Abstract:
Aiming the case that the value of performance evaluation indexes may be grey data, this paper designs nonlinear standardized transformation methods for the grey evaluation indexes of benefit type, cost type and suitable type, respectively. By transforming the grey evaluation indexes′ value into the standardized scope between 0 and 1, the grey evaluation indexes with different types and different scopes can be compared with each other. The designed concavetype (convex-type) transformation method attains the strict (loose) evaluation requirements for the performance evaluation indexes. The selection principles of the curvature coefficient α are given. The updating of the parameter α can nonlinearly alter the changing trend with the objective function for evaluation indexes of different types, e.g., cost-type, benefit-type and suitable-type. A numerical example is given to illustrate the proposed method.