<p>In contemporary studies about historical communities, there exists a notable lack of emphasis on the aspects of comfort and greenhouse gas emissions in both indoor and outdoor physical environments. Furthermore, a limited number of investigations have been conducted to exploring optimization strategies for estimating the interior and outdoor physical environment inside these old communities. Thus, this study implements a stochastic simulation of microenvironment comfort and greenhouse gas emission of residential building using Monte-Carlo and improved hybrid intelligence paradigms. The proposed method is an integrated paradigm of artificial neural network (ANN) and improved grey wolf optimizer (IGWO), i.e., ANN-IGWO. Based on the analysis of the thermal environment of low- and medium-rise old societies in five representative cities of China, hybrid intelligence paradigms are proposed for building energy modelling. The result of the ANN-IGWO was compared with other hybrid ANNs constructed with grey wolf optimizer, particle swarm optimization, ant colony optimization, salp swarm algorithm, whale optimization algorithm, and spotted hyena optimizer. Additionally, the performance of three standalone models was compared. As per results, it is evident that the proposed ANN-IGWO can be effectively employed as a robust tool for assessing the microenvironment comfort levels and greenhouse gas emissions in residential areas. The proposed model will additionally function as a scientific foundation for energy conservation and emission reduction, while also offering decision-making approaches for the advancement of old societies in China.</p>