2012 年 78 巻 789 号 p. 1325-1337
This paper investigates a feedforward control technique for saving the operating energy of a 2-DOF flexible manipulator with a point-to-point (PTP) motion, in which the residual vibration also can be suppressed. The 2-DOF manipulator has one prismatic joint and one revolute joint. The Lagrangian approach in conjunction with the assumed modes method is applied to derive the equations of motion of the manipulator system. For the PTP motion task, the trajectory of the translational motion is set to a cycloidal motion. On the other hand, the trajectory of the rotational motion is designed to simultaneously minimize the residual vibration and the operating energy. In the present method, we attempt to express the trajectory of the joint angle by an artificial neural network (ANN), and then a vector evaluated particle swarm optimization (VEPSO) algorithm, which is a multi-objective optimization algorithm, is used for learning the ANN. By operating the manipulator along the trajectory obtained by the proposed method, the residual vibrations can be suppressed under the minimum energy condition. The numerical simulation results are compared with the experimental results; this comparison reveals the applicability and effectiveness of the proposed method.