Application of the Max-Min Ant System on Flow Shop Scheduling Problems

Abstract

The paper investigates the effectiveness of one of the algorithms in the Ant Colony Optimization algorithm family, the MAX-MIN Ant System algorithm, on the Flow Shop Scheduling. The main feature of the algorithm is the pheromone constraint, which defines a lower and upper bound on the pheromone values. This prevents the search space from converging too quickly towards a single solution. Another important feature of MMAS is that usually only the best ant (globally best or iteration best) updates the pheromone values. This helps the algorithm to find optimal or near-optimal solutions faster, while reducing the chance of getting stuck in local minima.