On learning input shaping technique for suppressing residualvibrations in nonlinear time-varying systems

Abstract

It is well known that conventional Input Shaping Technique (IST) is not very effective in suppressing residual vibrations for nonlinear and time varying systems. In an effort to increase the effectiveness for such systems, this thesis presents Learning Input Shaping Technique (LIST) which iteratively updates the parameters of IST from previous trials. Simulations are presented for four different cases: (1) when the natural frequency or damping of a system is not estimated well; (2) when a system has time varying vibration; (3) when a system has nonlinear flexibility; and (4) when a closed loop system includes a saturation element in the loop. LIST is experimented on a six D.O.F industrial robot to evaluate its effectiveness. The results of the simulations and the experiment show that the residual vibrations become considerably smaller as iteration goes on, thereby demonstrating the effectiveness of LIST. The results of LIST is compared to that of time varying IST (TVIST), which changes the IST parameters in real time according to an estimated frequency.