Modelling Pedestrian Crossing Behaviour at a Midblock Section

Abstract

In general, pedestrians want to cross a road at a mid-block section instead of intersections because it is the shortest route. According to an Indian scenario, most of the crosswalks in urban areas are not signalised and controlled. In this study, the crossing behaviour of pedestrians in terms of speed at a midblock section was studied by considering several parameters. Statistical tests such as ANOVA and Pearson correlation tests were performed in order to know the effect of parameters on the crossing speed of a pedestrian. The pedestrian’s age and gender, crossing type, pattern, time, number of stops, number of two wheelers, three wheelers, light and heavy commercial vehicles, and groups significantly affect their crossing speed. A multiple linear regression model was developed by considering the most significant parameters. The mean absolute percentage error, root mean square error, and R2 values were calculated to validate the models.