在系統或產品的成本效益評估領域中，壽期成本的估算是相當重要的。而壽期成本的 内容包括：研發成本、生產成本、操作/維護成本三大部份；這些成本都可進一步加以分 解，並且運用預測的方法進行估算。本研究即針對操作/維護成本中之循環性維修器材成 本，以類神經網路建構「成本估算關係式」，再以五年的實證資料比較類神經估算模式 與回歸模式的適合性。

There have been a lot of researches conducted to estimate the life cycle cost of the new product and defense air systems to facilitate the management and cost-effectiveness analysis. Among the cost structure, the replenishment repair parts cost plays an important role in the defense air system. In this paper we try to incorporate the neural network methods into the Life Cycle Costing Model (LCCM) to construct the Cost Estimating Relationships (CERs) to estimate the replenishment repair parts cost. The computational results show that neural networks are comparable to the regression models in this application.