DEVS-based modeling and simulation of data-driven exploration algorithms of lentic water bodies with an ASV

Abstract

This paper presents a Discrete Event System Specification (DEVS) modular architecture for modeling and simulating the behavior of an Autonomous Surface Vehicle (ASV), guided by gradient-free optimization and contour detection algorithms, while searching the extreme points and constant curves of a variable of interest in a water body. The ASV model integrates its dynamic differential equations and is controlled by a Guidance, Navigation, and Control (GNC) system implemented through several atomic models to facilitate changes in the behavior of any of them and test different types of regulators and guidance approaches in the future. The paper results show that the system is working correctly for the already supported missions (seeking extreme points and contour curves) and that adapting the specification to each of them is possible by modifying one of the models of the GNC.