Overlap Algorithms in Flexible Job-shop Scheduling

Abstract

The flexible Job-shop Scheduling Problem (fJSP) considers the execution of jobs by a set of candidate resources while satisfying time and technological constraints. This work, that follows the hierarchical architecture, is based on an algorithm where each objective (resource allocation, start-time assignment) is solved by a genetic algorithm (GA) that optimizes a particular fitness function, and enhances the results by the execution of a set of heuristics that evaluate and repair each scheduling constraint on each operation. The aim of this work is to analyze the impact of some algorithmic features of the overlap constraint heuristics, in order to achieve the objectives at a highest degree. To demonstrate the efficiency of this approach, experimentation has been performed and compared with similar cases, tuning the GA parameters correctly.