JADE算法在星系团图像盲源分离中的应用

Blind Separation of Galaxy Clusters Based on Joint Approximative Diagonalization of Eigenmatrix Algorithm

星系团是研究宇宙变化的重要现象,也是宇宙在电离信号研究中的强前景。如何从强烈的银河系和河外离散源前景中分离出星系团一直是制约探索和观测宇宙的关键问题之一。实验通过利用扩展联合对角化(JADE)分离方法对模拟的未考虑仪器效应的前景图进行盲源分离探讨研究,获得了其中的星系团成分。结果表明,JADE分离方法能有效地对模拟的星系团成分进行分离。

Galaxy clusters are the important phenomenon in studying the changes of the universe and the bright foreground compo- nent of the redshifted cosmic reionization signals.It is one of the key issues in restricting the exploration and observation of the uni- verse the clusters from the bright foreground of the Galactic and to separate extragalactic discrete sources. In the experiment the Joint Approximative Diagonalization of Eigenmatrix （JADE） algorithm is used to disentangle the cluster component from the simulated bright foreground on no consideration of the effects of instrument. The experimental results illustrates that the JADE algorithm can be used to separate the cluster component from the bright foreground validly.