An Effective Ant Colony Optimization-Based Algorithm for Flow Shop Scheduling

This article presents a modified scheme named local search ant colony optimization algorithm on the basis of alternative ant colony optimization algorithm for solving flow shop scheduling problems. The flow shop problem (FSP) is confirmed to be an NP-hard sequencing scheduling problem, which has been studied by many researchers and applied to plenty of applications. Restated, the flow shop problem is hard to be solved in a reasonable time, therefore many meta-heuristics schemes proposed to obtain the optima or near optima solution efficiently. The ant colony optimization (ACO) is one of the well-applied meta-heuristics algorithms, nature inspired by the foraging behavior of real ants. Different implementations of state transition rules applied in ACO are studied in this work. Meanwhile, a local search mechanism was introduced to increase the probability of escaping from local optimal. Hence, this work integrates the local search mechanism into ant colony optimization algorithm for solving flow shop scheduling problem to improve the quality of solutions. Simulation results demonstrate that the applied "random order" state transition rule used in ACO with local search integrated is an effective scheme for the flow shop scheduling problems.