Multi-objective optimization design of self-adaptive thermal management smart glass window in office buildings across different climatic zones

Windows are critical pathways for heat exchange between buildings and the outdoor environment, significantly affecting energy consumption and indoor comfort. This study proposes a self-adaptive thermal management smart glass window (STMSGW) with a thermal rectification effect. It primarily integrates thermochromic hydrogel (TCH) glass with a silica aerogel insulation layer, enabling dual-mode photothermal regulation. This study aims to optimize the STMSGW design parameters and indoor operation strategies, and further evaluate its overall performance and application potential across different climatic zones. It constructs a multi-level coupled optimization framework by integrating EnergyPlus, jEplus, Latin hypercube sampling (LHS), a Back Propagation Neural Network (BPNN)-based surrogate model, Non-dominated Sorting Genetic Algorithm II (NSGA-II), and entropy weight TOPSIS method. Design variables including transition temperature ( T tran ), transmittance ( τ ), window-to-wall ratio ( WWR ), thickness of silica aerogel ( tsa ), heating setpoint ( HSP ), cooling setpoint ( CSP ), and illuminance setpoint ( lux ) were optimized to minimize energy use intensity ( EUI ), thermal discomfort hours ( TDH ), and visual discomfort hours ( VDH ). The framework was applied to eight representative cities across five typical climatic zones in China. Results show that the optimized STMSGW reduced EUI by 15.29%–49.64% and 2.38%–40.69%, TDH by 1.6%–47.94% and 3.2%–41%, and VDH by 56.92%–88.5% and 55.86%–88% compared with double glazing unit (DGU) and Low-Emissivity (Low-E) glass, respectively. The optimal parameters varied with climate and orientation, demonstrating the importance of climate-adaptive design for the STMSGW. Overall, the proposed STMSGW demonstrates improved energy efficiency as well as indoor thermal and visual comfort across different climatic zones.