Prediction on power energy mix of China based on neural network model

Power source planning is an important measure to guarantee reasonable development of energy mix. This paper proposed a novel method to predict power energy mix based on fuzzy mapping and neural network algorithm. Firstly factors which influence power energy mix were categorized into four kinds, and fuzzy mapping method was applied to quantize their influence on energy mix. Then installed generating capacity of different generation types in China were forecasted by neural network, and forecasting results were verified by historical data. Finally power energy mix from 2012 to 2030 were forecasted, which shows that in China the rate of hydropower and wind power will increase steadily, while the rate of thermal power will reduce gradually in the future at the premise of no significant changes of policy and no important technology breakthrough.