| 요약2 |
When designing space remodeling, it is essential to evaluate various scenarios that reflect the demands of project stakeholders and users in terms of cost and efficiency. Preliminary verifications are necessary to minimize the economic and time costs caused by rework and operational shutdowns due to design errors. Historically, preliminary reviews relied on personal experience, which lacked objectivity and scientific basis, resulting in low reliability.
However, advancements in dynamic system analysis theories and system simulation technologies now offer new potential as verification tools. This study serves as preliminary research to develop a multilayer simulation optimized for interior design practice. It compares and analyzes Discrete Event Simulation(DES), Agent based Modeling(AB), System Dynamics(SD), and Pedestrian based Event Model(PED), which are widely used in academia and industry, to identify their characteristics. A medium-sized hospital’s health check-up center was used as a case model to perform quantitative and qualitative analyses of the simulation results of these four models under the same conditions, thereby assessing the efficacy and limitations of each model. The study found that DES, PED, and AB showed consistent results, suggesting potential for integration. However, SD exhibited time discrepancies with the other models, indicating limitations in integration and its abstract nature posed challenges for practical application in architectural systems. AB excels in simulating interactions between entities but struggles with spatial object recognition, whereas PED can recognize spatial objects but has difficulty controlling individual entity behaviors.
Therefore, a multilayer design that combines AB and PED for mutual complementarity is deemed the most effective approach for realistic architectural space system modeling. Through this research, a theoretical foundation was established for the concept of multilayer system simulation and its application in the field of interior design. To practically apply such complex system simulations, it is essential to prepare accurate base data and set reasonable precision. The use of Monte Carlo simulations to generalize data could be one solution to this challenge. |