New heuristics for m-machine no-wait flowshop to minimize total completion time

This paper presents several new heuristics for the m-machine no-wait flowshop with total completion time as the criterion. The performance of the proposed heuristics is compared with those of three existing heuristics including a recently developed Genetic Algorithm. Computational experience with small and large number of jobs and processing time distributions of U(1,100) and U(1,10) demonstrates the superiority of the proposed heuristics with respect to error performance. For example, for U(1,100), number of jobs 400, and number of machines 25, the suggested proposed heuristics PH1(p) and PH3(p) yield an average percentage relative error of 0.006% and 0.257%. This is compared with an average percentage relative error of 2.764% for the best performing existing heuristic. (C) 2004 Elsevier Ltd. All rights reserved.