2011 Volume 24 Issue 1 Pages 1-8
In this paper we propose and evaluate some idea to improve an existing exact algorithm for Input-Output Scheduling Problem (IOSP) in automated warehouses. The existing algorithm is based on LP relaxation of IOSP, which is solved by the column generation method allowing relaxed columns (routes). Our idea is, expecting to enhance LP solution, to impliment the column generation using only exact routes, and to reduce consequently increasing calculation cost by dropping (pegging) unusable edges. The pegging test is done in the preprocessing phase by solving Lagrangian relaxation of IOSP formulated in node cover decision variables. The results of computational experiments show that the proposed algorithm can solve slightly large sized instances in less execution time than existing one.