Revision as of 10:29, 3 December 2024

Abstract

Urban underground public spaces are relatively enclosed, with high daily passenger flow and diverse directions. Enhancing evacuation capabilities through dynamic guidance systems is a crucial solution. Therefore, this paper proposes a double-layer computational framework based on meta-heuristic integration of a bi-objective evacuation path optimization method. Firstly, the outer layer computation explores the population by combining the genetic algorithm and simulated annealing. It uses temperature parameters to probabilistically accept worse solutions, enhancing global search capability and avoiding local optima. Secondly, the inner layer calculates the fitness value of each solution from the outer layer, considering both the shortest path and minimum risk objectives. The path optimization is carried out by A∗ algorithm on the basis of constructing the safety matrix by using breadth-first search algorithm, and the distance matrix by using Dijkstra’s algorithm. Finally, this method can search for the shortest safe path, effectively guiding people safely from the starting point to the exit. The experiments show that compared to traditional evacuation path planning methods, the proposed method significantly improves path optimization capabilities, quickly planning the shortest and safest evacuation route. It can provide guidance for fire safety and emergency plans for evacuations.

Full document