Abstract
The aim of proposed research is to build a representative simulation model and implement experimental design and response surface methodology (RSM) to analyze the effect between system parameters and key performance indicators. Subsequently, the optimal setting of footwear production line can be determined. A simulation optimization framework is proposed and validated by an empirical study. In the optimization process, transfer batch size, material arrival frequency, number of shoe lasts and the turnover of shoe last are the selected parameters. After screening factors, they are determined that they can affect throughput, cycle time and machine utilization significantly. The first-order model is built by experimental design in order to analyze the curvature effect of model. Furthermore, the second-order model is constructed by RSM and central composite design. An optimal setting and predicted response are evaluated through the contour, surface plots and analysis of response surface. Moreover, the stationary and canonical analysis are performed to identify whether predicted response is maximum, minimum or saddle point. The empirical study is divided into two parts, single-objective and multi-objective optimization. The multi-objective is optimized by using desirability function and optimization software. The experiment result shows that both of their predicted responses are better than current setting.