Modelling Human Decision in Fire Evacuation Simulation through BDI Based Cognitive Architecture

Abstract

Fire evacuation is important to save people live during disaster. During evacuation, people are unable to make adequate decision due to stress. Understanding human decision making in fire evacuation is needed to ensure people understand how to behave in an emergency when an unpredictable fire incident happens. Works have been done to model the human decision making in fire evacuation through mathematical modelling. We take the software model approach in which the model can reuse, ease of extension, ease of maintenance, understandable or comprehensiveness etc. Meanwhile, the models can reduce the complexity in human behaviour modelling. This paper presents the modelling of human decision making in fire evacuation through BDI based cognitive architecture. Cognitive architecture is based on cognitive sciences and aim at describing human cognitive processes as precisely as possible. To support the modelling of cognitive architecture for fire evacuation, agent-oriented approach is adopted. Specially, the Agent oriented modelling, AOM, is adopted and extended to model human decision making in fire evacuation. To validate the model, a case study of fire evacuation is presented. The case study reveals the potential usage of the agent oriented BDI modelling in modelling human decision during fire evacuation.