Scheduling Flexible Assembly Lines using variants of Differential Evolution

A schedule which optimally allocates scarce resources is important to improve the performance of a Flexible Assembly Line (FAL). This paper investigates the performance of different variants of a Differential Evolution-based algorithm in solving a FAL scheduling problem. The performance of the FAL is evaluated based on the weighted sum of Earliness/Tardiness (E/T) penalties and the balance of the FAL. Using a real production data set, the control parameters of the different variants are tuned extensively to improve the performance of the algorithm. A comparison between the different variants of the Differential Evolution algorithm with current work in literature shows the superiority of DE variants.