Steady-State Genetic Algorithm Applied to Scheduling with Wearing Tools

The Task Escalation Problem is Addressed Considering Tool Changes by Wear, an NP-difficult problem with wide industrial application. In order to minimize the total number of tool changes, a mathematical model and an implementation of the Steady State Genetic Algorithm (SSGA) are presented. Given the absence of a reference benchmark in the literature, a new set of instances composed of 120 problems and the adaptation of a Genetic Algorithm of Biased Random Keys to generate higher limiting factors is presented. The computational experiments performed demonstrated that the model is able to prove optimality for instances with up to 8 tasks. In addition to reporting solutions equivalent to all known optimal solutions, for larger instances, SSGA outperformed the meta-heuristics adapted from the literature in solution quality and computational time.