Data-Driven Robust Command Shaping for Vibration Suppression in Stacker Cranes: An Experience-Oriented Approach

Stacker cranes used in industrial logistics systems require fast and precise positioning; however, they often suffer from vibration responses owing to structural flexibility and variations in plant characteristics, degrading positioning accuracy and takt efficiency. This study proposes a data-driven robust command shaping (DD-RCS) method to suppress such vibrations. The DD-RCS method adopts an experience-oriented approach, achieving robustness simply by using multiple positioning data collected under varied plant conditions. Based on this experience data, it predicts responses under plant variations and generates position commands by solving a predicted response-based optimization problem. Experimental validation using a laboratory positioning device demonstrated that DD-RCS successfully reduced the residual vibrations and improved the positioning performance compared to a conventional S-shaped command, demonstrating its practicality for stacker crane applications.