Interference structure-based directional acoustic-band amplifier for enhanced sound sensing

Abstract

Listening to a desired sound from a cacophonous background remains a formidable challenge using a compact, single acoustic sensor. In this study, we developed a novel interference structure with a phase cancellation mechanism for directional bandpass filtering and amplification, dramatically enhancing target sound sensing in heavy-noise environments. The compact design (0.2 $L$), a log-scale reduction compared to leading methods, the proposed structure effectively amplifies from 1873 Hz to 22 kHz with an outstanding structural efficiency (21.60). In contrast to traditional acoustic metamaterials that require re-fabrication for tuning, the structure enables target frequency tuning simply by rotating its orientation. Under 100  dB noise, the structure improves the peak magnitude of the target frequency by up to 4.82 times. In fault diagnosis, under 84  dB noise factory, amplifies critical CNC tool features by up to 19.9 times. Notably, a CNC tool fault diagnosis AI model achieved a 78.6 % true positive rate under 84  dB noise using the structure, compared to 0.0 % in the free-field condition. This work offers a new paradigm for compact and high-performance acoustic sensing, combining sensitivity, selectivity, and robustness, with strong potential for deployment in advanced intelligent fault diagnostics in extreme noise environments.