Title :
Modified recursive extended least squares identification algorithms
Author :
Ai-guo Wu ; Zhi-Guang Wang
Author_Institution :
Shenzhen Grad. Sch., Harbin Inst. of Technol., Shenzhen, China
Abstract :
For ARMAX models, modified recursive extended least squares identification algorithms are presented. The basic idea lies in two aspects. One is to decompose the original system into two subsystems. The other is that the most recent information is used to update the parameters, which is different from the hierarchical principle. A simulation example is employed to test the effectiveness of the proposed algorithms.
Keywords :
autoregressive moving average processes; least squares approximations; ARMAX model; autoregressive moving average with exogenous input model; hierarchical principle; recursive extended least squares identification algorithm; Algorithm design and analysis; Autoregressive processes; Convergence; Indexes; Noise; Prediction algorithms; Vectors; ARMAX; Extended least squares identification; Hierarchical principle;
Conference_Titel :
Control and Decision Conference (CCDC), 2013 25th Chinese
Conference_Location :
Guiyang
Print_ISBN :
978-1-4673-5533-9
DOI :
10.1109/CCDC.2013.6561177
