Neural networks for optimization of an early warning system for moored ships in harbours

Abstract

Within the BlueSafePort project an Early Warning System (EWS) is being developed for forecasting and alerting emergency situations related to ship navigation in ports, as well as operational constraints. Port terminals downtime leads to large economic losses and largely affects the port’s overall competitiveness. So, the goal of such EWS is to reduce the port’s vulnerability by increasing its planning capacity and efficient response to emergency situations. As any EWS, its usefulness depends greatly on its reliability and accuracy. To achieve more accurate predictions a new method was developed to optimize forecasts produced by the system. Using available database from buoys, pressure sensors and meteorological stations, neural networks were trained to optimize numerical models results.