Optimal Excitation Trajectories for System Identification of Underwater Vehicles
Fotis Panetsos, Kostas Kyriakopoulos
AI summary
Problem
Existing system identification methods for underwater vehicles rely on heuristic inputs or large datasets, often failing to capture coupled dynamics and requiring expensive testing facilities.
Approach
The authors parameterize excitation trajectories using multi-segment Bézier curves and solve a constrained optimization problem to minimize the regressor matrix condition number while ensuring kinematic feasibility and safety.
Key results
- Developed a Bézier curve-based trajectory optimization framework that minimizes the regressor matrix condition number
- Enforced pose, velocity, acceleration, and continuity constraints directly through control point manipulation
- Experimentally validated on a VideoRay Defender in a water tank, achieving a regressor condition number of 160.10
- Demonstrated accurate dynamic parameter identification via cross-validation, successfully predicting velocity on unseen trajectories
Why it matters
Provides a practical, data-efficient framework for accurately modeling underwater vehicle dynamics, enabling better control, state estimation, and fault diagnosis for autonomous underwater missions.
Abstract
In this work, we propose a structured method- ology for the system identification of underwater vehicles through the design of optimal excitation trajectories. To this end, the trajectories are parameterized using B ́ezier curves, which ensure smooth and differentiable motion profiles while facilitating the enforcement of constraints through appropriate manipulation of the control points. An optimization problem is formulated to determine a dynamically feasible excitation trajectory that respects safety limits and maximizes the quality of the collected data, thereby enabling reliable estimation of the vehicle’s dynamic parameters using least squares. The proposed methodology is experimentally validated in a labo- ratory water tank, where the dynamic parameters, identified from the optimized trajectory, are evaluated by predicting the vehicle’s velocity through forward simulation on previously unseen trajectories.