Research Analyzer
← Back ICRA 2026

Data-Driven Control Optimization on Frequency Response for Fast and Precise Motion of Flexible Joint Robots

Deokjin Lee, Junho Song, Sehoon Oh

PDF

AI summary

Key figure (auto-extracted from paper)
A data-driven convex optimization framework using measured frequency response data automatically synthesizes stable, high-bandwidth controllers for flexible joint robots, outperforming traditional model-based methods in speed and precision.
Data-driven control Flexible joint robots Frequency response function Convex optimization Bandwidth maximization Controller synthesis

Problem

Flexible joint robots suffer from configuration-dependent resonance and inertia variations that severely limit control bandwidth and stability. Conventional model-based approaches struggle with accurate parameter identification and manual tuning, while existing FRF-based methods lack automation and multi-joint optimization capabilities.

Approach

The method directly utilizes measured frequency response data to formulate controller design as a convex optimization problem, automatically tuning parameters to maximize bandwidth and guarantee stability across varying configurations without explicit model identification.

Key results

  • Automated controller synthesis via convex optimization of measured FRF data
  • Maximization of control bandwidth while ensuring closed-loop stability across configurations
  • Superior tracking accuracy and vibration suppression compared to H∞ and heuristic FRF methods
  • Successful high-speed drumming task execution demonstrating robustness to impacts and inertia variations

Why it matters

Enables fast, precise, and robust motion control for flexible robotic systems in dynamic human-centric applications without relying on difficult-to-obtain dynamic models.

Abstract

This paper presents a data-driven control opti- mization framework for flexible joint robots (FJR) based on frequency response function (FRF) data, enabling automated controller synthesis without explicit model identification. Unlike conventional model-based approaches that rely on accurate parameter estimation, the proposed method directly utilizes measured FRF data and formulates the controller design as a convex optimization problem. The controller maximizes control bandwidth while ensuring stability across a wide range of configurations. Experimental validation on a FJR demonstrates superior tracking accuracy, vibration suppression, and robustness compared to model-based methods. Furthermore, a high-speed drumming task demonstrates the ability of the controller to handle repeated impacts and inertia variations, highlighting the potential of FRF-based control for the fast and precise operation of flexible robotic systems.

Index terms

Flexible Robotics Motion Control Optimization and Optimal Control

Related papers