Human crowd evacuation framework and analysis using look-ahead-based reinforcement learning algorithm
by Hyunsoo Lee
International Journal of the Digital Human (IJDH), Vol. 1, No. 3, 2016

Abstract: Human evacuation framework is one of representative applications using the digital human model. While the early human evacuation frameworks focused on the emergency modelling and evacuation simulation, contemporary evacuation models are evolved as control frameworks which can be used in real-time emergency situations. While many research studies propose several control methods for human evacuation, most of the models ignore the human crowdedness in the emergency situations, comparatively. As the ignorance of the evacuating crowdedness may lead to the more harmful situation, this paper proposes an effective method considering both objectives - the fastest evacuation routes and the crowd-less paths to the exit. In order to generate the shortest paths to the exit, a reinforcement learning approach is provided. The learning method generates candidate directions from current status, and then the less crowded directions are extracted using the proposed look-ahead crowded estimation method. The suggested method can be used as a real-time control algorithm for successful human crowd evacuations.

Online publication date: Thu, 20-Oct-2016

The full text of this article is only available to individual subscribers or to users at subscribing institutions.

Existing subscribers:
Go to Inderscience Online Journals to access the Full Text of this article.

Pay per view:
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.

Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of the Digital Human (IJDH):
Login with your Inderscience username and password:

    Username:        Password:         

Forgotten your password?

Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.

If you still need assistance, please email