Study On Target-oriented Bi-attribute User Equilibrium With Double Uncertainty

Nowadays,by using big data,Internet of Things,cloud computing and other means to carry out scientific operation on the macro flow distribution of urban traffic system to alleviate the problem of urban road traffic congestion has become a scheme that has gradually received attention in our country.And the microcosmic behavior of travelers is the key factor determining the macroscopic flow distribution of travelers in the transportation network.Therefore,the macroscopic distribution of traffic flow in stochastic transportation networks can be studied by analyzing the individual behavior of travelers to achieve the target of formulating reasonable policies.Empirical research has demonstrated that travelers’ micro-behavior is influenced by multiple attributes,and the acceptance level of these attributes further affects their path selection.In this paper,we aim to examine travelers’ responses to various attribute influences amidst the double uncertainty from the long-run perspective.Here the uncertainty comes from the supply-side risk of the traffic network and travelers’ perception errors.With this research aim,we propose a multi-attribute user equilibrium considering travel time,travel time reliability,and travel cost in the stochastic transportation network.That is,we discuss about the influences of these three attributes on travelers’ path selection.In the equilibrium,due to the incorporation of travelers’ perception errors,deterministic attributes become stochastic,and the stochastic correlation among these three attributes is captured by Gaussian Copula function with their marginal distributions: targets for the three attributes are all determined endogenously by travelers’ desirable target achievement probability.Finally,the equilibrium model is solved by method of successive averages.Numerical tests on the single OD pairs traffic network and the multiple pairs OD traffic network are conducted to show the convergence of the solution algorithm and to illustrate travelers’ target-oriented responses to different attributes.In general,the main contribution of this paper is to propose a goal-dependent analysis method and discuss the above empirical results.This research provides a theoretical foundation for stochastic traffic network design,particularly for congestion pricing design,and helps mitigate congestion in real-world traffic networks.