Modeling and solution for the ship stowage planning problem of coils in the steel industry
Lixin Tang,Jiyin Liu,Fei Yang,Feng Li,Kun Li
DOI: https://doi.org/10.1002/nav.21664
2015-10-01
Naval Research Logistics (NRL)
Abstract:We consider a ship stowage planning problem where steel coils with known destination ports are to be loaded onto a ship. The coils are to be stowed on the ship in rows. Due to their heavy weight and cylindrical shape, coils can be stowed in at most two levels. Different from stowage problems in previous studies, in this problem there are no fixed positions on the ship for the coils due to their different sizes. At a destination port, if a coil to be unloaded is not at a top position, those blocking it need to be shuffled. In addition, the stability of ship has to be maintained after unloading at each destination port. The objective for the stowage planning problem is to minimize a combination of ship instability throughout the entire voyage, the shuffles needed for unloading at the destination ports, and the dispersion of coils to be unloaded at the same destination port. We formulate the problem as a novel mixed integer linear programming model. Several valid inequalities are derived to help reducing solution time. A tabu search (TS) algorithm is developed for the problem with the initial solution generated using a construction heuristic. To evaluate the proposed TS algorithm, numerical experiments are carried out on problem instances of three different scales by comparing it with a model‐based decomposition heuristic, the classic TS algorithm, the particle swarm optimization algorithm, and the manual method used in practice. The results show that for small problems, the proposed algorithm can generate optimal solutions. For medium and large practical problems, the proposed algorithm outperforms other methods. © 2015 Wiley Periodicals, Inc. Naval Research Logistics 62: 564–581, 2015
operations research & management science