A study on pedestrian walking behavior
Date
2005
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
University of Delaware
Abstract
In this thesis, a study on pedestrian walking behavior is conducted. A pedestrian’s walking behavior is influenced by many factors. A person’s age, gender, physical disability, size, the group of which he is part of, the environment and location in which he is walking, whether he is carrying any luggage or if he is with a child, are some of the factors which influence his walking behavior i.e. the way his interactions take place with other pedestrians and his own walking speed. This thesis focuses on the influence of personal characteristics (age, gender, size, ethnicity, social status, attractiveness/ appearance and physical disability) and travel companionship characteristics (whether walking alone or as part of a group e.g. family, friends, couple, co-workers, etc) on the pedestrian’s walking behavior. Interaction rules are developed for pedestrians based on their personal and travel companionship characteristics. The interaction rules based on personal characteristics are incorporated in a pedestrian simulation model called FEET. ☐ The approach followed has been divided into two stages. Stage I is to determine the personal and travel companionship characteristics, which affect the pedestrian walking behavior, and determine the interaction rules between pedestrians based on these characteristics. The methods used under Stage I included literature review of existing pedestrian simulation models and the research done in studying pedestrian walking behavior, observing pedestrians from a video recording of people moving in an indoor auto-show obtained from Japan, calculating walking speed of different types of pedestrians from digitized video recording data obtained from Japan, observing pedestrians in different settings like university cafeteria, shopping mall, main street and university campus, and measuring the walking speed of different types of pedestrians on campus at the University of Delaware. ☐ Based on the literature review it was found that the pedestrian simulation models available at present did not incorporate the personal and travel companionship characteristics of pedestrians while modeling their waking behavior. All pedestrians are treated as identical. There is no consideration given to group movement. All pedestrians are assumed to be walking alone. Also, the pedestrian’s scanning area in front of them was limited to only the next walking step. Empirical research has found that the pedestrians scan the area in front of them up to 10 walking steps ahead while choosing their next step. ☐ Stage II of the thesis was to incorporate the pedestrian interaction rules developed in Stage I into the FEET simulation model. The interaction rules based on the personal characteristics were incorporated into the FEET simulation. The interaction rules based on travel companionship can be incorporated into FEET once its programming code has the objects/sub-routines for handling group movement. Also, the FEET simulation was modified so that the pedestrians scanned 3 walking steps ahead in front of them while choosing the next step.