Nonlinear Online Optimization for Vehicle-Home-Grid Integration including Household Load Prediction and Battery Degradation

This paper investigates the economic impact of vehicle-home-grid integration, by proposing an online energy management algorithm that optimizes energy flows between an electric vehicle (EV), a household, and the electrical grid. The algorithm leverages vehicle-to-home (V2H) for self-consumption and vehicle-to-grid (V2G) for energy trading, adapting to real-time conditions through a hybrid long short-term memory (LSTM) neural network for accurate household load prediction, alongside a comprehensive nonlinear battery degradation model accounting for both cycle and calendar aging. Simulation results reveal significant economic advantages: compared to smart unidirectional charging, the proposed method yields an annual economic benefit of up to EUR 3046.81, despite a modest 1.96% increase in battery degradation. Even under unfavorable market conditions, where V2G energy selling generates no revenue, V2H alone ensures yearly savings of EUR 425.48. A systematic sensitivity analysis investigates how variations in battery capacity, household load, and price ratios affect economic outcomes, confirming the consistent benefits of bidirectional energy exchange. These findings highlight the potential of EVs as active energy nodes, enabling sustainable energy management and cost-effective battery usage in real-world conditions.