Novel artificial intelligence driven model-based control framework for solar–electric vehicle home energy optimization

Rising energy demand leads to higher electricity prices, grid instability, and increased pollution from conventional sources. The synergistic deployment of solar Photovoltaic (PV) systems with Electric Vehicle (EV) infrastructure offers a sustainable solution to meet energy needs, enhance energy efficiency, and reduce carbon emissions in residential and transportation sectors. This study presents a novel engineering application of an artificial intelligence controller: the Manual Dispatch Algorithm–Trained Deep Q-Network with Dueling architecture (MDA-TD-DQN). It is integrated into the Artificial-Intelligence-Data-Driven (AIDD) model within the triple-tiered home energy management system to enhance demand response efficiency. The proposed approach optimizes electricity costs, reduces energy consumption, and lowers grid stress by enabling intelligent load and EV scheduling while prioritizing cost-effective energy sources in a bidirectional solar–EV–grid system. The first-tier estimates household load defined by the end user. The second-tier schedules shiftable appliances to reduce energy use and align with surplus PV generation. The third-tier coordinates power exchange among sources and loads, operating in both vehicle-to-grid and vehicle-to-home modes. This study also investigates the impact of flexible working modes on predicting energy consumption and associated costs. The AIDD model is evaluated using advanced control algorithms, demonstrating strong effectiveness and adaptability across multiple operating scenarios.