On the Solvability and Implementation of Mixed-Model Just-In-Time Production System

Abstract

Scheduling problems are most primitive problems in Computer Science and Industries. Obtaining an optimal sequence in mixed-model production system under the just-in-time philosophy is one of such a challenging problem. The problem in a multilevel facility are strongly NP-hard, however, the single-level problems are pseudo-polynomial solvable. In this dissertation, developments of mixed-model just-in-time production problems are studied thoroughly. Different purposed algorithms are tested for their solvability and implementation purpose. Lastly, more practical mixed-model just-in-time sequencing problem is considered with the given set of sequences as precedence constraints. An efficient algorithm, which obtains an optimal solution for the maximum deviation objective in single level is studied and is extended as a solution for overlapping sequences.