A Modelling and Tabu Search Heuristic for a Continuous Galvanizing Line Scheduling Problem

With good mechanical properties and corrosion-resistance, galvanized coils are used in many fields nowadays, such as auto industry and household appliance industry. Because of high additional value of galvanized product, steel plants pay much more attention on how to make continuous galvanizing lines (CGL) run efficiently. In this paper, we address a scheduling problem of continuous galvanizing line, which aims to reduce production cost and improve production efficiency. A steel plant usually consists of two parts, the primary steel making part and the finishing steel making part. In primary steel making part, slabs are rolled into hot coils in hot strip mills. Sequentially hot coils are transported to cold mills, where hot coils are shaped thinner and produced into cold coils. Cold coils have bad mechanical properties and are prone to rusting. In order to meet advanced requirement, it is necessary to process cold coils by annealing and galvanizing. These two processes are integrated into CGL. Coils are heated and cooled in a continuous annealing furnace to improve the mechanical properties. And zinc coats are coated on surfaces of coils to protect them from corrosion. The CGL we research in the paper produces principally coils used to make autos. The outer surfaces of autos are made of outer galvanized coils which have very high surface finish requirement, and the inner surfaces are made of inner galvanized coils with normal surface finish requirement. Because of very high requirement of outer coils on surface finish, it is required that outer coils must be produced at the beginning of schedules when rollers are smooth and the number of outer coils produced continuously can not exceed a specified number. In our best knowledge, no published paper has described or solved such a similar problem so far. The interrelation between scheduling two kinds of coils makes the problem difficult to solve. In this paper, we propose a heuristic based on tabu search. The procedure of generating initial solutions is divided into three subroutines, and Hungarian method which is the classic method to solve assignment problems are employed in one of subroutines. Five neighborhoods are designed in tabu search phase. Finally to improve the effectiveness of the heuristic, three additional strategies, intensification search, diversification search and path relinking, are used in the heuristic. Many researches have been published on scheduling in steel industry. Lee et al. described a scheduling problem A Modelling and Tabu Search Heuristic for a Continuous Galvanizing Line Scheduling Problem