摘要
针对信号源方向时变情况,提出一种新的跟踪方法.该方法利用性能优越的最大似然估计避免了子空间跟踪类方法需要不断重复的协方差矩阵分解.为有效解决最大似然估计巨大计算量的问题,引入粒子群算法并对其进行改进,使其能够自动跟踪目标,把目标锁定在一个很小的搜索范围之内.通过大幅度缩小搜索的范围和运用群智能搜索可以有效降低算法的计算量.仿真结果表明,与子空间类算法相比,该方法具备解相干的能力和较好的跟踪精度,而且能够保证算法的实时性.
A new method to estimate direction-of-arrival （DOA） of moving sources has been proposed. Making use of a maximum likelihood algorithm（MLE） , this method avoids decompositions of the covariance matrix which must be repeated in methods based on subspace tracking. Additionally, to avoid the huge computational costs of the MLE, a particle swarm algorithm was considered and improved. In this way targets were tracked and their DOA estimated using a very small space in which the maximum could be sought. As the searching space was greatly reduced and swarm intelligence was used in searching, computational costs were significantly reduced. Simulation resuits showed that DOA estimation using the improved particle swarm algorithm has the ability to track coherent sources and performs better than methods based on subspace tracking both in tracking precision and real-time effectiveness.
出处
《哈尔滨工程大学学报》
EI
CAS
CSCD
北大核心
2009年第7期843-846,共4页
Journal of Harbin Engineering University
基金
黑龙江省科技攻关项目(GZ08A101)
关键词
动态目标
DOA估计
最大似然估计
粒子群算法
moving source
DOA estimation
maximum likelihood estimation
particle swarm algorithm
