Search

The awareness of the need for environmental protection is increasing throughout the world. The focuses of green supplier selection are on considering environmental criteria and strengthening the competitiveness of the entire supply chain. The purpose of this study is to develop a green supplier selection procedure for the thin film transistor liquid crystal display (TFT-LCD) industry using polarizer suppliers as an example. First, a decision framework for green supplier selection is developed based on literatures and the supplier audit forms provided by an anonymous flat panel display manufacturer in Taiwan. Then, a hybrid multiple criteria group decision-making (MCGDM) method is proposed based on analytic hierarchy process (AHP), entropy, elimination and choice expressing the reality III (ELECTRE III), and the linear assignment method to assist the manufacturer in choosing among four polarizer suppliers. The final ranking results for green supplier selection and different opinions from each department are provided. An improvement report is suggested to enhance suppliers’ performance. For the evaluation procedure, most managers emphasize the importance of current capability and the capability of research and development. Furthermore, we found that the subsidiary supplier should improve quality control competence immediately to be considered as the potential candidate of primary supplier.

We propose a new objective weight method by using intuitionistic fuzzy (IF) entropy measures for multiple attribute decision making (MADM). We utilize the nature of IF entropy to assess the attribute weight based on the credibility of data, and the concept is totally different with the traditional one. Moreover, there were many IF entropy measures which were originated with different theories. We choose a geometric concept which was introduced by Szmidt and Kacprzyk in 2001, and we combine the concept with several distance measures to form several IF entropy measures. We investigate the differences among those varied measures through the experiment simulation. According to the experiment results, the differences undoubtedly exist among those measures. Besides, we also understand the number of attributes and alternatives would influence the degree of difference among those measures.