Optimizing Multi-Level Shopping Mall Design Using Agent Simulation Based on Visual Perception— Taking Taskin Square in Changsha as an Example

Abstract

In the case of Changsha Taskin Square, according to the theory of space syntax, the multi-level shopping mall is translated into a two-dimensional model by connecting each floor with models of vertical path in DEPTHMAP. Then a visibility analysis is made with the model. According to the theory of ecological approach to visual perception, an agent model is built to simulate the pedestrian distribution in the multi-level shopping mall, by using the result of the visibility analysis. By simulating the pedestrian distribution of initial plan and optimized plan of Taskin Square, and comparing the simulation results, several principles and measures for optimizing the multi-level shopping mall design are verified: (1) the form of horizontal circulation space should be concise and symmetry as possible; (2) a core space should be set to connect each floor in the vertical direction; (3) elevators and escalators should be set in the central area with high visual accessibility; (4) the atrium and vertical path, like escalators and elevators, should be set near the entrance of the building, and an aisle should be set to connect them to the entrance directly.