PRESENTING A BI-OBJECTIVE MODEL FOR DESIGNING A DISTRIBUTION NETWORK IN A THREE-LEVEL SUPPLY CHAIN WITH CONSIDERING INFLATION IN A FUZZY ENVIRONMENT

To strengthen management of supply chain and increase its eciency, how to step up the research on supply chain distribution network system has become a major topic on logistics. As traditional logistics can no longer suit the requirements for great circulation due to a waste of resources, it is necessary for us to establish distribution systems with network structure, which are composed of some nodes and lines, whose activities serve as a foundation of supply chain distribution system activity. Supply chains cover everything from production, to product development, to the information systems needed to direct these undertakings. One of the important loops supply chain is distribution network. Design and analysis of the distribution network is one of the most important issues facing the distribution companies. The importance of this issue from the fact that logistics costs is in the main line items cost of sales and distribution companies. In recent years, the main problem in the design of the distribution network, minimizing of all the costs associated with it which includes the cost of production, transportation, storage of inventory, and too locating of distribution centers and routing sending products between the loops of supply chain. This study develops a mixed integer nonlinear programming (MINLP) model to design a supply chain. In this paper a bi-objective model for a distribution network in a three-echelon supply chain of multi-product multi-period is considered. The network is consist of manufacturing plants, distribution centers and customer. In this model, some factors including of climate situation on distribution network in form of the concept of reliability is investigated. In addition, in this study, with considering some costs as fuzzy and studying the eect of in ation on costs of distribution centers, is tried the proposed model is closer to real-world problems. Since this model has high complexity, so for solving this model, is used a meta-heuristic method such as Genetic algorithm. Finally the obtained results is analyzed.