Parallel Genetic Algorithm for a Flow-Shop Problem with Multiprocessor Tasks

Abstract

Machine scheduling problems belong to the most difficult deterministic combinatorial optimization problems. Since most scheduling problems are NPhard, it is impossible to find the optimal schedule in reasonable time. In this paper, we consider a flow-shop scheduling problem with multiprocessor tasks. A parallel genetic algorithm using multithreaded programming technique is developed to obtain a quick but good solution to the problem. The performance of the parallel genetic algorithm under various conditions and parameters are studied and presented.