Study On Traffic Assignment Models Based On Travelers’ Decision Inertia

The early traffic assignment problems supposed that travelers were perfectly rational and always chose the shortest path. With the continuous development of behavioral science, psychology and other disciplines, travel decisioners have been identified as boundedly rational. They may not always be pursuit of utility maximization but exhibit behavioral characteristics of inertia in the repetitive travel selection process. This research has focused on the inertial behavior of travel decision and given the corresponding traffic assignment models with determined and undetermined travel time.Firstly, the concept of inertia and some relevant theories are reviewed. The current researches of travel decision behavior and traffic assignment models are also summarized. In those sum-ups, inertial behavior of travel choice has been analyzed in detail and some reasons are provided to explain it from the perspective of behavioral economics and psychology.Then, this research has considered the inertia behavior in the route choice process under deterministic network at first, namely, the behavior that travelers adhere to the rule of "acceptable path" but not the shortest one. And due to individual heterogeneity, the multi-class user equilibrium model based on the "acceptable path" rule is proposed, which has expanded the exsisting models and made them more actual. In addition, we analyze some properties of the equilibrium solution sets and the influence of heterogeneity to the equilibrium, also give the corresponding nonlinear complementarity problems. At last, an example is used to help understand the attributes of equilibrium solution setsThereafter, we analyze the inertial behavior of route choice in the case of stochastic traffic network. Travelers may consider the mean and the reliability of travel time at the same time when choose a path, from which we define the generalized travel cost and "reliability value" that represents travels’ evaluation of travel risk. Route choice context is considered as a reference point so that travelers on different paths regard the "gains" and "losses" distinctively from transferring to an alternative path; meanwhile, cognitive bias, satisfactory behavior etc. make travelers only choose the acceptable path, and on different paths the acceptable thresholds of travelers are different. Those factors indicate that the reliability value is adaptive with the context and the route choice behavior is inertial. Subsequently, the user equilibrium model based on adaptive reliability value under stochastic network is proposed and the analysis to the equilibrium conditions is provided. We also give the formulation of the equivalent nonlinear complementary problem, in addition, the existence and properties of this model’s solution sets are discussed.