UD分解在辨识算法中的应用  

An Application of UD Factorization to Identification Algorithms

作　　者：肖德云[1] 牛绍华[1] 

机构地区：[1]清华大学

出　　处：《控制与决策》1990年第4期12-18,11,共8页Control and Decision

摘　　要：本文利用信息压缩阵的概念和UD分解技术,实现了对最小二乘类辨识算法的统一处理。这种处理方法同样也适用于多变量过程的最小二乘辨识,它不仅可以同时递推辨识出模型的参数和阶次,而且揭示了各种辨识算法之间的内在联系。An unified identrfication-method, which can be applied to a kind of least-squares problem, is proposed in the paper, utilizing the matrix, called the condensed information matrix [1], and UD factorization technique. The method is also applicable to least-squares identification of multivariable systems. All the model parameters and orders can be obtained simutaheously using the method.And the internal relations of the identification algorithms can be delineated.

关 键 词：UD分解 算法 自动化系统 系统辨识 

分 类 号：TP11[自动化与计算机技术—控制理论与控制工程]