Cost-Effective Model-Predictive Framework for Residential Electric Vehicle Charge-Discharge Coordination

Many residences are adopting electric vehicles and installing photovoltaic systems on the rooftop of their house. Due to dynamic tariffs, it becomes complicated to manage the electricity. This paper develops a cost-effective and computationally efficient way to predict how to control the charge and discharge cycles of electric vehicles in smart homes with rooftop photovoltaic systems and changing tariff structures. The framework does not rely on complicated deeplearning architectures. Instead, it uses Holt-Winters exponential smoothing for short-term load and photovoltaic generation forecasting and trend-seasonal decomposition for electricity price prediction. This makes it possible to make accurate hourahead forecasts that can be utilized in real time. These predictions are used in a scheduling model based on linear programming that aims to lower net operating costs while still following battery state-of-charge limits, residential charger limits, bidirectional operation, and costs related to degradation. The suggested system uses real dataset to make it possible to charge at the same time when prices are low and solar output is high, and to discharge at the same time when prices are high. The results show a net daily profit of around Rs 177 and a big drop in reliance on the grid. This shows that using model predictive control-driven electric vehicle scheduling for home energy management is a practical option. The methodology offers a simple, scalable, and affordable alternative to optimization and deep learning-based approaches that require a lot of computation.