New models for the stochastic parallel assembly line balancing problem

This thesis investigates stochastic parallel assembly line balancing problem involving balancing of more than one identical or similar line. The thesis aims at two primary goals. The first goal (Type I), is the minimization of the number of station for a pre-determined cycle time and task incompletion probability. For this problem, new chance-constraint mathematical models have been developed. By decreasing the number of variables for an improvement on solution efficiency of models, some definitions (Kmin, Ei, Li) have been introduced. Also, a heuristic has been developed for the solution of medium and large sized problems, and its performance has been analyzed with test problems. The second goal (Type II) is the minimization of task incompletion cost. Also, a new heuristic has been developed for this problem and its performance has been analyzed with test problems for this goal. In this study, chance-constrained mathematical models have been solved with GAMS mathematical programming package, and the heuristics have been coded with the C++ language using Borland C++ Builder 6.0 compiler.