2026-11-092026-11-09https://scholars.csmu.edu.tw/handle/123456789/38623The objective of this proposal is to establish a simulation model of blood supply chain management. This framework of blood supply chain simulation model is based on the Taichung blood center of Taiwan Blood Services Foundation and its contracting hospitals. The suppliers are blood donors and the customers are hospitals’patients. This blood center is like the purchasing and distribution center of general supply chain system. The duty of purchasing center is demand forecasting and blood collection. The duty of distribution center is to receive the demand from the contracting hospitals and to assign the blood to the contracting hospitals. Because the related products of blood have many types (for instance, plasma, packed blood cells, serum, and so on), this proposal will select the packed RBC (Red Blood Cell Concentrate) as the research product in this blood supply chain system. The reason is that the packed RBC is the most frequently used products compared to the other related products of blood. Also, the packed RBC is generally used in transfusion of internal medicine/surgery. In addition, the packed RBC has the shelf life of 35 days so that the packed RBC belongs to a kind of perishable inventory. In order to obtain the parameters’ setting for the blood supply chain system based on the multi-response problem, the proposal will use the orthogonal array of Taguchi’s method to simulate the data, and then the dynamic Taguchi’s method will be used here. The blood supply chain system will compute the performances including total cost and inventory turnover rate. In this proposal, the signal factor is service level. The control factors are blood delivery policy, the maximized inventory level, the minimized inventory level, and the advertising frequency of blood donation advertisement (average arrival rate). The noise factor is demand variability. This research will use Neural network to construct the relationship among signal factor, noise factor, control factor with responses. Finally, the Genetic algorithm will be used to obtain the optimal combinational parameters’levels for setting the optimal multi-response performance. Comparison and conformation experiment will further be used to prove that the optimal combinational parameters’levels in the blood supply chain system are feasible and effective.本研究主要目的係建立一套血液供應鏈管理模擬系統，此模擬系統係以台灣血液基金會 台中血液中心所屬中區之契約醫院，進行血液供應鏈管理流程進行模擬。此血液中心類 似供應鏈系統中的聯合採購系統及分配中心，血液中心的上游供應商為捐血者，下游顧 客為中區各契約醫院。因血液供應成分多達11 種，本研究僅採紅血球濃厚液，進行計 算模擬。因此紅血球濃厚液幾乎用於內科及外科上的輸血使用，故使用上最多且最頻 繁，同時由於此產品有效期間為35 天，超過此一有效期間後即進行銷毀，故亦屬週期 性存貨。本研究將使用田口的直交表進行模擬，以獲得供應鏈的最佳參數水準設定，由 於系統模擬符合動態的供應鏈行為，所以動態的田口方法含多重績效的設定將被定義於 求解此一問題上面。本研究將採總成本及存貨周轉率當作績效評估指標，服務水準將被 定義為信號因子，血液使用政策、最大存貨量、最低存貨量、宣傳及活動次數所引起的 供給到達率為控制因子，需求率為雜音因子。本研究將以類神經網路建立信號因子、雜 音音子、控制因子及多重績效指標間的關係。最後本研究將以基因演算法進行求解已獲 得供應鏈管理的最佳參數設定，另外確認實驗及與一般田口動態的方法所得的結果也將 會被比較，以顯示本研究有獲的較佳的結果。Project