膜计算优化随机最大似然DOA快速估计方法  

DOA fast estimation method for membrane computational optimization of random maximum likelihood

作　　者：向长波[1] 于玮 宋华军[2] 刘芬 Xiang Changbo;Yu Wei;Song Huajun;Liu Fen(The 41st Research Institute of China Electronic Science and Technology Group Corporation,Qingdao 266555;College of Information and Control Engineering,China University of Petroleum,Qingdao 266580;Qingdao Hisense TransTech,Qingdao 266071)

机构地区：[1]中国电子科技集团公司第四十一研究所电子测试技术重点实验室,青岛266555 [2]中国石油大学(华东)信息与控制工程学院,青岛266580 [3]青岛海信网络科技股份有限公司,青岛266071

出　　处：《高技术通讯》2019年第9期833-840,共8页Chinese High Technology Letters

基　　金：国家自然科学基金(61602517);中央高校基本科研业务费专项资金(18CX02109A)资助项目

摘　　要：随机最大似然算法(SML)是一种优秀的波达方位(DOA)估计算法,但SML解析过程中极其繁重的计算复杂度制约了该算法在实际系统中的应用。针对SML计算复杂度高的问题,提出了一种融合膜计算(MC)的随机最大似然算法。首先利用膜计算的优化框架将SML算法的解空间进行膜划分,划分为基本膜和表层膜;然后在每个基本膜内并行采用粒子群算法(PSO)进行局部寻优,同时将基本膜区域内的局部最优解送至表层膜进行全局优化;最后在表层膜区域中采用人工蜂群优化算法进行全局最优解的搜索。实验结果表明,本文算法极大地降低了SML的解析复杂度,计算时间较常用的GA、AM和PSO算法提高了超过10倍,在收敛速度方面具有显著的优势,且测向精度优于传统空间谱算法。The stochastic maximum likelihood(SML)achieves exceptional performance of estimating direction-of-arrival(DOA).However,the high computational complexity of analytic method limits SML for further applications in real-time systems.Considering the high computational complexity of SML,we explore a membrane computing algorithm for SML estimation through membrane division and definition of evolution rule and communication mechanism.First of all,the whole searching space is divided into several basic membranes and a surface membrane.In each basic membrane,the particle swarm optimization(PSO)algorithm is adopted to find the local solution.All the local solutions are collected into the surface membrane and finally the artificial bee colony optimization algorithm is used to get the global solution.Experimental results show that the calculation time of the proposed algorithm is over 10 times more than that of conventional GA,AM,PSO algorithm,which greatly reduces the computational complexity of SML and the performance is better than the traditional algorithms,in addition,the proposed method achieves significant merit of decreased convergence speed.

关 键 词：波达方位(DOA)估计 随机最大似然算法(SML) 膜计算(MC) 粒子群算法(PSO) 人工蜂群算法(ABC) 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术]