Robust recurrent-neural-network control for X-Y table in CNC machine

• Main Authors: Po-Huang Shieh, 謝伯璜
• Other Authors: Faa-Jeng Lin
• Format: Others
• Language: zh-TW
• Published: 2004
• Online Access: http://ndltd.ncl.edu.tw/handle/35570432407524025776

碩士 === 國立東華大學 === 電機工程學系 === 92 === ABSTRACT The subject of this thesis is to develop a robust recurrent-neural network control system for the reference contours tracking of a biaxial motion mechanism in CNC machine. First, the dynamic model of a field-oriented PMSM drive is derived. Then, an adaptive recurrent-neural-network (ARNN) motion control system and a robust recurrent-neural-network (RRNN) sliding-mode motion control system are developed individually for the robust and precise position control of the PMSM. In the proposed ARNN control system, a recurrent-neural-network (RNN) with accurate approximation capability is employed to approximate an unknown dynamic function and the adaptive learning algorithms that can learn the parameters of the RNN on line are derived using Lyapunov stability theorem. Moreover, the proposed RRNN sliding-mode control system is proposed to effectively approximate the equivalent control of the sliding-mode control based on the derived motion dynamics. Using the proposed control, the motion tracking performance is significantly improved and the robustness to parameter variations, external disturbances, cross-coupled interference and friction torque can be obtained as well. Furthermore, some experimental results due to circle, four leaves, window and star reference contours are provided to show that the dynamic behaviors of the proposed control systems are robust with regard to uncertainties.