Research on power load forecasting based on combined model of Markov and BP neural networks

With the development of power systems, the accuracy of electric power load forecasting plays a more important role in the safe operation of power system and raising the level of the national economy. Load forecasting is a very important element of the power system operation scheduling, an important module of the energy management system (EMS), and is the basis to ensure safe and economical operation and achieve grid scientific management and scheduling. Whether Power load forecast is accurate will also affect power system planning, programming and other management departments´ works. This paper selected the point load data of some city, used BP neural network training, network simulation and prediction, and through the amendment model of Markov to error-correction and adjust it will further enhance point load forecasting accuracy based on the BP neural network prediction. Through point load forecasting accurately we got more accurate information for planning and operation of power system electricity generation and distribution.