A Data-Driven Approach for Generator Load Prediction in Shipboard Microgrid: The Chemical Tanker Case Study

Energy efficiency and operational safety practices on ships have gained more importance due to the rules set by International Maritime Organization in recent years. While approximately 70% of fuel consumed a commercial ship is utilized for propulsion load, significant portion remaining auxiliary generators responsible ship’s onboard load. It crucial comprehend impact electrical load generators, as it significantly assists maritime operators strategic energy planning minimize chance unexpected breakdowns during operation. However, an appropriate handling mechanism required when there are massive datasets varied input data involved. Thus, this study implements data-driven approaches estimate chemical tanker generator using 1000-day real dataset. Two case studies were performed, namely, single prediction each total all generators. The results show that DG1, DG2, DG3, decision tree model encountered least errors MAE (0.2364, 0.1306, 0.1532), RMSE (0.2455, 0.2069, 0.2182), MAPE (17.493, 5.1139, 7.7481). In contrast, deep neural network outperforms other models generation prediction, with values 1.0866, 2.6049, 14.728 MAE, RMSE, MAPE, respectively.