As an energy substitution analysis and prediction tool with flexible parameter structure, long-range energy alternatives planning system(LEAP)can provide strong support and guidance for guiding the work of electricity sub-stitution, and it has been widely used in a variety of energy planning and environmental analysis. Aiming at the basic application demand of accurate parameters of LEAP model, this paper proposes a specific parameter classification prediction method. First, a targeted data structure is established, and the parameters that need to be input into the LEAP model are divided into general parameters and scenario parameters according to the uncertainty of their development. Secondly, improved GM(1,1) model, an improved gray model based on background values and initial conditions, is used to predict the general parameters. Thirdly, a Gray-Monte Carlo model is proposed to predict scenario parameters and their occurrence probability. Finally, the correctness of the parameter classification and the parameter prediction model are verified by example analysis, and it is proved that the co-application of the two prediction methods improves the accuracy of the parameters and further improves the accuracy of the electric energy substitution prediction.