主动实时声定位系统的软件设计

发布时间：2018-06-15 05:32

本文选题：实时声定位 + TDOA　；参考：《中北大学》2017年硕士论文

【摘要】：陆地爆炸试验和军事演习时,为了进行爆炸威力技术参数测试,常在离爆炸点一定距离处随机布置众多冲击波传感器。为实现对大量冲击波传感器位置的快速、准确定位,本文提出了主动实时声定位的解决方案,并重点研究了主动实时声定位的软件设计实现与优化。论文构建了基于声达时间差(TDOA)定位原理的主动实时声定位的系统模型。主动实时声定位系统由无线监控终端、声信号发射系统、声信号接收处理定位系统三大子系统组成。设置了扬声器将发出声信号的频率及编码,并设计编写了前两个子系统的软件程序。考虑到处理定位系统的实时性和软件架构的优越性,声信号接收处理定位系统采用基于ARM9的μC/OS-II实时操作系统的软件编程。论文初步研究设计了声信号处理识别涉及的四个过程的算法,细致深入地研究了提高处理实时性、软件系统鲁棒性、定位精确性三方面性能的软件结构优化策略和部分算法的优化方案。在处理实时性方面,提出了快速频率检测算法及个别细节部分计算方法的优化方案。在软件系统鲁棒性方面,按能量特征对有效信号区间再细分段;确定特征信号起始点时,对与特征信号相关度较好对应的位置点均保存备选;信源信号快速测频时,对已识别的频率下次不再检测,且将首轮测频结果保存以防首轮测频有未识别频率时对此频率进行二轮检测。在定位精确性方面,用相位矫正方法纠正声信号起始点偏差并据环境温度精确计算声速来提高距离差的精度。代入实验数据并将优化前后的效果性能作比较,各方面性能优化后确实得到极大提升。多次反复实验证实,主动实时声定位系统定位大量传感器的方案是可行的。经软件程序结构和部分算法的深入优化,系统的鲁棒性、实时性、精确性极大提高。目前,定位精度达40cm,处理时间为5s。
[Abstract]:In order to test the technical parameters of explosion power, a large number of shock wave sensors are randomly arranged at a certain distance from the explosion site during the land explosion test and military exercises. In order to locate a large number of shock wave sensors quickly and accurately, this paper presents a solution of active real-time acoustic localization, and focuses on the software design and optimization of active real-time acoustic positioning. In this paper, a system model of active real time acoustic localization based on the principle of time difference of arrival (TDOA) is constructed. The active real-time acoustic positioning system consists of three subsystems: wireless monitoring terminal, acoustic signal transmitting system and acoustic signal receiving and processing positioning system. The frequency and coding of the sound signal will be set up by the loudspeaker, and the software program of the first two subsystems is designed and written. Considering the advantages of real-time processing and software architecture, the acoustic signal receiving and processing positioning system adopts the software programming of 渭 C / OS-II real-time operating system based on ARM9. In this paper, the four algorithms involved in acoustic signal processing and recognition are preliminarily studied, and the robustness of software system is studied in detail and deeply. The optimization strategy of software structure and the optimization scheme of some algorithms are presented in this paper. In the aspect of real-time processing, the optimization scheme of fast frequency detection algorithm and the calculation method of some details is put forward. In the aspect of software system robustness, the effective signal interval is subdivided according to the energy feature; when the starting point of the feature signal is determined, the location corresponding to the feature signal correlation degree is preserved. The identified frequency will not be detected next time, and the first frequency measurement results will be saved in case there is an unrecognized frequency in the first frequency measurement, and the frequency will be detected by two rounds. In the aspect of location accuracy, phase correction is used to correct the initial deviation of acoustic signal and the sound velocity is calculated accurately according to ambient temperature to improve the accuracy of distance difference. With the experimental data and the comparison of the performance before and after optimization, the performance of all aspects of the optimization has been greatly improved. Repeated experiments have proved that the active real-time acoustic positioning system is feasible to locate a large number of sensors. The robustness, real-time and accuracy of the system are greatly improved by deep optimization of the software program structure and some algorithms. At present, the positioning accuracy is up to 40 cm and the processing time is 5 s.
【学位授予单位】：中北大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN912.3

【参考文献】