Hammerstein模型的记忆效应盲辨识方法  

Blind Identification Method of Memory Effect for Hammerstein Model

作　　者：胡啸[1] 马洪[1] 

机构地区：[1]华中科技大学电子与信息工程系武汉国家光电实验室,武汉430074

出　　处：《计算机工程》2011年第6期18-20,23,共4页Computer Engineering

基　　金：国家自然科学基金资助项目(10975056);航天科技创新基金资助重点项目(CASC200904);武汉光电国家实验室创新基金资助项目(Z080005)

摘　　要：针对未知记忆深度的Hammerstein模型,提出一种基于高阶累积量的Hammerstein模型记忆效应盲辨识方法。将Hammerstein模型中对记忆深度的确定转换为对模型输出信号高阶累积量扩展矩阵的求秩问题,给出对角元素乘积(NPODE)方法以确定记忆深度,分别比较该方法与GM直接定阶法、拐点法的鲁棒性。结合提出的记忆深度估计算法,给出线性记忆模块系数的提取方法。理论推导与仿真结果表明,线性记忆模块系数的提取过程不受无记忆非线性效应的影响。Aiming at the Hammerstein model with unknown memory depth,this paper proposes a blind identification method of memory effect for Hammerstein model based on high cumulant.The memory depth determination is converted into finding the rank of extended matrix constructed by cumulants of system output.In order to yield the rank of cumulant matrix.It proposes Nonlinear Product Of Diagonal Entry（NPODE） method,and simulation verifies its robustness by comparing with GM method and inflexion method.Linear block coefficients extraction method is given,the theoretical derivation and simulation result indicates that the extraction process is not affected by the strength of nonlinearity of Hammerstein model.

关 键 词：对角元素乘积方法 HAMMERSTEIN模型 记忆效应 累积量 记忆深度 

分 类 号：TN911.7[电子电信—通信与信息系统]