Lightweight and Intelligent Real-Time Fire Evacuation on Mobile-WebVR Building

Abstract

Mobile-WebVR based real-time fire evacuation suffers from the limitation of the cache and the computing of the Web-browser. This paper proposes a lightweight and intelligent solution for popular fire evacuation. Firstly, lightweighting the scenario objects; secondly, lightweighting the collision detection of avatars; thirdly, lightweighting the optimal path planning for real-time evacuation. At last, a lightweight evacuation system is created with fire flattering the Mobile-WebVR (Lightweighting Fire Evacuation system, LFE system). The system uses the real data (the static scenario metro station BIM data, and the dynamic smoke data). The object data are reduced to 0.3%; 220 avatars as agents take part the fire evacuation on Mobile-Web showing the FPS value is 26; the heuristic ACO (Ant Colony Optimization) is used in the virtual reality metro station for each agent to plane path, and the LFE path planning is real-time and an effective solution.