This thesis describes a deterministic capacity expansion planning problem in which a firm has to meet demands for the services of two distinct but related equipments over a finite number of discrete time periods. Each equipment supplies two kinds of service demand devided into a fixed ratio. These service demands occured at both equipments are similar so that capacity conversion from only one of the two equipments into the other can be made to fulfill the similar service roll. Therefore, such demands can be met by dircet capacity addition (installation) or capacity conversion from the other equipment in idle(exess). In other words, capacity expansions can be initiated by either installation or conversion at the associated costs. And once converted, the capacity becomes an integral part of the new equipment. All the cost functions including installation cost, conversion cost and holding cost are assumed to be nondecreasing and concave. The objective is to find a policy(plan) of capacity installaton and conversion at each period over the whole planning horizon such that the total cost is minimized.
Using a network flow approach, the optimal solution properties are characterized. And a dynamic programming algorithm is then developed, which can be used to solve the problem efficiently. Thereafter the model is extended to the cases with capacity bounds and short-term capacity leasing allowed, respectively.
본 논문은 부분적인 용도변경이 허용되는 두종류의 설비에 대한 운영체계의 용량확장계획에 관한 모델을 다루었다. 모델내의 두종류 설비는 각각 두종류의 서비스 수요를 만족하며 각시점에서의 수요는 알려져 있는 경우로 하였고 설비의 용도변경은 한가지 설비만 가능토록 고려하였다. 그래서 총비용(설치비, 용도변경을 위한 전환비, 여유설비의 유지비)의 최소로 되는 상태에서의 설치용량 및 전환량을 결정하는 계획을 찾는것을 본 논문의 목적으로 하였다.
최적해를 구하기 위해서 네트 흐름을 이용한 특징들을 분석하였으며 최적해를 찾을 수 있는 동적 계획법을 제시하였다.