What question did this study set out to answer?

The research aims to develop an AI-driven strategy to optimize electric vehicle (EV) charging while ensuring power quality improvements in the grid.

March 30, 2026Open Access

Artificial intelligence-driven optimal charging strategy for EV with integrated power quality enhancement in electric power grids

Puntos clave

The research aims to develop an AI-driven strategy to optimize electric vehicle (EV) charging while ensuring power quality improvements in the grid.
Utilized a Temporal Fusion Transformer (TFT) for forecasting EV charging demands.
Employed a Proximal Policy Optimization (PPO)-based deep reinforcement learning agent for coordination of smart charging.
Integrated a multi-objective power quality optimizer with D-STATCOM capabilities for real-time power quality adjustments.
Validated the AI solution using MATLAB simulations on a distribution feeder with multiple EV chargers.
Reduced energy losses by 59.7% and load-shedding instances by 75.4%.
Improved power factor from 0.910 to 0.969.
Decreased Total Harmonic Distortion (THD) from 6.8% to 4.6%, below IEEE-519 standards.
Achieved a reduction in voltage deviation from 7.2% to 4.1%.

Resumen

The rapid adoption of electric vehicles (EVs) presents significant challenges regarding the stability of the power grid, involving increasing the peak demand as well as voltage deviation, power quality (PQ) degradation, harmonic distortion, and reactive power mismatch. This paper proposes an integrated artificial intelligence solution to optimize EV charging, utilizing predictive forecasting and adaptive control, to ensure compliance with the power quality standards of grid power. In the artificial intelligence solution, we employ a Temporal Fusion Transformer (TFT) to forecast multi-horizon charging demand, with a Proximal Policy Optimization (PPO)-based deep reinforcement learning agent to coordinate smart charging. The proposed artificial intelligence solution comprises a multi-objective power quality optimizer with Distribution Static Compensator (D-STATCOM) capabilities, allowing for real-time harmonic filtering and reactive power compensation. The proposed artificial intelligence solution was validated using the MATLAB platform, utilizing simulations comprising a 10 MVA distribution feeder with 20 EV chargers, with power ratings ranging from 7 to 150 kW. The simulation results confirmed substantial improvements in the grid’s performance, including a reduction in energy losses by 59.7%, a decrease in instances of load-shedding by 75.4%, and an improved power factor from 0.910 to 0.969. The performance measures used to indicate power quality observed substantial improvements, with the Total Harmonic Distortion (THD) reduced from 6.8% to 4.6%, limited to a maximum of 3.03% (as prescribed by IEEE-519), and an observed 7.2% to 4.1% decrease in voltage deviation. The proposed AI-based strategy optimally balances charging efficiencies, cost minimization, and power quality improvements. The proposed AI-driven approach successfully balances charging efficiency, cost optimization, and power quality enhancement, providing a scalable solution for large-scale EV integration in smart grid environments.

Preguntar a la IA

Me gusta

Guardar

Ver artículo completo