Solving multi-objective permutation flowshop scheduling problem using CUDA

In this paper we propose a parallel tabu search algorithm for the bi-criteria scheduling problem implemented on CUDA platform. The idea presented in this paper, refers to bi-criteria permutation flow shop case: minimization of total completion time (makespan) and total flow time. Proposed parallel Tabu Search algorithm uses multi-start with varying criteria weights in order to improve algorithms effectiveness. For the set of common benchmarks, proposed approach finds superior approximation of the Pareto front to other methods and obtains it in significantly shorter computation time compared to sequential methods.