Viability-based computation of spatially constrained minimum time trajectories for an autonomous underwater vehicle: implementation and experiments

A viability algorithm is developed to compute the constrained minimum time function for general dynamical systems. The algorithm is instantiated for a specific dynamics (Dubin's vehicle forced by a flow field) in order to numerically solve the minimum time problem. With the specific dynamics considered, the framework of hybrid systems enables us to solve the problem efficiently. The algorithm is implemented in C using epigraphical techniques to reduce the dimension of the problem. The feasibility of this optimal trajectory algorithm is tested in an experiment with a Light Autonomous Underwater Vehicle (LAUV) system. The hydrodynamics of the LAUV are analyzed in order to develop a low-dimension vehicle model. Deployment results from experiments performed in the Sacramento River in California are presented, which show good performance of the algorithm.