This paper introduced the source of electric power big data, the steps and key links of big data technology application, and the current research on load forecasting of big data technology application on smart grid demand side were reviewed. Then combined with the example, the artificial neural network was used for load forecasting, and the flow of algorithm and the training method of model were introduced in detail. The types of influencing factors were concluded, with the new influencing factors brought by the ubiquitous Internet of Things, the smart grid demand side forecasting services were analyzed based on this example. Finally, an energy Internet architecture that integrated the ubiquitous power internet of things and smart grid in a big data environment was proposed, and a research route for applying big data technology for multi-energy demand forecasting in the future was proposed.