Integrating a Human Behavior Model within an Agent-Based Approach for Blasting Evacuation

Authored by Gian Paolo Cimellaro, Stephen Mahin, Marco Domaneschi

Date Published: 2019

DOI: 10.1111/mice.12364

Sponsors: European Research Council (ERC)

Platforms: NetLogo

Model Documentation: Other Narrative

Model Code URLs: Model code not found

Abstract

Several studies on Emergency Management are available in the literature, but most of them do not consider how the human behavior during an emergency can affect the evacuation process. Therefore, the novel contribution of this article is the implementation of an agent-based model to describe the evacuation, due to a blast in a public area, integrated with a human behavior analytical model. Each agent has its own behavior that is described in a layered framework. The first layer simulates the ``agent's features{''} function. Then, an ``individual module{''} describes dynamically the emotional aspects using (i) the Decision Field Theory, (ii) a stationary stochastic model, and (iii) the results coming from a questionnaire. An agent-based model with integrated human behavior is proposed to test critical infrastructures in emergency conditions without performing full scale evacuation tests. Analyses could be performed both in real time with a hazard scenario and at the design level to predict the system response to identify the optimal configuration. Therefore, the development of the proposed methodology could support both designers and policy makers in the decision-making process.

Tags

Simulation pedestrian evacuation Social desirability Emergency response