Optimizing building layout to minimize the level of danger in panic evacuation using genetic algorithm and agent-based crowd simulation Conference Paper uri icon


  • Evacuation of architectural and urban spaces could in times be fatal. Current crowd simulation models are not capable of estimating the level of danger in design platforms. This paper suggests an agent-based evacuation model in Rhino-Grasshopper environment which offers a parametric design platform. This model calculates the rate of possible injuries as a function of the building layout as well as the intended evacuation speed. The model is validated by comparing its results with the benchmarks of the previous studies. Then, genetic algorithm was used to alter a parametric building layout to reach the minimum level of danger. The success of this procedure shows that it might be a good alternative to using building codes.

published proceedings

  • European Group for Intelligent Computing in Engineering, EG-ICE 2013 - 20th International Workshop: Intelligent Computing in Engineering

author list (cited authors)

  • Zarrinmehr, S., Asl, M. R., Yan, W., & Clayton, M. J

publication date

  • January 1, 2013 11:11 AM