This study numerically explores the flow physics associated with nanosecond pulsed plasma actuators that are designed to control shock-wave induced boundary-layer separation in a Mach 2.8 supersonic flow. By using two dielectric barrier surface discharge actuator configurations, parallel and canted with respect to the flow velocity vector, a previous experiment suggested that the actuator worked in two ways to influence the interaction: boundary layer heating and vorticity production. The heating effect was enhanced with the parallel electrode and made the boundary-layer separation stronger, while the canted electrode produced vorticity and suppressed the boundary-layer separation due to the momentum transfer from the core flow. Because the detailed physical processes are still unclear, in this paper a numerical investigation is undertaken with a large eddy simulation and an energy deposition model for the plasma actuation, in which the dielectric barrier discharge produced plasma is approximated as a high temperature region. The flow characteristics without the plasma actuation correspond to the experimental observation, indicating that the numerical method successfully resolves the shock-wave/boundary-layer interaction. With the plasma actuation, complete agreement between the experiment and calculation has not been obtained in the size of shock-wave/boundary-layer interaction region. Nevertheless, as with the experiment, the calculation successfully demonstrates definite difference between the parallel and canted electrodes: the parallel electrode causes excess heating and increases the strength of the interaction, while the canted electrode leads to a reduction of the interaction strength, with a corresponding thinning of the boundary layer due to the momentum transfer. The counter flow created by the canted actuator plays an important role in the vortex generation, transferring momentum to the boundary layer and, consequently, mitigating the shock induced boundary layer separation.