A TOC-TOB dual-track aggregation framework for heterogeneous V2G resources and its deep reinforcement learning-based decision-making

Résumé fourni par la source

Abstract Against the backdrop of high renewable energy penetration, new power systems are witnessing a surging demand for high-frequency, controllable, flexible resources. However, the actual deployment of vehicle-to-grid (V2G) technologies is characterized by significant heterogeneity and uncertainty. To balance the incentive compatibility and dispatchability for both Customer-Oriented (To-Customer, TOC) and Business-Oriented (To-Business, TOB) resources, this paper proposes a dual-track TOC-TOB aggregation mechanism. We formulate the dispatch decision-making process as a Partially Observable Markov Decision Process (POMDP) and employ a Proximal Policy Optimization (PPO) algorithm integrated with Gated Recurrent Units (GRU). By capturing temporal features, this approach maximizes profit while minimizing constraint violations. Experimental results demonstrate that the proposed dual-track mechanism effectively adapts to the distinct characteristics and preferences of both resource types. It significantly outperforms the rolling dispatch requirements of the electricity spot market, providing theoretical support and practical reference for the market-oriented utilization of grid flexibility resources and the low-carbon energy transition.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.