滑动平均噪声干扰双输入多率系统最小二乘迭代辨识  

Extended least-squares based iterative identification for two-input multirate systems with moving average noises

作　　者：陆静[1] 张彩霞[1] 丁锋[1] 

机构地区：[1]江南大学控制科学与工程研究中心,无锡214122

出　　处：《东南大学学报（自然科学版）》2008年第A02期77-80,共4页Journal of Southeast University：Natural Science Edition

基　　金：国家自然科学基金资助项目(60574051);江苏省自然科学基金资助项目(BK2007017)

摘　　要：研究了滑动平均噪声干扰的双输入多率系统最小二乘迭代辨识算法.首先推导出2个输入通道采样周期不相等的多率系统的离散时间状态空间模型,得出对应的传递函数模型.针对辨识模型信息向量中存在不可测噪声项的困难,利用最小二乘迭代原理,将未知噪声变量用其迭代估计值来代替,提出了这类双输入多率采样数据系统的最小二乘迭代辨识算法.最后通过仿真例子比较了最小二乘迭代辨识算法与递推增广最小二乘算法的辨识效果,说明了所提出算法的参数估计精度较高.The extended least-squares based iterative （LSI） identification algorithm is studied for two-input multirate systems with moving average noises. The state-space models are derived for multirate systems with two different input sampling periods and further the corresponding transfer function models are obtained. To solve the difficulty of having unmeasurable variables in the information vector of identification models, the LSI algorithm is presented by replacing the unmeasurable varia- bles with their iterative estimates based on the iterative least-squares principle. Finally, the simulation results indicate that the proposed algorithm has highly accurate parameter estimation, compared with the recursive extended least squares identification algorithm.

关 键 词：迭代辨识 递推辨识 参数估计 最小二乘 多率系统 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]