An evidential reasoning-based stochastic multi-attribute acceptability analysis method for uncertain and heterogeneous multi-attribute reverse auction

Multi-attribute reverse auction has been frequently adopted by manufacturers or governments to purchase goods or services. In order to address the multi-attribute reverse auction problem with imprecise and heterogeneous information, this study introduces an evidential reasoning-based stochastic multi-attribute acceptability analysis (ER-SMAA) method. Firstly, quantitative evaluations are transformed to belief degrees on a pre-defined set of evaluation grades using the imprecise Simos-Roy Figueira (SRF) method. The SRF method is also adopted to sample different sets of attribute weights compatible with the preferences of experts. Then, the evidential reasoning approach is used to fuse evaluations. Regarding the plurality of rankings obtianed by possible transformed performances and possible sets of attribute weights, the stochastic multi-attribute acceptability analysis (SMAA) is applied to draw robust conclusions about the ranking of providers. A numerical example concerning the winner determination for clean energy device procurement is given to illustrate the effectiveness and robustness of the proposed ER-SMAA method.