Research On Decision-Making Met Hod Of Dynamic Evacuation

With the continuous increasing of modern large-scale public buildings, fire risks and fire incident frequently happen, which bring a big challenge to the fire control system of constructions. Nowadays the fire evacuation design needs to focus on safety evacuation efficiency's improvement and fire casualties'reduction.Based on the large-scale and complex of modern architecture, a new intelligent evacuation method which was called dynamic evacuation has been suggested. It is different from the traditional static evacuation depending on alarm devices and independent and fixed evacuation lights to guide people escaping. Dynamic evacuation is just that collect a series of dynamic information at the fire scene, and then get the degree of risk of all nodes through doing the information analysis and calculation work, and find an optimal evacuation being the shortest distance of minimum risk. So evacuation path changes with the environment changing.This paper focuses on the application of a method calculating an optimal path effectively, which can replace the optimal path in time and greatly enhanced the evacuation efficiency. This paper analyzes the risk factors of fire scene and the behavior characteristics of attendee, and deeply does the study of the classic route guidance algorithms like dijkstra algorithm, simulated annealing algorithm, genetic algorithm, ant colony algorithm and so on. As the ant colony algorithm can find better solutions and not focus local optimum, its performance is better than others in all matters. Therefore, this article determines the ant colony algorithm to be dynamic path evacuation optimization method ultimately. And examples prove that the method can track environmental changes in real time and output a new global optimal path in time...