Research on supplier selection model based on an improved TOPSIS

摘要

The traditional TOPSIS method to initialize the initial decision matrix are all used the linear transformation methods on [0,1] interval, these methods has the problem of "only award, no punishment". So this paper uses the principle of "reward good and punishment bad", puts forward a kind of "reward good and punishment bad" transform operator which is easy calculation and practical for benefit type, cost type and interval type decision indicators. Then it proposed an improved TOPSIS decision model. Finally this evaluation model is used to make an empirical study on the supplier selection problem, and the result of the empirical study verified the validity and feasibility of this method.