ABSTRACT
The scope of the present study is to develop a real-time vessel performance analysis tool to evaluate the fouling condition of ships. During operation, the vessel will experience an increase of resistance due to several factors, linked to the sailing conditions, which should be modeled. Added wave resistance, wind resistance, shallow water and trim effect, are examples of parameters, which affect the power needed to propel the vessel. Any increase in resistance will result to the increase of fuel consumption and thus the increase of harmful emissions to the environment. In order to develop a robust and reliable analysis module to assess ship performance, a machine learning approach was selected, framed by naval architecture principles. The goal of the developed system is to investigate the potential use of Artificial Neural Networks (ANNs) in estimating the shaft power needed to propel the vessel at any operational, environmental and loading state. Clean hull and propeller condition were considered, in order to isolate the effect of fouling at any stage, by comparing the estimation of the ANN with the shaft power measured by the torque meter and logged through the automated data transmission system.
