Uncontrolled charging and discharging process for a large number of electric vehicles(EVs)could have significant negative impacts on the secure and economic operation of a power system concerned. On the other hand

the vehicle-to-grid(V2G)technology can support the EVs to participate in ancillary service provision

but the effects on the behavior of the interactive power systems for a different scheduling policy are different. Users can make decisions assisted by mobile social networking to select the scheduling strategy with beneficial effects. Based on the mobile social network analysis

a charging and discharging behavior prediction method of EVs with time-of-use(TOU)constraints is studied

which views per EV as a single individual in a social network. Subjected to the overall influence

its spatial and temporal characteristics are similar to the cross in the elementary particle group. Combined with the influence factor of mobile social networks

a predictive model of EVs charging and discharging behavior with constrains of safe operation for power system and the best interests for EVs users is established. Simulation results show that the method is valid. This work is supported by National Natural Science Foundation of China(No. 51477151

No. 51407076)

Natural Science Foundation of Hebei Province (No. F2014502050)

and the Fundamental Research Funds for the Central Universities (No. 2015ZD28).