A model of pedestrian activity and destination choices using observational data

The aim of this study is to investigate the pedestrian activity and destination choices at congested areas in Hong Kong. A modeling system is framed to determine the relationship between travel demand and pedestrian activity utility. Four decision models are recommended namely: Stay or leave choice, activity choice, shop type decision and destination choice models. Pedestrian stay or leave decision model is calibrated in this study using observational data collected from the survey. Factors affecting pedestrian travel choices are identified such as relative and absolute times. A binary logit model has been adopted to determine how the two types of travel and time of day affect pedestrian decision. In order to collect relevant data to calibrate the proposed model, Causeway Bay, being one of the most congested areas in Hong Kong is selected as a case study. A tracking survey has been conducted in this study area, in which the walking paths of sample pedestrians are traced. Data collected during the tracking survey are used to calibrate the utility functions of pedestrian activity and destination choices. The calibration results obtained indicate that the utility of staying in the area decreases with increasing both relative and absolute times. This also stimulates the increment of leaving the study area. It is felt that by using this approach, urban planners will be able to predict pedestrian travel choice and subsequently be able to offer a more effective estimation of pedestrian origin-destination matrices and link flows in congested urban areas.