Tight lower bounds by semidefinite relaxations for the discrete lot-sizing and scheduling problem with sequence-dependent changeover costs

We study a production planning problem known as the discrete lot-sizing and scheduling problem with sequence-dependent changeover costs. We discuss two alternative QIP formulations for this problem and propose to compute lower bounds using a semidefinite relaxation of the problem rather than a standard linear relaxation. Our computational results show that for a certain class of instances, the proposed solution approach provides lower bounds of significantly better quality.