Optimization Scheduling Strategy for Integrated Energy Systems with Multi-Stack Fuel Cells Based on Deep Reinforcement Learning

Fuel cells play an increasingly important role in integrated energy systems as a crucial component in hydrogen-to-electricity conversion. However, traditional single-stack fuel cells suffer from insufficient capacity, low efficiency, and poor reliability, making them inadequate for supporting system energy demands. Consequently, introducing multi-stack fuel cells is essential for the economic and efficient operation of hydrogen-integrated energy systems. This study first establishes variable working condition models for electrolyzer and multi-stack fuel cells, an electrochemical equipment degradation model, and linear models for other components. Then, an integrated energy system optimization operation model is formulated, aiming to minimize economic costs while considering the characteristics of coupled equipment as constraints. Finally, to address the mixed-integer nonlinear model, a deep reinforcement learning algorithm is employed to achieve rapid model resolution and system optimization. The results demonstrate that the proposed method effectively enhances the economic efficiency of hydrogen-integrated energy systems.