A Risk-based Model of Evacuation Route Optimization under Fire

Emergency evacuation plan plays a key role for fire risk management and successful evacuation. In this work, a topological model of evacuation routes is established and the corresponding matrix function is also proposed in order to evaluate evacuation ability. Simultaneously, risk assessment of fire scenarios is made based on numerical simulation. And on this basis, the variation laws of risk indicators such as temperature, thermal radiation, the concentration of toxic gas are analyzed in details and dynamic risk assessment of evacuation routes is made. Introducing the concept of equivalent routes, the scheme of the best route for evacuee at each location is the one along with the shortest time and minimal risk and suggested based on the Dijkstra algorithm. And then, one case is presented and result indicates that this model can aid people to avoid crowdedness and evacuate as soon as possible under fire accident. The risk-based model is also useful for the evacuation planning.