An Autonomous Charge Controller for Electric Vehicles Using Online Sensitivity Estimation

Abstract

Despite their sustainable benefits, large-scale adoption of electric vehicles (EVs) into the distribution system is challenging. Uncontrolled charging of EVs could increase voltage violations, system power losses, and feeder overloads. In this article, an autonomous EV charge control strategy is proposed to control EV charging in a way that mitigates their negative impacts. To ensure fair contribution from EVs connected to different nodes, both the local nodal voltage and sensitivity of voltage to load changes are used as inputs to the proposed controller. This is so because these two inputs have a complementary relationship. That is, nodes with lower voltages are generally more sensitive to load changes than those with higher voltages. In addition, a novel approach for locally estimating the voltage sensitivities online is proposed. This ensures the robustness of the proposed control strategy to changes in loading conditions and system configurations without the need for communication. An EV-rich test distribution system is modeled in the DIgSILENT PowerFactory environment and used to validate the proposed control strategy. The test results demonstrate the effectiveness and robustness of the proposed strategy considering different loading conditions, system reconfiguration, distributed generators, and voltage control devices.