Agent-based charging scheduling of electric vehicles

Abstract

The electric vehicle technology intends to mitigate negative impacts of the energy challenge on the current transportation infrastructure. However, integrating a large number of such vehicles imposes a significant additional load to the grid and may overload it. This paper proposes a hierarchical event-driven multi-agent system framework for coordinated charging scheduling of electric vehicles. Household agents negotiate temporal travel patterns with substation agents to decide when electric vehicles should charge their batteries. A scalable load scheduling algorithm is proposed to schedule charging process of electric vehicles in real-time regardless of using any forecasting method. It aims to permit as many electric vehicles as possible to operate while keeping their aggregated charging energy consumption below continuous electricity-price-dependent thresholds over time. Simulations confirm that the framework benefits from charging flexibilities, reduces the charging cost, and shaves the grid's peak.