Digital Twin-Based Intelligent HVAC Control for Underground Metro Stations under Piston Effect

This study addresses the challenge of maintaining thermal comfort and energy efficiency in underground metro stations under dynamic piston effects induced by train operations. Conventional HVAC control strategies often fail to adapt to rapid environmental fluctuations, resulting in inefficient energy consumption and unstable thermal conditions. To overcome these limitations, this paper proposes a digital twin-based intelligent control framework integrating Computational Fluid Dynamics (CFD), Artificial Intelligence of Things (AIoT), and edge computing technologies. The proposed system enables real-time environmental sensing, on-site estimation of Predicted Mean Vote (PMV) and HVAC loads, and predictive control through a digital twin model. CFD simulations calibrated with field measurement data reveal that increasing train speed from 40 km/h to 60 km/h significantly elevates tunnel pressure, leading to increased air infiltration through Platform Screen Doors (PSD) and higher ventilation loads. Based on this insight, a multi-objective optimization strategy is developed to balance thermal comfort and energy consumption. Simulation results demonstrate that the proposed framework successfully maintains PMV within the ISO 7730 recommended range of ±0.5 while achieving improved energy efficiency during peak operation periods. The proposed approach provides an effective solution for enhancing environmental control precision in underground metro systems.