Trajectory following using model predictive control for SeaDrone collision avoidance using Robot Operating System

Abstract

Navigation of unmanned surface vehicle (USV) is much important now a days due to much improvement in communication system. Details survey information can be collected using USV by passing it over wireless communication. Even remote and natural disaster effected places can be greatly helped by this. This paper focuses on collision avoidance self-control of USV using model predictive control. For testing and simulation GAZEBO is used. ROS is also used to communicate with USV sensor.

This experiment is worked out mainly in two steps. First stability of control and mathematical model is checked using mathematical simulation. Later it is tested in GAZEBO which simulates real life effect on experiment object in terms of forces acting on object.

Satisfactory outcome is achieved in these steps. Using CasAdi for MPC was stable.