Simulation of autonomous surface vehicles using smoothed particle hydrodynamics

Abstract

The study of Autonomous Surface Vehicles (ASVs) is a growing area of research as traditional methods of ocean observation and mapping are expensive. Oceanographic buoys require buoy tenders with appropriate lift capacity, and research vessels are high-cost assets that require an experienced and knowledgeable crew. ASVs are attractive as they can be operated from a distance, can be smaller than traditional research vessels, and are able to navigate to areas of interest, unlike moored buoys while often providing equivalent oceanographic data with lower operational costs. Unfortunately, the development of ASVs is costly. Field trials often require teams of people, chase boats, sensors to monitor environmental conditions, and operations centres. Model testing also requires specialized facilities and experienced personnel. Simulations can be a complementary and cost-e ective approach to assist in the development of ASVs. Simulations accelerate the development of new hull designs and testing of algorithms before the use of scale or eld testing. This thesis focuses on the simulation of ASVs with the use of Smoothed Particle Hydrodynamics (SPH). SPH is a computational uid dynamics method that discretizes the uid into a set of particles which use a Lagrangian coordinate system. Although SPH was originally developed to solve astrophysics problems, it is rapidly growing in popularity in the ocean engineering elds. This thesis explores the use of SPH to simulate ASVs in water to help understand the relationship between ASVs dynamics and the dynamics of the water. A challenge to ASVs development is producing algorithms that are robust to environmental disturbance. This thesis also explores methods for developing numerical wave tanks so that ASVs can be tested under environmental disturbances to help developers validate control algorithms before scale or eld testing.