Multi-stage group scheduling problems with sequence dependent setups

The challenges faced by manufacturing companies have forced them to become more efficient. Cellular manufacturing is a concept that has been accepted as a technique for increasing manufacturing productivity in batch type production by efficient grouping of parts (jobs) with some similarities in processing operations into groups and sequentially matching machine cell capabilities for performing these operations. In each cell, fording the best sequence of processing the assigned groups to the cell and the jobs in each group by considering some measure of effectiveness, improves the efficiency of production.; In this research, it is assumed that n groups are assigned to a cell that has m machines. Each group includes bi jobs (i = 1, 2, ..., n). The set-up time of a group for each machine depends on the immediately preceding group that is processed on that machine (i.e., sequence dependent set-up time). The goal is to find the best sequence of processing jobs and groups by considering minimization of makespan or minimization of sum of the completion times.; The proposed problems are proven to be NP-hard. Thus, three heuristic algorithms based on tabu search are developed to solve problems. Also, two different initial solution generators are developed to aid in the application of the tabu search-based algorithms.; The lower bounding techniques are developed to evaluate the quality of solutions of the heuristic algorithms. For minimizing makespan, a lower bounding technique based on relaxing a few constraints of the mathematical model is developed. For minimizing sum of the completion times, a lower bounding approach based on Branch-and-Price (B&P) technique is developed.; Because several versions of tabu search are used to solve the problem, to find the best heuristic algorithm, random test problems, ranging in size from small, medium, to large are created and solved by the heuristic algorithms. A detailed statistical experiment, based on nested split-plot design, is performed to find the best heuristic algorithm and the best initial solution generator. The results of the experiment show that the tabu search-based algorithms can provide good quality solutions for the problems with an average percentage error of 8.15%.