AI-Guided Passive Design Optimization for Solar Responsive Residential Heating Across Distinct Climate Zones

Space heating remains a consequential component of residential energy demand across many climates and persists as a seasonal load even in regions where cooling dominates annual consumption. This study examines the extent to which AI-guided passive design optimization can reduce residential heating demand when envelope and solar-responsive parameters are considered in isolation. A standardized single-story residential prototype is simulated across three climatic contexts: (a) Riyadh, representing a hot-dry environment; (b) Barcelona, representing a temperate environment; and (c) Toronto, representing a cold-humid environment. The analysis combines dynamic building energy simulation with multi-generation parametric optimization based on evolutionary search. The research objective is to minimize annual space heating demand under fixed comfort conditions. Cooling is intentionally excluded, and heating demand is modeled through an ideal loads approach to focus on effects related to the building's envelope and solar gains. Under these controlled assumptions, the optimization leads to substantial reductions in heating demand across all climates, ranging from approximately 43% in cold conditions to high relative reductions in the hot and dry case. The resulting optimal solutions demonstrate how passive design strategies vary by climate. The findings support AI-guided passive optimization as a transparent decision-support approach in the residential early design stage.