Electric Vehicle Coordination through Dynamic Virtual Power Plants

The integration of electric vehicles (EVs) in distribution grids is a possible solution to reaching the goal of a reliable and sustainable environment and electrifying the transportation system. EV integration is widely implemented by introducing the virtual power plant (VPP) concept in which EVs can be clustered and controlled together. In this way, one single VPP or aggregator model can be used to solve challenges in the grid such as issues related to power quality, system losses, and peak demand management. 

This thesis will analyse the conventional single VPP model and show the limitations of conventional models, which have inadequate use of EVs to solve grid issues. To overcome issues associated with conventional models, this thesis proposed a dynamic VPP algorithm that can cluster EVs into several different VPPs based on the EVs’ present state of charge and plug-out time. After the formation of different VPP clusters, the EV coordination and V2G optimization of each VPP cluster is formulated as a mixed integer nonlinear optimization model to maximize customer satisfaction while subjected to grid constraints. 

The proposed methodology was evaluated by MATLAB and Open-DSS simulation, and the results indicated that the proposed methodology has better grid performance than the results of the conventional single fixed VPP model.