Abstract: This paper formulates an energy-saving index system of pumped storage power stations, and develops a dynamic algorithm of comprehensive energy level evaluation on the stations using a comprehensive method of objective and subjective weighting. Using historical data, this method can achieve a comprehensive evaluation on the energy levels over multiple operation periods. We demonstrate the energy flow structure and operation characteristics of a pumped storage station to formulate the index system, then determine synthesized weights for each index using G1 method and vertical scatter degree method. A margin method is adopted for the observed index values, and a linear projection method is used to nondimensionalize the indicators that are quantified with a triangle fuzzy matrix. Finally, we can calculate the linearly-weighted indexes of the energy level for the power station operated in each period. The performance of this model and its evaluation results are verified via an application to a pumped storage power station.