arXiv subjects

Yeyson A. Becerra-Mora

Publications and source records attributed to Yeyson A. Becerra-Mora.

3 recordsLinked to original sources

arXiv2026

Dynamical System-Based Imitation Learning and Neuroadaptive Control for Trajectory Recovery in Autonomous Ships

Repetitive maritime operations can be effectively learned using the Imitation Learning (IL) paradigm, which transfers human expertise directly to Unmanned Surface Vehicle (USV) control systems. Dynamical Systems (DS) are widely used to model non-linear human demonstrations while offering inherent stability guarantees. However, real-world execution under persistent marine perturbations reveals a critical trade-off: standard DS-based IL approaches prioritize global target convergence at the expense of localized trajectory reproduction fidelity. To address this limitation, we present a hybrid learning-control architecture that integrates a DS-based IL reference generator with a neuroadaptive controller. Our approach introduces a control action that drives the USV back to the demonstrated path following exogenous disturbances, enabling dynamic human-like reactive alignment-termed behavioral tracking. The proposed methodology is validated using the Marine Systems Simulator (MSS) toolbox. Simulation results confirm that the framework generalizes complex maneuvering tasks while substantially improving trajectory tracking fidelity under disturbances compared to alternative control strategies.

eess.SY↗

arXiv2025

Learning port maneuvers from data for automatic guidance of Unmanned Surface Vehicles

At shipping ports, some repetitive maneuvering tasks such as entering/leaving port, transporting goods inside it or just making surveillance activities, can be efficiently and quickly carried out by a domestic pilot according to his experience. This know-how can be seized by Unmanned Surface Vehicles (USV) in order to autonomously replicate the same tasks. However, the inherent nonlinearity of ship trajectories and environmental perturbations as wind or marine currents make it difficult to learn a model and its respective control. We therefore present a data-driven learning and control methodology for USV, which is based on Gaussian Mixture Model, Gaussian Mixture Regression and the Sontag's universal formula. Our approach is capable to learn the nonlinear dynamics as well as guarantee the convergence toward the target with a robust controller. Real data have been collected through experiments with a vessel at the port of Ceuta. The complex trajectories followed by an expert have been learned including the robust controller. The effect of the controller over noise/perturbations are presented, a measure of error is used to compare estimates and real data trajectories, and finally, an analysis of computational complexity is performed.

eess.SY↗

arXiv2023

Robust data-driven learning and control of nonlinear systems. A Sontag's formula approach

An interlaced method to learn and control nonlinear system dynamics from a set of demonstrations is proposed, under a constrained optimization framework for the unsupervised learning process. The nonlinear system is modelled as a mixture of Gaussians and the Sontag's formula together with its associated Control Lyapunov Function is proposed for learning and control. Lyapunov stability and robustness in noisy data environments are guaranteed, as a result of the inclusion of control in the learning-optimization problem. The performances are validated through a well-known dataset of demonstrations with handwriting complex trajectories, succeeding in all of them and outperforming previous methods under bounded disturbances, possibly coming from inaccuracies, imperfect demonstrations or noisy datasets. As a result, the proposed interlaced solution yields a good performance trade-off between reproductions and robustness. The proposed method can be used to program nonlinear trajectories in robotic systems through human demonstrations.

eess.SY↗