Inertia-Emulation-Based Fast Frequency Response From EVs: A Multi-Level Framework With Game-Theoretic Incentives and DRL
Yuyang Wan, Ning Wang, Xueshan Liu, Yanbo Wang, Frede Blaabjerg, Zhe Chen
Abstract
As the modern power grids increasingly integrate intermittent renewable energy sources, the reduction in system inertia creates challenges for frequency stability. Electric vehicles (EVs), with their rapid response capabilities and growing penetration, offer promising potential for providing virtual inertia services. In this paper, a novel fast frequency response framework with virtual inertia of EVs is proposed in a coupled power-transportation network. The proposed framework addresses the complex coordination required across multiple stakeholders through a hierarchical approach encompassing grid operators, charging stations (CSs), and EV users. At the grid level, the optimal inertia scheduling is developed to reduce system costs, considering the spatialtemporal characteristics of EVs and power fluctuations. For coordination between CSs, a game-theoretic incentive mechanism is designed to maximize collective benefits while encouraging EV participation in inertia services. In each CS, a deep reinforcement learning (DRL) based inertia control strategy can achieve the optimal inertia distribution, considering both EV battery conditions and user requirements. The simulation results illustrate the effectiveness and superiority of the proposed method. This approach provides a feasible operation solution for providing EV-based inertia in modern low-inertia grids.