Abstract:
Bulk cargos maritime transportation represents significant risks to the vessel, its crew, and the environment, and is duly regulated by IMO (International Maritime Organization), who created the IMSBC (International Maritime Solid Bulk Cargoes Code). The application of empirical models for moisture prediction takes huge importance in this context, supporting due time decisions to guarantee the overall safety cargo and regulatory requirements compliance. Over 980 iron ore fines cargoes database, with chemical quality, moisture and size distribution
were studied. In this work tree different models were developed to cargo moisture prediction (time series, regression, and artificial neural network) and outcomes compared against each other. The obtained results showed artificial neural network models were able to explain higher variance, over 90%, thus being more suitable to industrial application.