Power Demand Forecasting and Application based on SVR

Abstract

Power as one of the most important energy to promote the development of national economy, plays an essential role in the normal operation of all aspects of society. Because of its production and use are difficult to store in large quantities, it is necessary to forecast future demand, which will become an important basis for making power development plans. Considering the complex non-linear relationship between power demand and its influencing factors, it is difficult to describe it accurately with the traditional mathematical models. In this paper, we select six major influencing factors and use the support vector machine to predict future power demand. The prediction accuracy is improved by parameter optimizing. At the same time, the simulation experiment of Shandong Province is conducted to further verify the validity and feasibility of the model.