In electricity market

the prediction of kW·h is very important for decision-makers. It is because that the prediction kW·h is the base of total vendition kW·h to be determined

and the base of decomposing the total vendition kW·h into small ones related to individual customers. In this paper

Changchun City

located at the northeast of People’s Republic of China

is employed to be studied. Based on historical vendition kW·h data of the city and the theory of Back Propagation Artificial Neural Network

the forecasting model is deduced. Simulation results show that the forecasting precision of ANN is higher than that of regression approach.