Abstract: In order to improve the sufficiency of power grid energy utilization and ensure the sustainable development of power grids, this paper proposes a comprehensive energy allocation optimization method based on demand uncertainty.The basic probability distribution interval is used to quantify the demand variable of cognitive uncertainty,and the probability density function is used to refer to the demand variable of random uncertainty,and by dividing the discrete probability density function,its representation method is made the same as the representation form of the cognitive uncertainty variable. The Cartesian product operation is used to construct a joint probability interval, and the relationship between demand variables with mixed uncertainties is analyzed. The full-cycle investment cost and operating cost are set as the configuration optimization goals,and the augmented Lagrangian function is decomposed into two configuration optimization goals by means of the alternating direction multiplier algorithm, and the configuration of uncertain demand variables is iteratively optimized. In the case verification analysis, the proposed method can make the comprehensive energy of the power grid fully complementary,the resources can be used efficiently,and it has a good economic optimization ability through verification of the optimization results of the electric energy resources,the amount of wind abandonment and the daily routine cost.