Data-driven design of elevated subway station constrained by transient thermal comfort

Elevated subway station buildings are usually semi-open spaces with lightweight transparent envelope, creating a contradiction between daylighting and thermal comfort in summer and winter. In this study, the typical elevated subway station building in the cold zone of China was selected. A multi-objective optimisation of daylighting and thermal performance for the elevated subway station buildings was implemented, and the transient comfort of the passengers under the non-dominated solution set was analysed. Finally, the optimal solution based on clustering analysis was proposed. The results show that, compared with the prototypical model, the maximum change in the station hall is the eastward window-to-wall ratio, increasing from 0.10 to 0.26 (160%). The long-axis spacing of the skylight in the connecting corridor is expanded from 3 m to 9.78 m (226%). In the middle of the platform, the long-axis window spacing of the skylight needs to be adjusted from 10 m to 37.74 m. Interestingly, the transient thermal comfort of passengers has greatly improved, especially in the connecting corridors. Meanwhile, the universal thermal climate indexes in winter and summer are increased by 18.7% and 9.22% respectively, while the useful daylight illuminance is increased by 62.73%.