An integrated model of production and purchase with returned defective items in lot-splitting delivery environment

• Main Authors: Heisan-kai Fang, 方賢凱
• Other Authors: Shiow-yun Chang
• Format: Others
• Language: zh-TW
• Published: 2007
• Online Access: http://ndltd.ncl.edu.tw/handle/22445716224237720727

碩士 === 國立成功大學 === 工業與資訊管理學系碩博士班 === 95 === In the past, most scholars studied just-in-time lot-splitting delivery problems in well quality environment. Aiming single supplier and single buyer, scholars determined optimal production and order quantity or developed a buyer-supplier coordination model to determine optimal production/order quantity. In practice, suppliers usually produce some defective items during production process. Therefore, on the basis of coordination model, this research considers buyers take complement inspection to receive delivery size. When they find failure or flaw items, they will return defective items back to suppliers and ask suppliers renew and repair flaw items. This research assumes failure follows Exponential Distribution or Weibull Distribution and solves delivery size, the number of deliveries and production/order quantity. In the bargain, we compare decision variables of returned model with that Kim and Ha did at 2003. After that, this research studies factors that influence the value of decision variables. In conclusion, decision variables can be solved individually. In parameter analysis, when defective rate increases, delivery size will increase and the number of deliveries will decrease. In comparison, delivery size of returned model is larger than that of without returned model. If failure follows Exponential Distribution, the number of deliveries in returned model is smaller than that without returned model. On the other side, when failure follows Weibull Distribution, the number of deliveries in returned model is larger than that without returned model. Finally, the research studies how the shape and scare parameters influence decision variables. If the shape parameter increases and the scare parameter decreases, delivery size will decrease, production/order quantity and the number of deliveries will increase. On the other side, when the shape parameter decreases and the scare parameter increases, delivery size will increase, production/order quantity and the number of deliveries will decrease. If the scare parameter is fixed, as the shape parameter increase, delivery size will decrease, production/order quantity and the number of deliveries will increase.