de Schaetzen, Rodrigue2024-08-122024-08-122024-08-122024-07-29https://hdl.handle.net/10012/20785The interest in ship navigation in the Polar regions, particularly the Arctic, is rapidly growing as a result of the increasing accessibility to these remote areas caused by rising global temperatures. The reduction in the extent of Arctic sea ice is gradually providing alternative shipping routes for global trade, which are significantly shorter in distance and can lead to lower emissions and lower costs associated with fuel consumption and transit time. Despite the range of potential benefits, ice navigation introduces numerous risks and challenges to safe ship operation due to the distinctly hazardous environment. Ice conditions often require ships to reduce speed and deviate from the main course to avoid maneuvers that are more likely to cause damage to the ship. In addition, broken ice fields are becoming the dominant ice conditions encountered in the Arctic, where the effects of collisions with ice on the ship are highly dependent on where contact occurs and on the particular features of the collided ice floes, including ice thickness, size, and shape. While existing works have shown success in computing global routes that minimize costs associated with ice navigation such as the average ice-induced forces exerted on the ship, none have considered the problem of optimizing ship navigation at the local level where costs are computed based on individual ship-ice collisions. In this thesis, we present a framework for autonomous navigation of ships operating in ice floe fields. Trajectories are computed in a receding-horizon manner, where we frequently perform planning updates given updated ice-field data. During a planning iteration, we simplify the trajectory planning problem by assuming a fixed nominal speed that is safe with respect to the current ice conditions, and compute a smooth reference path that minimizes a collision cost. We formulate a novel cost function that minimizes the kinetic energy loss of the ship from ship-ice collisions and incorporate this cost as part of our lattice-based path planner. Solutions computed by the lattice planning stage are then used as an initial guess in our proposed optimization-based improvement step to compute a locally optimal path. We conducted extensive experiments to validate our approach both in simulation and in a physical test-bed managed by the National Research Council Canada (NRC).enarcticpath planningroboticsoptimizationlocal navigationship navigationMaster Thesis