|
It is important to study navigation behaviors and effects of
interferences of multiple ocean-going vessels when they share the same
sailing course like a transport convoy, with or without constantly
exchanging navigation decisions among these vessels. Especially a
large number of ships, such as pleasure craft, fishing and commercial
vessels, etc., are passing through a narrow channel or maneuvering a
congested port entry and departure. Each ship needs to make an
autonomous decision and control to avoid collisions with other ships,
and to maintain the planned course without deviating much from the
group. The motivation of this research is to identify a computation
model suitable to capture the dynamics of this kind of group-based
navigation, in order to assess the maneuverability of ships and to
predict hazardous operation conditions. A model called a ship
cluster behavior model is proposed to analyze a group behavior
using a set of abstract forces, the dynamics defined over them, and
synthetic effect of these forces. This model attempts to capture the
correlations and interactions of the member ships in a group and to
predict group-wise movement as a whole. This simulation system is
built using a Java programming language to demonstrate our
contributions in this study such that (1) the ship cluster behavior
model is revised to incorporates kinematics that will characterize
movement of various type of vessels, which will in turn enable us to
analyze more realistic cluster model, and (2) the model is extended to
incorporate the way to identify obstacles in the sea and to make
detour decisions to avoid contacting them.
|