Guided local search algorithm for hot strip mill scheduling problem with considering hot charge rolling

Abstract

This study investigates the hot strip mill scheduling problem which is one of the most important planning problems in the steel industry. The problem is formulated using the prize collecting vehicle routing problem. The new proposed formulation considers more details and more realistic constraints than those used in previous studies. The hot charge technique leads to considerable savings in energy and other benefits in the process of steel production. In our proposed formulation, the necessary provisions required for obtaining an initial level of hot charge have been taken into consideration. A search algorithm has been developed that consists of three major phases including separation of slabs that can be scheduled, generation of an initial solution, and improvement of the solution. Generation of the initial solution is accomplished using a greedy constraint satisfaction algorithm and solution improvement through a guided local search. Proposed model and search algorithm have been tested on random and collected instances from practical production data in Mobarakeh Steel Complex. The experimental results show the high accuracy and efficiency of the proposed model and search algorithm.