Optimal path planning and high level control of an autonomous gliding underwater vehicle

• Main Author: Galea, Anna M., 1976-
• Other Authors: James G. Bellingham.
• Format: Others
• Language: English
• Published: Massachusetts Institute of Technology 2007
• Subjects: Electrical Engineering and Computer Science.
• Online Access: http://hdl.handle.net/1721.1/36674

Thesis (S.M.)--Massachusetts Institute of Technology, Dept. of Electrical Engineering and Computer Science, 1999. === Includes bibliographical references (leaves 65-67). === Recent oceanographic interests have focussed on the littoral ocean, where regions of shallow water and high current are prevalent. Long the domain of remotely operated vehicles and, more recently, autonomous propeller-driven vehicles, glider vehicles are being designed and tested for these conditions. These vehicles face special problems in these environments because of their slow water speed and their depth-dependent energy function which requires them to use more energy to move the same horizontal distance through shallow water than in deeper waters. Missions run in shallow waters and with high-speed, time-variable currents thus face propulsion energy consumption difficulties as well as time efficiency concerns. A solution to the optimal path for a typical glider vehicle was sought in a simulated environment without vehicle dynamics considerations, where the minimization criteria were time and energy required to travel between two waypoints. Dynamic programming, analytic methods, and numerical programming methods were applied to the problem, with numerical programming generating the most general results and the greatest insight into the minimization problem. Under the assumption of constant water speed, time optimal paths were insensitive to depth variations, and were instead a compromise between a short path and one that accommodated high currents. For energy-minimization runs, the optimal paths deflected towards deeper regions, and were much less sensitive to current than to depth. To run missions with these strategies, a layered control architecture was implemented for these vehicles. Special behaviors were written to accommodate glider operations, and a dynamic controller written which incorporated improvements suggested by working with its Odyssey model. The specific control strategy can be transported to any glider vehicle, regardless of the hardware used to affect dynamic changes. === by Anna M. Galea. === S.M.