基于STM32的麦克风阵列声源定位系统的研究

发布时间：2018-06-28 16:35

本文选题：声源定位 + 麦克风阵列　；参考：《燕山大学》2014年硕士论文

【摘要】：基于麦克风阵列的声源定位技术在视频会议、车载电话、数字助听器等领域有广泛的应用。随着科技的进步，ARM以其高性能、低成本、小体积显示出了极大的优势。如何在ARM上实现高精度的声源定位受到研究者们的普遍关注。在各种声源定位算法中，多重信号分类技术精度最高，性能较好，，但受到计算量的限制，其在ARM上较难实现。本文主要研究了该方法，并对其改进使之能够在ARM上实现。首先，概要的阐述了基于麦克风阵列的声源定位方法，比较其优缺点。其次，详细介绍了麦克风阵列的信号模型，阐述了多重信号分类技术的原理并对其进行仿真分析。以此为基础，研究了改进的MUSIC算法，该方法结合最大功率的频点定位和免特征分解算法，解决了MUSIC算法计算量大的问题，并实现了对多频信号的定位。再次，以STM32F103RBT6为核心处理器件，设计了一个基于ARM的实时声源定位系统，详细介绍了该系统的硬件结构和软件程序设计。整个系统由前端麦克风阵列声音采集模块、模拟信号预处理模块以及声音信号识别测向模块构成，其中改进的MUSIC算法为本文的核心。最后，在实际环境中对本文设计的声源定位系统进行了大量的测试，实验结果表明，本文的改进算法具有更好的定位性能，估计误差和均方根误差能够控制在一个较小的范围内，可用在实时定位系统中。
[Abstract]:The sound source location technology based on microphone array has been widely used in video conference, on-board telephone, digital hearing aid and so on. With the progress of science and technology, arm has shown great advantages with its high performance, low cost and small volume. How to achieve high accuracy sound source location on arm has been paid more attention by researchers. Among all kinds of acoustic source location algorithms, multi-signal classification technology has the highest accuracy and better performance, but it is difficult to implement on arm because of the limitation of computation. This paper mainly studies this method and improves it so that it can be implemented on arm. Firstly, the sound source location method based on microphone array is briefly described, and its advantages and disadvantages are compared. Secondly, the signal model of microphone array is introduced in detail. Based on this, the improved music algorithm is studied, which combines the maximum power frequency location algorithm and the feature free decomposition algorithm to solve the problem of large computation of music algorithm, and realize the localization of multi-frequency signal. Thirdly, with STM32F103RBT6 as the core processing device, a real-time sound source location system based on arm is designed. The hardware structure and software program design of the system are introduced in detail. The whole system is composed of front microphone array sound acquisition module, analog signal preprocessing module and sound signal recognition and direction finding module, among which the improved music algorithm is the core of this paper. Finally, a large number of tests on the sound source location system designed in this paper are carried out in the actual environment. The experimental results show that the improved algorithm has better localization performance. Estimation error and root mean square error can be controlled in a small range and can be used in real time positioning system.
【学位授予单位】：燕山大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN912.3

【引证文献】