Abstract:
This study explores how smart charging with dynamic hourly pricing can address grid
congestion concerns arising from the growing adoption of electric vehicles (EVs) in
Pakistan. We propose an optimization technique to establish an hourly pricing model for
Pakistani distribution companies, promoting off-peak charging behavior among EV
owners. An agent-based energy management system is then introduced to facilitate
coordination between EV aggregators and the grid. This system employs machine learning
to accurately predict battery state-of-charge and integrates both grid-to-vehicle (G2V) and
vehicle-to-grid (V2G) functionalities for optimized energy flow. The model is evaluated
using real distribution network data with seasonal load variations. The results reveal that
with 10% EV penetration, hourly pricing can significantly reduce charging costs for EV
owners (up to 28% and 31% during summer and winter, respectively). Additionally, it
offers substantial relief for the grid by considerably reducing peak transformer load
compared to flat or 2-part tariffs. This research demonstrates the potential of smart
charging with dynamic pricing as a cost-effective and efficient solution for promoting EVs
in Pakistan while mitigating grid congestion.
