Strategic Study of CAE >> 2006, Volume 8, Issue 2
Identification of Time-delay Systems and Correction of NARMA Model
1. School of Electrical and Automation Engineering , Tianjin University , Tianjin 300072 , China
2. College of Automation Engineering , Qingdao University , Qingdao , Shandong 266071 , China
Next Previous
Abstract
The paper introduces complemental explanation and analysis of existing time-delay identification scheme of nonlinear systems; meanwhile, points out the impropriety of NARMA model correction strategy for time-delay system^. The neural networks predictive control of time-delay system is taken as an example to present the correct NARMA model correction methods. Simulation results verify that the advanced correction methods can obtain good control performance and anti-disturbance capability.
Keywords
identification ; NARMA model ; neural network ; predictive control
References
[ 1 ] Hornik K, Stinchcombe M, White H. Multilayer feedforward networks are universal approximators[J].Neural Networks,1989,2(5):359~366 link1
[ 2 ] Tan Y, Cauwenberghe A V. Neural-network-based d-step-ahead predictors for nonlinear systems with time delay[J].Engineering Applications of Artificial Intelligence,1999,12:21~35 link1
[ 3 ] Tan Y,Cauwenberghe A V.Nonlinear one-step-ahead control using neural networks:control strategy and stability design[J].Automatica,1996,32(12):1701~1706 link1
[ 4 ] Tan Y,de Keyser R.Neural network based predictive control for nonlinear processes with time-delay[A].Systems,Man,and Cybemetics,‘Humans,Information and Technology´,1994 IEEE Intemational Conference on,Vol2[c],1994.1116~1121
[ 5 ] 陆燕,杜继宏,李春文.延迟时间未知的时延系统神经网络补偿控制[J].清华大学学报.1998,38(9):67~69 link1
[ 6 ] de Jesus O,Pukrittayakamee A,Haga M T.A comparison of neural network control algorithms[A].Neural Networks,2001,Proceedings.UCNN´01.Intemational Joint Conference on Vol 1[C],2001.521~526
[ 7 ] Siegelmann H T,Home B G,Giles C L.Computational capabilities of recurrent NARX neural networks[J].IEEE Trans Systems,Man,and Cybemetics B,1997,27:208~215 link1