Maneuvering Performance of Autonomous Underwater Vehicles

摘要

High-fidelity linear pitch and yaw dynamic models were developed for a class of autonomous underwater vehicles manufactured by Bluefin Robotics Corporation, based on analysis and field data. Classical control design was then carried out and successfully tested. An algorithm for automatic identification of these models, followed by automatic specification of gains, was developed and delivered to Bluefin. We also created modeling and control schemes for the ONR hovering autonomous underwater vehicle (HAUV), that was developed jointly by MIT and Bluefin; this vehicle has now successfully performed hull surveys on five vessels. Finally, we have applied stochastic simulation techniques to general control problems (which include vehicle maneuvering), illustrating new methods for assessing basic stability and for performing control gain selection in nonlinear systems.