Online Parameter Identification for Dynamical Modeling of an Autonomous Catamaran

This paper presents an online parameter identification approach to enhance navigational autonomy in Unmanned Surface Vehicles (USVs). The proposed method addresses the critical challenge of accurately estimating the effects of external forces on vehicle motion in marine environments, which is essential for precise control and operation of USVs. Additionally, by integrating our model fitting framework with ROS2, the middleware used to interface with the deployed USV, we enable real-time estimation of these disturbances impacting its dynamics. The work leverages computationally efficient and optimal parameter identification techniques to improve state estimation, demonstrated through simulations and real-world deployments. Results indicate that the method effectively reduces error in predicting USV states in dynamic marine environments.