movement. The innovation in this paper is that the ANC algorithm uses signals of moving error microphones to tune the controller without re-estimating the secondary path transfer functions. Since, the phase error of secondary path estimates at different head positions remains within ±90°, the secondary paths estimated once, when the head is positioned at the center, can be used for other head positions. This is established through real-time experiments using dSpace 1104 ACE system (dSpace, Germany).
主动式头枕是主动噪声控制（ANC）的一项重要应用，使用过程中能为驾乘人员提供一个安静区域。然而，其性能会因头部移动而下降，这是一个重要问题。与传统的主动式头枕算法不同，在本文中，两个误差传声器被安装在如帽子或头带等头部附件上，这样两个误差传声器都能随头部移动，保持耳朵与传声器之间的相对距离不变。两个次级扬声器靠近头部放置，朝向耳朵。通过考虑头部移动的实时实验研究了主动式头枕的性能。本文的创新之处在于，ANC算法使用移动误差传声器的信号来调整控制器，而无需重新估计次级路径传递函数。由于在不同头部位置的次级路径估计的相位误差保持在±90°以内，因此当头部位于中心位置时一次性估计的次级路径可用于其他头部位置。这是通过使用dSpace 1104 ACE系统（德国dSpace公司）进行实时实验得以验证的。
