Shahriar Khandoker

Shahriar Khandoker

Auto-generated profile, based on publicly available information

Shahriar Khandoker


Research Focus & Metrics

Academic Career Length

1Year (2026-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
25 April 2026 in SSRN Electronic Journal
Verified
Preprint
Open access
RERubayat Ecolo
Rubayat Ecolo
SKShahriar Khandoker
Shahriar Khandoker
IFIstiaq Ahmed Fahim
Istiaq Ahmed Fahim
ANAnindya Niloy
Anindya Niloy
TJTaskin Jamal
Taskin Jamal
TATareq Aziz
Tareq Aziz
The intensive diffusion of electric-vehicles (EVs) has posed serious challenges to power distribution networks, especially in urban charging settings where uncoordinated charging may cause peak-hour congestion, transformer overloading, and poor service delivery. A major challenge in EV charging management is the uncertainty in arrival state-of-charge (SOC), charging duration, and energy demand, which complicates both day-ahead planning and real-time decision-making. Existing scheduling approaches primarily focus on optimization and policy design but often assume that these parameters are known or can be approximated using simple heuristics, limiting their effectiveness under real-world variability. To fill this gap, this paper suggests a combined machine-learning-based EV charging scheduling system that explicitly links predictive intelligence and policy optimization. The proposed solution would involve using supervised machine learning models to effectively forecast the SOC of EVs at arrival, charging time, and energy consumption with reference to past charging data. Such predicted values are subsequently inscribed into a multi-layered scheduling that facilitates not only a route-based charging station shifting but also provides user-preference-conscious slot allocation while processing priority of emergency vehicles in real-time. Additionally, a demand-based time-of-use (TOU) tariff is introduced to reduce peak-hour grid stress. Simulation results demonstrate that the proposed framework significantly outperforms non-predictive methods, achieving approximately 98.9% higher energy throughput and 40% more vehicle allocation, while also reducing peak load, shortening waiting times, and improving emergency response. The most significant outcome of this work is a single framework that can be easily scaled and applied to create an effective solution for managing intelligent EV charging.

16 September 2026 in SSRN Electronic Journal
Verified
Preprint
Open access
RERubayat Ecolo
Rubayat Ecolo
SKShahriar Khandoker
Shahriar Khandoker
IFIstiaq Ahmed Fahim
Istiaq Ahmed Fahim
ANAnindya Niloy
Anindya Niloy
TJTaskin Jamal
Taskin Jamal
TATareq Aziz
Tareq Aziz
The intensive diffusion of electric-vehicles (EVs) has posed serious challenges to power distribution networks, especially in urban charging settings where uncoordinated charging may cause peak-hour congestion, transformer overloading, and poor service delivery. A major challenge in EV charging management is the uncertainty in arrival state-of-charge (SOC), charging duration, and energy demand, which complicates both day-ahead planning and real-time decision-making. Existing scheduling approaches primarily focus on optimization and policy design but often assume that these parameters are known or can be approximated using simple heuristics, limiting their effectiveness under real-world variability. To fill this gap, this paper suggests a combined machine-learning-based EV charging scheduling system that explicitly links predictive intelligence and policy optimization. The proposed solution would involve using supervised machine learning models to effectively forecast the SOC of EVs at arrival, charging time, and energy consumption with reference to past charging data. Such predicted values are subsequently inscribed into a multi-layered scheduling that facilitates not only a route-based charging station shifting but also provides user-preference-conscious slot allocation while processing priority of emergency vehicles in real-time. Additionally, a demand-based time-of-use (TOU) tariff is introduced to reduce peak-hour grid stress. Simulation results demonstrate that the proposed framework significantly outperforms non-predictive methods, achieving approximately 98.9% higher energy throughput and 40% more vehicle allocation, while also reducing peak load, shortening waiting times, and improving emergency response. The most significant outcome of this work is a single framework that can be easily scaled and applied to create an effective solution for managing intelligent EV charging.