Hybrid genetic algorithm for multi-objective flow shop scheduling problem with sequence dependent setup time: parameter design using Taguchi's robust design method

Abstract

This paper determines optimal parameters of the proposed hybrid genetic algorithm for solving the multi-objective flow shop scheduling problem with sequence dependent setup time. The objectives considered in this study are minimisation of makespan and mean tardiness. In order to achieve these objectives, genetic algorithm is used in combination with a local search method to obtain Pareto-optimal solutions. The best parameters of the proposed algorithm are determined using the Taguchi's robust design method and the concept of utility index value. The set of parameters corresponding to the highest utility value is selected as the optimal parameters for the proposed algorithm. The analysis of results reveals that crossover type is the most influential parameter. The other parameters in the order of importance are mutation probability, crossover probability, mutation type and initial population. The hybrid genetic algorithm is applied to the benchmark problems of flow shop scheduling with sequence dependent setup time.