Optimizing autonomous underwater vehicle routes with the aid of high resolution ocean models

Abstract

In underwater vehicle operations in areas such as estuaries, vehicles may face currents with magnitudes equal to or exceeding the vehicle's maximum forward speed. We propose a method which generates vehicle routes taking into account ocean current forecasts from high resolution ocean models, in order to both take advantage of the ocean current velocity and avoid its negative effects. We formulate the problem in an optimal control setting and derive the associated Hamilton-Jacobi-Bellman partial differential equation (PDE). We solve this PDE using a parallelized C++ implementation of a numerical method which allows us to obtain the solution in a few minutes on a mainstream computer. After obtaining the solution of the PDE, optimal trajectories with any initial condition can be computed efficiently. The method is illustrated using data from high-resolution ocean models of the Sado river estuary in Portugal. Two mission scenarios are analyzed, which highlight the influence of ocean currents on optimal trajectories and the benefits of considering ocean current forecasts in mission planning.