Abstract: In a complex building, the traditional fire evacuation system cannot plan an optimal evacuation direction in real time, and sometimes misleads people to dangerous areas. To solve this problem, this paper proposes an A*-IACO algorithm that combined A* algorithm with an improved ant colony optimization （IACO）. The A*-IACO introduces a new heuristic function to weaken the influence of heuristic values on path planning. It uses the improved pheromone increment and pheromone range to overcome the local optimal problem and at the meanwhile utilizes segmentation rule to avoid the searching suspension phenomenon. Thereby it can enhance the path planning ability in fire environments. In addition, it adopts a hybrid optimal control strategy to further improve the results. The experiment results show that the proposed A*-IACO obtains the shortest evacuation path with fewest inflection point and performs high accuracy in various fire environments. According to fire environment 4, A*-IACO achieves not only best evacuation path selection but also lowest iteration times comparing to ACO and IACO. As a result, it illustrates good performance on path planning ability.

Key words: fire, people evacuation, path planning, A*-IACO algorithm