Designing liner shipping networks

In this paper the combined fleet-design, ship-scheduling and cargo-routing problem with limited availability of ships in liner shipping is considered. A genetic algorithm based solution method is proposed in which the ports are first aggregated into port cluster to reduce the problem size. When the cargo flows are disaggregated, a feeder service network is introduced to ship the cargo within a port cluster. The solution method is tested on a problem instance containing 58 ports on the Asia-Europe trade lane of Maersk. The best obtained profit gives an improvement of almost 20% compared to the reference network based on the Maersk network.