Voyage optimization combining genetic algorithm and dynamic programming for fuel/emissions reduction
Helong Wang, Xiao Lang, Wengang Mao
Abstract
Deterministic optimization algorithms generate optimal routes/paths and speeds along ship voyages. However, a ship can rarely follow pre-defined speeds because dynamic sea environments lead to continuous speed variation. In this paper, a voyage optimization method is proposed to optimize ship engine power to reduce fuel and air emissions. It is a combination of dynamic programming and genetic algorithm to solve voyage planning in three-dimensions. In this method, the engine power is discretized into several levels. The potential benefit of using this algorithm is investigated by a medium-size chemical tanker. A ship's actual sailing is used to demonstrate benefits of the proposed method. On average 3.4% of fuel-saving and emission reduction can be achieved than state-of-the-art deterministic methods. If compared with the actual full-scale measurements, on average 5.6% reduction of fuel consumption and GHG emissions (about 275 tons) can be expected by the proposed method for the six case study voyages.