Multiple crossovers on multiple parents for the multiobjective flow shop problem

Abstract

The Flow Shop Scheduling Problem have been tackled using different techniques which goes from mathematical techniques like Branch and Bound to metaheuristics like evolutionary algorithms (EAs). Although in the real world this problem will be found more frequently with more than one objective, most work been done is based on a single objective. Evolutionary algorithms are very promising in this area because the outcome of a multiobjective problem is a set of optimal solutions (the Pareto Front) which EAs can provide in a single run. Yet another advantage of EA’s over other techniques is that they are less liable to the shape or continuity of the Pareto Front. In this work, we show three implementations of multiobjective Evolutionary Algorithms. The first one uses Single Crossover Per Couple (SCPC), while the other two use Multiple Crossover on Multiple Parents (MCMP), continuing with previous works[7, 8]. These two methods show an enhancement on the performance of the first method. Details of implementation and results are discussed.