Geometry-Based Virtual Simulation for Fire Escape in Emergency Environment

摘要

The paper research the choosing of the optimal escape routes in fire emergency simulation. Aiming at the lack of fire emergency behavior description and its path pre-processing mechanism without dynamic adapting, this paper presented a novel algorithm based on geometry method in order to make the simulation according with the facts. We abstracted complex scenes as the graph model, and established an adaptive dynamic path query table to create the dynamically generated map, and proposed a model of population density to restrict the escape expectation velocity, which reflects the population movement of social group behavior in emergency situations. Experimental results show that the algorithm can not only access the dynamic optimal escape route, but also realistically describe the crowd social escape behavior under fire emergency situations.