Computational analysis and generation of traditional Chinese private gardens through space syntax and parametric design

Abstract

This research develops a methodological framework for computational analysis and generation of traditional Chinese private gardens, powered by two well-known algorithms in the field, with Space Syntax for analysis and parametric design for generation. Applying this framework, the paper commences with an analysis of two different categories of Chinese private gardens using selected Space Syntax techniques. Next, mathematical measurements derived from the analysis are used as the basis to capture essential spatial patterns in these two garden types. These quantitative results are then used to direct the development of a parametric design system to generate new design instances that share the same spatial patterns of the original traditional Chinese private gardens. The effectiveness of this computational approach is demonstrated through two case studies, Yuyuan Garden and Wangshiyuan Garden, both located in Southeast China and each representing a different category of traditional Chinese private gardens with a typical planning structure. The outcomes of the paper contribute to potential new insights about these important heritage sites, and demonstrate a formal approach to their computational analysis and generation.