Based on the electric load, macroeconomic data and meteorological observation data in Huaibei of Anhui from 2012 to 2018, some equations are built by using correlation analysis, multiple regression analysis and curve fitting method. The actual power load is divided into trend load and fluctuating load. The trend load forecast model is established based on economic index after the relationship between trend load and economic data is analyzed. The fitting rate of the model reaches 99.3% . The forecast model of fluctuating load is built based on meteorological factors after the relationship between fluctuating load and meteorological factors are analyzed. The fitting rate of the model reaches 97.8%.The application of these results provides scientific basis for operation and dispatch of Huaibei Power Grid, and it has achieved obvious social and economic benefits.