With the advancement of smart grid construction, the power data generated during power production and use is becoming more and more abundant. The use of big data technology for load forecasting is of great significance in guiding the planning and operation of power systems. This paper comprehensively considers the impact of economic and meteorological factors on the load characteristics, decomposes the total load into the basic load affected by the economy and the meteorological sensitive load affected by meteorological factors. Then the linear regression method and the random forest regression (RFR) in big data technology were used to model the two. Finally, the wavelet neural network (WNN) algorithm is used to intelligently correct the prediction results. Comparing the above method with the prediction results of a region without wavelet neural network method and support vector machine (SVM) method, the proposed method has higher prediction accuracy.