A recovering beam search algorithm for the single machine Just-in-Time scheduling problem

We consider the Just-in-Time scheduling problem where the Just-in-Time notion is captured by means of multiple conflicting criteria. The calculation of any non-dominated solution for these criteria is achieved by solving an extension of the single machine problem of minimising the mean weighted deviation from distinct due dates. In the extended problem each job to schedule is also constrained by a release date and a deadline. This problem is NPNP-Hard in the strong sense and we propose heuristic algorithms to solve it. Computational experiments show that, among those algorithms, the most effective heuristic, in terms of quality, is a Recovering Beam Search algorithm.