Parameter Estimation for Hammerstein Nonlinear Controlled Auto-Regression Models

摘要：

A recursive least-squares identi.cation algorithm is developed for Hammerstein nonlinear models, which consist of memoryless nonlinear blocks followed by linear dynamical systems described by controlled auto-regression (CAR) models. Convergence analysis of the proposed algorithms indicates that the parameter estimation error consistently converges to zero under proper conditions. The simulation results show that the proposed algorithm is effective.

关键词： Recursive identification least squares Hammerstein models parameter estimation martingale convergence theorem

作者: Wei Fan Feng Ding Yang Shi

作者单位: Control Science and Engineering Research Center Jiangnan (Southern Yangtze) University Wuxi, P.R. Ch Department of Mechanical Engineering University of Saskatchewan Saskatchewan, Canada S7N 5A9

会议类型: 国际会议

会议名称: 2007 IEEE International Conference on Automation and Lofistics

会议地点: 山东济南

会议语种:英文

在线出版日期: 2007-08-18（万方平台首次上网日期，不代表论文的发表时间）