Routing Mobile Charging Stations to Alleviate Congestion in Restructured Power Markets

This paper presents a novel approach to alleviate network congestion in restructured power markets by dynamically routing Mobile Charging Stations (MCS) to critical grid locations. Traditional congestion management methods, such as Locational Marginal Pricing (LMP) and static infrastructure reinforcements, need more flexibility to address real-time congestion arising from variable demand and renewable output. Our approach leverages a Deep Q-Network (DQN) reinforcement learning framework to optimize MCS routing, dynamically positioning charging stations where they can best relieve congestion. We present the mathematical model incorporating the dynamic pricing for MCS deployment with market prices, enhancing costeffectiveness while reducing congestion. Simulation results on a test power system validate the effectiveness of the proposed model, showing significant reductions in congestion costs, improvements in Locational Marginal Prices, and lower operational costs associated with MCS deployment.