Simulation of unorganized group behavior in case of emergency

• Published in: Nauchno-tekhnicheskiĭ vestnik informat͡s︡ionnykh tekhnologiĭ, mekhaniki i optiki Vol. 19; no. 6; pp. 1130 - 1138
• Main Authors: Viksnin, I.I., Lyakhovenko, J.A., Tursukov, N.O.
• Format: Journal Article
• Language: English
• Published: Saint Petersburg National Research University of Information Technologies, Mechanics and Optics (ITMO University) 01.12.2019
• Subjects: emergency simulation; multi-agent systems; poorly organized groups; simulation
• ISSN: 2226-1494; 2500-0373
• DOI: 10.17586/2226-1494-2019-19-6-1130-1138

Subject of Research. The paper presents an advanced model and experimental study results considering an unorganized group behavior in case of emergency. An approach to simulation of human behavior is based on the Dirk Helbing and Peter Molnar model of social forces (Social force model). Such model takes into account the informational impact on the individual behavior of group agents. The developed model is adapted for crowd behavior description in emergency situations. Method. The essence of the proposed model of behavior lies in the idea that at every instant the agent solves the problem of finding the optimal transition direction from a current location to some safer place. The implementation of AnyLogic simulation tool is proposed as a software platform for the simulator with the built-in Pedestrian library, which gives the possibility to build models with a large amount of information about pedestrian movements (traffic). The tool used has a graphical interface, provides a description of the interaction between agents within the group and the possibility for the Java programming language application. Main Results. The proposed model is implemented in the form of a simulator using open information about the fire in the Lame Horse Club (December 5, 2009) in the city of Perm. Comparison of information about a real fire with the simulation results by the proposed method is performed, as well as according to the existing method of group behavior in a multi-level branched room, and by the method of moving agents along the shortest path. It is shown that the results obtained by the proposed model most closely match the real results. Practical Relevance. The presented model of an unorganized group behavior makes it possible to analyze the emergencies that have occurred and test existing and planned premises in order to ensure their safety.