Abstract
Semiconductor manufacturing is a capital intensive industry that capacity utilization significantly affect the capacity effectiveness and final profit. Capacity portfolio planning strategy must coordinate capacity expansion and migration decisions in balance for profit maximization. Manufacturers have to determine capital investments based on various demand forecasts of products in advance for long-lead time in manufacturing process. The rapid rate of change in semiconductor technology makes it difficult for companies to estimate future tool need especially in new-generation products planning decisions. Considering human prediction uncertainty that makes predictive belief degrees far away from probability frequency when lack-of-data, this study aims to develop an uncertain regret decision (URD) strategy framework based on uncertainty theory for capacity expansion and migration problem. The objective of proposed model is to minimize the potential regret of capacity surplus/shortage under demand uncertainties. Besides, an intelligent algorithm is designed for solving the uncertain regret model. A numerical experiment was conducted to estimate the effectiveness and robustness in uncertain environment.