| In this research, two lot-streaming flow shop problems are addressed. The first problem is the n-job, m-machine scheduling problem with distinct job due dates. The objective is to minimize the mean weighted absolute deviation (MWAD) from due dates. This performance criterion has been shown to be non-regular and requires a search among schedules with inserted idle times to find an optimal solution. For a given job sequence, linear programming (LP) formulations are presented to obtain optimal sublot completion times for cases where buffers between successive machines have limited or infinite capacities, and sublots have equal-size or are consistent. A no-wait flow shop problem is also considered. Pairwise interchange (PI) methods based on several rules to generate different initial sequences and neighborhood search mechanisms are used to search for the best sequence.; PI methods have been widely used to find best sequences. They often produce good solutions for small-sized scheduling problems with a regular performance measure. However, as the size of scheduling problems becomes larger, they depend mainly on an initial sequence and mostly fall on local optimum solutions. In this research, a hybrid genetic algorithm (HGA) is developed to escape local optimum solutions and to search for the near-optimum solutions. HGA incorporates a genetic algorithm and PI method.; Genetic algorithms (GAs) have produced good solutions in various application areas, but they have not succeeded in certain cases due to premature convergence, which results from the selection operator of GAs. In this research, a new genetic algorithm (NGA) is developed to overcome the premature convergence. NGA replaces the GAs' selection and mating operators by new operators (marriage and pregnancy operators). NGA also adopts the idea of inter-chromosomal dominance and individuals' similarity.; The second problem is the n-job, m-machine, equal-size sublot, infinite capacity buffer, lot-streaming flow shop scheduling problem with due date tolerances. The objective is to minimize the mean absolute deviation (MAD) of job completion times from their distinct due dates with tolerances. For a given job sequence, an LP formulation is presented to obtain optimal sublot completion times. The NGA is used to search for the best sequence. Finally, some future research problems are presented. (Abstract shortened by UMI.)... |
