Abstract
The single machine early/tardy job scheduling problem (SMETP) consists of determining the best processing sequence for a set of jobs in order to minimize total costs. Minimizing both earliness and tardiness costs pushes the completion of each job to as close to its due date as possible in accordance with the just-in-time philosophy of production planning. As the problem is NP-hard and local optimality conditions may be well away from global optimality conditions new methods are needed to generate near optimal solutions. We propose a new heuristic for the SMETP that alternates search techniques performed at different neighborhood ranges using its own results to guide the alternation in a dynamic way. The computational results we obtained on randomly generated test problems compared very favorably with the results obtained with tabu search and simulated annealing algorithms.The single machine early/tardy job scheduling problem (SMETP) is a NP-hard problem for which most properties of optimal solutions for single machine problems with regular objective function do not hold. In this paper we present a new composite heuristic for the SMETP that combines tabu search, simulated annealing and steepest descent techniques to generate near optimal schedules. Computational experience is reported for a set of randomly generated test problems.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call