In this paper, we proposed a pedestrian behavior model and estimated it using precise walking trajectory data measured by laser scanner sensors, and demonstrated that the proposed model is effective for evaluating pedestrian spaces and can be used for a planning support tool.
First, a route selection model for describing pedestrians' macro walking behavior was described by routes, which minimize the walking distance to the destinations. In addition, by incorporating the psychological stress received from other pedestrians and obstacles such as walls and pillars during walking, we extended the walking model which was proposed in the previous research. By combining these two models, a pedestrian behavior model was constructed.
Next, we conducted a behavior monitoring survey of pedestrians using laser sensor scanners in a reception area for outpatients in a hospital, and measured the walking trajectory of pedestrians with high accuracy. Furthermore, the unknown parameters of the pedestrian behavior model were estimated using the precise walking trajectory data, and the walking characteristics which vary according to pedestrians' attributes (sex, patient/staff, assistant tool) were discussed. Furthermore, we validated the description accuracy of the pedestrian behavior model using the walking trajectory data.
Next, we proposed a method to evaluate the comfort and efficiency of pedestrian space from the view point of psychological stress of pedestrians. In general, it is difficult to estimate the degree of psychological stress of pedestrians by using only the frequency of pedestrian traffics of pedestrians. However, by using the pedestrian behavior model and by comparing the distribution of psychological stress values at different time periods and places, we demonstrated that it is possible to grasp the distribution of comfort and efficiency in pedestrian spaces.
Finally, another behavior monitoring survey of pedestrians was conducted in the same hospital after renovation, and we compared the measured trajectory data with the execution result of walk simulation based on the pedestrian behavior model. Specifically, by comparing the spatial distribution of pedestrian traffic frequency and psychological stress, we demonstrated that the pedestrian behavior simulation is effective and available for a building planning support tool.
In this paper, we demonstrated the characteristics of pedestrian behavior which vary according to the attributes of pedestrians. However, it is not yet confident whether the proposed model is applicable to other type of facilities where the characteristics of pedestrians and their walking behaviors are totally different from that in hospitals. Construction of a pedestrian behavior model in various facility types is our future work. Furthermore, the descriptive power of the pedestrian behavior model is still insufficient for grouped pedestrians walking together with multiple people. We would like to discuss it and modify the model in our future work.
