A Real- Time Vessel- Buoy Collision Risk Estimation Model for Inland Waterway Navigation

The inland waterway is the navigation passage in inland waters including rivers, lakes, reservoirs and channels, in which the buoy plays an important role in the pilotage for guiding vessels (especially in large size) to avoid dangerous water areas. However, the inland waterway buoys can easily be drifted or even damaged by a vessel crash, which may lead to a serious waterway traffic accident. In order to increase the security of buoys, the collision risk estimation is seriously required to detect potential Vessel-Buoy collision events ahead of time. Motivated by this demand, we propose a real-time Vessel-Buoy collision risk estimation model to reduce the collision probability. In this study the Kalman Filtering (KF) is firstly introduced to predict trajectories. Then an improved KUNZI model is built to estimate collision risks. The experiment results show a real-time dynamic Vessel-Buoy collision risk estimation and reveal the advantages compared to a traditional estimation model.