:: Volume 5, Issue 4 (3-2018) ::
jgit 2018, 5(4): 67-91 Back to browse issues page
An Agent-based Pedestrian Evacuation Modeling from a Rectangular Room Using Cellular Automata and Fuzzy Inference System
Seyed Mohammad Homayooni, Parham Pahlevani *
University of Tehran
Abstract:   (3240 Views)
In this paper, a model has been proposed for simulating the evacuation process from a rectangular room with one exit by changing the rush and the agents’ velocity, as well as changing the width of the exit and the dimensions of the room and the achieved results were compared to those of similar former researches. For a dynamic simulation in the study area, the study space was changed to be discrete and the cellular automata was used. To calculate the tendency of every agent for moving to its surrounding cells, several evacuation parameters were considered. The effective evacuation parameters of this study are the distance from exit, the distance of the agents from each other, the angle between the agent’s chosen path and the straight path toward exit, and the agent’s rush. Each of these parameters enters into a fuzzy inference system. The outputs of these systems are the tendency of the agent to move into that cell based on the considered parameter. Hence, there would be one tendency per parameter. Finally, for each cell, these values were combined based on their weights and accordingly, there would be one tendency for each surrounding cell of the agent. To obtain the results, at first, the weights of the distance from exit, as well as the angle between the agent’s chosen path and the straight path toward exit were changed and the results were discussed. Then, the rush of the agents were changed and the results were shown. Moreover, in this paper, the results of changing the exit’s width were discussed and compared with those reported by the similar practical research. Finally, by changing the dimensions of the room, our results were compared to those of the former similar research and consequently, the optimum conditions for the evacuation process have been proposed. In the situation that the agents were in medium rush and the agents were evacuating from a place of 4×9 m2, i.e., similar to the practical research done by Kretz, our results were 94.54% in line with the real evacuation data results.
Keywords: Evacuation, Rectangular Room, Agent-based Model, Cellular Automaton, Fuzzy Inference System
Full-Text [PDF 2500 kb]   (848 Downloads)    
Type of Study: Research | Subject: GIS
Received: 2016/10/8 | Accepted: 2017/08/13 | Published: 2018/03/19

XML   Persian Abstract   Print

Rights and permissions
Creative Commons License This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
Volume 5, Issue 4 (3-2018) Back to browse issues page