|
Abstract
|
The widespread adaptation of electric vehicles will extend their impact from power systems to natural
gas and heating networks since these energy systems are delicately interconnected. This study investigates
uncoordinated and coordinated charging control strategies of electric vehicle fleets (EVFs) in network-
constrained multi-energy systems (NMES). A mixed-integer second-order cone programming (MISOCP) model
is presented to capture the nonlinearities in EVFs battery degradation cost, power flow equations of the electric
distribution system (EDS) and natural gas network (NGN). To deal with uncertain renewable energy production,
a Wasserstein-based distributionally robust optimization (DRO) framework is applied. Moreover, the behavioral
uncertainties in EVFs’ traveled distance and arrival/departure times are also taken into account. The model is
implemented on two different NMES sizes (one small and one large benchmark systems) to assess the models
functionality and scalability in real-world applications. The outcomes illustrate how smart and uncoordinated
EVF charging strategies can impact each of these networks, while the proposed DRO model leads a conservative
estimation over the probability distribution functions of renewable energy production. Most notably, using the
smart charging strategy led to 50.55% reduction in the operational costs of the small test system and 27.83%
for the large test system. Furthermore, despite the lowered risk and heightened reliability, the application with
DRO model showed 23.67% and 5.88% higher cost for the small and large systems, respectively.
|