Function space assignment determines the function of all spaces and is essential to the design of a remodeled building. Optimization of function space assignment that relies upon users' activities is difficult due to the lack of quantifiable and detailed data about users' activities. In this paper, an activity-based simulation process model consisting of five phases is proposed and applied to an actual project. The phases include activity data collection, activity simulation, space movement pattern mining, function-space assignment optimization, movement simulation and assessment. The process also allows the building administrator to adjust the optimized assignment to accommodate hard-to-quantify preferences via a variety of indexes such as movement distance, corridor flow, and other space usage-related indexes. Results from application of the proposed model to an actual remodeling project indicated that the model offered the administrator a better assignment than the contracted architect, with a 5.79% increase in the objective function and a 23.92% decrease in movement distance. The adjusted assignment based on service-related indices, such as noise interference, further improved service by accounting for parameters that were not easily included in the objective function for optimization.
|頁（從 - 到）||935-950|
|期刊||Computer-Aided Civil and Infrastructure Engineering|
|出版狀態||Published - 1 12月 2015|