مدل‌سازي انتخاب ساخت براي سفارش و ساخت به‌منظور انبارش در يك محيط تركيبي

MAKE TO ORDER, MAKE TO STOCK SELECTION MODEL IN A COMBINED SYSTEM

در اين نوشتار يك سيستم توليد انبارش چند‌محصولي در دو حالت ظرفيتِ توليدِ نا‌محدود و محدود، با تقاضا و زمان توليد تصادفي در نظر گرفته شده است. هدف يافتن شرايط بهينه در انتخاب حالت ساخت براي سفارش (MTO) يا ساخت براي انبارش (MTS) براي هر‌يك از محصولات، با كمينه‌سازي مجموع هزينه هاي نگه‌داري و كمبود موجودي است. فرضيات جديدي كه منطبق با شرايط دنياي واقعي در نظر گرفته شده‌اند عبارت‌اند از: امكان توليد محصولات معيوب با بازرسي بدون تاخير، توليد محصولات معيوب همراه با بازرسي تاخير‌دار. همچنين سيستم تك‌مرحله‌يي را به حالت توليد چند‌مرحله‌يي به‌صورت شبكه‌يي از ماشين‌آلات تعميم داده‌ايم. هريك از مسايل مورد بررسي با كمك سيستم‌هاي صف مدل‌سازي شده و با استخراج پارامترهاي لازم، نسبت به استخراج شرايط بهينه‌ي ساخت براي سفارش و ساخت براي انبارش اقدام شده است.

In this article, a multiproduct production-inventory system with stochastic demand and production time, with and without capacity constraint, is considered. Our purpose is to determine the optimal conditions for each product when we must select between make to stock and make to order production systems. The objective is minimizing total inventory and backlog costs. We include the new assumptions regarding the models found in the literature by considering imperfect quality and delayed inspection. Also, the results of a single stage are extended to a multi stage production system, where there is a network of machines. The problems are modeled using queue theory principals. Consider a manufacturer that produces different items with stochastic demand. The processing times of an item at each stage of the production process are random variables. If there are a great number of parallel processors at each stage, then, an infinite capacity for the manufacturing system is assumed. In this paper, first, we considered a single stage production system with infinite capacity. There is a given probability to produce a defective item. The detected defective items are not recycled into the system and are waste items. The inspection time is very short, compared with processing time, and is ignorable. The processing times and demands follow the Poisson probability distribution. The second developed model is similar to the first, but, with one difference. Here we limited the capacity of the production. The third model is an extension of the second one, where we consider a non-zero inspection time of the final product. The results show that inspection time has no influence on the optimal policy. Finally, we developed multi stage models in two cases: infinite and finite capacity. In multi stage systems, there is a probability for a given item to continue the process from one stage to the next. Also, the number of necessary stages to complete an item is a random variable, and, at each stage, there is a probability that an item will go directly to the inventory store. Consideration of machine breakdown probability and limited storage capacity could be suitable topics for further research.