Keywords: evacuation, cellular automata, underground shopping street.

Overcrowding in urban cities is one of urban problems in the present society. In order to solve this problem, we need to utilize urban space highly efficiently. Especially positive development and use of underground space are expected as effective measures to solve this problem. Currently underground space is used in various forms such as underground shopping areas, subway stations and underground parking lots. Also compared with the space on the ground, underground space with no weather condition effects can provide a comfortable space. Underground space, not disturbed by cars or bicycles, has also developed as a safety zone.

This study is to set up a simulation system for crowd behaviour in emergency evacuations for an underground shopping street. The aim of this study is to contribute to disaster prevention measures in underground spaces by clearly estimating or examining various incidents which cannot be predicted in an evacuation drill. Human evacuation behaviour is defined by diverse factors including walking speed, range of vision, decision making ability and impacts of disasters such as fire, floods, and others. It is considered that some of these factors have been unable to be fully addressed by our past studies [1,2,3]. Based on those points, cellular automata are adopted in this study which can effectively resolve the complexity using local interaction rules. This method enables several factors related to evacuation behaviour of a crowd to be collectively incorporated into local interaction rule at all times. Thus complicated evacuation behaviour can be reproduced. In this study a simulation system was built adopting the behaviour of selecting an exit, the impact of evacuation guides, human individuality with respect to an emergency and the impact of surrounding people and so forth. Then it was led to estimating evacuation behaviour of a crowd on an underground shopping street. The characteristic of cellular automata is that it enables complicated phenomena to be reproduced with a simple local interaction in cells. Cellular automata are not a top-down analysis, but a bottom-up analysis where complicated phenomena are analyzed by cumulating local interactions from a micro stage. This means that cellular automata have an advantage in self-organizing and reproducing various irregular and complicated phenomena.

In this method based on cellular automata, firstly a subject space is equally divided into grid cells. One cell occupied by one person has a square shape with 0.5 meters in each side. Then it is classified into nine categories as an obstruction, a man, an evacuation guide, a store, an exit of store, a passage, an exit, a source of fire and an emergency exit sign. Each category has own of state quantity. The method employed in this study to reproduce human evacuation behaviour in two ways. One is to consider factors affecting the human evacuation as situation index and another is to place restrictions on human evacuation. The factors taken in this study are the factors of basic behaviours such as the influence of distance to an exit, of the surroundings of obstructions, of going round repeatedly, of the environment, of an evacuation sign which is an outer factor, of a fire source and of the direction others face which is related to human interaction. Furthermore this study incorporated the influence of an evacuation guide as a factor to evaluate state quantity. A simulation system of crowd evacuation behaviour with cellular automaton was developed. The program developed in this study can simulate evacuation behaviors caused by the fire that occurred at a restaurant in Tenjinchikagai underground shopping street. By considering random location of fire sources, different evacuation behaviours were confirmed.

1

T. Madza, H. Otsuka, T. Chishaki and H. Uchida, "A study on simulation of evacuation behaviour at underground shopping street using Cellular automata", Journal of Social Safety Science, No.2, 95-100, 2000.

2

T. Madza, H. Otsuka, T. Chishaki and M. Ohno, "A study on simulation of evacuation behaviour considering human individual specificity and interaction at Tenjin underground shopping street", "Proceedings of the Symposium on Underground Space", Vol.8, 19-28, 2003.

3

T. Madza, H. Otsuka, T. Chishaki and M. Ohno, "A study on simulation of evacuation behaviour considering beginning time of evacuation and the influence of a guide during evacuation at underground shopping street", "Proceedings of the Symposium on Underground Space", Vol.10, 11-20, 2005.

purchase the full-text of this paper (price £20)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description
purchase this book (price £105 +P&P)