Heuristic Methods for Weather Routing

Abstract

Weather routing involves optimizing routes for vessels to reduce travel time and fuel consumption by utilizing meteorological data. It is an interdisciplinary problem that requires insights ranging from naval engineering to mathematics. This thesis proposes three optimization algorithm for weather routing: Hybrid Search, Super A∗ Search, and Bézier Evaluation with Evolutionary Strategy (BEES), paired with a discrete Newton-Jacobi smoothing method. Each method leverages different mathematical principles: differential equations, graph theory, and evolutionary algorithms. We will evaluate their performance in various meteorological conditions and analyze their output across different sailing speeds, ocean regions, and times of the year on a common benchmarking platform.