Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
482034 | European Journal of Operational Research | 2008 | 21 Pages |
In this paper we consider the scheduling problem of minimizing the weighted number of late jobs on a single machine (1|rj|∑wjUj)1|rj|∑wjUj. A branch-and-check algorithm is proposed, where a relaxed integer programming formulation is solved by branch-and-bound and infeasible solutions are cut off using infeasibility cuts. We suggest two ways to generate cuts. First, tightened “no-good” cuts are derived using a modification of the algorithm by Carlier (1982, EJOR, v.11, 42–47) which was developed for the problem of minimizing maximum lateness on a single machine. Secondly we show how to create cuts by using constraint propagation. The proposed algorithm is implemented in the Mosel modelling and optimization language. Computational experiments on instances with up to 140 jobs are reported. A comparison is presented with the exact approach of Péridy at al. (2003, EJOR, v.148, 591–603).