인간객체 데이터 구축 및 건축계획 지능형 시뮬레이션

초록

While previous studies proposed varying frameworks and commercial applications that simulate occupants’ behaviors, a study on developing an empirical database of autonomous human objects, called Virtual-Users who respond to built environments, is not conducted much yet. Therefore, this present study aims to investigate a framework of database that generates and orchestrates autonomous behaviors of human objects and its applicability for further intelligent architectural design. As a method, this study collected authentic occupants’ behaviors in public facilities, and sorted them to each database of physical (physiological and psychological), semantic, activity, and behavioral rule (model), which relies on the previous FFU model (Kalay et al., 2014). Such database of human objects responds to physical (geometrical), semantic objects and spaces of BIM. In result, the simulation system generated quantitative measure data of circulation distances and patterns, bottleneck points, occupation duration. The measure data were also evaluated and interpreted as qualitative performances, such as accessibility, profitability, safety, and workability. This study expects that this simulation supports architects’ evidence-based exploration and decision-making on design alternatives and prototypes. Future study will adopt recent advancing AI technologies to develop valid and reliable behavioral models and predict optimal performances amongst genitive simulation results.