Adaptive Control of Soft Continuum Robots

First-author adaptive dynamic sliding mode control validated on soft continuum manipulators under payload variations.

Outcome

I developed and validated a robust adaptive dynamic sliding mode control pipeline for soft continuum manipulators, achieving 38% more accurate tracking than inverse-dynamics baselines under varying payloads.

Problem

Soft continuum manipulators are highly nonlinear and sensitive to payload changes, making accurate control difficult with standard model-based methods alone. This project improved robustness and tracking under realistic manipulation conditions.

System

  • Modeling: Euler-Lagrange dynamics for soft continuum manipulator behavior
  • Control: adaptive dynamic sliding mode controller for robustness to uncertainty and payload changes
  • Validation: comparative experiments against inverse-dynamics baselines
  • Deployment: hardware-oriented implementation for real manipulator testing scenarios

Contribution

  • Led control architecture development as first author
  • Dynamic modeling and controller derivation
  • Benchmark design and experimental validation
  • Manuscript preparation and conference publication

Technical Stack

  • Nonlinear dynamics modeling
  • Adaptive sliding mode control
  • Soft continuum manipulation
  • Experimental benchmarking

Key Results

  • 38% more accurate tracking than inverse-dynamics baselines under varying payloads
  • Presented at IEEE ICRA (2022)

Media

Video demo: adaptive control of a soft continuum manipulator under real-world uncertainties.

Skills

nonlinear modeling adaptive control sliding mode control soft robotics experimental validation