Analysis of Crowd Behavior in Route Choice Based on Dynamic Programming

Pedestrian security has been becoming increasingly important in urban traffic because of the explosion and intensive activities of urban population. Considering the complexity of crowds' behavior, it is necessary to analyze the characteristics of the crowd and the law of crowds' route choice. The macro and micro characteristics data were obtained from video observation at the study sites and the researchers' survey, including the flow rate, speed, density of the crowd, etc. Based on the analysis of the relationship between speed and density of the crowd, a crowds' route choice model was derived according to the basic theory of dynamic programming by dividing the route choice process into different stages, selecting the state/decision variables, analyzing the recurrence relation and the expression of the optimal value function. The model was verified by the field data collected in observations and the comparison between the calculation results and field data shows the model proposed can accurately describe the pedestrian route. The model provides a mathematic method of the crowds' simulation. And the understanding of pedestrian behavior would help to improve the level-of-service of pedestrian traffic facilities in planning and design.