匹配场声源定位的并行计算方法研究

摘要

Through studying and analyzing the inherent parallelism which existed in algorithm of the match field sound source localization, frame of parallel computing algorithm had constructed in the match field localization in order to shorten the running time and accelerate the speed of localization algorithm effectively. Parallelization grid method was introduced in detail , which was applied to the match field localization algorithm. And parallel algorithm had been carried on the test and the corresponding analysis in the MPI parallel environment, in the way of interconnecting binuclear PC machines as the parallel computing hardware platform through fast internet, choosing Windows operate system and FORTRAN compiler, and using TCP/IP as the standard communication protocols. The study shows the proposed parallel algorithm running results are better,so questions of the load big computation and long time-consuming etc. Can be settled effectively to exploiting parallelization in matched field processing methods, in order to provide settle path for fast sound source localization.%为了达到有效缩短算法运行时间、加快定位速度的目的,研究并分析了匹配场声源定位算法中存在的固有并行性,构建了匹配场声源定位的并行计算算法框架.详细介绍了匹配场声源定位算法的网格法并行化方法；并以互连的双核PC机作为并行计算硬件平台,选用Windows操作系统、FORTRAN编译器,使用TCP/IP作为标准的通信协议,在MPI并行环境中对并行算法进行了测试和相应分析.研究表明,提出的并行算法运行效果较好,可以通过开发匹配场处理方法的并行性,有效解决信号匹配处理中遇到的计算量大、耗时长等问题,为实现快速声源定位提供解决途径.