Suppr超能文献

一种用于在未知各向异性结构材料特性的情况下精确定位声源的两步混合技术。

A two-step hybrid technique for accurately localizing acoustic source in anisotropic structures without knowing their material properties.

作者信息

Kundu T, Yang X, Nakatani H, Takeda N

机构信息

Department of Civil Engineering and Engineering Mechanics, University of Arizona, Tucson, AZ 85721, USA.

State Key Laboratory of Hydraulics and Mountain River Engineering, College of Water Resource & Hydropower, Sichuan University, Chengdu 610065, China.

出版信息

Ultrasonics. 2015 Feb;56:271-8. doi: 10.1016/j.ultras.2014.08.009. Epub 2014 Sep 7.

Abstract

Acoustic source localization techniques generally assume straight line propagation of waves from the acoustic source to the sensor. However, it is well-known that in anisotropic plates the acoustic energy does not always propagate in straight lines. Even for isotropic plates containing a cavity or an inclusion between the acoustic source and the sensor the straight line propagation assumption is violated. In such cases only options available in localizing acoustic source is to use relatively expensive distributed sensor systems, or to follow time reversal techniques based on the impulse response functions which is labor intensive and computationally demanding. A two-step hybrid technique is proposed in this paper for predicting acoustic source in anisotropic plates. During the first step it was assumed that the waves propagated along straight lines from the acoustic source to the sensor. The source was localized with this simplifying assumption. Then this first prediction was improved in the second step by solving an optimization problem. Experimental results showed that the second step always moved the estimates towards the actual source location. Thus it always reduced the prediction error irrespective of whether the final prediction coincided with the actual source location or not.

摘要

声源定位技术通常假定声波从声源到传感器是直线传播的。然而,众所周知,在各向异性板中,声能并非总是沿直线传播。即使对于在声源和传感器之间包含一个空腔或内含物的各向同性板,直线传播假设也不成立。在这种情况下,声源定位的唯一可用选项是使用相对昂贵的分布式传感器系统,或者采用基于脉冲响应函数的时间反转技术,而这既费力又对计算要求很高。本文提出了一种两步混合技术来预测各向异性板中的声源。在第一步中,假定波从声源沿直线传播到传感器。基于这个简化假设对声源进行定位。然后在第二步中,通过求解一个优化问题对第一次预测进行改进。实验结果表明,第二步总是使估计值朝着实际声源位置移动。因此,无论最终预测是否与实际声源位置重合,它总是能减小预测误差。

文献AI研究员

20分钟写一篇综述,助力文献阅读效率提升50倍。

立即体验

用中文搜PubMed

大模型驱动的PubMed中文搜索引擎

马上搜索

文档翻译

学术文献翻译模型,支持多种主流文档格式。

立即体验