Analyzing and Optimizing Pedestrian Flow through a Single Route in a Topological Network

Abstract

In emergency cases, people are typically recommended to use the shortest route to minimize their travelling time. This recommendation may however not yield the optimal performance in the long run since the route may be over utilized after a certain point of time and this situation eventually causes heavy blockages. This paper thus measures the pedestrian flow performance through all available single routes in a topological network based on relevant arrival rates. The performance was measured using an M/G/C/C state dependent queuing approach which dynamically models pedestrians' walking speed in relation to their current density in a route. The analysis was based on an imaginary network consisting of various routes and topologies. For each route, its performance in terms of the throughput, blocking probability, expected number of pedestrians and expected travel time was first evaluated. The performance was then compared to each other and also compared to the flow performance if all available routes were utilized. The results indicated that the shortest route did not necessarily generate the optimal throughput and that the utilization of all available routes to flow pedestrians generated better performance. The optimal performance could be obtained if the arrival rate was controlled at a certain level.