共 49 条
Multi-innovation Stochastic Gradient Parameter Estimation for Input Nonlinear Controlled Autoregressive Models
被引:26
作者:
Xiao, Yongsong
[4
]
Song, Guanglei
[3
]
Liao, Yuwu
[2
]
Ding, Ruifeng
[1
]
机构:
[1] Jiangnan Univ, Sch Internet Things Engn, Wuxi 214122, Peoples R China
[2] Xiangfan Univ, Sch Phys & Elect Engn, Xiangfan 441053, Peoples R China
[3] Wuxi Inst Technol, Sch Mech Technol, Wuxi 214121, Peoples R China
[4] Jiangnan Univ, Key Lab Adv Proc Control Light Ind, Minist Educ, Wuxi 214122, Peoples R China
关键词:
Hammerstein model;
multi-innovation identification;
parameter estimation;
recursive identification;
stochastic gradient;
system modeling;
DUAL-RATE SYSTEMS;
IDENTIFICATION METHODS;
AUXILIARY MODEL;
PERFORMANCE ANALYSIS;
ITERATIVE SOLUTIONS;
OUTPUT ESTIMATION;
MATRIX EQUATIONS;
ARMAX SYSTEMS;
ALGORITHMS;
D O I:
10.1007/s12555-012-0322-8
中图分类号:
TP [自动化技术、计算机技术];
学科分类号:
0812 ;
摘要:
This paper proposes a multi-innovation stochastic gradient (MISG) parameter estimation algorithm for an input nonlinear controlled autoregressive (IN-CAR) model, i.e., a Hammerstein nonlinear CAR system, by expanding the innovation length. The analysis and simulation results indicate that the proposed MISG algorithm can generate more accurate parameter estimates for IN-CAR systems compared with the stochastic gradient algorithm.
引用
收藏
页码:639 / 643
页数:5
相关论文