期刊名称：International Journal of Advances in Soft Computing and Its Applications
印刷版ISSN：2074-8523
出版年度：2019
卷号：11
期号：2
页码：132-149
出版社：International Center for Scientific Research and Studies
摘要：Emotions are the common properties involved during emergencies. Various emotions are observed during emergency such as panic, fear, sad and anxiety which will lead to the behaviour and trajectories during the evacuation. Even there are many studies of computational simulation model which focus on crowd behavioural, yet there are still gaps of real evacuation due to the scarcity of real data and mostly assumption is made on the simulated model. In this paper, investigation on how the key emotions turn into behaviour are studied from the real emergency video. The key emotions extracted are panic and confuse which is observed from the real emergency video. These emotions are then map to psychological theories adopt from Lazarus Theory of emotions and stress. Due to the uncertainties in human behaviour, fuzzy approach is chosen to map linguistic value to speed (numerical) which will then lead to behavioural results. The simulation results show that the inclusion of emotions via fuzzy rules has resulted 63.14% accuracy compare to non-fuzzy 56.01% to real emergency trajectories data.
