چكيده به لاتين
Past research in the build to order supply chain (BTOSC) has focused on uncertainty in demand. Due to the nature of these chains, there is no final product warehouse and the product must be delivered to the customer as soon as it is produced. The delivery time to the customer is one of the most important parameters to be considered, and the existence of sub-piece warehouse in addition to helping to optimize this parameter, also has a significant effect on cost reduction, which has been less well researched.
Uncertainties in the model of most authors did not take into account the uncertainty in supply. This uncertainty in supply occurs more often in the presence or absence of a supplier in the real world. In previous studies, the supplier's choice has been made unilaterally; it suggests a supplier that has the lowest cost for the supply chain. It is necessary to enter the uncertainty of the type of acceptance or refusal of the supplier to participate in production in a particular time period.
In the present essay, which is based on the case study of the Islamic Republic of Iran's Ministry of Defense, a BTOSC model has been developed, consisting of three levels of supplier, manufacturer, and customer. At the level of the manufacturer, there is a sub-piece warehouse.
In order to deal with uncertainty in demand, a scenario-based optimization method has been used. In the case study, the uncertainty in demand and its tolerance caused the suppliers not to participate in the chain due to losses. Therefore, by Mulvey's method, we solved the model in such a way that the output of the model of the demand situation in each period is shown based on different scenarios. The model in general describes the time horizon of the planning whether its participation or non-participation by each supplier is his or her own. Also, the model with two objective function of reducing costs and increasing supplier participation in each scenario is converted and optimized using the augmented ꜫ-constraint to a single objective function model. The major scenario modeling problem is the high complexity of the model in this method, which is used to solve the Data envelopment analysis (DEA) method to reduce the number of scenarios.
