This paper takes the total coal energy consumption from 2000 to 2016 as the original data, analyzes the improved grey prediction model, and introduces the logistic prediction model. The weight coefficient of the combined model is determined by the minimum sum of the residuals of the combined model and the original data. The combination forecasting model is established. By verifying the prediction results of the three, the combined model is better than the single forecasting model and has better applicability. It has good guiding significanc...