GA-LSTM and NSGA-III based collaborative optimization of ship energy efficiency for low-carbon shipping.

Improving ship energy efficiency has been a crucial part for energy preservation and emission control of the shipping industry. However, the current optimization technologies mainly concentrate on a single optimization of navigation speed, route and trim. There is still a shortage of an effective collaborative optimization method to enhance the ship energy efficiency. In this regard, it is necessary to carry out more efficient optimization approach to further enhance the ship fuel efficiency. Therefore, a new collaborative optimization approach for energy efficiency optimization considering the coupling effects of navigation route, speed, trim and various environmental variables is proposed in this study. Firstly, a predictive model for ship energy consumption, which considers the sailing route, speed, trim and various environmental factors, is established by using Genetic Algorithm (GA) improved Long Short-Term Memory (LSTM) approach. On these bases, a collaborative optimization method based on the Non-dominated Sorting Genetic Algorithm III (NSGA-III) is proposed. The results of a case study show that the proposed collaborative optimization strategy can save fuel consumption by as much as 4.54%, compared with the original operational mode. Therefore, it holds significant importance for further enhancing ship fuel efficiency and promoting the advancement of low-carbon shipping. Improving ship energy efficiency has been a crucial part for energy preservation and emission control of the shipping industry. However, the current optimization technologies mainly concentrate on a single optimization of navigation speed, route and trim. There is still a shortage of an effective collaborative optimization method to enhance the ship energy efficiency. In this regard, it is necessary to carry out more efficient optimization approach to further enhance the ship fuel efficiency. Therefore, a new collaborative optimization approach for energy efficiency optimization considering the coupling effects of navigation route, speed, trim and various environmental variables is proposed in this study. Firstly, a predictive model for ship energy consumption, which considers the sailing route, speed, trim and various environmental factors, is established by using Genetic Algorithm (GA) improved Long Short-Term Memory (LSTM) approach. On these bases, a collaborative optimization method based on the Non-dominated Sorting Genetic Algorithm III is proposed. The results of a case study show that the proposed collaborative optimization strategy can save fuel consumption by as much as 4.54%, compared with the original operational mode. Therefore, it holds significant importance for further enhancing ship fuel efficiency and promoting the advancement of low-carbon shipping. [Display omitted] • A collaborative optimization model of sailing route, speed and trim based on the GA-LSTM and NSGA-III is established. • A collaborative optimization method for the ship energy efficiency improvement is proposed. • The proposed collaborative optimization method can reduce fuel usage as much as by 4.54%. • The proposed method can be regarded as an essential guidance for the ship energy efficiency improvement, thereby promoting the low-carbon shipping development. [ABSTRACT FROM AUTHOR]