Explicit Multiobjective Evolutionary Algorithms for Flow Shop Scheduling with Missing Operations
Diego Gabriel Rossit, Daniel Alejandro Rossit, Sergio Nesmachnow
Abstract
The impact of Industry 4.0 on production systems has significantly enhanced personalized production services for products customization, implying that production processes end up being customized as well. In this scenario, scheduling in flow shop configurations faces new challenges, since some of the products may require operations that other products do not, and the interest on problems with missing operation is renewed. This work addresses a multi-objective flow shop problem with missing operations, aimed at minimizing makespan and total tardiness. Two multi-objective evolutionary algorithms based on NSGA-II and SPEA2 are proposed to solve the problem, The experimental evaluation demonstrates that the proposed multiobjective evolutionary algorithms are able to compute accurate solutions to the problem, properly approximating the Pareto front for the studied instances. In turn, the multiobjective approach improved over a single-objective evolutionary algorithm previously developed for the problem.