Development of new training algorithms for neuro wavelet systems on the robust control of induction servo motor drive Industrial Electronics, IEEE Transactions on英文电子书.pdf

IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 49, NO. 6, DECEMBER 2002 1323 Development of New Training Algorithms for Neuro-Wavelet Systems on the Robust Control of Induction Servo Motor Drive Rong-Jong Wai, Member, IEEE Abstract—A robust wavelet neural network control (RWNNC) the PID parameters obtained through the aforecited approaches system is proposed to control the rotor position of an induction usually need manual retuning before being transferred to the servo motor drive in this paper. In the proposed RWNNC system, a process under control. wavelet neural network controller is the main tracking controller that is used to mimic a computed torque control law, and a robust There has been considerable interest in the past several years controller is designed to recover the residual approximation for in exploring the applications of neural network to deal with ensuring the stable control performance. Moreover, to relax the nonlinearities and uncertainties of the real-time control system requirement for a known bound on lumped uncertainty, which [4]–[6]. It has been proven that artificial neural network can comprises a minimum approximation error, optimal network approximate a wide range of nonlinear functions to any desired parameters and higher order terms in a Taylor series expansion of the wavelet functions, an RWNNC system with adaptive bound degree of accuracy under certain conditions [4]. It is generally estimation was investigated for the control of an induction servo understood that the selection of the neural network training motor drive. In this control system, a simple adaptive a