In the smart grid environment

the real-time forecasting and pricing of power demand based on demand side management(DSM)is crucial for maintaining supply-demand balance and reducing the peak of power. This paper adopts the Bayesian method to forecast and update the real-time power demand information of users in smart grid. Then

a leader-follower game model is established to model the real-time price and the interaction behavior of power demand strategy between one retailer and multiple customers

and the game equilibrium is analyzed. Numerical simulation analysis shows that

compared with the real-time pricing mechanism in smart grid without updating of electricity demand forecasting

the proposed real-time pricing mechanism with forecasting update of electricity demand is more advantageous to increase users' satisfaction of electricity consumption and facilitate users to participate in DSM

while increase the profit of power retailer.