Multi-objective evolutionary algorithm for a ship routing problem in maritime logistics collaboration

This paper proposes a multi‐objective evolutionary algorithm in maritime logistics collaboration of two liner shipping companies in joint‐routing network design. The model is called the ship routing problem and two objectives being minimised are total cost and deviation in fair cost proportion. The method combines NSGA‐II and the principles of effective genetic algorithms from the literature, and an example of application with data background from the Indonesian archipelago is demonstrated. Both the method and its application in real‐life problems have never been encountered in academic publication, therefore this research has significant contribution and practical values on those fronts. Three dispersal mechanisms are tested with two different mutation probabilities and the results suggest that different rate supports different mechanism. Running times are longer in higher mutation rate, but in general the DV(1) mechanism is faster than both DL mechanisms. Non‐dominated solutions are found and translated to joint routings of both carriers.