AN INTEGRATED APPROACH TO SUSTAINABLE SUPPLIER SELECTION BASED ON EVIDENTIAL REASONING APPROACH AND BEST-WORST METHOD

In recent years, supplier selection is one of the vital issues in supply chain. In the past, supplier selection was only based on costs. Years after that, further than costs, factors such as quality, minimum delivery time, etc. have become important. So, it has been shown that sustainable supplier selection concerns environmental and social factors besides the economic factors that have to be considered simultaneously. Hence, supplier selection is a multiple criteria problem in which qualitative and quantitative factors must be considered. One of the causes of environmental criteria importance is the pollution which is increasing by the activities of companies. Also decreasing the available resources is another cause of such. Finally, to create a sustainable supply chain, sustainable supplier selection is very important. In this paper, a sustainable supplier selection is considered. A method is presented for the first time to evaluate suppliers with an Evidential Reasoning (ER) approach which considers the uncertainties of personal judgments. Also, in conditions that there is an absence of information in some of the sub-criteria during the comparison of suppliers, they can be compared based on the minimum and maximum utility function of the suppliers' value. At first, criteria and sub-criteria are chosen in order to evaluate and select the suppliers in sustainable supply chain based on company strategies, expertise and literature review in this field. Then, using the BMW, an effective multiple criteria decision-making method, weights of criteria and sub-criteria determined and in the end to evaluate and rank the suppliers an ER method is employed. ER approach is an algorithm based on a multi-criteria evaluation matrix and combination of forecasts Dempster---Shafer and by using that, factors of multi-level structures can be combined. Imperfect evaluation resulted from the lack of information or the inability of the assessor in performing precise judgments or the assessor's error in group decision making, is considered via the error degree of the combined information and utility intervals to determine the imperfect degree of the primary information in decision analysis are provided. In the end, a case study is provided to investigate the procedure in details.