基于FPGA的声源定位系统软硬件设计

发布时间：2018-03-25 02:09

本文选题：声源定位　切入点：FPGA　出处：《哈尔滨工业大学》2017年硕士论文

【摘要】：基于麦克风阵列的声源定位是阵列信号处理的一个重要应用,在单兵作战、狙击手方位判定等军事领域和视频会议、智能机器人等民用领域都有着广泛的应用。针对当前声源定位系统多是基于PC平台完成数据的处理和声源方位的判定,具有价格昂贵,操作复杂,功耗高,不易携带等缺点,本文提出了基于FPGA的麦克风阵列声源定位系统,以FPGA为数据处理平台,实现了小体积,便携,低功耗,价格合适的系统特性,在视频会议,野外探测和军事作战中,有着重要的应用价值。本文主要从理论研究、软硬件设计和定位实验三部分对声源定位系统展开了研究。首先,本文对波束形成、高分辨率谱估计和到达时间差三种定位算法进行了比较和研究,并对三元麦克风线性模型和六麦克风空间模型进行理论推导,分析定位误差产生原因。使用MATLAB对TDOA算法和MUSIC算法进行了仿真比较,针对FPGA定位系统的特性,选择计算量小,实时性好TDOA算法作为本文的定位算法。然后,针对系统性能要求,完成了基于FPGA的声源定位系统软硬件设计。其中硬件设计包含了信号调理电路,电源供电电路,AD采集电路,SD存储电路的设计,实现了八路信号的同步采集,存储和分析,具有便携,低功耗的特性。软件设计包含了硬件电路控制和定位算法实现两大部分,基于硬件基础,采用了模块复用、缩减数据位数、使能控制、流水结构等方式完成了程序优化并使用Modelsim对各个模块进行了仿真验证,给出了系统逻辑资源消耗量和定位程序仿真耗时量。最后,针对设计的定位系统,进行了相关的性能测试和定位实验。结果表明,系统设计较为合理,能够较好的实现声源定位。针对实验过程中产生的误差,给出了系统误差分析,并提出了相应的改进方案,为后续的研究提供了思路。最终,本文在FPGA上实现了对声源信息的采集,存储,分析和定位。
[Abstract]:Sound source location based on microphone array is an important application of array signal processing. It is used in military fields such as single combat, sniper orientation determination and video conference. Intelligent robots and other civil fields have been widely used. In view of the current sound source location system is mostly based on PC platform to complete the data processing and sound source orientation determination, with expensive, complex operation, high power consumption, difficult to carry and other shortcomings. In this paper, a microphone array sound source location system based on FPGA is proposed. With FPGA as the data processing platform, the system features of small volume, portable, low power consumption and suitable price are realized, which can be used in video conferencing, field exploration and military operations. It has important application value. In this paper, the acoustic source positioning system is studied mainly from three parts: theoretical research, hardware and software design and localization experiment. First, this paper focuses on beamforming. The high resolution spectrum estimation and time of arrival (DOA) localization algorithms are compared and studied, and the linear and six-microphone space models are derived theoretically. The reason of location error is analyzed. The TDOA algorithm and MUSIC algorithm are simulated and compared by using MATLAB. According to the characteristics of FPGA positioning system, the algorithm of TDOA is chosen as the location algorithm of this paper, which is small in computation and good in real time. Then, according to the performance requirements of the system, the algorithm is introduced. The hardware and software design of sound source location system based on FPGA is completed. The hardware design includes signal conditioning circuit, power supply circuit, AD acquisition circuit and SD storage circuit, and realizes the synchronous acquisition, storage and analysis of eight signals. The software design includes two parts: hardware circuit control and location algorithm realization. Based on hardware, module reuse is adopted to reduce the number of data bits, so that it can be controlled. Income structure and other ways to complete the optimization of the program and use Modelsim to verify the simulation of each module, given the system logical resource consumption and positioning program simulation time. Finally, for the design of the positioning system, The results show that the design of the system is reasonable and the sound source location can be realized well. The error analysis of the system is given, and the corresponding improvement scheme is put forward. Finally, this paper realizes the acquisition, storage, analysis and localization of sound source information on FPGA.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN912.3

【相似文献】