Pedestrian Flow Modeling by Adaptive Control

Assessment tools for pedestrian walking speeds, comfort, and travel times are important for the planning and geometric design of infrastructural facilities under regular and safety-critical circumstances (evacuations). This contribution focuses on pedestrian walking behavior theory and modeling. It is assumed that pedestrians are adaptive controllers who minimize the subjective predicted cost of walking, given changing walking objectives and pedestrian attitudes. The theory asserts that pedestrians predict the behavior of other pedestrians on the basis of their observations and the assumed walking strategies of nearby pedestrians. The approach yields closed-form mathematical expressions predicting the acceleration and directional change of a pedestrian. It was found that the model was able to predict the behavior of pedestrians accurately (lane formation, walking positions, faster-is-slower effect, etc.), as shown by several application examples.