摘要
针对相干信号源方向时变的情况,分析了样本协方差矩阵的更新,在此基础上提出了一种基于粒子群算法的跟踪方法.该方法直接利用性能优越的最大似然估计器,避免了子空间跟踪类方法需要运用空间平滑等解相干技术对数据协方差矩阵进行的预处理和数据协方差矩阵不断分解的过程,同时通过锁定目标、大幅度缩小搜索的范围和运用群智能搜索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,并且能够保证算法的实时性.
As the direction ot coherent signal source varies with time, we proposed a tracking method based on particle swarm algorithm after an analysis of renewal of sample＇s covarianee matrix. With the direct application of maximum likelihood algorithm, it can avoid the pretreatment of the covariance matrix with the spatial smoothing method, and the decompositions of the covariance matrix which otherwise should be repeated in the methods based on subspaee tracking. Furthermore, it can substantially reduce the computational effort of maximum likelihood algorithm by loeking the targets, shrinking the search region and using group intelligent searching. Simulation results show that, compared with the methods based on subspaee tracking, the proposed method has excellent ability to track coherent sources, with high precision performance and real-time property.
出处
《应用科技》
CAS
2008年第11期26-30,共5页
Applied Science and Technology
