Zone-based many-objective building decarbonization considering outdoor temperature and occupation uncertainty

Abstract

Operational building decarbonization is challenging due to complex weather conditions and occupation uncertainty. This paper introduces a precise optimization framework integrating the building information modelling technique and intelligent algorithms to dynamically predict multi-scenario building carbon emissions and optimize the building emissions performance considering building zones and weather conditions. Firstly, the on-site data and building information modeling-supported building emissions simulation data are collected for model training. Secondly, using intelligent algorithms, zone-based hourly carbon emissions and multi-scenario carbon emissions prediction are conducted simultaneously. Thirdly, the optimal decarbonization strategies are conducted using intelligent algorithms under various weather conditions. This framework has been verified for decarbonization in a high-rise operational building. The results reveal that: (1) The carbon emissions prediction is highly consistent with the ground truth after considering the sub-zone correlations; the R² for the west, south, and east zones are 0.900, 0.900, and 0.942, respectively; (2) The surrogate models can accurately predict carbon emissions and thermal comforts with all R² are higher than 0.912. The optimization rate of the building reaches 59.2 % while the outdoor temperature is above 35 °C. (3) In the many-objective optimization model, considering occupation uncertainty makes the strategy close to the actual situation, reaching the decarbonization by 6815.23 kg compared to the empirical operation for the cooling period. This work provides a new path for operational building precise management and control-oriented optimization decarbonization.