A learning model for traffic assignment: incorporating Bayesian inference within the strategic user equilibrium model

摘要: This paper addresses adjusted travel route choice in the context of new transport developments and incremental traveller learning. It is assumed that can impact perceptions adjustments multiple ways. For instance, if travellers expect a project to significantly increase or decrease overall demand they may change their daily based on those expectations. Further, over time, will learn actual network demand, adapt accordingly. In particular, this employs methodological framework model day-to-day learning process road users, corresponding system performance time with focus specific developments. Travellers assume an initial distribution, incrementally update it experiences. Bayesian Inference used strategic user equilibrium compute underlying traffic assignment pattern. Numerical analysis conducted test demonstrate terms perceived path choice, times.