Parametric multi-objective optimization of rural building envelopes using NSGA-II and Honeybee–EnergyPlus simulation framework
Tian Congxiang, Nur Azfahani Ahmad, An Nisha Nur Welliana Abd Rased, Hongchun Li, Wenqing Li, Ahmed N. Abdalla
Abstract
Improving the energy performance of rural residential buildings in hot-summer/cold-winter (HSCW) regions are essential for reducing energy demand, enhancing thermal comfort, and ensuring the affordability of retrofit strategies. Many existing dwellings in these climates suffer from poor insulation, inefficient envelopes, and excessive operational energy use. This study proposes a parametric multi-objective optimization framework that integrates NSGA-II and the Honeybee–EnergyPlus simulation environment to identify optimal passive retrofit solutions for rural buildings. A representative two-story rural dwelling in Hubei Province, China, was modeled using Rhino–Grasshopper for parametric geometry, Honeybee–Ladybug for climate-responsive input generation, and EnergyPlus for dynamic thermal simulation. The model was calibrated using hourly indoor temperature data, with validation results satisfying ASHRAE Guideline 14 (NMBE <5 %, CV(RMSE) < 15 %), ensuring simulation accuracy. A global sensitivity analysis (GSA) was performed across 15 passive design variables, identifying 11 key parameters—primarily insulation thickness, window type, and orientation ratios—as critical influencers of energy use intensity (EUI), predicted percentage dissatisfied (PPD), and retrofit cost. These variables were then optimized using the NSGA-II evolutionary algorithm via the Wallacei X plugin to explore trade-offs across multiple objectives. The optimization process generated 161 Pareto-optimal solutions, balancing energy efficiency, comfort, and economic feasibility. The most well-rounded solution achieved an annual EUI of 24.3 kWh/m 2 , maintained PPD at 46.7 %, and limited retrofit investment to approximately 49,000 CNY, resulting in a simple payback period of less than eight years.