Kitlesel özelleştirme ortamında bileşen ikamesi

An optimal inventory acquisition-allocation problem for the assemble-to-order manufacturing systems is investigated in this study. This problem is formulated as a two-stage stochastic programming model with fixed recourse by taking into account the concepts of mass customization, component substitution, product modularity and shortage. According to different viewpoints and assumptions made, several variants of the base model are developed. In particular, to model the behavior of the risk-averse decision-maker, the conditional value-at-risk of the recourse actions is considered when taking the first stage component purchasing decisions. Various computational experiments are later carried out to provide insights into these inventory planning problems. Based on the experimental results, the product shortage costs, the component substitution costs, the uncertainty of total demand and market proportions for the component options, the number of components and modules in the end product structure are all identified as the key factors affecting the performance metrics such as the average supply ratio, the expected value of perfect information, and the value of the stochastic solution. On the other hand, it is shown that the optimal inventory allocation model can be converted to a minimum cost proportional flow model with disconnected networks, and hopefully, a fast executing exact solution algorithm can be devised after a careful study of the theoretical aspects of this model. In particular, it is aimed to benefit from the underlying network flow structure and modular decomposition as much as possible with this primal simplex algorithm.