首页> 外文会议>SAE Noise and Vibration Conference and Exhibition >Advanced Source Localization Techniques Using Microphone Arrays
【24h】

Advanced Source Localization Techniques Using Microphone Arrays

机译:使用麦克风阵列的高级源本地化技术

获取原文

摘要

Microphone arrays used in vehicle acoustics are mainly designed for fast setup and basic evaluation (e.g. using delay-and-sum beamforming) resulting in a restriction to free field environments. Applications in vehicle interiors require advanced source localization techniques taking into account the reflections at the different panels appearing as mirror sources. Coherence filtering techniques allow for the detection of these mirror sources. An additional sensor is placed as a reference close to the main source. This reference signal is used to filter the array signals increasing the overall dynamic range of the acoustic source mapping. The discrimination of the original source and the reflections is obtained by manipulating the impulse responses between the reference signal and all microphone signals. Other advanced signal processing techniques can be used to increase the limited dynamic range of conventional beamforming such as principal component analysis removing the dominant sources. An implementation allowing for online application based on spatial filters will be presented. In addition the production requirements of large-scale arrays, for example wind tunnel measurements and pass-by or fly-over measurements, are discussed. Manufacturing a large-scale array with an appropriate precision for the microphone positions and the video system would be very expensive. A very efficient solution turns out to be a combination of a small-calibrated camera module (consisting of a video system plus a small number of microphones with high-precision positioning) and a large-scale array containing microphones with low-precision positioning. Array measurements using a standard loudspeaker for excitation serve as input for nonlinear optimization techniques to identify the positions of all microphones based on the orientation of the camera module. The techniques also allow for combining small arrays to increase the spatial resolution. This paper describes the theoretical background of different source localization techniques based on microphone arrays. Application examples are used to illustrate not only their advantages but also their potential drawbacks.
机译:车辆声学中使用的麦克风阵列主要设计用于快速设置和基本评估(例如,使用延迟和和波束成形)导致对自由场环境的限制。车辆内部的应用需要高级源本地化技术考虑到出现作为镜子源的不同面板的反射。相干滤波技术允许检测这些镜子源。额外的传感器被放置为接近主源的参考。该参考信号用于过滤阵列信号,从而增加声源映射的整体动态范围。通过操纵参考信号和所有麦克风信号之间的脉冲响应来获得原始源和反射的辨别。其他先进的信号处理技术可用于增加传统波束形成的有限动态范围,例如去除主导来源的主成分分析。将呈现允许基于空间过滤器的在线应用程序的实现。此外,还讨论了大规模阵列的生产要求,例如风隧道测量和通过或飞越或飞越测量。为麦克风位置的适当精度制造大型阵列,视频系统将非常昂贵。一个非常有效的解决方案旨在成为小校准相机模块(由视频系统组成的少量麦克风,具有高精度定位的少量麦克风)以及包含低精度定位的麦克风的大规模阵列。使用标准扬声器进行激励的阵列测量用作非线性优化技术的输入,以基于相机模块的方向识别所有麦克风的位置。该技术还允许组合小阵列来增加空间分辨率。本文介绍了基于麦克风阵列的不同源定位技术的理论背景。应用例子用于不仅说明它们的优点,还用于说明它们的潜在缺点。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号