Evaluation of productivity, efficiency and ranking of thermal power plants: an approach based on stochastic data envelopment analysis

In data envelopment analysis, different models are developed in different fields with different data for evaluation and ranking of DMUs. While in many applications issues, unit managers are faced with stochastic data, and they need a method to evaluate their supervised units in a way that can evaluate and rank such DMUs. When working with stochastic data, considering the probability of occurrence of unpredictable states (the level of error) provided by managers, the DMUs are evaluated. In this paper using Probability statistics techniques and normal distribution and the BCC model with undesirable outputs and a specific risk ofSpecified,a new stochastic model called Expected Ranking Criterion is introduced. Based on this,the stochastic cross-efficiency evaluation. Given the non-uniqueness of resulting optimal solutions, a model is introduced for rating priorities by which cross-efficiency is performed using aggressive method. The proposed model is implemented for 32 thermal power plants with stochastic inputs and undesirable outputs.