An Inventory Management Model for Blood Product with Chaotic Time Series Demand

• Main Authors: Huei-RuChen, 陳慧如
• Other Authors: Tai-Yue Wang
• Format: Others
• Language: zh-TW
• Published: 2015
• Online Access: http://ndltd.ncl.edu.tw/handle/67685012369594816121

碩士 === 國立成功大學 === 工業與資訊管理學系 === 103 === Blood is not only a crucial resource in hospital but also a perishable product. Therefore blood inventory management is a trade-off between shortage and wastage (expired blood). Nowadays, demographic structures for many countries have changed. Good inventory management for blood products is more and more important because blood products demand is increasing and blood donation is decreasing. The main reasons for this phenomenon are the aging population and decreasing birth rate. This is especially important in the area using blood donation as main blood sources such as Taiwan. Since the blood demand is uncertain, researchers use probability distribution to describe the blood demand in previous studies. As the deterministic nonlinear chaotic dynamic system evolved, many researchers found that there are many chaotic phenomena around our environment. In this study, the chaos characteristics of a real blood product demand is examined by the value of maximal Lyapunov exponent. Besides, a comparison study between chaotic time series method and ARIMA model is conducted for their forecasting performance. Furthermore, a decision model includes shortage, expired and holding cost for blood product inventory management in a hospital is implemented to minimize the cost of blood product inventory. In this proposed model, the state of inventory, future demand and freshness of blood which supplied by blood center are taken into account to set up the optimal order-up-to level. Finally, a real demand of platelet from a hospital is provided to verify the appropriation of this model. In addition, the sensitivity analysis is conducted to determine the influences of different parameters with regard to total cost and the optimal order-up-to level. The results show that there is no significant difference between two methods of demand forecasting. The freshness of supplied blood product is very sensitive to the optimal order-up-to level. If the freshness of blood product are the same from the supply side, the holding cost per unit is more sensitive than the shortage and expired cost per unit to the optimal order-up-to level.