Neural network modeling and generalized predictive control for Giant Magnetostrictive actuators

In the application of the Giant Magnetostrictive actuators (GMA), hysteresis of the GMA is particularly significant and causes undesired effect in the control system. This paper investigates the application of neural network based on generalized predictive control to eliminate the hysteresic effect of GMA. The modified Elman neural network is used as the multi-step predictive model, the fused identification model is proposed to improve the predictive and control precision. The modified Elman neural network on-line learning improves the control system adaptability to the unpredicted operating envirnment for GMA. Simulations on GMA position cotrol are included to illustrate the effectiveness of the proposed control scheme.

[1]  David W. Clarke,et al.  Generalized predictive control - Part I. The basic algorithm , 1987, Autom..

[2]  D I Soloway,et al.  Neural Generalized Predictive Control: A Newton-Raphson Implementation , 1997 .

[3]  Wenmei Huang,et al.  Hysteresis compensation for giant magnetostrictive actuators using dynamic recurrent neural network , 2006, IEEE Transactions on Magnetics.

[4]  C. Natale,et al.  Effects of hysteresis compensation in feedback control systems , 2003 .

[5]  John S. Baras,et al.  Modeling and control of hysteresis in magnetostrictive actuators , 2004, Autom..

[6]  Vincent Hayward,et al.  Phase control approach to hysteresis reduction , 2001, IEEE Trans. Control. Syst. Technol..

[7]  C. Natale,et al.  Feedback control systems for micropositioning tasks with hysteresis compensation , 2004, IEEE Transactions on Magnetics.

[8]  Fevzullah Temurtas,et al.  Application of neural generalized predictive control to robotic manipulators with a cubic trajectory and random disturbances , 2006, Robotics Auton. Syst..

[9]  Weng Ling.  DYNAMIC MODEL WITH HYSTERETIC NONLINEARITY FOR A GIANT MAGNETOSTRICTIVE ACTUATOR , 2003 .

[10]  Kumpati S. Narendra,et al.  Identification and control of dynamical systems using neural networks , 1990, IEEE Trans. Neural Networks.

[11]  John S. Baras,et al.  Modeling and control of a magnetostrictive actuator , 2002, Proceedings of the 41st IEEE Conference on Decision and Control, 2002..

[12]  Seung-Woo Kim,et al.  Improvement of scanning accuracy of PZT piezoelectric actuators by feed-forward model-reference control , 1994 .

[13]  Ye-Hwa Chen,et al.  Piezomechanics using intelligent variable-structure control , 2001, IEEE Trans. Ind. Electron..

[14]  Ciro Visone,et al.  Identification and compensation of Preisach hysteresis models for magnetostrictive actuators , 2001 .

[15]  Jun Ni,et al.  Dynamic neural network modeling for nonlinear, nonstationary machine tool thermally induced error , 2005 .

[16]  Alison B. Flatau,et al.  An Energy-Based Hysteresis Model for Magnetostrictive Transducers , 1997 .