Abstract
Our research studies the multi-generation capacity planning problem. Multi-generation capacity planning has become an interesting research subject, because many uncertainty factors affect the accuracy of planning. These factors include price uncertainties, demand fluctuation and uncertain product life cycle. The objective of this research is find a multi-generation capacity portfolio in each time period that is robust to demand, price, product life cycle uncertainties. This research models the multi-generation capacity planning problem as a Markov decision processes (MDP) problem. In this MDP model, we consider two generation of manufacturing technology. The new generation capacity serves as a flexible resource that can be used to downward fulfill the deficiency of old generation capacity. The MDP model is solved by value iteration algorithm, and the solution provides a robust capacity expansion strategy for both types of capacity. The objective of this MDP model is to maximize the expected profit under demand and price uncertainties. An efficient algorithm is also developed by action elimination techniques to reduce the computation complexity of the problem. We verify our results under different types of product life cycle setting.