Abstract

This paper presents an application of Meta-RaPS (Meta-heuristic for Randomized Priority Search) to the unrelated parallel machine scheduling problem with sequence-and-machine dependent setup times, where the objective is to minimize the makespan (Rm|Sijk|Cmax)- Meta-RaPS constructs and improves feasible solutions at each iteration through the utilization of randomized rules. After a number of iterations the best solution is kept. In this application, Meta-RaPS uses a self-adjusting mechanism to update its parameters as the search goes on. This mechanism is based on the gradient search method and explores the two-parameter space systematically until the best setting is found. The mechanism used as well as Meta-RaPS application is explained in detail, tested, and compared with results from literature.