Data Driven Approach to Input Shaping for Vibration Suppression in a Flexible Robot Arm

2025-04-22

International Conference on Soft Robotics

AI Summary

To suppress residual vibration at the tip of flexible manipulators, this paper proposes a data-driven, adaptive input shaper parameter tuning method. The approach constructs an interpolation model over the workspace using experimentally measured natural frequencies and damping ratios, enabling real-time generation of optimal input shaper parameters without requiring precise dynamic modeling. This is the first work to introduce a data-driven interpolation strategy for input shaper parameter tuning, overcoming the limitations of conventional methods reliant on system identification or offline parameter optimization. The method is model-free and hardware-compatible, making it suitable for complex configurations such as multi-material 3D-printed flexible arms. Experimental results demonstrate that the proposed method significantly reduces residual vibration amplitude—achieving an average attenuation exceeding 70%—while improving end-effector positioning accuracy and operational stability.

Technology Category

• Intelligent Robots: Manipulation
• Planning, Routing, and Scheduling: Replanning and Plan Repair
• Humans and AI: Interaction Techniques and Devices

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

This paper presents a simple and effective method for setting parameters for an input shaper to suppress the residual vibrations in flexible robot arms using a data-driven approach. The parameters are adaptively tuned in the workspace of the robot by interpolating previously measured data of the robot’s residual vibrations. Input shaping is a simple and robust technique to generate vibration-reduced shaped commands by a convolution of an impulse sequence with the desired input command. The generated impulses create waves in the material countering the natural vibrations of the system. The method is demonstrated with a flexible 3D-printed robot arm with multiple different materials, achieving a significant reduction in the residual vibrations.

Problem

• Adaptively tune input shaper parameters for vibration suppression
• Use data-driven interpolation to reduce residual vibrations
• Test method on flexible 3D-printed robot arms

Innovation

• Data-driven adaptive tuning of input shaper
• Interpolation of measured residual vibration data
• Vibration-reduced commands via impulse convolution