Controle de orientação e planejamento de caminho de curta distância para o veleiro robótico NBoat II

Abstract

The main challenges in the development of autonomous sailboats are: control, path and trajectory planning, sensor data acquitision, and power supply. Towards this direction, this paper introduces a study on the problems of control and path planning commonly found during the autonomous sailing projects. The methods developed here are to be applied in the sailboats used in the N-Boat project, allowing them to reach targets points accurately and quickly, and to perform one of the must difficult tasks in sailing that is navigating against the wind. To achieve these goals, at first, a method to find the most appropriate low level controller for the desired application is developed. This method uses a dynamic PI controller, coming up with a table that contains the best proportional and integrative parameters that are appropriate to each situation according to the model used. A method for generation of paths in situations contrary to wind is also modeled, implemented and tested (in simulation). To generate the way points, this method takes into account two parameters: the distance available for the maneuvering and the desired orientation of the boat during the maneuver. An optimization method is proposed, based on genethic algorithm, implemented, and also tested (in simulation) for getting the controller best parameters. The method manipulate some defined parameters for generating paths, finding the ones that generate the path in which the boat achieves the minimum time to destination. Results of various simulation experiments are shown to demonstrate the validity and robustness of the methods developed.