Comparison and optimization of control policies in automobile manufacturing systems by simulation

Abstract

This thesis studies material flow control policies for automobile manufacturing systems. Various control policies are implemented in simulations of manufacturing systems to test whether they increase the efficiencies of the systems in terms of specific performance measures of interest. Among the control policies, Control Point Policy (CPP) is deeply studied, because this policy is designed for controlling complex manufacturing system with multiple product types. First, fundamental research in CPP is presented to understand the effects of the parameters on single product type manufacturing systems. Then, multiple product type, assembly-disassembly systems are studied with various control policies, including hybrid policies. Finally, a real automobile manufacturing system case study is presented, and various control policies are experimented on in the simulation model. Because the evaluations of performances are done by simulations, the speed of simulation becomes a very important problem. This thesis therefore presents a new approach to accelerating the speed of simulation.