ALGORITHM PERFORMANCE COMPARISON OF CROSS ENTROPY (CE) AND ALGORITHM PARTICLE SWARM OPTIMIZATION (PSO) ON COMPLETION ISSUES FLOWSHOP SCHEDULING

Abstract

Flowshop scheduling is scheduling the production of n-job sequence that has the same production process. There are various techniques of this problem, one of which is the cross entropy algorithm. Modified cross entropy algorithm to reduce the number of samples for the completion of NP-hard problems in crew scheduling problems showed a relatively good performance. To that end, cross entropy algorithm modifications can provide a new alternative in solving flowshop scheduling problems. This study aimed to assess the performance of cross entropy by reducing the number of samples to complete flowshop scheduling problem for m-machine n-job results from trials with a number of different job - different and the same number of machines, resulting in that the solutions resulting from the cross entropy algorithm with a decrease in the number of samples is relatively good results in problems that are large when compared with the PSO algorithms and the computing time required for problem resolution shorter than the cross entropy algorithm.

Keywords: Flowshop Scheduling, tardiness, particle swarm optimization, cross entropy