Fuzzy neural network based scenario features extraction and mapping model for crowd evacuation stability analysis

Abstract

The scenario features of large-scale crowd evacuation are complex and uncertain, leading to difficulties in making intelligent decisions for crowd evacuation stability. Therefore, this paper utilizes a fuzzy neural network method to analyze and extract main crowd scenario features. Subsequently, the stability characteristics as evacuation historical backtracking can be the attributes for evacuation stability judgment. Finally, we can obtain mechanism causes of instability and state evolution law of crowd evacuation, to intervene and guide the large-scale crowd evacuation.