Real-Time Egress Model for Multiplex Buildings under Fire Based on Artificial Neural Network

When fire occurs in a large multiplex building, the direction of smoke and flames is often similar to that evacuation building occupants. This causes bottlenecks specific compartment, especially when occupant density very high, which unfortunately leads many fatalities injuries. Thus, development an egress model can ensure safe occupants required minimize number casualties. In this study, correlations between temperature with visibility toxic gas concentration were investigated through simulation on from databases for training artificial neural networks (ANN) created. Based this, ANN predict available time was developed, it estimated (ASET) accurately. addition, guide rapid routes proposed, rationality proposed verified detail application example. The provided optimal route longest margin safety consideration both ASET movement under fire.