Household plug-in electric vehicles (PEVs) are attracting much attention in demand-side management (DSM) for shifting loads from peak hours. This paper proposes a scenario for DSM programs to schedule household energy consumption considering bidirectional energy trading of PEVs. Residential community is deployed as an agent of native users in the scenario, and also the charging cost and discharging profit of PEVs are not public information among different communities. To optimize energy consumption scheduling among communities with incomplete information, a Bayesian game approach is formulated. The existence of Bayesian Nash equilibrium is proved mathematically. To implement the proposed Bayesian game approach, an effective iterative algorithm is presented. Simulation results show that the proposed Bayesian game approach can benefit all participated users, reduce the peak-to-average ratio of the overall energy demand.