A new algorithm is proposed for joint diagonalization. With a modified objective function, the new algorithm not only excludes trivial and unbalanced solutions successfully, but is also easily optimized. In addition, ...A new algorithm is proposed for joint diagonalization. With a modified objective function, the new algorithm not only excludes trivial and unbalanced solutions successfully, but is also easily optimized. In addition, with the new objective function, the proposed algorithm can work well in online blind source separation (BSS) for the first time, although this family of algorithms is always thought to be valid only in batch-mode BSS by far. Simulations show that it is a very competitive joint diagonalization algorithm.展开更多
基金supported partly by the Key Program of National Natural Science Foundation of China (U0635001U0835003)+3 种基金the National Natural Science Foundation of China (60505005 60674033 60774094)the Natural Science Fundof Guangdong Province (05006508).
文摘A new algorithm is proposed for joint diagonalization. With a modified objective function, the new algorithm not only excludes trivial and unbalanced solutions successfully, but is also easily optimized. In addition, with the new objective function, the proposed algorithm can work well in online blind source separation (BSS) for the first time, although this family of algorithms is always thought to be valid only in batch-mode BSS by far. Simulations show that it is a very competitive joint diagonalization algorithm.
文摘联合对角化方法是求解盲源分离问题的有力工具.但是现存的联合对角化算法大都只能求解实数域盲源分离问题,且对目标矩阵有诸多限制.为了求解更具一般性的复数域盲源分离问题,提出了一种基于结构特点的联合对角化(Structural Traits Based Joint Diagonalization,STBJD)算法,既取消了预白化操作解除了对目标矩阵的正定性限制,又允许目标矩阵组为复值,具有极广的适用性.首先,引入矩阵变换,将待联合对角化的复数域目标矩阵组转化为新的具有鲜明结构特点的实对称目标矩阵组.随后,构建联合对角化最小二乘代价函数,引入交替最小二乘迭代算法求解代价函数,并在优化过程中充分挖掘所涉参量的结构特点加以利用.最终,求得混迭矩阵的估计并据此恢复源信号.仿真实验证明与现存的有代表性的对目标矩阵无特殊限制的复数域联合对角化算法FAJD算法及CVFFDIAG算法相比,STBJD算法具有更高的收敛精度,能有效地解决盲源分离问题.
文摘针对多分量调频信号源混合相交时频分布盲分离,提出白化-均匀加权非正交联合对角化(Whitening-Uniformly Weighted Exhaustive Diagonalization using Gauss iteration,简称W-UWEDGE)方法估计混合矩阵。白化对相关信号去冗余处理,无需约束源信号概率密度形式,仅限制源之间整个时频面上无完全重合成分,非正交联合对角化则针对复数域。首先将非正交联合对角化可辨识性从时延平面推广至二次型时频平面,然后利用基于白化处理的梯度范数选择自项时频点(auto-time frequency point),进而利用均匀加权近似联合对角化方法估计混合矩阵,分析Amari error值随信噪比及时频矩阵个数的变化规律,与针对混合信号时间历程及时频分布的两类分离方法进行性能比较,显示出所提盲分离方法的优越性。最后应用于转子运行状态识别与齿轮复合故障源分离。仿真与实验数据分析表明所提出方法分离多分量调频相关源的有效性。