Dynamic Path Optimization with Real-Time Information for Emergency Evacuation

In order to find the optimal path for emergency evacuation, this paper proposes a dynamic optimization algorithm based on real-time information search and it takes fire accident as an example introduce principle. Before accidents, uses Dijkstra get prior evacuation network which includes paths from each node exit port. When accidents occur, evacuees are unable pass through passage where point blocking located, then proposed method breadth-first strategy solve problem network, dynamically updates according information. Because global port, only searches local avoid blockage or dangerous area. online solves pathfinding entire topology is offline calculation, can in short time when situation changes. The simulation tests performances of with different situations building, results show that effective faces changes caused by factors such evacuee size, people distribution, location, points.