Constructing prototypical building models based on the similarity theory coupled with entropy weight method

Because of the lack of established prototypical models in China, evaluations or predictions of the impact of changes in building codes, appliance and equipment standards, new building technologies, materials, or even design solutions are normally based on particular selected or customized models that vary greatly in different studies, thus jeopardizing the validity of the results. As a step to address this, this study developed a coupled method of the similarity theory and the entropy weight coefficient method to construct prototypical building models that represent the building stock. To present such a constructing method, we adopted apartment buildings in cold climates in China as an example. This research firstly collected a series of architectural design documents of apartment buildings in cold climates in China and performed a statistical analysis of their apartment layout features and other associated design data to achieve a statistical building model. Subsequently, a statistical model was introduced to further simplify the statistical building model’s dimensions and thermal properties of the building envelopes and attain the prototypical building models. The effectiveness of the extracted prototypical model for general building energy analysis is presented. Given the availability of design data and surveys, this method of constructing prototypical building models based on the similarity theory can be implemented into other types of building prototypical model development (e.g., office building, single-family house, etc.).