Impact and control of network QoS on smart grid controlled electrical vehicle charging

Novel, networked information-rich control systems are emerging to provide a stable and cost-efficient operation of future electricity distribution grids. However, the dependence on fault-prone, low-cost, and heterogeneous network technologies and architectures challenges the grid control quality. In this work, we study the impact of varying network QoS on the low voltage grid operation in an electrical vehicle charging scenario. The analyzed charging control system relies on: a) grid power sensing using smart meters via high latency power line communication and, b) charging point actuation commands disseminated via unreliable wireless links (IEEE 802.11). Based on emulation results, we quantify the maximum acceptable meter reading delay from network transmission that sufficiently minimizes load prediction error. Further, based on the introduction of a timed reliable communication protocol, it is shown how changing the trade-off in QoS parameters of delay, loss and information inconsistency can be applied to overcome degradation of controller performance.