基于Zynq和蜜蜂进化遗传算法的声源实时定向系统  

Real⁃time sound source orientation system based on Zynq and Bee Evolutionary Genetic Algorithm

作　　者：陆智辉 兰昀弢 郑郁正[1] 刘凯 唐国璇 LU Zhihui;LAN Yuntao;ZHENG Yuzheng;LIU Kai;TANG Guoxuan(College of Communication Engineering,Chengdu University of Information Technology,Chengdu 610225,China)

机构地区：[1]成都信息工程大学通信工程学院,四川成都610225

出　　处：《电子设计工程》2024年第1期164-169,174,共7页Electronic Design Engineering

基　　金：青年科学基金项目(6210012231)。

摘　　要：针对声源定向系统利用多重信号分类(MUSIC)算法测向时存在精准度高,但实时性偏低的问题,基于Zynq平台和蜜蜂进化遗传算法(BEGA),设计了一款软硬件协同工作的声源实时定向系统。系统利用MEMS麦克风均匀圆阵和Zynq采集和传输声源数据,并引入BEGA提升MUSIC算法搜索谱峰的速度。实验表明,硬件平台具有不掉帧、延迟低的优良性能,同时数据处理单元利用BEGA大幅缩短了谱峰搜索的时间,并且具备精准的定向性能。因此,该系统满足实时性要求,也保留了MUSIC算法精准度高的优点。Aiming at the problem of high accuracy but low real⁃time when using Multiple Signal Classif⁃ication(MUSIC)algorithm direction finding in the sound source orientation system,a real⁃time sound source orientation system based on Zynq platform and Bee Evolutionary Genetic Algorithm(BEGA)is designed.The system uses MEMS microphone uniform circular array and Zynq to collect and transmit sound source data,and introduces BEGA to improve the speed of MUSIC algorithm to search spectral peaks.Experiments show that the hardware platform has excellent performance of no frame dropping and low delay.Besides,the data processing unit uses BEGA to greatly shorten the time of spectral peak search,and has accurate orientation performance.Therefore,the system meets the requirements of real⁃time and retains the advantages of high accuracy of music algorithm.

关 键 词：声源定向 MUSIC Zynq 蜜蜂进化 MEMS 

分 类 号：TN971[电子电信—信号与信息处理]