Path planning in multi-scale ocean flows: Coordination and dynamic obstacles
Section snippets
Introduction and motivation
The simultaneous use of multiple agents enables complex tasks to be completed in a coordinated and robust manner. In nature, coordination is commonly exhibited by bees, birds, ants, and other animals when they perform activities such as foraging, navigation and hunting. Coordination also improves the overall performance of teams of autonomous robots. Multi-agent systems are then also versatile and effective in completing tasks such as exploring, mapping and monitoring. To execute such tasks,
Time-optimal path planning using level-set equations
Our results on time-optimal navigation in dynamic ocean flows with moving obstacles and time-optimal coordinated formation control are based on a recent level-set approach (Lolla, Haley, Lermusiaux, 2014a, Lolla, Lermusiaux, Ueckermann, Haley, 2014b, Lolla, Ueckermann, Yigit, Haley Jr., Lermusiaux, 2012). These level-set equations, which can be directly solved for in ocean models, are now outlined. The presentation is limited to two dimensions; the extension to higher dimensions is
Time-optimal path planning in the presence of dynamic obstacles
The methodology for time-optimal path planning (Section 2) is now extended to handle moving obstacles and forbidden regions, without any increase in the computational cost. Unlike solid obstacles, forbidden regions (e.g. fishing zones, polluted areas, regulated territories) do not affect the flow-field. It is only the AUVs that cannot operate in such regions. Henceforth, both these types of impediments will usually be simply referred to as obstacles. Our objective then is to rigorously compute
Coordination of underwater vehicles using level-set methods
We now derive a rigorous methodology for distance-based coordination of AUVs operating in minimum-time, within strong and dynamic flows. To do so, we integrate a leader–follower coordination technique with ocean predictions and the above level-set approach (Section 2). The goal is to enable safe, coordinated motion of vehicles in dynamic currents, where the coordination constraints are formulated in the form of relative positions of the vehicles or inter-vehicle spacing during the mission. The
Applications
In what follows, we exemplify our schemes for dynamic formation control and path planning with dynamic obstacles, and analyze their respective results, using several ocean flows. We first consider time-optimal coordinated vehicle motion within an idealized wind-driven double-gyre (Section 5.1). We then demonstrate time-optimal path planning in the presence of moving obstacles within currents exiting an idealized strait or estuary (Section 5.2). Finally, we analyze the results of time-optimal
Summary and conclusions
For many ocean applications, the coordinated operation of multiple AUVs provides significant advantages such as robustness and efficiency, thereby enhancing the utility of individual vehicles. Most coordination techniques usefully assume that the kinematics of vehicles are unaffected by the benign environment in which they operate. However, in the ocean, longer-range underwater vehicles, especially gliders, possess limited relative operating speeds. It is thus important to account for the
Acknowledgments
We acknowledge the Office of Naval Research for supporting this research under grants N00014-09-1-0676 (Science of Autonomy, A-MISSION), N00014-14-1-0476 (LEARNS), N00014-07-1-0473 (PhilEx) and N00014-12-1-0944 (ONR6.2) to the Massachusetts Institute of Technology (MIT). We are thankful to the members of the MSEAS group and to the PhilEx team for several helpful discussions. We also thank the anonymous reviewers for their useful comments.
References (100)
- et al.
A fast level set method for propagating interfaces
J. Comput. Phys.
(1995) - et al.
The fast construction of extension velocities in level set methods
J. Comput. Phys.
(1999) - et al.
Statistical field estimation for complex coastal regions and archipelagos
Ocean Model.
(2011) - et al.
Coupled physical and biogeochemical data-driven simulations of Massachusetts Bay in late summer: real-time and postcruise data assimilation
J. Mar. Syst.
(2003) An oceanic general circulation model framed in hybrid isopycnic-cartesian coordinates
Ocean Model.
(2002)- et al.
A characteristic based volume penalization method for general evolution problems applied to compressible viscous flows
J. Comput. Phys.
(2014) - et al.
Continuous trajectory planning of mobile sensors for informative forecasting
Automatica
(2010) - et al.
Progress toward autonomous ocean sampling networks
Deep Sea Res. Part II: Top. Stud. Oceanogr.
(2009) - et al.
Leader–follower cooperative attitude control of multiple rigid bodies
Syst. Control Lett.
(2009) - et al.
The California Current System: A multiscale overview and the development of a feature-oriented regional modeling system (FORMS)
Dyn. Atmos. Oceans
(2011)