Improving flexibility of agent‘s path selection in cellular pedestrian flow model

Publication Type:

Conference Paper


Pedestrian and Evacuation Dynamics PED (2012)


The continuous social force model and cellular automata model are the two main approaches used in simulation of pedestrian flow. In cellular automata models a static field, which stores the distance to a destination for each cell, is normally used for agent’s orientation in space. This means that agent’s next move is to the neighboring cell that is closest to the destination.

However, this approach may result in unrealistic behavior. Agents tend to select the shortest path and ignore alternative paths even if the shortest path is crowded and an alternative path would be better with respect to, e.g., time to destination. The reason is that agents’ decisions are based only on the information about the static field in the nearest cells, and information about alternative paths is simply unavailable for them. Hence, they act like alternative paths do not exist.
This work suggests a new orientation model to make agent behavior more realistic. The idea is to select the next neighboring cell to move to by using a linear combination of the static field value and a value of direction. The value of direction for a given neighborhood cell is determined by the static field value at a look-ahead distance in the direction defined by the neighborhood cell. Agent behavior is studied using different predefined look-ahead distances, and by assigning different weights to the static field value and the value of direction. The model is calibrated using data from an experimental study. With the new orientation model, the agents behave in a more natural way: the shortest path is still preferred if it is free but alternative paths start to attract the agents if the shortest path gets crowded.