IJCAS

Stewart platform is a six degrees of freedom (DOF) parallel robot manipulator with rather high powerto-weight ratio and more accurate positioning that of conventional serial-link robots. Nevertheless, modelling and control of parallel Stewart platform is so abstruse and significant for vibration isolation, especially in the case with base excitation in stabilizers. In this paper a clear nonlinear dynamic model of hydraulically actuated Stewart platform with 6-DOF is presented via Kane’'s method. The dynamics of Stewart platform embedded with hydraulic actuator equations is utilized to design a modern control scheme with the aim of vibration isolation. In order to attenuate upper platform vibration under uncertain base excitation profile, a novel neural network based selftuning PID (NN-PID) control scheme is implemented. Following numerical simulation results, the proposed control method satisfies vibration isolation of the upper platform under uncertain base excitation and shows well robustness against external disturbances. It is concluded that the controller is of fine adjustability and can isolate a wide range of external disturbances. "

Keywords: Control, hydraulic actuators, Kane’'s method, neural network, self-tuning PID controller, Stewart platform.