Problem
Research questions and friction points this paper is trying to address.
aquatic robots
hydrodynamic model identification
model-based control
vessel dynamics
MOSAIC-SV: Real-Time Adaptive Identification of Vessel Dynamics for the Control and Deployment of Aquatic Robots
🤖 AI Summary
This study addresses the limited control adaptability of aquatic robots caused by reliance on expensive hydrodynamic models by proposing a real-time adaptive control system. The method identifies control-sufficient models using engineering priors and employs a physically admissible unscented Kalman filter combined with command-dependent projection correction to achieve frame-by-frame parameter re-estimation during closed-loop tasks without dedicated identification experiments. A model predictive path integral controller is further integrated to optimize planning and control. Simulations demonstrate recovery of calibrated model performance, while real-vessel tests reveal over 25% improvement in cruising speed and more than 56% reduction in motion prediction error. Successfully deployed on multi-ton vessels, the proposed approach significantly enhances robustness against unmodeled dynamics.
📝 Abstract
Model-based control of an aquatic robotic platform depends on a hydrodynamic model that is costly to identify and specific to the hull, payload, and conditions it was measured in. Here, we present MOSAIC-SV, a deployable real-time adaptive dynamics identification and control system that identifies a control-sufficient dynamics model from a spec-sheet engineering prior, without dedicated identification trials, and re-estimates it at every control step of a closed-loop mission while the controller plans on it. A physically admissible unscented Kalman filter re-estimates hydrodynamic, disturbance, and actuator parameters at every control step of the closed-loop mission; while a command-dependent consider projection withholds corrections the current command cannot attribute between actuator effectiveness and external force; and a model predictive path integral controller plans on the current estimate. In simulation on a CyberShip II plant, MOSAIC-SV recovers the transit performance of the calibrated model under static mismatch and transient changes, and stays within 20% of its own transit time at the unscaled prior when its inertia or damping prior is wrong by an order of magnitude. In on-water field trials on the Blue Robotics BlueBoat, a twin-thruster catamaran, MOSAIC-SV transits at least 25% faster and predicts its own motion with at least 56% less error than its frozen engineering prior, including under an unmodeled payload. The same MOSAIC-SV system concept was also feasibly deployed on a 6.3-tonne dual outboard monohull.
Innovation
Methods, ideas, or system contributions that make the work stand out.
Adaptive Dynamics Identification
Unscented Kalman Filter
Model Predictive Path Integral Control
Aquatic Robots
Real-Time Estimation
🔎 Similar Papers
No similar papers found.
