A classical theoretical problem in traffic is to describe the consequences of conflicting behaviour between crossing traffic flows on capacity and safety. Theories dealing with two vehicle flows based on the `accepted time gap' concept is well developed. However, this is not directly applicable for the interaction between pedestrians and vehicles, where in practice one of the two parties give way or force entry. In the absence of a well established theory, describing these interactions and conflicts, and due to the shortage of empirical results, such a concept has not been included in the computation methods of capacity for unsignalised intersections. To address to this, the project as outlined here has the objective to describe the interaction between pedestrians and vehicles as a choice between `to cross' or `not to cross' based on whether pedestrian or vehicle influences the choice under what situations or circumstances towards crossing efficiently and by taking into consideration the pedestrian safety. A probabilistic model is proposed to be developed describing the traffic behaviour at the conflict point, based on a `symmetrical' definition of the `accepted time gap'. A database is intended to be developed registering the conflict behaviour between pedestrians and vehicles besides other measurable and observable aspects in nine unsignalised urban intersections, through video based survey methods. The model is proposed to be tested based on empirical data to illustrate how the accepted time gap distribution in the conflict between pedestrians and vehicles depend on various factors of efficiency and pedestrian safety. The project is composed of two parts. Part I concerns with the theory and Part II is to be on empirical data testing and validating of the model. This report presents Part I.
Abstract