Self-sensing active artificial hair cells inspired by the cochlear amplifier, Part I: Theoretical and numerical realization

Abstract

The mammalian cochlear amplifier displays a unique nonlinear characteristic of amplifying or compressing the acoustic stimuli based on their level. Designing control algorithms for artificial hair cells (AHCs) with active sensing capabilities could mimic such dynamics. The present work focuses on developing a novel self-sensing active AHC scheme that implements the amplification/compression function of the AHCs and extends the authors’ previous designs for future cochlear implants or sensor design applications. The AHC functions near a Hopf bifurcation and varies the output piezoelectric voltage by a power-law relationship with the input. It is the first time that a self-sensing AHC is modeled as a quadmorph cantilever controlled by a phenomenological cubic damping controller. The AHC’s control signal is supplied to a pair of its piezoelectric layers and the output of the AHC is the sensed voltage of the second piezoelectric pair. This is in contrast to the previous studies where the AHC’s tip-velocity was measured with external sensors. The requirement of permanent external sensors in the system was a strict limitation that is eliminated in this work. An in-depth study of this novel AHC is conducted in this paper and AHC’s implementation is examined experimentally in Part II of the paper.

Self-sensing active artificial hair cells inspired by the cochlear amplifier, Part II: Experimental validation

“…Miniaturization of this active AHC and fabrication of a Microelectromechanical System (MEMS) for cochlear implants (Davaria and Tarazaga, 2017; Davaria et al, 2020) or sensor applications is only possible if external sensors such as laser vibrometers are omitted from the system. Therefore, to tackle this limitation, a self-sensing active AHC was proposed in Part I of this paper (Davaria and Tarazaga, 2023). A macro-scale quadmorph piezoelectric beam was adopted as the AHC for the first time and the piezoelectric voltage was used in the feedback loop.…”