Analysis and Prediction of the Real-time Load Characteristics of Power Grid with SVM Algorithm

Through the analysis and prediction of grid load characteristics, we can grasp the status of local power grid load operation effectively. According to the law of power grid load characteristics, the power enterprises would take effective dispatching measures to adjust power grid and avoiding the energy consumption. Based on the status of the power system and the load characteristics of the power grid, this paper analyzed the short-term load and long-term load. Using the support vector machine algorithm for the prediction and analysis of 168 data. Analyzed by the RMS error of 1.8448 and 2.4492, respectively, the simulation results show the short-term load characteristics changes are similar to the recent load data. Analyzed the annual load characteristic curve, we can see that is the "V" type, and the fluctuation of the annual load curve decreases steadily.