Research On Intelligent Adaptive Vibration Control Of Two-link Flexible Manipulator

Compared to the conventional rigid manipulator,flexible link manipulators have the potential advantage of higher operational speed,smaller actuators,lower energy consumption,better maneuverability and safer operation due to the reduced inertia.The motion of a two-link flexible manipulator is combined with a wide range of rigid body motion and vibration of the flexible manipulator.The elastic vibration between the two flexible arms is also coupled,which makes the flexible manipulator a nonlinear coupling system.The flexible strcutrue has the characteristics of low damping and light wight,under the external disturbances,the continuous low frequency vibration will be easily excited,which will affect its normal work.The elastic vibration also decreases the positioning accuracy and affects the stability of the system.Therefore,research on how to reduce the vibration in the process of movement is particularly necessary.Based on the vibration and structural mechanics,the elastic deformation of a two-link flexible manipulator is analyzed with assumed modal method.The modal analysis,experimental identification and theoretical analysis are carried out under the condition of equilibrium position.The mathematical model of a two-link flexible manipulator is derived based on the assumed mode method and the Lagrange principle.In order to suppress the vibration of a two-link flexible manipulator,the feed-forward and feedback control algorithm are investigated.The feed-forward control strategy is based on the input shaping technology;the input of the flexible system is shaped by a sequence of impulses to reduce the residual vibration.The signal measured by piezoelectric sensor is used as the feedback signal,and three feedback control strategy including proportional derivative(PD)algorithm,fuzzy neural networks(FNN)control algorithm and generalized minimum variance self-tuning(GMVSTC)control algorithm are designed.The alternating current(AC)servo motor is used to attenuate the vibrations.A two-link flexible manipulator experimental setup is constructed.For the two-link rigid-flexible coupling manipulator,the experimental study of active vibration control is carried out by using the input shaping algorithm,PD algorithm and self-tuning control algorithm.The experimental results show that the performance of feedforward control algorithm based on the input shaping and the feedback control algorithm based on self-tuning control is better than PD algorithm.The PD algorithm,fuzzy neural networks algorithm and self-tuning algorithm are applied to suppress the vibration of two-link flexible manipulator,the experimental results demonstrate that the investigated fuzzy neural networks algorithm and self-tuning algorithm can suppress the large-amplitude and small-amplitude vibration of both link more quickly than the traditional PD algorithm.At last,the accceleartion detected by the accelerator is used as the feedback signal,the vibration control experiment is carried out by using the acceleration signal of flexible manipulator and a kind of nonlinear control algorithm.