Multi-operator hybrid genetic algorithm-simulated annealing for reentrant permutation flow-shop scheduling

Abstract

This study develops an improved hybrid genetic algorithm-simulated annealing (IGASA) algorithm to solve the reentrant flow-shop scheduling problem with permutation characteristics. The reentrant permutation flow-shop (RPFS) allows the jobs to visit certain machines more than once and has been proven to be an-hard problem. The proposed improved hybrid algorithm integrates the simulated annealing (SA) and genetic algorithm (GA) to obtain the near-optimal solutions by considering three objectives: Minimizing the makespan, the average completion time, and total tardiness. The multi-operator mechanism is proposed for the crossover and mutation operations to improve and maintain the diversity of individuals throughout the generation. The effectiveness and robustness of the proposed method are examined in the data sets of various-sized instances with different degrees of complexity. The results highlight that the proposed hybrid algorithm is a promising alternative in solving the RPFS scheduling problem.