Performance-driven Multi-Objective Optimization for Nursing Unit Design

Amid evolving public health challenges and increasing demands for carbon reduction, the optimization of hospital building performance has become a critical concern. This study focuses on nursing units in general hospital wards, with the aim of balancing natural ventilation, daylighting, and energy efficiency. A parametric generation and multi-objective optimization framework was developed on the Grasshopper platform, integrating spatial parameter modeling with performance simulations of natural ventilation (Butterfly), daylighting (Honeybee-Radiance), and energy consumption (Honeybee-Energy). Coupled with the Wallacei evolutionary algorithm, the framework enabled an automated workflow that facilitated coordinated optimization of multiple objectives. The results revealed pronounced trade-offs among the three indicators. Relative to the original scheme, the Average Optimal Solution improved ventilation from 0.37 to 0.51 m/s and daylighting from 58.72 % to 63.04 %, with only a slight increase in energy use intensity (about 0.7 %). The AWS-Optimal Solution achieved the highest ventilation (0.52 m/s), the AUDI-Optimal Solution delivered the best daylighting (64.29 %), and the EUIOptimal Solution minimized energy consumption (211.62 kWh/(m2 & sdot;a)) at the expense of other indicators. These findings confirm that single-objective optimization undermines overall performance, whereas multi-objective optimization ensures more balanced outcomes. The proposed approach contributes methodological and theoretical foundations for performance-driven, lowcarbon hospital design, thereby supporting the development of sustainable healthcare environments.