Deep Reinforcement Learning for Residential EV Charging: A Comprehensive Survey

The rapid adoption of electric vehicles (EVs) introduces new challenges in managing residential charging demand, where uncoordinated behavior can increase energy costs, stress local grids, and reduce system efficiency. Deep Reinforcement Learning (DRL) has emerged as a powerful framework for developing intelligent, data-driven charging strategies that adapt to dynamic user behavior, time-varying electricity prices, and renewable generation. This paper presents a comprehensive survey of DRL techniques applied to home EV charging. We categorize existing research across three core dimensions\u2014feature engineering, algorithm selection, and environment modeling\u2014forming a unified analytical framework for evaluating and comparing DRL-based approaches. Key advancements in state representation, reward design, and control strategies are reviewed, alongside open challenges in scalability, interpretability, and generalization. Finally, the paper highlights future research opportunities including transformer-based forecasting, multi-objective DRL, and standardized simulation environments. The survey aims to guide researchers and practitioners toward developing robust, user-centric, and commercially viable DRL-based residential EV charging systems.