A Substantially Improved New Algorithm for Flowshop Scheduling Problem with Uncertain Processing Times

Abstract

The performance measure of total completion time (TCT) plays a key role in manufacturing to improve performance, e.g., reducing inventory levels. Moreover, since uncertainty is an inevitable part of certain manufacturing environments, it is especially important to address cases with uncertain processing times. This paper addresses the four-machine flowshop scheduling problem to minimize TCT with uncertain processing times. Due to the NP-hardness of the problem, different algorithms were presented as solutions in scheduling literature. In this paper, a new substantially improved algorithm is proposed and parameters of the algorithm are fine tuned. The proposed algorithm is compared to the best existing algorithm (RAIRO Operations Research 54, 529-553, 2020) in scheduling literature using extensive computational experiments and statistical analysis. Computational methods using the programming language python, along with statistical inference, is used to confirm the effectiveness of the proposed algorithm over the existing ones. Computational methods reveal that the proposed algorithm is, on average, 86.8% more effective than the best existing one in literature with similar computational times. A test of hypothesis further confirms the effectiveness of the proposed algorithm with a p-value of less than 0.00001, which is practically zero.