Hierarchical Energy Management and Charging Scheduling in the PV–CS–EV Integrated System

The integration of photovoltaic (PV) systems, electric vehicles (EVs), and charging stations (CSs) faces critical challenges, including PV intermittency, uncertain EV charging demand, and inefficient energy management. Existing strategies often overlook the forecasting precision of PV generation, the economic risks of electricity trading, and the diverse demands of EVs, leading to suboptimal performance. To address these drawbacks, we propose a two-tier management framework for PV–CS–EV systems, optimizing the CS operations by balancing profit maximization and risk minimization. The first tier employs accurate PV forecasting and power trading strategies between CS, PV farms, and the grid to mitigate economic risks from PV intermittency and market fluctuations. The second tier provides diverse charging strategies to maximize user satisfaction and profit. A key challenge lies in the complex interdependencies between the two tiers, requiring simultaneous optimization of power trading and user-specific charging scheduling under uncertainties. To tackle this, we introduce a hierarchical multiobjective reinforcement learning (MORL) algorithm, which efficiently coordinates decision tasks of both tiers through partial environment information interaction. Experimental results demonstrate the effectiveness of the proposed framework in enhancing the economic performance of PV–CS–EV systems.