基于麦克风阵列的声源定位算法研究

发布时间：2018-07-05 18:29

本文选题：麦克风阵列 + 声源定位　；参考：《南京大学》2015年硕士论文

【摘要】：在阵列信号处理领域,基于麦克风阵列的声源定位技术是一项研究热点,其在军事和民用等领域都得到了广泛应用。在实际应用中,由于存在环境噪声和混响干扰,尤其是在室内环境中(如小型会场、教室、办公室),实现精确定位是当前声源定位技术的主要难点。此外,声源定位还存在运算量较大,系统成本较高等问题。针对这些问题,本文致力于室内环境下麦克风阵列声源定位系统的算法研究和系统实现工作,重点研究了基于时延估计的声源定位算法,提出了基于GCC改进算法的分数时延估计算法,并在实际环境中搭建了定位系统,采集了实测数据,验证了算法的有效性。本文主要工作如下：1.对当前三种声源定位算法进行理论分析,基于时延估计的声源定位算法具有算法实用性高,实时性好等优点,选择其作为本文的研究重点；2.分析了麦克风阵列的空间模型(远场模型和近场模型)以及信号接收模型(理想模型和实际模型),根据IMAGE模型,给出了房间冲击响应的表达式和模型仿真；3.对时延估计算法进行系统研究,重点介绍了广义互相关和自适应最小均方两种时延估计算法；在广义互相关算法的基础上,针对低信噪比、高混响的语音环境,利用混响能量比例系数,提出了一种基于GCC改进算法的分数时延估计算法,进一步提高了时延估计的精确度；4.对几何定位算法做了系统的研究,重点研究了最大似然估计算法和最小二乘估计算法；5.设计开发一套基于十字形麦克风阵列的声源定位系统,详细介绍了声源定位系统整体结构,并对整个定位系统做了相关测试实验,分析了可能导致误差产生的原因。
[Abstract]:In the field of array signal processing, sound source location based on microphone array is a research hotspot, which has been widely used in military and civilian fields. In practical applications, due to the presence of environmental noise and reverberation interference, especially in the indoor environment (such as small meeting room, classroom, office), accurate location is the main difficulty of the current sound source location technology. In addition, there are many problems in sound source location, such as large computation and high system cost. In order to solve these problems, this paper focuses on the algorithm research and system implementation of microphone array sound source location system in indoor environment, and focuses on the acoustic source location algorithm based on time delay estimation. A fractional time delay estimation algorithm based on GCC improved algorithm is proposed, and a positioning system is built in the actual environment, and the measured data are collected to verify the validity of the algorithm. The main work of this paper is as follows: 1. Based on the theoretical analysis of three current acoustic source location algorithms, the acoustic source location algorithm based on time delay estimation has the advantages of high practicability and good real-time performance, so it is chosen as the focus of this paper. The space model (far-field model and near-field model) and the signal receiving model (ideal model and actual model) of microphone array are analyzed. According to the image model, the expression of room impulse response and the simulation of the model are given. Based on the generalized cross-correlation algorithm and the adaptive minimum mean square algorithm, the generalized cross-correlation algorithm and the adaptive least mean square algorithm are introduced systematically, and then, for the low signal-to-noise ratio and high reverberation environment, the time delay estimation algorithm is studied systematically. A fractional time delay estimation algorithm based on GCC improved algorithm is proposed based on reverberation energy proportional coefficient, which improves the accuracy of time delay estimation. The geometric location algorithm is studied systematically, and the maximum likelihood estimation algorithm and the least square estimation algorithm are studied emphatically. A sound source location system based on cross-shaped microphone array is designed and developed. The overall structure of sound source positioning system is introduced in detail.
【学位授予单位】：南京大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TN641;TN911.7

【参考文献】