An Automatic Collision Detection and Avoidance Module for Inland Navigation

Abstract

Based on the positional data of other inland vessels which is available through the fusion of radar object tracking and AIS (Automatic Identification System) a model-based prediction of their kinematic states is presented. Information about the river shape is fed into the prediction model according to the type of vessel and its navigational situation. Together with the future states of our own vessel, which are known because it is automatically steared along a guiding line, ranges of encounter are calculated. Furthermore, a path planning algorithm based on nonlinear dynamic optimization is presented that is capable of calculating collision-free paths in real-time that account for the vessel's constrained dynamics. Practical results of the discussed automatic collision detection and avoidance module as part of an integrated navigation system for inland vessels are shown concluding the contribution.