Low-carbon retrofit of existing buildings for vulnerable older occupants: hierarchically validated surrogate optimization reveals energy-comfort trade-offs and radiative cold-discomfort limits

Description

This file includes the core model and the original (raw) data used in this study.

Steps to reproduce

1、Extract the dataset and install OpenStudio 3.7.0, EnergyPlus 23.2.0, and Python 3.11 or later. Install the Python dependencies from 02_workflow/requirements.txt. Set the HV_SAGO_ROOT, ENERGYPLUS_EXE, and OPENSTUDIO_EXE environment variables, and then open 02_workflow/scripts. 2、Run 00_smoke_and_extremes.py, 01_sample_and_run.py, and 01b_supplement_h25.py sequentially to check the baseline model, generate the Sobol samples, and perform the annual EnergyPlus simulations. 3、Run 02_surrogate_search_verify.py to train the surrogate models for eight outputs, screen the 960,000 feasible designs, and verify the selected Pareto candidates using exact simulations. 4、Run 03_robustness.py, 04_winter_root_cause.py, and 05_single_factor_g1_g8.py to reproduce the robustness tests, winter discomfort diagnosis, and G1-G8 single-factor analyses. 5、Run 07_synthesize_evidence.py to consolidate the results. Verify the main outputs in 04_data, particularly surrogate_holdout_audit_final.json, route_a_recommendation.json, route_a_robustness_summary.json, winter_root_cause_summary.json, and manuscript_key_metrics.csv. The complete workflow comprises 6,347 annual simulations; runtime depends on the available hardware and number of parallel workers.

Institutions

• Yangtze University

  Hubei, Jingzhou

Categories

Funders

• Ministry of Education of the People's Republic of China

  State Council of the People's Republic of China

  Beijing, Beijing
• Hubei Provincial Department of Education

  Hubei, Wuhan

Licence