This paper addresses the one-machine scheduling problem with earliness-tardiness penalties. We propose a new branch-and-bound algorithm that can solve instances with up to 50 jobs and that can solve problems with even more general non-convex cost functions. The algorithm is based on the combination of a Lagrangean relaxation of resource constraints and new dominance rules.
机构:
E China Inst Technol, State Key Lab Breeding Base Nucl Resources & Envi, Nanchang 330013, Jiangxi, Peoples R China
E China Inst Technol, Sch Math & Informat Sci, Fuzhou 344000, Jiangxi, Peoples R ChinaYuanpei Univ, Dept Healthcare Management, Hsinchu, Taiwan