Predicting Preliminary Structural Strength Requirements of Cargo Vessels using Artificial Neural Network

Abstract

In the preliminary stage of a ship design, different classification societies rules are usually followed for predicting the structural strength after finalizing the principal particulars of the ship. Most of the formula for evaluating the requirements of structural strength of a ship using classification societies rules are empirical and the time required is very significant. In present study, an artificial neural network (ANN)-based method is proposed to predict the structural strength requirements for cargo vessels. Keel Plate Weight (KPW), Bottom Plate Weight (BPW), Inner Bottom Plate Weight (IBPW), Side Shell Plate Weight (SSPW), Bulkhead Weight (BW) and Main Deck Weight (MDW) is predicted as a function of ships’ rule length (L), breadth (B) and draft (T). An ANN model was trained to achieve a root mean square error (RMSE) of less than 0.13. The R² of the trained model used to evaluate the new data is 0.998, which indicates that the various requirements of weights calculated by ANN model is in good agreement with the classification societies results.