Power load forecasting is an important basis for power system generation planning. Accurate load forecasting is beneficial to improve the economy and reliability of power system operation. The load of power system is related to many factors, and there is a complex nonlinear relationship between the future load and the available influence variables. 

Support vector machine (SVM) is a new machine learning algorithm, which can solve some practical problems such as small samples, nonlinearity, high dimension and local minima. Compared with short-term power load data, medium-and long-term power load data has the characteristics of small samples.