Key Laboratory of Underwater Acoustic Communication and Marine Information Technology of the Ministry of Education, College of Ocean and Earth Sciences, Xiamen University, Xiamen 361005, People's Republic of China.
Acoustic Research Laboratory, Tropical Marine Science Institute, National University of Singapore, 12A Kent Ridge Road, Singapore, 119222, Singapore.
J Acoust Soc Am. 2018 Aug;144(2):1018. doi: 10.1121/1.5051330.
Atlantic bottlenose dolphins () can effectively discriminate between water-filled cylinders with different wall thicknesses. The dolphins' performance may be particularly good when the cylinders are thinner. The dolphins' performance is also asymmetric, in that the discrimination accuracy is not equal when the target thickness increases and decreases by the same amount. Inspired by this, a finite element model is proposed here to mimic a dolphin actively transmitting sound to discriminate between targets using acoustic echoes. The waveforms and frequency spectra of acoustic echoes from a standard cylinder and comparison cylinders with wall thickness differences of ±0.3 mm were compared. The employed model simulations show good agreement with previous experimental measurements by Au and Pawloski [(1992). J. Compar. Physiol. A (1), 41-47]. Asymmetric arrival time shifts were found for the echo peaks and troughs with the same sequence numbers when the wall thicknesses were increased and decreased by the same amount. This asymmetry became more significant for echo peaks and troughs with higher sequence numbers. Apart from these asymmetric arrival time shifts of the acoustic echoes, the patterns of echo waveforms, the spatial distributions of sound pressures in the water, and the particle vibratory displacements in the cylinders were also found to vary with cylinder thickness. The physical origin of this asymmetric discrimination by the dolphins was explored using both geometric acoustics and wave acoustics. The asymmetry observed might be caused by the circumferential surface (Lamb) wave in the cylinder wall, which is a wave acoustics phenomenon that cannot be derived from geometric acoustics. The findings in this paper might be valuable not just for understanding the mechanism of the effect described, but also for helping the development of biomimetic intelligence for robust signal processing in underwater target discrimination.
大西洋宽吻海豚()能够有效地分辨出具有不同壁厚的充水圆柱形容器。当圆柱形容器较薄时,海豚的表现可能特别出色。海豚的表现也存在不对称性,即当目标厚度增加和减少相同量时,其分辨精度并不相等。受此启发,提出了一种有限元模型来模拟海豚主动发射声音,通过声回波来区分目标。比较了标准圆柱形容器和壁厚差为±0.3mm 的比较圆柱形容器的声回波的波形和频谱。所采用的模型模拟结果与 Au 和 Pawloski [(1992). J. Compar. Physiol. A (1), 41-47] 的先前实验测量结果吻合较好。当壁厚以相同的量增加和减少时,对于具有相同序号的回波峰和波谷,发现了不对称的到达时间偏移。对于具有较高序号的回波峰和波谷,这种不对称性变得更加显著。除了这些声回波的不对称到达时间偏移外,还发现了回波波形的模式、水中声压的空间分布以及圆柱体内质点振动位移随圆柱厚度的变化。使用几何声学和波动声学探索了海豚这种不对称分辨的物理起源。观察到的不对称性可能是由圆柱壁中的圆周表面(Lamb)波引起的,这是一种波动声学现象,无法从几何声学中推导出来。本文的研究结果不仅有助于理解所描述效应的机制,还有助于开发仿生智能,以实现水下目标识别中的稳健信号处理。