本研究主要是探討台灣血液供應鏈之區位定址問題，依據退化性產品(血液)之特性區分為兩階段討論本研究的數學模型，策略階段為整合供應鏈管理議題中定址與運輸問題；戰術階段則是庫存控制問題，以兩階層的供應鏈為例，由捐血中心將血液產品運送至血液倉儲，血液倉儲再將血液產品配送至各層級醫院。 本模式所考量的目標為總供應鏈成本最小與服務回應率最大，因研究模式為多目標模式，所以本研究使用基因遺傳演算法（NSGA-Ⅱ）求解混合非線性整數規劃問題，並以兩個貪婪法則運用基因遺傳演算法求取柏拉圖最佳解。 依據所欲研究問題之混合式基因遺傳演算法所建立的MATLAB程式進行相關模式求解與數據分析，瞭解數學規劃模式中相關參數與不同因子之間的變化，對整個整合性血液供應鏈網路選址問題中最適解的影響。 最後，本研究從台北捐血中心所負責之各層級醫院中找出15家潛在血液倉儲的實際位置，由設計三種個案情境將不同目標函數賦予三個不同權重數值，得知不同目標函數權重之情境明顯影響血液倉儲地點設置與數量的決策，求解出每個血液倉儲最適補貨週期與經濟訂購量，並以敏感度分析來探討各項成本的影響程度，而此研究結果可提供於決策者最適選擇方案。 The study is focused on the design of Facility Location Problem (FLP) in a logistics network for a blood supply chain in Taiwan. According to the blood characteristic of deterioration, we design a two-staged model to depict our problem. In the strategic level, we discuss the issues of facility location and transportation decisions in supply chain management. In the tactic level, we consider about the inventory control problem. In our two-staged blood supply chain, the blood center, where the blood is collected from donors, sends whole blood to a blood warehouse and then it is distributed to different levels of hospitals. Our goal of using two objectives to minimize the total supply chain cost and the maximize responsiveness level. In order to find the Pareto optimal solutions, we use Non-dominated Sorting Genetic Algorithm (NSGA-Ⅱ) to solve a mixed nonlinear integer programming problem with two greedy heuristics. We use the software of MATLAB to solve our established model. According to the computational results, we realize the effects of the optimal Pareto solutions in our problem. We also design three different scenarios to understand how different conditions affect both strategic and tactic decisions so that, the decision makers could make decisions from the computational solutions. Finally, we perform sensitivity analysis to understand the impacts on changing model parameters.
