Visual Observer-Based State-Feedback Autonomous Motion Control on SE(3)

Abstract

This paper introduces a visual observer-based state-feedback control tailored for autonomous motion on SE(3) to address challenges posed by modeling uncertainties and measurement noise. The proposed approach unfolds in two fundamental phases. In the initial stage, the visual observer, based on modeling information, visual sensors, and pre-deployed landmarks, along with the state-feedback controller, are developed independently, underpinning their individual semi-global practical asymptotic (SPA) stability. This modular approach ensures a robust foundation for the subsequent synthesis. The latter stage applies the well-established Small Gain Principle to regulate observer and controller parameters to guarantee the SPA stability of the closed-loop system with the visual-observer based state feedback control. The effectiveness of the proposed method is validated through simulation and experiments.