Design and multi-objective optimization of a novel robust command shaping technique for the tolerable level of residual vibration

Abstract

This manuscript addresses the problem of designing a robust input shaper capable of suppressing residual vibrations in flexible robotic and mechanical systems under modeling errors and parameter uncertainties, while also providing smooth reference commands.The proposed approach integrates Cycloid, Ramped Versine and Ramp (CPRVPR) functions with Zero Vibration (ZV, ZVD, and ZVDD) input shapers, optimizing their parameters using the Vibrating Particle System (VPS) Algorithm. Furthermore, the study proposes a novel multi-objective function that accounts for critical parameters of input shaping techniques in flexible robotic systems and the robustness constraints of Extra-Insensitive input shapers. The theoretical outcomes of the proposed command shaping approaches were experimentally validated through their application to a linear crane system. The effectiveness of the three proposed methods was demonstrated by comparing them against fifteen well-known input shaping techniques. The novel intelligent command shaping design was shown to effectively mitigate or eliminate residual vibrations in flexible systems, even under high levels of uncertainty.