Scaling on Identical Parallel Machines with Reentry and Tool Constraints: An Evolutionary Approach

This work addresses an extension of the Task Escalation Problem in Identical Parallel Flexible Machines with Tool Constraints. The variant studied incorporates additional industrial constraints, such as task re-entry, release times, delivery dates, and tool changes in indivisible subsets of heterogeneous sizes. In order to simultaneously minimize the total delay and the setup time, a first-of-its-kind implementation of the Steady-State Genetic Algorithm is proposed. The method was evaluated using a set of reference instances from the literature and the results were compared to the reference matheuristic for the problem. The computational experiments, corroborated by statistical tests, attest to the superiority and efficiency of the proposed approach. The algorithm established new best solutions for 63.34% of the evaluated instances, including all the larger problems.