Path intelligent optimization for dense crowd emergency evacuation in heritage buildings

Abstract

There is a class of heritage buildings, such as museums, palaces and castles, which are closed buildings with multiple layers and exits. There are a lot of precious heritage and cultural relics in heritage buildings, and a large number of tourists need to be received every day. When suffering from emergencies such as fires, heritage buildings will be confronted with serious consequences if individuals cannot evacuate in time. In this paper, the path intelligent optimization for dense crowd emergency evacuation is proposed in heritage buildings. Firstly, based on Artificial Fish Swarm Algorithm (AFSA), agglomeration model is used when there is no guidance of guide personnel. Secondly, single-layer evacuation model is used to classify individuals into different categories based on improved K-Nearest Neighbor (KNN) algorithm. The model can help individuals choose the nearest stair and save escape time. Thirdly, based on Dijkstra shortest path algorithm, multi-layer evacuation model is used to find the shortest distance and the optimal path for individuals to reach the exit. These models are applied to the Louvre in order to verify the rationality by calculating the approximate evacuation time. In case of emergency, the algorithm proposed in this paper can be modified according to the actual heritage buildings.