Empirical prediction methods for rudder forces of a novel integrated propeller-rudder system

Features of the energy saving integrated propeller-rudder system are discussed. Both conventional a,ld artificial neural networks empirical methods for prediction of rudder forces are introduced. These are based on experimental data obtained during cavitation tunnel tests with various configurations of integrated system coupled with known empirical and theoretical models. Experiments with the integrated system are described. Measured data together with results from both conventional and artificial natural networks approaches are presented. A comparative investigation of both methods is undertaken, both with regard to accuracy and development costs.