Abstract: Aiming at the problems of large public buildings having complex structures and difficult fire evacuation, a dynamic fire evacuation system was built based on improved ant colony algorithm. The three different stages of the fire were analyzed. The optimal dynamic fire evacuation path was obtained. By mathematical simulation, the algorithm was compared with other algorithms, and the effectiveness was verified. Combining with the Internet of Things, the mobile terminal fire evacuation was designed on the Android platform. The system can guide the users to evacuate in real time and reach the safety exit quickly. The experimental results showed that the large- scale public building fire evacuation system designed can quickly and accurately find the optimal evacuation path and improve the fire evacuation efficiency.

Key words: improved ant colony algorithm, Internet of Things, evacuation, large public buildings