Main engine power prediction with artificial neural network for general cargo ships

Before starting the physical construction of ships, it is very important to determine several basic parameters in the preliminary design phase. Traditionally, hydrodynamic model tests are used to determine the main engine power and other basic ship parameters. Although these tests are reliable, they are time-consuming and expensive in nature. In this study, the estimation of the main engine power for general cargo ships was carried out using the artificial neural network (ANN) method as an alternative to the traditional approach. The ANN model used a comprehensive set of input parameters including the ship length overall, breadth, gross tonnage, deadweight tonnage (DWT) and service speed. In the study where a rich data set was used, 70% of the data was allocated for model training, 15% for validation and the remaining 15% for testing. In order to increase the reliability and generalization ability of the ANN model, a total of 1000 alternative models were created by randomly assigning different numbers of hidden neurons in each study. The R values for the training, validation and testing data sets in the best model were calculated as 0,992, 0,988 and 0,986, respectively. Moreover, the Mean Absolute Error (MAE) and Root Mean Square Error (RMSE) values remained consistently low in all normalized datasets, ranging from 0,0128 to 0,0148 for MAE and 0,0178 to 0,0238 for RMSE. These results highlight the high predictive ability of the ANN model in predicting the main engine power, indicating that it can be used as a feasible and efficient alternative to traditional hydrodynamic model tests.